" Technical tool insight: Price breakout" BY ACTIVE TRADER STAFF (Active Trader, March 2001

)

Breakout Trading Technique article collections: BASIC and ADVANCED ..............................

2 5 9

"More bang for your buck: Patterns within patterns" BY ACTIVE TRADER STAFF (Active Trader, October 2000) "Anticipating breakouts and beating slippage" BY STEVE WENDLANDT (Active Trader, August 2000)

.............................

............................... .......................... ......................

"Trading System Lab: 100-20 channel breakout system" BY DION KURCZEK (Active Trader, June 2003) . . . . . . .

13 16 18 20 23 28 32 35 39 41 42 47 53 58 60 62 64 66
1

"Futures Trading System Lab: 100-20 channel breakout system" BY DION KURCZEK (Active Trader, June 2003) . . . . . . . . . . . "Futures Trading System Lab: 60-minute breakout system" BY VOLKER KNAPP (Active Trader, January 2004) . . . . . . .

.........................

"Futures Trading System Lab: Four-percent breakout system" BY VOLKER KNAPP (Active Trader, September 2004) . . . . . . . "Broadening patterns: Clues to breakout direction" BY THOMAS N. BULKOWSKI (Active Trader, April 2004)

.......................

.............................

"High, tight flag helps squeeze out profits" BY THOMAS N. BULKOWSKI (Active Trader, December 2004) "Mastering two-minute breakouts" BY KEN CALHOUN (Active Trader, September 2001) "Swing trading 10-day channel breakouts" BY KEN CALHOUN (Active Trader, March 2002)

..........................

..............................

.................................

"Trading System Lab: Volatility breakout system" BY THOMAS STRIDSMAN (Active Trader, October 2002)

............................. .....................

"Futures Trading System Lab: Futures volatility breakout system" BY THOMAS STRIDSMAN (Active Trader, October 2002) . . . . . . . . "Better breakout trading: The noise channel system" BY DENNIS MEYERS, PH.D. (Active Trader, September 2001)

.......................... ......................

"The long and short of it: The noise channel breakout system 2" BY DENNIS MEYERS, PH.D (Active Trader, October 2001) . . . . . . "The multibar range breakout system" BY DENNIS MEYERS, PH.D (Active Trader, January 2004) "Trading System Lab: DeMark variation" BY THOMAS STRIDSMAN (Active Trader, September 2001) "Trading System Lab: Dynamic breakout system" BY THOMAS STRIDSMAN (Active Trader, February 2003)

............................

........................... ............................ ..........................

"Futures Trading System Lab: Dynamic breakout system" BY THOMAS STRIDSMAN (Active Trader, February 2003) . . "Futures Trading System Lab: Experimenting with exits" By VOLKER KNAPP (Active Trader, June 2004) . . . . . .

...........................

"Futures Trading System Lab: Monthly breakout" BY DION KURCZEK AND VOLKER KNAPP (Active Trader, March 2004) . . . . . . . . . . . . . . . . . . . . . . . . . . "Trading System Lab: 60-minute breakout system" BY VOLKER KNAPP (Active Trader, January 2004) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ACTIVE TRADER • www.activetradermag.com

68

TRADING Basics

Technical tool insight:

Price breakout
ished its effectiveness to the point that many traders look for false breakouts (when price pushes through a breakout level, only to reverse back through it) at these levels, to take positions against the direction of the initial breakout (referred to as “fading” the breakout). Breakouts are not limited strictly to moves to new highs of a certain number of bars (i.e., 10-bar, 20-bar or 40-bar breakouts). As mentioned, price can also “break out” through the support and resistance levels of trading ranges, or other past technical milestones such as long-standing highs or lows. Figure 1 shows 40-day breakout levels on a daily chart. Figure 2 shows 20-bar breakout levels on a 10-minute chart. Figure 3 shows a breakout above the resistance level defined by a past significant high.

Price breakouts are the basis of many of the most successful trading approaches. We explain the basics of this trading technique.

T

he price “breakout” is one of the simplest — and most powerful — concepts in trading. It occurs when price moves forcefully out of a consolidation or trading range (a period of relatively narrow, sideways price movement) or pushes above or below an established price level (support or resistance), initiating either temporary followthrough or a sustained trend. The act of pushing to new highs or lows (especially if the price level in question has been repeatedly tested in the past) is evidence of strong momentum and suggests the market has the potential to continue in that direction. In other words, the basic logic behind price breakouts is that a market making new highs (and with potential for further price gain) is exhibiting strength and should be bought, while a market making new lows (and with potential for further price decline) is exhibiting weakness and should be sold. For example, the reason new 52-week highs or lows in stocks are so commonly referenced is because of the implied significance of price breaking through these levels. This concept of price movement is valid on intraday time frames as well as daily or monthly ones.

The four-week highs or lows simply represent natural resistance and support levels. This kind of trading system is often referred to as stop-and-reverse (SAR), because when a trade signal is generated, the existing position is liquidated (stopped out) and a new position (a reverse of the previous one) is established. This basic trading rule — which gained widespread popularity as the “20-day breakout” — was integral to many popular mechanized trading strategies, most famously those of futures trader Richard Dennis group of trend-followers known as the “Turtles.” Trend-following traders (especially in the futures markets) used this simple technique, or a variation of it, to exploit strong trends in the 1970s and ’80s. However, the widespread popularity of the 20-day breakout level has dimin-

FIGURE 1 DONCHIAN BREAKOUT CHANNELS, DAILY 40-day Donchian breakout levels, both high and low. A basic breakout approach is to buy when price exceeds the n-bar (in this case, the 40-day) high and sell when it falls below the n-bar low.
Highest price of last 40 bars Oracle Corporation (ORCL), daily
45

40 37.00 35

Donchian breakout levels
The term “breakout” is often associated with Richard Donchian, the first person to popularize the systematic use of breakout levels. His basic approach was called the Donchian “four-week rule,” which consisted of the following: 1. Go long (and cover short positions) when the market makes a new fourweek high (that is, when price exceeds the highest price of the previous four weeks). 2. Go short (and cover long positions) when the market makes a new fourweek low (that is, when price drops below the lowest price of the previous four weeks).
2

30 26 9⁄16 25

Lowest price of last 40 bars

21.50 20

15

27 3 10 24 31 7 14 28 6 13 20 27 3 10 24 1 8 15 22 30 5 12 19 26 3 10 17 24 31 7 14 21 28 5 11 18 25 2 9 16 23 30 6 13 27 4 Jan. 2000 Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec.

Source: QCharts by Quote.com

www.activetradermag.com • March 2001 • ACTIVE TRADER

Glossary
FIGURE 2 DONCHIAN BREAKOUT CHANNELS, INTRADAY The breakout concept is applicable to any time frame. Here, the highest 20-bar highs and lowest 20-bar lows are shown by the channel lines.
Oracle Corporation (ORCL), 10-minute
27.25 27 26 9⁄16 26

Highest price of last 20 bars

25

24 23.81 23

A false breakout occurs when price pushes through a support or resistance level in the anticipated direction, suggesting a new price thrust or trend, only to (relatively) quickly reverse direction when no real followthrough materializes. Because traders who bought or sold on the initial breakout may all scramble at once to get out of their trades when the market fails to follow through, the reversal can be quite forceful. For this reason, contrarian traders sometimes fade initial breakouts to capitalize on these short-term reversals. Stop-and-reverse (SAR) refers to a trading approach that is always in the market, long or short. The existing position is liquidated (stopped out) and a new position (a reverse of the previous one) is established, using the same signal in the opposite direction. For example, a simple 40-day SAR breakout system would buy when price exceeds the highest high of the last 40 days and sell when price falls below the lowest low of the last 40 days. Support and resistance. Support is a price level that acts as a “floor,” preventing prices from dropping below that level. Resistance is the opposite: a price level that acts as a “ceiling;” a barrier that prevents prices from rising higher.

Lowest price of last 20 bars
14 15 10 11 12 11/28 Tuesday 13 14 15 10 11 12 11/29 Wednesday 13 14 15 10 11 12 11/30 Thursday 13 14 15

22

10 12/1 Friday

Source: QCharts by Quote.com

The Donchian-type breakout is also commonly referred to as a “price channel” breakout.

Application
Traders using breakouts are basing their trades on the following principle: If price momentum is strong enough (either up or down) to push through a significant technical level, there is a good chance price will continue in that direction for at least a while. As a result, these price levels represent logical trade entry and exit levels with well-defined risk, both for traders who expect follow through in the direction of the breakout and, as will be described shortly, traders who are looking to fade breakouts.

ing range, traders who go long on the breakout can place protective stops in a number of technically logical places, in relation to the range. First, the stop could be placed below the low of the trading range. Second, a more conservative stop placement would be in the middle of the trading range (or in the upper 25 percent of the trading range, etc.). Finally, the most conservative alternative is a stop just below the original breakout level, which might be used by

FIGURE 3 BREAKOUT ABOVE PRIOR HIGH A prior high creates a resistance level that is tested multiple times before price breaks out to the upside. A significant trending move follows.
Sun Microsystems Inc. (SUNW), Weekly
28 25 59⁄64 24

Key points
Price breakouts are typically used as trend-following signals. The greater the number of days (or price bars) used to determine the breakout, the longer-term trend the trading system will reflect and attempt to exploit. For example, a 20-day (or 20-bar) breakout would capture shorter trends than a 40-day breakout, which in turn would reflect shorter trends than an 80-day breakout. Generally, in terms of trend-following approaches, the longerterm the breakout, the more significant the price move and the greater the likelihood of sustained follow through. Breakout trading can also simplify risk control because stop-loss levels are often easy to identify. For example, if price breaks out of the upside of a trad-

Breakout above previously tested high

20

16

12

8

Jan. 1997

Apr.

July

Oct.

Jan. 1998

Apr.

July

Oct.

Jan. 1999

Source: QCharts by Quote.com

ACTIVE TRADER • March 2001 • www.activetradermag.com

3

FIGURE 4 TRADING RANGE BREAKOUT WITH STOP LEVELS The boundaries of a trading range provide logical stop levels for a breakout trade. Connors M. Similarly. only to reverse back into the trading range and eventually break out through the top of the range.com 4 www. For example. 2-minute Far side of trading range (stop 1) Midpoint (stop 2) 54 55 53 Trading range Breakout level (stop 3) 51 15⁄16 51 10:00 10:30 11:00 11:30 12:00 12:30 13:00 13:30 14:00 14:30 15:00 15:30 9:30 10:00 10:30 11:00 11/21 Tuesday trading range and possible stop points. Apr. May June July Source: QCharts by Quote. 1993 Trading Systems and Methods by Perry Kaufman 3rd edition. daily 38 36 35 23⁄64 34 32 Additional research: Trading for a Living by Alexander Elder John Wiley & Sons.com • March 2001 • ACTIVE TRADER . Again. but this turns out to be a false breakout. 1996 Far side of trading range Trading range 30 28 26 Midpoint 24 22 20 False breakout Original breakout level 25 2 9 16 23 30 6 13 20 27 3 10 18 24 3 10 17 24 31 7 14 21 28 5 12 19 27 2 9 16 23 30 7 14 21 Dec. a very short-term trader. Whenever the original reason for a trade is nullified. Jan. Source: QCharts by Quote. or penetrate it by a certain percentage. and selling markets showing weakness (downside breakouts) with further potential for downside movement is the basis of many trading plans and systems on many time frames. Figure 5 shows the reverse situation. (Note also. After a downside breakout of the range. The breakout concept is also easily mechanized for traders interested in a systematic approach. The stock first breaks out to the downside of the trading range. the midpoint of the range (or some other point within the range) or the upper level of the range. In either case.com Bottom line The breakout concept is one of the most important in technical trading.activetradermag. depending on how conservative he was. John Wiley & Sons. All these choices have one thing in common: The placement of the stop corresponds to a price move that negates the validity (to varying degrees) of the original breakout. the second and third options would be likely short entry points for traders looking to fade the upside breakout. the stock first breaks out below the bottom of the trading range. that position should be eliminated. 1995 Schwager on Futures: Technical Analysis by Jack Schwager John Wiley & Sons. could place a stop-loss order at the original breakout level.) Figure 4 shows a downside breakout out of a FIGURE 5 FALSE BREAKOUT AND REVERSAL In this case. Mar. Because of the possibility of false moves at popular breakout levels. the trader may wait for the market to stay above the breakout level (or close above it) for a certain number of bars. but they can also cut down on false signals. the boundaries (and the midpoint) of the trading range provide logical stop levels — both for the initial downside breakout and the subsequent upside breakout. The stock reverses back into the trading range and eventually breaks out through the upside of the trading range. 1998 Technical Analysis of the Financial Markets by John Murphy New York Institute of Finance. the stop-loss levels are again easily identified. a trader. Buying markets showing strength (upside breakouts) with further potential for upside movement. Gordon Publishing Group. false breakouts are the foundation of some counter-trend trading techniques. American Express Inc. Such techniques delay entry and limit profit potential (and will result in some missed trades). (AXP). 1997 Feb. Microsoft Corporation (MSFT). after an initial upside breakout. 1999 Street Smarts by Linda Raschke and Laurence A. traders looking to capture trending moves sometimes use confirming signals to improve the likelihood of success.

Figure 1 provides an example. but alas. flags and pennants is that these formations allow you to clearly define the risk on your trades.com that determines when and where you’ll exit with either a loss or a profit.TRADING Strategies More bang for your buck: PATTERNS WITHIN PATTERNS hat makes a good trade? Well. In late June. After an initial upside breakout. Microsoft Corporation (MSFT). daily Upside breakout 82 80 78 76 W Support level used as initial stop Stopped out 74 72 5⁄8 72 70 68 12 19 26 3 July 10 17 24 31 Aug.com • October 2000 • ACTIVE TRADER . It moved sideways to lower over the next several days before. stopping out the long position. by having a plan FIGURE 1 FALSE BREAKOUT A trading range develops in the aftermath of a sharp rally. The stock broke out of the upside of the range (around 80 1⁄8) on July 6. triangles. We’d all like to know a trade will be good in advance. For example. The risk on this trade was a moderate 3 5⁄8 points. That’s exactly what happened. because a downside reversal through the support of the range would be a bearish development. A move back below this level would suggest the upside thrust was actually a false breakout and that the trade should be exited. But what do you do when a trading range is much wider and a stop based on either the support or resistance level represents too large a risk? Figure 2 5 www. But when you’re putting a position on. you may look to buy an upside breakout of the range in anticipation of a continuation of the uptrend. the outcome is unpredictable. Two days after entry the stock had pulled back into the trading range. on July 19. around 76 1⁄2. Microsoft (MSFT) established a relatively narrow trading range after approximately a 16-point rally. Then. The logical place to put an initial protective stop is below the low of the trading range. 7 Source: Qcharts by Quote.activetradermag. the stock reverses to the downside. The advantage of trading breakouts of congestion patterns such as trading ranges. The initial protective stop would have been placed just below the support level of the trading range. What you look for when you’re getting in a trade is an entry point where the odds of a move in your favor are better than average. penetrating the downside of the range and stopping out the long trade. most traders would say a nice profit makes a good trade. the markets are not so accommodating. if a stock moves into a trading range after a rally. you try to structure a trade where the potential reward is greater than the known risk. in retrospect.

Dec. Oracle Corporation (ORCL). FIGURE 2 RANGE RISK Using the opposite side of a trading range as a stop for a breakout trade can result in large initial risk if the trading range is wide. daily 130 125 120 15⁄16 120 115 shows a much more volatile trading range than that in Figure 1. This more conservative method is based on the idea that a strong breakout move should follow through immediately and not reverse back into the trading 110 105 100 95 90 4 11 18 25 1 8 15 29 6 13 27 3 10 24 31 7 14 28 6 13 20 27 3 10 24 1 8 15 22 30 5 12 19 26 3 10 17 24 31 7 14 Nov. Jan. Mar. Source: Qcharts by Quote. A trader looking to enter long on an upside breakout of this range would have to accept a risk of more than 16 points.com { Narrow range 60 50 40 30 4 11 18 25 1 8 15 22 29 6 13 20 27 3 10 18 24 31 7 14 22 29 6 13 20 27 3 10 17 24 1 8 15 Mar. 2000 Feb.How to create trade opportunities with increased reward and decreased risk by trading patterns within patterns. narrower trading range forms just at the resistance level of a larger range. Apr. May June July Aug. some traders place the initial stop in the middle of the trading range. you can look for smaller congestion patterns near the support or resistance levels of the range.com . 2000 Feb. Using this range as the basis of an upside breakout trade would have offered the same entry 6 80 77 Wider trading range 70 { Oct. Using the support level of the smaller range as a protective level for an upside breakout substantially reduces the trade’s initial risk.com FIGURE 3 CONGESTION WITHIN CONGESTION A shorter. International Business Machine Corp. However. As a result. (IBM). assuming the bottom of the range was used for the initial stop-loss. Another way to reduce risk on breakout trades is to look for shorterterm patterns within larger patterns that allow you to place your initial stop-loss closer to your entry point.activetradermag. Source: Qcharts by Quote. Dec. Figure 3 shows the formation of a wide trading range in Oracle (ORCL) at the beginning of this year. May ACTIVE TRADER • October 2000 • www. Basing entry and stop points on the levels defined by the smaller pattern can reduce the risk on the trade as well as provide the opportunity for early entry into the position. Patterns within patterns When the risk implied by a particular trading range is exceptionally large. Nov. daily 90 range. Apr. Using the same approach as in the previous example — buying on an upside breakout of the trading range and placing an initial protective stop below the low of the range — would represent considerable risk. Jan. a much narrower trading range developed in February.

this represents a large stop. Nasdaq 100 Index (QQQ). FIGURE 5 NARROW FLAG A narrow flag consolidation forms near the resistance level of an intraday head-and-shoulders bottom pattern. Source: Qcharts by Quote. we’ll look at the practical risk-reward impact this can have on a trade. triangles. In this case. Figure 5 shows a 15-minute chart of the Nasdaq 100 tracking stock (QQQ). point but a much closer stop. flags and pennants is that these formations allow you to clearly define the risk on your trades. Mar. Jan. daily 90 1⁄2 Resistance 80 Flag 70 60 50 40 The advantage of trading breakouts of congestion patterns such as trading ranges. May June July Aug. (Later.com 7 www. For a short-term trader. Nov.com • October 2000 • ACTIVE TRADER . (EMC) repeatedly pulled back from resistance around 72 1⁄2. The low of the flag provides a lower-risk stop level than the most recent swing low.activetradermag. the most recent swing low around 51 would be the reference point for the initial stop-loss — a risk of more than 20 points. In this case. Dec. The upside breakout of this flag provided an early entry to the subsequent surge that pushed the stock past the 72 1⁄2 resistance level to new highs. However.) Figure 4 provides another example. offering early entry into the upside thrust move. placing a stop one tick below the low of the narrower trading range would have reduced the risk to 6 3⁄4 points. The stock formed a large bottoming pat- 30 27 4 11 18 25 1 8 15 29 6 13 27 3 10 24 31 7 14 28 6 13 20 27 3 10 24 1 8 15 22 30 5 12 19 26 3 10 17 24 31 7 Oct. but it’s still a dramatic improvement and the profit potential for the move out of the larger trading range is still intact. EMC Corp. Because a well-defined horizontal trading range did not develop (the stock swung back and forth in an increasingly wider range).com Apr. it formed a flag consolidation from June 7 to June 12 with a high around 69 7⁄8 (the highs of the bars in the flags were within 1⁄16 of each other) and a low around 66 13⁄16. 15-minute 93 3⁄8 92 Resistance 88 84 Narrow flag 80 S H 19 May 22 23 24 25 S 76 26 30 31 1 June 2 Source: Qcharts by Quote. as the stock bounced off that low and made another run at the resistance level. 2000 Feb.FIGURE 4 FLAG NEAR RESISTANCE A small flag forms just below a well-defined resistance level. EMC Corporation (EMC).

This level would mark a good spot to take at least partial profits on the position and raise the stop on the balance of the position. As was the case with Figure 4. January 2000 high was 41 7⁄32. you can look for smaller congestion patterns near the support or resistance levels of the range.activetradermag.com Structuring a trade Figure 3 provides a good example of how this approach can work in the context of a complete trade plan. it consolidated in a narrow flag pattern with resistance around 82 7⁄32 and support around 81 5⁄8. (MOT). Even if you are stopped out on a move through the support of the smaller congestion pattern. A final example is shown in Figure 6. Nov. As the stock approached the resistance level for the second time. This consolidation marked an opportunity to exit part of the position with a profit. Playing an upside breakout of this pattern and using its support level for the initial stop (rather than the most recent swing low around 76) reduced the risk on a long trade to less than a point. 2000 Source: Qcharts by Quote. but they fulfill the primary goals of smart trading: They allow you to establish trades with shorter-term risk and longer-term profit potential.” p. whereby the size of the previous price move is added to the current price. The rally from the late-October 1999 low to the early- Using the measured move approach on the smaller price swing from Jan. Even though price gapped above the flag. Adding the size of the price move preceding the trading range to the low of the larger trading range (around 46 5⁄8) results in an upside target of 87 27⁄32. daily 52 49 171⁄256 48 44 Trading range 40 36 32 Flag 28 3 10 17 24 1 7 14 21 28 6 12 19 26 2 9 16 23 30 7 13 20 27 4 11 18 26 1 8 15 22 29 6 13 20 27 3 May June July Aug. this loss is much smaller than the one that would have occurred had the stop been placed below the low of the larger trading range. a trader who went long on the intraday upside thrust above resistance (say. 42. Oct. the stop on the remainder of the position could then be moved up to the breakeven point. Motorola. 28. in the middle of a larger trading range with resistance around 32 3⁄8. at 62 5⁄8) on Feb. Jan. 14 high of 64 3⁄4 (18 1⁄8 points) sets up a shorter-term price target of 77 7⁄16. forming the larger trading range. 14 and used the low of the smaller trading range (around 58 5⁄8) as the stop level. Inc. shorter-term price pattern. which was nearly 12 points lower. locking in a profit on the trade. Dec. The stock actually formed When the risk implied by a particular trading range is exceptionally large. as mentioned earlier. another flag after hitting a high of 76 1⁄2 on Feb. playing the upside of this smaller pattern offered early entry and a tighter stop on a long-side trade. longer-term price pattern with a risk based on the smaller. Placing a stop just below the flag support at 29 15⁄16 would have reduced the initial risk on the trade to less than two points. Motorola (MOT) formed a flag consolidation in late-October 1999 that offered the opportunity to trade an upside move with lower risk. you can still re-enter a long position if the market reverses again and breaks out above resistance a second time. see “Opening day opportunities. on May 30. A trader looking to buy on an upside breakout of the range could use the measured move approach.) The bottom line: The development of the smaller trading range allowed the estab- lishment of a trade with a price target based on the larger.com . 8 ACTIVE TRADER • October 2000 • www. 22. would have been stopped out on the intraday downside thrust on Feb. However. Here. In future articles we’ll expand on these ideas by looking at additional measuring objectives and ways to put breakouts into context in relation to underlying trends of different magnitudes. These patterns may develop relatively infrequently. A flag forms in the middle of a larger trading range. the smaller pattern allowed you to both use a tighter stop and get in earlier on an upside breakout. Another general advantage of this approach is that it increases your flexibility. the preceding sell-off is not shown) with resistance around 82 5⁄8. For example. (For more information on taking profits and moving stops. The stock gapped out of the flag (a bullish sign) above 31 1⁄2 and continued to run past the resistance of the larger trading range. The stock then moved sideways. 28 low of 46 5⁄8 to the Feb. to project a price target.FIGURE 6 EARLY ENTRY tern (a head-and-shoulders bottom pattern. Sept.

TRADING Strategies Anticipating BREAKOUTS and beating SLIPPAGE Trading breakouts is a tried-and-true approach on all time frames. But intraday and other short-term traders can sometimes give up precious points because of slippage.activetradermag. 9 www. Here’s one trader’s take on finding setups that allow you to enter early and beat the breakout crowd.com • August 2000 • ACTIVE TRADER .

it follows the Repeated tests of a support level increase the odds of a downside breakout. he explained what probably is one of the most significant discoveries in the markets: the TD One-Tick. The day it finally broke that support level (March 30. the odds are extremely good that price level will be broken shortly. Let’s look at a second example. 5-minute 10:00 11:00 12:00 13:00 14:00 15:00 16:0010:00 11:00 12:00 13:00 14:00 15:00 16:00 61 59 Stop placed at most recent swing high (50 3⁄4) 57 55 53 CMGI repeatedly tests support at 50 in a weak market 51 1⁄8 51 49 47 350. that level gets weaker and weaker. because of repeated tests of the support level. Figure 1 is a five-minute chart of CMGI. while breakout traders view tests of support as fuel to propel an eventual breakout. In layman’s terms. It makes sense that the opposite also is true: If a price prints more than once at a certain high or low. try to enter a market order with a browser-based online broker the first day of a hot IPO and see what happens. there are also many traders waiting to short the stock once it does break down. In this case. In this example. if you wait for the stock to trade at 49 15⁄16 and then try to establish a short position. the weaker it becomes. In fact. One-Time Rule. more a particular level is tested.500 Source: CyberTrader by CyberCorp. Netro Corp. most people view support levels as opportunities to go long. slippage is probably the most overlooked and significant cost in trading. Here’s the question: If. ACTIVE TRADER • August 2000 • www. not only are traders establishing new long positions with their stops just below the support level at 50. to be filled on a trade and where your order is actually executed. if a stock continually prints or finds support or resistance at a certain price. let’s look at why it works. A short position can be established in anticipation. why wait for the breakout? Doing so increases the odds of having to chase the market or missing the trade. New York). From that. or fast market conditions. most significant highs and lows only print one time at the extreme price. But through a little-known tendency. Every time a stock tests a support or resistance level. 2000) was a very weak day in the broader market indices. you can make slippage work for you instead of bleeding you dry. Don’t forget that all the people FIGURE 1 CHISELING AT SUPPORT who bought the stock around $50 will either be stopped out or will wait for an opportunity to breakeven on their trades. But first. The stock bounced off support at 50 six times (and who knows how many prints actually occurred at that level). That’s slippage! Slippage can be caused by a number of factors: Poor execution by a broker. In Figure 2. In it. if most of your trading techniques are breakout related. you can use this trick on almost every trade you enter. the odds are very good the 50 level will be broken (and the broader market indices support this view). Paul Tudor Jones and Steve Cohen. CMGI (CMGI). That is invaluable information for any trader who uses breakouts as part of his or her strategy. then that high or low will be broken in short order almost every time. John Wiley & Sons. In fact. you’ll probably end up missing the trade waiting for an uptick.com 10 . that new high or low should hold for a significant period of time. If you don’t understand this concept. as if a hammer and chisel were chipping away at it. which helped the stock to finally break down. wrote a book (his second) called New Market Timing Techniques: Innovative Studies in Market Rhythm and Price Exhaustion (1997. a highly regarded trading system developer who has worked with such top traders as George Soros. Slippage is the difference between where you expect.activetradermag. Fortunately. While it’s true that we all try to keep our costs down to the bare minimum without sacrificing service or technology.BY STEVE WENDLANDT O ne of the most important aspects of short-term stock trading is something you almost never hear about: Slippage. or want. communication failure or other technical problems. A good opportunity to short NTRO came at the prior day’s close when NTRO closed right at the support One tick at a time Tom DeMark. This rule states if a market makes a new high or low just once (a single print) and backs off from that point. with a stop just above the most recent swing high to protect against an upside reversal. The bottom line is that when support at 50 is penetrated it quickly turns into significant resistance. (NTRO) was bouncing off the 82 1⁄2 level for about two weeks.

daily Jan.FIGURE 2 EARLY OPPORTUNITIES A close at the low of the bar preceding the downside breakout. establishing a position in advance of a breakout above the trendline was supported by strength in the S&P 500 and Nasdaq indices. a buy order was entered at 11 www. before entering the CMGI trade on the short side. For example. The weakness of these indices would help pull the stock below the support level. Figure 3 shows one last example. If you wait until the stock breaks out you are almost always forced to pay the spread — if you can get it at all. difficult to get short once a stock breaks through support.434. Most times. offers an early entry opportunity for a short position.000 Source: CyberTrader by CyberCorp. As soon as WLA began to move toward the trendline. But you must use caution when entering breakout trades early. The pre-opening call was for the Nasdaq and S&P 500 to go higher that morning. Warner Lambert (WLA) opened for trading at 121 1⁄2. Netro Corp. Entered long at 122 1⁄4 (before the breakout). But. or should you exit the position on the close? One approach to reduce risk is to use the last swing low or high as your initial stop-loss point. In this case. May 130 125 9⁄64 1231⁄2 Stock tests trendline resistance in strong overall market. The only reason not to hold the position is if the overall market begins to move counter to the trade (i. May 122 109 1⁄2 Stock tests support in weak market. On May 25. All breakout traders know it’s very FIGURE 3 GOING WITH THE MARKET Pre-breakout entry should be confirmed by the broader market indices. and they both began to rally from the open. Feb. you could have placed an initial stop loss at 50 3⁄4 which was the last swing high on the fiveminute chart. and the market begins to drop precipitously). It would have been difficult to get short after the market opened for trading on the day of the breakdown (although. 117 1101⁄2 104 971⁄2 91 841⁄2 8. what if the stock never breaks out? Should you hold the position until it does. you should have checked to make sure the Nasdaq and S&P 500 were both weak on the day and trending lower. you can simply wait for the breakout to materialize. if the trade is any good. just at the support level. you should enter before the breakout.. level for the second day in a row.900 Source: CyberTrader by CyberCorp. The next morning NTRO gapped lower and continued to drop dramatically. Warner Lambert (WLA). you never want to enter a trade that is counter to the overall market momentum. Mar. This created a setup to go long before the actual breakout above the trendline. waiting for the breakout. daily Jan. you may ask. and the stock repeatedly tests a support or resistance level. you’re long. In the CMGI example. With a stop in place. Mar. just under the down trendline of a nice triangle pattern.e. (NTRO). Feb.820. You must either wait for an uptick (which may not happen) or offer it short 1 ⁄16 higher than the inside bid (for Nasdaq stocks). there were some upticks in the pre-market). Short trade entered at 82 9⁄16 97 841⁄2 72 591⁄2 47 341⁄2 25 11⁄16 5. But if the stock is dropping like a rock. who is going to hit your offer? The bottom line is that if you want to trade a stock when the overall market is trending in the direction of your potential trade. Apr.com • August 2000 • ACTIVE TRADER . Apr. you even can avoid paying the spread because the stock will be whipsawing back and forth between the bid and offer.activetradermag.

com . but in certain situations entering early can yield excellent trading results. they only think about the losers. And don’t forget about the trades you missed completely because the stock just ripped through the support or resistance level and you couldn’t even get a partial fill. slippage affects you whether or not you make a profit on the trade. but those are usually the most potentially profitable trades because the stock is moving so forcefully. Most traders don’t even think about the effect of slippage on their winning trades.activetradermag. scratch a trade from these false breakouts. This approach will also help you on the breakout trades that don’t materialize because you’ll have a better entry price and may even be able to still garner a small profit or. Very rarely should you wait for the actual breakouts to materialize on any of these patterns. including breakouts from daily and intraday cup-and-handle patterns. trendline breakouts and spike and ledge patterns (see Figure 4). Remember. They provide well-defined support or resistance levels you can use to anticipate breakouts. 12 ACTIVE TRADER • August 2000 • www. at worst. Try the following experiment: Multiply 50 percent of all the shares you have traded over a given time period by 1⁄8 and see what you come up with. Those extra fractions add up quickly. That’s being conservative.FIGURE 4 BREAKOUT PATTERNS A sampling of the breakout patterns short-term traders can use on any time frame. Cup and handle breakout Trendline breakout Spike and ledge breakout Triangle breakout 122 1⁄4. You can usually gain an extra 1⁄8 (sometimes as much as a point) simply by realizing that support and resistance almost always get broken. Not long after. triangles. it continued to rally for the rest of the day. You can use this entry technique on any breakout-related trade in any timeframe. well before the 123 1⁄16 breakout point. you would have been filled at a minimum of 13⁄16 worse than the early entry price. We tend to forget about those missed opportunities completely. No approach is without risk. Had you waited for WLA to print at 123 1⁄16. the overall market strength helped pull WLA through the trendline.

The 100. Starbucks (SBUX). Rules: 1. The number of days used to calculate the breakout level is called the “channel length.000 140. Microsoft (MSFT). Sears (S).00 36. Test period: January 1993 through February 2003.00 44. 50. Enter long on the next bar at the highest 100-day high. The exit strategy allows the system to follow large moves until price makes a significant reversal. Disney (DIS). We will also examine the results of using a range of channel lengths and how a “walk-forward optimization” could improve the results of the system for the most recent year. Deduct $10 slippage and commission per trade. at best: It returned only 12. System results: The system’s performance was mediocre. 2.00 M August 2002 September Source for all figures: Wealth-Lab Inc. Test data: The system was tested on the Active Trader Standard Stock Portfolio.000 50.00 10. along with the overall equity curve. which contains the following 18 stocks: Apple Computer (AAPL). FIGURE 2 SAMPLE TRADES FIGURE 1 EQUITY CURVE The long. a market is demonstrating it has the momentum to establish a trend. and a short channel length (20 days) was used for exits. International Paper (IP).000 120.00 42.00 40.00 M 5.com • June 2003 • ACTIVE TRADER This short trade was triggered when price crossed below the 100-day low.00 34.000 40.000 130. Starting equity: $100. the system was exposed to the market nearly 75 perwww. In this test.100-20 channel breakout system System concept: This is a classic trend-following system that buys when price moves above the highest high of the last x days and sells when price falls below the lowest low of the last y days.” Breakout systems are based on the logic that by making a new price high (or low). (www.000 100. JP Morgan Chase (JPM). Boeing (BA).61 percent over 10 years.000 110.00 Boeing (BA). one long channel length (100 days) was used for entries. Caterpillar (CAT).000 160.000 180. Exit long on the next bar at the lowest 20-day low. Money management: Risk a maximum of 2 percent of total account equity per trade. Citibank (C).000 60. International Business Machines (IBM).00 46.com) October November 13 . Cisco (CSCO).and 20-day high/low channels are plotted as gray lines.000 70.activetradermag. The long side of the system substantially outperformed the short side during the 10-year test period 190. Intel (INTC).wealth-lab.000 10.000.and short-only equity curves. General Motors (GM).000 80.00 32. Trade the number of shares that would result in a 2-percent loss of account equity if the stop level were hit. and price will likely continue in that direction. AT&T (T) and Wal-Mart (WMT). Coke (KO). 3.00 Short Cover Volume 30.000 90.000 30. Exit short on the next bar at the highest 20-day high. The exit occurred when price crossed above the 20-bar high. Furthermore. are shown here. Hewlett Packard (HPQ). Enter short on the next bar at the lowest 100-day low. The position size is based on the difference between the entry price and the initial stop level. daily 48.000 20.000 170. while buy and hold would have returned more than 253 percent.000 0 3/3/93 3/2/94 Equity Account balance ($) 3/1/95 2/6/96 Cash 2/3/97 2/2/98 1/7/99 1/3/00 Linear reg Long 1/2/01 1/2/02 Short 1/2/03 4.000 150.00 38.

19% 0. the system is between two equity highs • No.15 13.47% 49. If you have a system you’d like to see tested. it was also accompanied by extreme volatility. please send the trading and money-management rules to editorial@activetradermag. return ratio return return profitable consec.15 22. It is interesting to note.17 33.91 6/14 LEGEND: Net profit — profit at end of test period.activetradermag. hold time (losers): Max.15 11. the number of days used to determine the channel lengths).com.766 12. profitable — the largest number of consecutive profitable periods • Max. unprofitable — the largest number of consecutive unprofitable periods Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc. consec.31% 50. less commission • Exposure — the area of the equity curve exposed to long or short positions. Trade statistics No. the odds that you would have known to use those specific parameter values PERIODIC RETURNS Avg.’s testing platform. periods profitable unprofitable Weekly 0. These results confirm short trading in equities can be tricky.51% 0. profit (winners): Avg. and although this period did produce a small profit.78% 5 4 Annually 2.04% 0. consec. System parameters: One way many traders attempt to improve a system is to “optimize” its parameters (in this case.79 0. It is not meant to recommend or promote any trading system or approach. hold time (winners): Avg. which means we are squeezing just about as much performance out of this system as possible.61 73.00% 3 2 LEGEND: Avg.608 12. short of using margin or some other form of leverage.28% 50. Past performance does not guarantee future results. in days. consec. Maximum drawdown for the long side of the system was only 18 percent. win/loss — the maximum number of consecutive winning and losing trades Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.91% -10. historical testing may not reflect a system’s behavior in real-time trading. Sharpe Best Worst Percentage Max. while buy and hold experienced a devastating 66 percent maximum drawdown. DD (%): Longest flat days: 35. Traders are advised to do their own research and testing to determine the validity of a trading idea.cent of the time. profit (winners) — the average profit for winning trades • Avg. gain/loss (%): Avg.09 34. ACTIVE TRADER • June 2003 • www.42% 11 9 Monthly 0.33 -7. with only 38-percent market exposure. holding time: Avg.05 0. We measured the results of the short side of the system after the broad marSTRATEGY SUMMARY Profitability Net profit ($): Net profit (%): Exposure (%): Profit factor: Payoff ratio: Recovery factor: Drawdown Max. however.09 12. trades: Win/loss (%): Avg.com 14 . trades — number of trades generated by the system • Win/Loss (%) — the percentage of trades that were profitable • Avg.19% 0. as opposed to cash • Profit factor — gross profit divided by gross loss • Payoff ratio — average profit of winning trades divided by average loss of losing trades • Recovery factor — net profit divided by max. The net return for long trades was 56 percent. consec. hold time (winners) — the average holding time for winning trades • Avg. Although this technique can result in a system that shows tremendous profit over a historical testing period. ket began to fall in the year 2000.09% -13.49% -8.36 1. consec. loss (losers) %: Avg.83% 6 6 Quarterly 0. hold time (losers) — the average holding time for losing trades • Max. Max. This involves testing various parameter combinations to find a range of values that result in the greatest profit over a given period. drawdown • Max DD (%) — largest percentage decline in equity • Longest flat days — longest period. hold time — the average holding period for all trades • Avg.29 1.67 53.25 0. return — the average percentage for the period • Sharpe ratio — average return divided by standard deviation of returns (annualized) • Best return — best return for the period • Worst return — worst return for the period • % Profitable periods — the percentage of periods that were profitable • Max.88 21. profit — the average profit for all trades • Avg. that the long side of the system performed much better than the short side.35 TABLE 1 BEST PARAMETER VALUES FOR EACH STOCK Symbol AAPL BA C CAT CSCO DIS GM HPQ IBM INTC IP JPM KO MSFT S SBUX T WMT Long period 70 70 130 130 80 70 80 90 90 70 130 130 130 120 70 120 90 110 Short period 16 14 26 14 24 18 18 24 18 16 14 26 16 18 14 16 26 14 FIGURE 3 DRAWDOWN CURVE 0% -5% -10% -15% -20% -25% -30% -35% 3/3/93 3/3/94 3/1/95 2/9/96 2/3/97 2/2/98 1/8/99 1/3/00 1/2/01 1/2/02 1/2/03 The system was never able to overcome the drawdown that began in mid-1995.53% -8.63% 48. win/loss: 330 38. loss (losers) — the average loss for losing trades • Avg.

000 55. Second. without cheating by using hindsight.000 50.54 percent during the one-year period. 15 www. the best-performing parameters are used to execute the system on a new.000 105.000 25. However. we tested the parameters on the most recent year of data. optimization must be used FIGURE 4 WALK-FORWARD OPTIMIZATION RESULTS After finding the optimal long and short channel lengths for each stock over the first nine years of historical data. 120.000 15.000 20. We performed a walk-forward optimization on the 100-20 channel breakout system by first optimizing the long and the short channel periods for the first nine years of historical price data. The walk-forward technique described here can help you find more realistic optimized parameters that have a better chance of performing well in real trading.000 110. The walk-forward optimized system lost 1.000 80.com • June 2003 • ACTIVE TRADER .000 75. Although it may be possible to improve the system’s performance by optimizing the channel periods for each stock. The parameters that worked best in the past years are unlikely to be those that work best in the future.000 65. historical (“out-of-sample”) data period after the sample period. there are ways to use optimization effectively.” First.000 100.000 35. (The system lost nearly 9 percent during this same year using the default parameter values of 100 and 20. system parameters are optimized on an initial (“sample”) data period.57 percent.000 95.000 85. but buy and hold lost 30.000 30.000 70.) The walk-forward optimization was effective in this case. One technique is called “walk-forward optimization.at the start of the period are about the same as picking the winning lottery numbers for tomorrow.000 45.000 90. The 100-20 channel breakout performs much better on the long side than on the short side in stocks.000 10.000 60.000 40. We then used the best-performing parameter values for each stock in the portfolio (see Table 1) and applied them to the last year of historical price data. — Compiled by Dion Kurczek of Wealth-Lab Inc.000 115.000 0 3/1/02 4/3/02 Equity Account balance ($) 5/7/02 Cash 6/12/02 7/22/02 8/28/02 Long 10/7/02 11/15/02 Short 1/2/03 2/6/03 Linear reg Buy & holds with caution. This allows you to find out if the optimized parameters would have improved the results going forward.activetradermag. Figure 4 is the equity curve for this optimized system. The system outperformed buy and hold (as well as the un-optimized parameters).000 5.

for better or worse. S&P 500 16 Linear reg Short ACTIVE TRADER • June 2003 • www. lean hogs (LH). multiply the difference between the entry price and the stop-loss price by the dollar value of a one-point move in the contract. 2.) Money management 1.200.000 1.000 1. Rules 1.800.000 150.000 Account balance ($) 140.200. Exit long on the next bar at the lowest 20-day low.000 90. The channel lengths are relatively long.400. 4.com) FIGURE 2 EQUITY CURVE: 6 PERCENT MAXIMUM RISK The equity curve using a 6-percent maximum per-trade loss highlights large returns accompanied by high volatility and large drawdowns. silver (SI). gold (GC).000 180. German bund (DT). (All trades are executed as stop orders.000 2. crude oil (CL).000 170.000 50.000 30.000.000 70. copper (HG).000 3. Japanese yen (JY).000 Account balance ($) 2. Deduct $10 slippage/commission per trade. Risk a maximum of 2 percent of account equity Linear reg Short Source for all figures: Wealth-Lab Inc.000. coffee (KC).com . For example.000 1. Eurodollar (ED). 46). live cattle (LC).000 210. In this case. soybeans (S).000 100. corn (C).000 3.000 20. natural gas (NG). Nasdaq 100 (ND). Exit short on the next bar at the highest 20-day high.000. This system goes long and short. 3.000 10.000 1. and divide the result by the contract’s minimum margin. the system would not be able to take this position.000 200. you would buy five [{(100 – 80)* $250}/$1000 = 5] contracts. As a result.000 40.000 2.200.600.000 600.activetradermag. which contains the following 20 futures: DAX30 (AX).000.000 400.000 0 3/25/93 Equity 3/1/94 2/1/95 Cash 1/4/96 1/2/97 1/2/98 1/4/99 Long 1/3/00 1/2/01 1/2/02 100-20 channel breakout system System concept: The channel breakout is probably one of the oldest trend-following systems around (see the stock Trading System Lab on p. unless the account equity is in excess of $250.600. sugar (SB). The $5.000 80.000 800. The system results published here are based on a 100-day channel length for trade entries and a 20day channel length for exits.000 110. To determine the position size (number of contracts to trade). assume the contract being traded has a point value of $250 and a $1.000 160. Enter long on the next bar at the highest 100-day high.000 0 3/25/93 Equity 3/1/94 2/2/95 Cash 2/1/96 1/8/97 1/2/98 1/4/99 Long 1/3/00 1/2/01 1/2/02 per trade.FUTURES Trading System Lab FIGURE 1 EQUITY CURVE: 2 PERCENT MAXIMUM RISK The system equity curve with the 2-percent maximum loss setting has a relatively stable uptrend. 3.) 2.000.000 60.800. 220.000 190.400. Enter short on the next bar at the lowest 100-day low.000 120. Starting equity: $100.000 130.400.000 margin requirement.000 200.wealth-lab. Test data: The system was tested on the Active Trader standard futures portfolio. assume the initial entry buy stop is at $100 (the value of the 100-day high) and the initial stop-loss level is at 80 (the lowest 20day low). (www.000 1. and one that has been especially popular in futures markets over the years.000 maximum loss this five-contract trade represents should not be more than 2 percent of the current portfolio equity. Next.000 2. The stop levels for both long trades and short trades play an important role.000 2. because the system is intended to catch long-term moves. Euro Forex (FX). (Results will also be discussed for a 6percent maximum risk version of the system. because they are used to calculate the position sizes in the different contracts.

hold time (losers): Max. return ratio return return profitable consec. hold time (winners) — the average holding time for winning trades • Avg.00% -10. return — the average percentage for the period • Sharpe ratio — average return divided by standard deviation of returns (annualized) • Best return — best return for the period • Worst return — worst return for the period • % Profitable periods — the percentage of periods that were profitable • Max. trades — number of trades generated by the system • Win/Loss (%) — the percentage of trades that were profitable • Avg.58% 0.’s testing platform.21 Drawdown Max. the system is between two equity highs • No. gain — the average profit for all trades • Avg. drawdown • Max DD (%) — largest percentage decline in equity • Longest flat days — longest period. The effect is dramatic: The drawdown increased to 50 per- 3/1/94 2/1/95 1/9/96 1/2/97 1/2/98 1/4/99 1/3/00 1/2/01 1/2/02 STRATEGY SUMMARY Profitability Net profit ($): 99.FIGURE 3 DRAWDOWN CURVE: 2 PERCENT MAXIMUM RISK The maximum drawdown was about 18 percent. as opposed to cash • Profit factor — gross profit divided by gross loss • Payoff ratio — average profit of winning trades divided by average loss of losing trades • Recovery factor — net profit divided by max.997. compare 3/25/93 the drawdown curves in Figures 3 and 4. DD (%): Longest flat days: 19. with -30. you can experiment with the system by optimizing the channel periods for each market.73 Avg. consec.64 0.64 Recovery factor: 3.32% -9. loss (losers) %: Avg. Figure 4 shows the result of increasing the maximum per-trade risk to 6 percent. especially because there is still plenty of margin available. The system returned an average profit of 8. Traders are advised to do their own research and testing to determine the validity of a trading idea. and the longest wait between new equity highs was more than 750 trading days. consec. hold time (winners): 56. loss (losers) — the average loss for losing trades • Avg. gain (winners) %: 6. and the ratio of winning to losing -15. unprofitable — the largest number of consecutive unprofitable periods Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc. steady uptrend. hold time — the average holding period for all trades • Avg.91% 52.99 Profit factor: 1.59 685 Trade statistics No.60 Avg.61 0.25% 29.00% trades was fairly balanced. 0.00% The position-sizing method keeps the system out of -40.67% 6 6 4 3 9 6 6 2 Quarterly 1.00% -14. profitable — the largest number of consecutive profitable periods • Max.29% -6.22 Avg.08% 40. -35.58 7. If you have a system you’d like to see tested. in days. Past performance does not guarantee future results.00% much higher volatility.48 Net profit (%): 100.66% 8.00% -18.00% 3/25/93 3/1/94 2/1/95 1/9/96 1/2/97 1/2/98 1/4/99 1/3/00 1/2/01 1/2/02 cent.00% System results: Both the long and short sides of the sys-10. periods profitable unprofitable Weekly Monthly Annually 0. consec. Figure 3 is the drawdown curve using a maximum risk of 2 percent.00% -2.09% -8. Max. the idea of increasing the risk and taking more contracts for each signal might sound like a good idea. As discussed in the stock Trading System Lab. although it resulted in no trade sig-45.00% 66.00 Exposure (%): 29. The 100-20 channel breakout performed fairly well in this test. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts. The maximum drawdown during this period was approximately 18 percent.99% -8.99% LEGEND: Net profit — profit at end of test period. Even though the system reached an account value of more than $3 million (refer (SP) and10 year T-Notes (TY). hold time (losers) — the average holding time for losing trades • Max. the accompanying drawdown would have been nearly impossible to stomach. The test used Ratio Adjusted data from Pinnacle Data Corp FIGURE 4 DRAWDOWN CURVE: 6 PERCENT MAXIMUM RISK The drawdown increased both in depth and length in this version of the system.com.07% 55. with the largest losing year being -8.00% -12. trades: 292 Win/loss (%): 45.00% To show the effect of the amount of risk taken.00% tem were profitable.00% many risky positions.14% 12. consec.25% 24. win/loss: -3. The equity curve (Figure 1) using the 2-percent maximum loss setting shows a rela-20. gain (winners) — the average profit for winning trades • Avg. The system’s market exposure was low — on average.activetradermag. 17 www.00% -16.15% 0.00% loss version (Figure 2) posts a much larger gain.55 0. PERIODIC RETURNS Avg.00% Test period: August 1993 to November 2002. The 6-percent maximum -25. -50.00% -4. Sharpe Best Worst Percentage Max.00% -6.80 22. -5.50 Payoff ratio: 1.42 percent per year.00% nals in some markets because the risk was too high throughout the entire test period. win/loss — the maximum number of consecutive winning and losing trades LEGEND: Avg.30 Avg. The system results on the futures portfolio were fairly good when the maximum risk was set to 2 percent per trade.17 5/7 to Figure 2). — Compiled by Dion Kurczek of Wealth-Lab Inc.00% -8.com • June 2003 • ACTIVE TRADER . about 30 percent. holding time: 37. It is not meant to recommend or promote any trading system or approach. historical testing may not reflect a system’s behavior in real-time trading.91 percent. Exposure climbed near 70 percent. 0. gain/loss (%): 0. please send the trading and money-management rules to editorial@activetradermag.21 Avg. Based on this information. consec.00% tively smooth. less commission • Exposure — the area of the equity curve exposed to long or short positions.

wealth-lab.88 percent on the starting capital in two years. This is important because we will close all positions not triggered by an opposite signal at the close of the day. Buy and hold’s largest drawdown (on Oct.000 0 10/15/01 Equity 1/8/02 4/1/02 Cash 6/24/02 9/23/02 Long 12/30/02 Short 4/2/03 6/26/03 9/25/03 Buy & hold Source for all figures: Wealth-Lab Inc.000 22. with long trades outperforming shorts. Exit: Exit all positions on signals in the opposite direction or at the end of the day. Test results: The results for the two years are very encouraging: a profit of 19. The system’s biggest string of losing trades was seven. For a detailed explanation of the strategy please read the stock Trading System Lab on p. (www.m.780 for the next signal. Money management: To equalize the weight of both markets.000 6. Entry rules: Long trades: Buy if the closing price of the third 30-minute bar is above the high of the first 60 minutes of the day.52 percent) on Feb. The intention was to see how the system performed on stock index futures as opposed to individual stocks. Please keep in mind that we use the portfolio result and not the individual result. to 10:30 a. the QQQ is designed to trade at one-fortieth of the Nasdaq 100.000 4.000 18. We use 49 percent to give us some leeway for commission.000 12. and for the closing time we use 4:15 p. QQQ and SPY can be traded intraday but have the advantage that no rollover occurs every three months. For example. This is very important and should always be used since only this method reflects what you would actually experience later in your trading. Test period: October 2001 until October 2003.000 2. 2002 (see Figure 2).000 8.000 20.87 percent. 21. 50. 49 percent of the current portfolio capital is allocated for every trade.000 Account balance ($) 60-minute breakout system Market: Futures (indices).m. There are a few interesting things to note. 24.activetradermag. 16. System concept: This is an intraday system that trades on breakouts of the range established in the first hour of trading. In this test the S&P 500 (SPY) and Nasdaq 100 (QQQ) index-tracking stocks were used as proxies for the S&P 500 and Nasdaq 100 futures. Like futures.com • January 2004 • ACTIVE TRADER 18 .000 and our strategy generates a new signal. 9.000 (nominal).com) FIGURE 2 DRAWDOWN CURVE The largest drawdown occurred early in the test period.000 14. if the total equity moves up to $22. 0% -1% -2% -3% -4% -5% -6% -7% -8% -9% -10% -11% -12% -13% 3/14/02 5/30/02 8/9/02 10/28/02 1/21/03 4/2/03 6/13/03 8/29/03 10/15/01 1/2/02 Test data: SPY and QQQ. compared to an unchanged result for the combined equities of the two indices (see Figure 1).000 10. The system generated its largest drawdown (-13. The SPY is designed to trade at one-tenth the level of the S&P 500. For the first hour range we take the prices from 9:30 a. Deduct $0.m. www. We downloaded more than two years of 30-minute bars from the QCharts historical intraday database for SPY and QQQ. we would invest $10. the uptick rule to enter short positions does not apply to these instruments. 2002) was -44. Short trades: Sell short if the closing price of the third 30-minute bar is below the low of the first 60 minutes of the day. Starting equity: $20.FUTURES Trading System Lab FIGURE 1 EQUITY CURVE The system produced a modest profit.01 per share slippage and commissions.

consec. STRATEGY SUMMARY Profitability Net profit ($): Net profit (%): Exposure (%): Profit factor: Payoff ratio: Recovery factor: Drawdown Max.88 1. trades: Win/loss (%): Avg.742 Trade statistics No.40 33. gain/loss (%): Avg.04 10/7 PERIODIC RETURNS Avg.activetradermag.00% 66.com 19 . This might be one reason the 60-minute breakout system performs much better on the ETFs than it does on the individual stocks. unprofitable — The largest number of consecutive unprofitable periods Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc. periods profitable unprofitable Weekly Monthly 0. trades — Number of trades generated by the system • Win/Loss (%) — The percentage of trades that were profitable • Avg. ACTIVE TRADER • January 2004 • www. profitable — The largest number of consecutive profitable periods • Max. as opposed to cash • Profit factor — Gross profit divided by gross loss • Payoff ratio — Average profit of winning trades divided by average loss of losing trades • Recovery factor — Net profit divided by max.’s testing platform. Nasdaq 100 index-tracking stock (QQQ).80 Sell 33. win/loss: 892 54.50 Buy 33. hold time — The average holding period for all trades • Avg. hold time (winners) — The average holding time for winning trades • Avg.93 1. win/loss — The maximum number of consecutive winning and losing trades LEGEND: Avg. On the downside.07% LEGEND: Net profit — Profit at end of test period. We received a short signal but got stopped out after the market bounced back. return ratio return return profitable consec. only 102 were stopped out by the opposite signal while the rest stayed with the initial direction.10 34.97% -3. return — The average percentage for the period • Sharpe ratio — Average return divided by standard deviation of returns (annualized) • Best return — Best return for the period • Worst return — Worst return for the period • Percentage profitable periods — The percentage of periods that were profitable • Max.67% 7 2 6 8 4 2 Quarterly 2. hold time: Avg. gain (winners) — The average profit for winning trades • Avg. consec. DD (%): Longest flat days: 3. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.00 Bottom line: There is a big difference between indices and stocks in regard to this system. consec. If you have a system you’d like to see tested. loss (losers) %: Avg. hold time (winners): Avg. hold time (losers) — The average holding time for losing trades • Max. Sharpe Best Worst Percentage Max.976.85% 52. It seems that in most cases.05 percent. Looking at the statistics it is interesting to note that out of the 892 trades over the last two years. consec. the system requires more fine tuning.21 0.36% 6. DD (%) — Largest percentage decline in equity • Longest flat days — Longest period. drawdown • Max. once the market begins an intraday trend. Nevertheless.com.88 44. the average profit per trade was just 0. On the following day it appeared the market was continuing its down move.80 19.70 33.00 33. Past performance does not guarantee future results. historical testing may not reflect a system’s behavior in real-time trading.46.35 -0. 9/10/03 with an index you hardly see large overnight gaps.60 33. it continues in that direction throughout the day. the ratio of trades that kept their original position for the whole day makes this strategy worthy of further investigation. that was exited at the close of the day.31% -3. in days.90 33.38 -13. 30-minute 34.86 7.80 0. profit (winners) %: Avg. also. Because of the low average profit per trade. gain — The average profit for all trades • Avg. hold time (losers): Max.39 7. This may be too little to really trade the system.30 Sell 33.95 1. less commission • Exposure — The area of the equity curve exposed to long or short positions. please send the trading and money-management rules to editorial@activetradermag. consec. It is not meant to recommend or promote any trading system or approach. — Volker Knapp of Wealth-Lab Inc. the system is between two equity highs • No.77% 0. loss (losers) — The average loss for losing trades • Avg.20 33.17% 6.FIGURE 3 SAMPLE TRADES The average hold time for both winning and losing trades was around seven days. 2003. Individual stocks tend to be much more volatile than an index.52 1.11 0.05 7.19% 0.60 0. Figure 3 shows a short trade on Sept.10 Buy 9/11/03 33.19% -2.19% 53. 10. Traders are advised to do their own research and testing to determine the validity of a trading idea. or $4. Max.80 7.

The four-percent breakout system is an attempt to quantify and profit from this market scenario. Caterpillar (CAT). It was also tested on the Active Trader Standard Stock Portfolio.000 20. Microsoft (MSFT). Traders take short positions anticipating a reversal.80 24.Four-percent breakout system Market: Nasdaq 100 index-tracking stock (QQQ).20 24. trade in the opposite direction of) sharp intraday moves.000 160.000 80.000 70.00 23.000.00 25.000 170. Rules: Entry — Buy today if price gains four percent from the previous trading day’s closing price. The system goes long when price rises four percent from the previous close — in this case..60 24. Boeing (BA).com • September 2004 • ACTIVE TRADER .20 27.com) Buy Buy Volume FIGURE 2 EQUITY CURVE The equity grew steadily from 2000 through 2002. the market ignores the contrarians and continues to rise. Cisco Systems (CSCO).000 190. (www.000 230.40 24. Sears (S). Intel (INTC). HewlettPackard (HPQ). and enters only one position at a time. Starbucks (SBUX). July 1994 to June 2004 for the Active Trader portfolio.000 50.40 25. assumed to be the point at which short sellers must concede they were wrong and cover their positions. In some situations.000 220.00 26. 2004) by James Altucher.40 26.000 100.000 200.60 25.000 30. Test period: March 1999 through June 2004 for the QQQ test.000 150.000 240.000 10.000 110. 20 FIGURE 1 SAMPLE TRADES March and April 2003 were very active months for the four-percent breakout system.80 25. Test data: The system was initially tested only on the QQQ.P.e.000 90. Morgan Chase (JPM). daily Sell Sell Sell Sell Buy Buy 27. Traders who fade the up move must cover their short positions. The thought behind this type of trading is price is unlikely to go much higher after an extreme up move. Exit — Exit on the open of the next trading day. Coca-Cola (KO).wealth-lab.activetradermag. driving prices even higher during the trading day. System concept: This system is from the book Trade like a Hedge Fund (John Wiley & Sons.000 120.40 27. The system is long only. General Motors (GM).000 140. however.60 26.000 180. two mid-size winners and one large losing trade.80 23.000 40.80 26.000 130. International Business Machines (IBM). so 100 percent of equity should be tied up on each trade. Risk control and money management: This system tests only one market.000 210. Citigroup (C). which contains the following 18 stocks: Apple Computers (AAPL). Disney (DIS). Nasdaq 100 index-tracking stock (QQQ).20 25.40 100M 50M April 2003 Source for all figures: Wealth-Lab Inc.000 0 3/10/99 9/1/99 3/1/00 9/1/00 3/1/01 9/4/01 3/8/02 9/5/02 3/6/03 9/3/03 3/3/04 Equity Cash Account balance ($) www.000 60. AT&T (T) and Wal-Mart (WMT).60 23.00 24. J. but the system has been stagnating since mid-2002. International Paper (IP). Both day traders and hedge-fund managers love to “fade” (i. Deduct $20 per round-turn trade for slippage and commissions. Figure 1 shows sample trades in QQQ from March and April 2003. There was one large winner.20 26. no short trades are made. Starting equity: $100. which leads to panic buying and further upward momentum. 250.

profits began in early 2000 and lasted until mid. The system is more or less flat from April 2003 forward.900 -11. 0% -1% -2% -3% -4% -5% -6% -7% -8% -9% -10% -11% 3/10/99 9/1/99 3/2/00 9/1/00 3/5/01 9/4/01 3/8/02 9/6/02 3/7/03 9/5/03 3/5/04 Test results: Figure 1 shows the first trade was a success.52% 50.56 11. This equity curve (Figure 4. hold time (winners) — The average holding time for winning trades • Avg.08 percent. Max.00 11/5 Avg. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.18 1. in days. the next trade wiped out the previous profit and then some. trades: Win/loss (%): Avg. were not as successful but nonetheless booked modest profit.82 1.09% 66. System variation: James Altucher publishes a variation of the system that adds one additional entry rule: Price must be down two percent on the day before entering a trade. However.023 121. loser — The average loss for losing trades • Avg. trade (%): Avg. the fact that the system was profitable on a portfolio of stocks (8. Sharpe Best return ratio return Weekly Monthly 0. Price rose 40 cents after entry and the market made a small gap open the following day for a twopercent profit. Drawdown the designers knew of previous QQQ price movement). LEGEND: Net profit — Profit at end of test period.00% 59. less commission • Exposure — The area of the equity curve exposed to long or short positions. 60) shows fairly steady growth from the beginning of the test period through mid2002. hold time — The average holding period for all trades •Avg. However. consec.05 Trade statistics No.34% 25. periods* profitable unprofitable -9. consec.76 0.30% 1.000. win/loss: 105 64. it was tested on other markets in an attempt to determine its validity. This equity curve mirrors the QQQ equity curve. consec. Our starting equity for the Active Trader portfolio test was also $100. DD (%): Longest flat days: 121. DD (%) — Largest percentage decline in equity • Longest flat days — Longest period.30% 1. and allows the system to capture solid rebound PERIODIC RETURNS Portfolio test results: While it is still too early to tell if this system is worth trading on the QQQ (because it’s possible the system was subconsciously designed to take advantage of what STRATEGY SUMMARY Profitability Net profit ($): Net profit (%): Exposure (%): Profit factor: Payoff ratio: Recovery factor: Drawdown ($): Max.10% 4 63 -6.com 21 . drawdown • Max.activetradermag.00 1. After a small loss in 1999. beyond the four-percent threshold.38 -7.95 percent annualized gain) and not just one stock is evidence the system is based on a valid core assumption.13 4. A flat period is considered unprofitable for purposes of this report.00 1. please send the trading and money-management rules to editorial@activetradermag.FIGURE 3 DRAWDOWN CURVE The drawdown phase from mid-2002 to the present dominates the drawdown curve. This rule is intended to avoid entering when a price move is nearly exhausted. winner (%): Avg.09 7. loser (%): Avg. there is a slight decline in capital and a general stagnation as fewer trades take place. If you have a system you’d like to see tested. The fact that the initial losing trade occurred on a day when prices gapped above the entry level on the open suggests the system might benefit from a filter that ignores the signal if price opens with a greater than four-percent gain. win/loss — The maximum number of consecutive winning and losing trades LEGEND: Avg. hold time: Avg. The only trade after that was in July 2003.98% -2.81 1. From that point.74% 75.to late 2002. the system is between two equity highs • No. which occurred only a few days later.93 420 *The system remains flat much of the time. and the result was a large loss upon the exit the following day. hold time (losers): Max.03 0. p. consec. historical testing may not reflect a system’s behavior in real-time trading. trade — The average profit/loss for all trades • Avg. although only 10 percent of equity was committed per trade. return — The average percentage for the period • Sharpe ratio — Average return divided by standard deviation of returns (annualized) • Best return — Best return for the period • Worst return — Worst return for the period • Percentage profitable periods — The percentage of periods that were profitable • Max. winner — The average profit for winning trades • Avg.40% -6. as opposed to cash • Profit factor — Gross profit divided by gross loss • Payoff ratio — Average profit of winning trades divided by average loss of losing trades • Recovery factor — Net profit divided by max. as the 12-percent drawdown began in late 2002.35 1. Past performance does not guarantee future results.67% 10 8 4 15 5 2 Quarterly 3.23% 22.93% Annually 16. The next two trades. The equity curve (Figure 2) provides a better indication of the system’s overall performance. profitable — The largest number of consecutive profitable periods • Max.89% 13. resulting in a loss of 0. unprofitable — The largest number of consecutive unprofitable periods Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc.71% 29. consec.55 1.com. Traders are advised to do their own research and testing to determine the validity of a trading idea. hold time (losers) — The average holding time for losing trades • Max. The drawdown curve (Figure 3) confirms this. Price gapped up at the market open. trades — Number of trades generated by the system • Win/Loss (%) — the percentage of trades that were profitable • Avg. Price then suddenly reversed.’s testing platform. return profitable consec.80 2. hold time (winners): Avg. It is not meant to recommend or promote any trading system or approach. and the entry order was filled (this particular trade would have probably been subject to negative slippage because of the volatility at the open). ACTIVE TRADER • September 2004 • www.51% Worst Percentage Max.

Bottom line: The four-percent breakout system could not be much simpler. but actual profit is higher and the system does not enter as many losing trades during the stagnation period of mid2002 to present. The bottom equity curve in Figure 4 shows the results of the system variation on the Active Trader portfolio.000 0 250. and this one is no exception.000 200. — Compiled by Volker Knapp of Wealth-Lab FIGURE 4 EQUITY CURVE: ACTIVE TRADER PORTFOLIO The upper equity curve shows the results of the four-percent breakout system on the Active Trader standard stock portfolio using 10 percent of equity per trade. Simpler systems are often the most effective.000 50. Since the new filter reduces the number of trades. The shape of the equity curve is similar to the previous run.activetradermag.com • September 2004 • ACTIVE TRADER . the position size was changed to 25 percent for each trade.000 100. This system is from James Altucher’s Trade like a Hedge Fund.000 0 7/15/94 7/31/95 6/7/96 6/2/97 6/1/98 5/6/99 5/1/00 5/1/01 5/1/02 5/1/03 4/5/04 Equity Cash A B 22 www. This generally increases the efficiency of the system while reducing the number of actual trades. 200.000 100. The lower equity curve is a system variation that enters after a down move and uses 25 percent of equity per trade. there needs to be sufficient “post-publication” data to provide a reliable test for the QQQs.Account balance ($) moves. However.000 50.000 150.000 150.

Partial rise 12 34 ACTIVE TRADER • April 2004 • www. How close is “close”? Use the figures in this article and your common sense as guides. because a PR or PD often slows overall momentum. Broadening tops and bottoms Figure 1 shows two broadening bottom patterns. Oct. can be difficult. The July pattern shows a PR. the July broadening bottom has three top trendline “touch- The July broadening bottom pattern appeared midway through the down move. However. there were at least two touches of each trendline before the PR. Learn to use these signals to increase profits when trading broadening patterns.activetradermag. too.TRADING Strategies BROADENING PATTERNS: Clues to breakout direction A partial rise or decline can predict the direction of a breakout. These are different from broadening tops because price enters the pattern from the top.com 23 . In both patterns. These signals predict immediate breakouts and indicate their direction. daily 34 32 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 1998 Mar. BY THOMAS N. Sept. A partial rise accurately signaled a downside breakout from the pattern. the size of the eventual breakout is not as large as when a PR or PD does not appear. Milacron Inc. BULKOWSKI T rying to determine when a breakout will occur in broadening chart patterns. which occurs after the pattern is established — that is. which are expanding rather than contracting price formations. May June July Source: Proprietary software (Thomas Bulkowski) Aug. That is. allowing you to increase your profits and reduce your losses. However. price touches each trendline at least two times and swings in a progressively wider range. Dec. Price makes a partial rise when it leaves the bottom trendline and works its way higher but fails to touch or come too close to the top trendline before turning away. the minor highs get higher and the minor lows get lower. For example. partial rises (PRs) or FIGURE 1 PARTIAL RISE partial declines (PDs) can improve the odds of making a correct decision. Nov. (MZ). Apr.

activetradermag. 24 www. Dec. An upward breakout cent of the time. which predicts an upward breakpredicted an upward breakout of the pattern. For PRs. The percentages reflect how often partial rises and partial declines predicted Notice how the July pattern is midbreakout direction. Figure 2 includes two broadening tops with PDs. without a PR. Again. out 65 percent of the time. Newport Corporation (NEWP). Analysis of 77 broadening bottoms on PDs predicting upside breakouts was broadening top pattern must touch each trendline at least two times before a PD signal can occur. In a partial decline. price es.com • April 2004 • ACTIVE TRADER TABLE 1 PARTIAL RISES AND DECLINES: SUCCESS RATES .” not two or four: The second minor 500 stocks from mid-1991 to mid-1996. not as continuation patRight-angled and ascending Not measured Not measured terns within those trends. the rise measured 36 percent. descending 76% Not measured ing bottoms usually function as reversals in a downtrend. the post-breakout decline measured 15 percent. daily while partial rises were 86percent accurate in predict7 ing downside breakouts. the post-breakPartial decline out up move was 32 percent. The accuracy rate of usually follows immediately. the not. The middle of the pattern is Broadening bottom 80% 67% around 23. without a PD. Mar. In a larger combined study of broadening tops 6 and bottoms. way between the price at the start of the downtrend (around 32) and its low Chart pattern Partial decline Partial rise (around 14). PDs worked 77 percent of the time. ascending Not measured 84% form in the middle of moves. showed that a PR correctly but does not come close to or touch the it a touch. Apr. the center of the 32-14 range. broadenBroadening wedge. the declines averaged 17 percent. for more statistics). Although broadening patterns someBroadening top 65% 86% times act as “half-staff patterns” that Broadening wedge.bottom trendline. 1998 Feb. Sept. the 4 PD affected momentum by reducing the eventual rally. top right. Nov. May June gy from the resulting down Source: Proprietary software (Thomas Bulkowski) move. but the third (point 3) does predicted a downward breakout 67 per. PDs not resulting in breakouts occurred just nine perPartial decline cent of the time. a leaves the top trendline and descends high (point 2) comes close enough to call bull market.even better — 80 percent (see Table 1. Oct. as they do in Right-angled and descending 78% 58% Figure 1. Thus. FIGURE 2 BROADENING TOPS Table 1 shows PDs in broadening tops correctly Both of these broadening tops included partial declines. When a 5 PD occurred. which 3 means false signals are comparatively rare. indicat2 ing a partial rise steals ener1997 June July Aug.

Like other broadening patterns. descending. Price touches at point 1 but it is not a minor high or low.” ascending broadening formation. ACTIVE TRADER • April 2004 • www. Apr. daily Partial rise 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 Sept. as shown in Figure 2. Why? Because the pattern at that point did not have at least two minor touches of each trendline. so it does not count as a touch.FIGURE 3 TRENDLINE TOUCHES Look for a partial rise or decline only after price touches each trendline of the broadening pattern at least twice. as is point 3. Sept.. right-angled broadening formations like the one shown here usually act as reversal patterns. A broadening top usually acts as a continuation pattern within the prevailing price trend. 1 2 Not a partial decline 3 Right-angled broadening formations Figure 3 shows a “rightangled. Oct. Point 2 is valid. when a partial rise occurred inside a broadening top pattern after price touched each trendline twice without triggering a breakout) occurred just 11 percent of the time in the 350 patterns examined. May June July Aug. The figure shows this. (CDI). 1995 Feb. CDI Corp. Dec. Source: Proprietary software (Thomas Bulkowski) Apr. daily Partial rise 37 35 33 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 1998 Feb. the breakout can occur in any direction. False breakout signals for PRs (i. Mar.activetradermag. right-angled broadening formation. Here.com 25 . Tommy Hilfiger (TOM). as prices rise into the pattern and exit out the bottom. but this pattern usually reverses the trend. Mar. By that time. Figure 4 shows a descending right-angled broadening FIGURE 4 REVERSAL PATTERN Although broadening formations are often continuation patterns. 1994 Nov. Only after price touches point 3 can you draw the horizontal trendline. After two touches of each trendline occur. look for a partial rise or decline. The partial rise that follows correctly predicts a downward breakout.e. May Source: Proprietary software (Thomas Bulkowski) June July Aug. a partial rise formed in this ascending. The top trendline slopes upward (ascends) and the bottom trendline is horizontal or nearly so. the three touches on the top connect an up-sloping trendline. The late-May decline in Figure 3 does not show a partial decline.

Later. Oct. The top trendline is horizontal and the bottom one slopes down. Mar. right-angled broadening pattern usually acts as a price reversal. Nov. Source: Proprietary software (Thomas Bulkowski) Nov. Broadening wedges Figure 5 shows a descending broadening wedge. Prices in the October wedge were trending downward into the pattern and exited out its top. As is the case with this example. Dec. right-angled broadening patterns. a partial decline fails to correctly predict an upward breakout. The rules for wedges are the same as other broadening patterns: There must be at least two minor high touches of the top trendline and at least two minor low touches of the bottom trendline. prices climbed into the pattern and broke out to the upside.com • April 2004 • ACTIVE TRADER . partial rises worked just 58 percent of the time and partial declines worked 78 percent of the time in descending. rather than a reversal. the two wedges shown in Figure 5 are reversal patterns. of the prevailing price trend. (WPS). WPS Resources Corp. Sept. May 26 www. Sept. The trend after the pattern ends is predominantly upward. In the August pattern. The first partial decline in the October pattern fails to predict an immediate upward breakout. (RIG). In the descending pattern. In the August pattern example. 2000 Feb. Mar.FIGURE 5 DESCENDING BROADENING WEDGE The August pattern did not produce a valid partial decline because price must drop from the top trendline and curl around. daily 32 31 30 Partial rise 29 28 27 26 25 24 23 Failed partial decline 22 21 20 June 1999 Aug. but is correct in the longer term. Source: Proprietary software (Thomas Bulkowski) 2002 Feb. the descending. Apr. Here. a partial rise precedes a downside breakout. However. price rose from the bottom trendline. Only then is the pattern valid and only then should you look for a partial rise or decline. bounces up but does not come close to or touch the top trendline before retracing its gains. This PR predicted a downward breakout. Price touches the bottom trendline. daily 52 50 48 46 44 42 40 38 36 34 32 31 30 29 28 27 26 25 24 23 22 21 Dec. 20 Not a partial decline pattern.activetradermag. FIGURE 6 ASCENDING BROADENING WEDGE After the pattern is established. which consists of two down-sloping trendlines (think of a downward-tilting megaphone). Transocean Inc. the slight dip in Partial declines 1999 June July Aug. but the overall trend (except for a few days after the breakout) was downward after the pattern. The pattern usually acts as a continuation. Oct.

In the October pattern. p.” November 2003. price leaves the bottom trendline. 28) shows an ascending broadening wedge. 32 ACTIVE TRADER • April 2004 • www.” July 2003. p. A PR correctly predicts a downward breakout 84 percent of the time. the first partial decline is a failure because price does not breakout upward immediately after touching the top trendline. p. 32 “The high-low game.” September 2002. 46 “Head-and-shoulders bottoms: More than meets the eye.early September was not a partial decline. 32 “The three rising valleys pattern. bow downward (without coming close to or touching the bottom trendline) and rejoin the top trendline.” February 2004. Price in a PD must start from the top trendline. p. 28 “Grabbing the bull by the horns.” August 2003. Not enough samples were found for partial rises in descending broadening wedges.com 27 . 28 “Tom Bulkowski’s scientific approach. In this case. Additional research Books by Thomas Bulkowski: Encyclopedia of Chart Patterns (John Wiley & Sons. 2002) Active Trader articles: “Technicals meet fundamentals in the earnings flag. A partial decline correctly predicts an upward breakout 76 percent of the time. p. Figure 6 (p. p. then reversed. p. 30 “A different breed of scallop. not the top one. Both trendlines slope upward and minor highs and minor lows touch each trendline at least twice.” December 2003.” September 2003. p. Instead. The partial rise does better when it leaves the bottom trendline. 2000) Trading Classic Chart Patterns (John Wiley & Sons. bounces up and then plunges through the bottom trendline.” January 2004.activetradermag. 28 “Pipe bottom reversals. price drops down again and finally shoots out the top. The January partial decline failed because price did not break out to the upside — it touched the top trendline.

William O’Neil.activetradermag. BY THOMAS N. Although it is sometimes difficult to buy high and sell higher. Flag criteria What should you look for when selecting HTFs? That depends on whom you ask.com • December 2004 • ACTIVE TRADER . Although many HTFs have irregular shapes. these patterns did. has several selection criteria (see “Additional reading. In this example. the price moves following HTFs show how such an approach can work. 33). who popularized the pattern. When price breaks out above the pattern.” p.) The flag shown in Figure 1 actually does not meet O’Neil’s criteria because it retraced 52 percent of the prior rise (most flags failed O’Neil’s filter because they retraced more than 20 percent) and the flag duration lasted more than seven weeks . BULKOWSKI A high. Finally. Applying these rules to 252 patterns found in price data of approximately 500 stocks between mid-1991 and early 2004 filtered out all of them! (An earlier study found only six of 81 patterns met his criteria. He has written the rally preceding the pattern should measure 100 to 120 percent and take less than two months.35 at the HTF’s starting point — a doubling of price in less than a month. Figure 1 shows an example of an HTF that formed in January-February 2000.50 and reached a high of 11. Volume slopes downward over the course of the flag.TRADING Strategies HIGH. you can usually draw a trendline along the top of the pattern to signal a breakout. as it did in 90 percent of HTFs in a recent study. tight flag (HTF) is a consolidation pattern that forms after a stock’s price doubles. 28 www. the stock rallied 52 percent from the closing price the day after price pierced the HTF’s upper trendline to the ultimate high. In Figure 1. it signals the rise is not over. The uptrend started in October at a low of 5. the flag should retrace no more than 20 percent of the preceding rally. The basic HTF trade strategy is to buy at the close of the day after price breaks out above the pattern’s upper trendline. TIGHT FLAG helps squeeze out profits This bullish formation boasts excellent post-breakout performance and a low failure rate — exactly the type of pattern traders should look for in bull markets. produce average gains of 69 percent. parallel trendlines mark the flag’s upper and lower boundaries. however. the flag should move sideways for three to five weeks.

but the post-breakout rise was 52 percent. After the breakout. 1998 Aug. price climbed 54 percent to the ultimate high. then searching for a nearby consolidation region. The study also ignored the size of the retracement. Source: Proprietary software (Thomas Bulkowski) Nov. HTF examples Figure 2 shows two HTFs identified in the study. The trend start point was determined by finding a 20-percent reversal of the existing trend. Sept. Apr. tight flag 9 8 7 6 Trend start 5 Another study that used different selection criteria for HTFs also showed an average post-breakout gain of 69 percent. Dec. measured from a prior high to the recent close. although HTFs equal to or shorter than the 14-day median length performed better (71 percent) than those that were longer (66 percent). Sept. broadening wedge and the second like a falling wedge. which is easy to find by looking for stocks that have moved up sharply in less than two months. May June 4 Source: Proprietary software (Thomas Bulkowski) FIGURE 2 TWO HTFS The first flag looks like a descending. measured from a prior low to the most recent close. Nov. Price rose 121 percent leading up to HTF 2. HTF 1 was preceded by a 156-percent rally that lasted 40 days. Both HTFs show good gains after the breakout with the first pattern hitting overhead resistance at the ultimate high. taking 51 days to make the climb. Mar. The study placed no limit on flag length.FIGURE 1 A LARGE HIGH. which occurred at a resistance area established by price peaks as far back as mid-1995 (not shown). Alkermes (ALKS).com 29 . Oct. daily Ultimate high 84 76 68 62 56 50 44 40 36 32 28 24 22 20 18 16 14 12 10 Trend start 8 6 1999 May June July Aug. The flag retraced 38 percent of the rally and lasted 15 days. National Semiconductor Corp. HTF 2 Ultimate high Trend start HTF 1 ACTIVE TRADER • December 2004 • www. The criteria for this study was simply a near doubling (a rise of 90 percent or more in less than two months) of the stock price. Mar. 2000 Feb. 1999 Feb. (NSM). The ultimate high was identified by finding a subsequent 20-percent trend change. TIGHT FLAG This pattern does not meet William O’Neil’s HTF criteria. daily Ultimate high 17 16 15 14 13 12 11 10 High. Oct. Such a well-defined flag shape is unusual.activetradermag. Dec.

Noven Pharmaceuticals (NOVN). Source: Proprietary software (Thomas Bulkowski) FIGURE 4 HTF FAILURE This HTF fails to travel far due to overhead resistance and a change in company fundamentals. After the breakout. Oct. For all 252 patterns in the study. Apr. daily 93 85 77 71 65 59 53 49 45 41 37 33 29 25 23 21 19 17 15 13 Trend start 11 May June July Aug. 9 Ultimate high Ultimate high HTF 2 HTF 1 Trend start 1999 Dec. near HTF 1’s ultimate high). price climbs 116 percent leading to HTF 1 and soars 129 percent afterward. Vertex Pharmaceuticals (VRTX). That may explain why price nearly doubled before trending downward. The second HTF pattern was preceded by a 136-percent price rise and followed by a 61-percent rally after the breakout. To determine an approximate target. the climb leading to the pattern averaged 124 percent.FIGURE 3 TWO MORE HTFS The first HTF launches from a flat base and price soars to the ultimate high. Mar. 2000 Feb. Volume trends downward in both. The stock tumbles on an earnings warning. The second trade is more typical with the rise to the ultimate high about half the distance. but the post-breakout gain was just 69 percent. 12 Nov. Figure 3 shows two more examples. Starting from a flat base in late 1999. Notice the irregular shapes of these two HTFs. daily Resistance Ultimate high 44 42 40 38 36 34 32 30 Dead-cat bounce 28 26 24 23 22 21 20 19 18 17 16 15 14 13 2000 Feb. Sept. Sept. The stock did not perform as well after HTF 2 because the market changed from bull to bear between the two patterns (the bear market started in March 2000. The flag retraced 45 percent of the rally and was 29 days long. from the trend start to the flag. May June July Aug. The first looks like a small broadening wedge and the second looks like a regular falling wedge. Apr. the move from the flag’s lowest low should measure approximately half this Trend start Source: Proprietary software (Thomas Bulkowski) 30 www.com • December 2004 • ACTIVE TRADER . price rallied 92 percent. After the breakout. This pattern did not have overhead resistance to overcome on its way to the ultimate high. Mar. on a percentage basis. HTF The measure rule The second pattern in Figure 3 is typical of the rise you can expect after an HTF in a bull market.activetradermag. compute the percentage change from the low of the trend start point to the high at the top of the flag.

thus. p. p. p. 32 “Tom Bulkowski’s scientific approach.5 times the average intraday trading range over the prior month. p. and none failed to climb less than 5 percent.com 31 . 46 “Head-and-shoulders bottoms: More than meets the eye. Only 10 percent of the 252 patterns in the study failed to climb at least 20 percent. Also. To trade the pattern.” November 2004. Underlying support or overhead resistance (look for a solid mass of horizontal price movement or peaks and valleys stopping near the same price area) spells death to most chart-pattern breakout trades. p. on average) will come in the first week. p. 28 “Grabbing the bull by the horns. 32 “The three rising valleys pattern. p. once price makes a new Pattern failures Figure 4 illustrates two types of pattern failures. 2000) Trading Classic Chart Patterns (John Wiley & Sons. For protection. use progressive stops. otherwise use 1. 42 “Half-staff patterns: Profiting from flags and pennants. but it fails to come close to the 64-percent gain using the measure rule. so enter as soon as you get the signal.” August 2004. Avoid a stock showing a DCB for at least six months — preferably a year. which is 1. Three months later. Most of the gains (35 percent. In this example. 36 “Technicals meet fundamentals in the earnings flag. amount. buying high and selling higher. if the HTF has an irregular shape. You need a good dose of courage to take the plunge. This is a momentum play.5 times the daily volatility. 48 “Three falling peaks: Bearish trend change pattern. Those are very low failure rates. p. Price bounced up during the next month before “rounding over” and making a lower low in a classic “dead-cat bounce” (DCB) pattern. 30 “A different breed of scallop. This is important: If you buy before the breakout. 2002) Active Trader articles: “Trading ‘busted’ patterns. 28 “Pipe bottom reversals. p. p. That will give the company time to get its act together.” November 2003. The stock tumbled 43 percent in one session. Problems cannot always be fixed in one quarter.” September 2004. p.” September 2003. The second failure comes from the fundamentals. high. p. Trading the pattern An HTF triggers a buy signal after a stock has made a significant up move and. raise the stop to just below the prior minor low.An HTF triggers a buy signal after a stock has made a significant up move and. will appear “overbought” to many traders. For example. thus. will appear overbought to many traders. projected upward from the flag low). 44 “Broadening patterns: Clues to breakout direction.” January 2004.” August 2003. the stock dropped another 32 percent on a warning about flat annual revenues.” February 2004. the time from the trend start to the flag start will be slightly less (by six days on average) than the time from the flag’s end to the ultimate high.” May 2004. close above the highest high in the pattern as the buy signal. Trading an HTF is like standing on the edge of the cliff and jumping off. Keep raising the stop as price climbs. The first is a rise blocked by overhead resistance. The company issued an earnings warning for the quarter and said fullyear earnings would suffer as well. use a ACTIVE TRADER • December 2004 • www. That is a significant rally by most standards. wait for price to either close above the flag trendline or. provided it is not too far away.” September 2002. 32 You can purchase past articles at www.” June 2004. p.com/purchase_articles.” April 2004. hoping the water at the bottom is deep enough.htm and download them to your computer.activetradermag. price climbed 31 percent (the pre-pattern rally was 128 percent) after the HTF breakout. price might drop instead. 1988) by William O’Neil Books by Thomas Bulkowski: Encyclopedia of Chart Patterns (John Wiley & Sons.activetradermag.” December 2003. 44 “Trading disaster: the dead-cat bounce. Additional reading Books: How to Make Money in Stocks (McGraw-Hill. 32 “Chart patterns: Does size matter. Use the measure rule to find a target price (half the price rise leading to the HTF. This “measure rule” works 90 percent of the time in a bull market.

We also use the time and sales window to confirm that any large block trades are going our way and that most trades are executed at the ask price for long trades. to 11 a. Moving average crossovers provide trend confirmation but generally lag price action. www. avoiding trades in the middle of the day’s range. use a two-minute candlestick chart encompassing a two-day time horizon (today and yesterday).. but these tools also are often unreliable. To make sure. Entering 0. the breakout. market makers frequently disguise their intentions via Electronic Communications Networks (ECNs) or Level II head fakes. Trail stop at this interval if market moves in direction of trade. many traders use specific candlestick chart patterns to indicate likely price direction. or the bid price for short sales. Further.3 to . buy the stock once it has cleared the whole number closest to the previous day’s high. we want to buy a breakout above the previous day’s high. respectively. the nearest whole number) and when time and sales shows that most trades are being executed at the ask price. low and closing prices.m. In addition.TRADING Strategies Mastering TWO-MINUTE breakouts How can you find consistent trade opportunities? One way is to trade breakouts through yesterday’s high and low — but only after the stock has shown its true colors. what’s a trader to do? Watch price action. Risk control: Stop-loss of no more than 0. We don’t want to buy a double top. The reason for placing the entry a certain amount above the previous day’s high — in this case. (EST) time period. or coinciding with. you would enter a buy order when the stock hits 48. A relatively consistent short-term. The tools For this approach. For example. For a long trade.5 — is to make sure the trade safely “clears the hurdle” of the previous day’s trading range. some institutional buy programs also factor in the open. and you cannot count on sustained trends in consolidating markets. accounting for any market noise that may be present. assume a stock made a high of 47.5 above the whole number helps avoid false breakouts.3 to 0. one that implies a move either up or down — confirms the breakout. pattern-based 32 I stock. which render the Level II screen more or less useless. In this case.com • September 2001 • ACTIVE TRADER . So. Also. Trades typically last several to 20 minutes.5 (having cleared 48. Here are stepby-step guidelines for applying this Strategy snapshot Strategy: “Two-day” breakout Market: Stocks Entry: Go long (short) on move . The rules The best time to use this method is the profitable and volatile 9:40 a. which would suggest strong demand for the stock. Exit: Exit with trailing stop or on close. This approach works because many professional traders and institutional buyers buy such breakouts. When such programs trigger buy signals and money starts flowing into a BY KEN CALHOUN t is often a struggle to find the most appropriate indicator for a given trading situation.activetradermag. you can ride the coattails of the large money on the way up. 48. Momentum oscillators or the Nasdaq and S&P 500 futures may provide early signals of shifts in the stock market. don’t enter the trade until the stock also has cleared the required noise level. A simple example is successive closes at the high (or low) of the price bars leading up to.5-points above (below) whole number closest to previous day’s high (low).e. however. trade is to buy upside or downside breakouts of the previous day’s high or low. Reverse the logic for short trades. high.9 yesterday. We’ll show how to apply this technique using twominute charts. It also is good if a directional chart pattern — i. Trading breakouts and breakdowns of chart patterns is a reliable and simple trading technique that can help you limit risk. A tool that works in one environment may not be appropriate in another.4 points.m.

5.3 to 0. p. depending on the circumstances) the immediately preceding price. although the principle is the same.50 Previous day (compressed) 4/27/01 9:30 4/30/01 10:00 Current day 10:30 11:00 11:30 12:00 45.e.00 49. Make sure you start charting by 8:30 a.technique.m. 2001.00 48.” Active Trader. it is useful to tighten the stop during this time to three or four “spreads” (the colored bands of bid and ask levels on the Level II screen) behind the current inside bid. so you can spot any pre-market top or bottom formations. Glossary Time and sales: The real-time. each day. Set up one of your trading screens to plot a single.50 47. the stock moves FIGURE 1 BUY SIGNAL back into the trading range near the whole number and the entry price level is violated. Based on the guideline to place the entry points 0.3 to 0.5 points above or below the previous day’s high and low prices.00 Buy signal is generated when price exceeds previous day’s high +.3 to 0. EBay Corp. ACTIVE TRADER • September 2001 • www.m.00 46.50 0 50. Combined with entering 0.5 (0. 1./Feb.e.” Active Trader. With decimal trading. this allows active traders to keep even tighter stops than was previously possible. 32.5 points above the previous day’s high (for long trades) or low (for short trades).com 33 . i. Ebay (EBAY) made a high of 48 and a low of 45. A buy signal occurs in EBAY when the stock moves .00 49.00 45.00 47. Define the day’s breakout and breakdown entry levels before each market open. July 2000. 4. Trail the stop to protect profits.activetradermag. on April 30 we set long entry at 48. this provides an excellent risk management tool. 51.5. (EBAY). (The rule varies slightly for NYSE and Nasdaq stocks. See “Program trading and fair value.. Jan.50 Previous day’s high: 48.) See “A walk on the short side. which was a whole number). Enter 0. p. buying S&P stocks and selling S&P futures).5 points. Because the market often reverses around 10 a. Trade examples Figure 1 shows that on April 27.50 51. 2. official record of executed trades (as opposed to bids and offers) throughout the day.50 48. Most trading platforms include a time and sales window to monitor this activity.5 points above the previous day’s high of 48. meaningless price fluctuations that can knock traders out of the market. Buy programs (program trading): Computer-based trading approach whereby institutions or large trading operations execute large volume in related markets to take advantage of discrepancies between them (i.5 points above the whole number nearest to yesterday’s high. All intraday trades should have a maximum stop-loss of 0. for more information. In effect..00 50. as shown in Figure 1.50 46. 3. large two-minute candlestick chart covering two days (today and the previous trading day) of trading activity. we will exit if the reason for the trade is negated. two-minute 52. Noise: Random.5 points above the previous day’s high. Uptick rule: Securities and Exchange Commission rule that requires short sales to be executed when the last recorded price in a stock is higher than (or equal to. price gaps and trends.00 Source: Data Broadcasting Corp. for more information. Identify the previous day’s high and low. 28.4 points.

Therefore. Most trades entered before 10 a. “Tight stops — no exceptions!” Short signal is generated at 41. it may take several attempts to execute a short trade. This trade was covered at 40. buying bottoms and shorting tops is largely a failing method.00 39.m. In time.60 41. despite the amazing predisposition of most new traders to attempt these types of trades. you can learn to avoid low-potential situations and focus on entries based on specific chart pattern breakouts and breakdowns.875 points in less than 20 minutes. 34 www.activetradermag.4 to clear the gap with as small a distance as possible. which will happen at 41.40.375.m.com • September 2001 • ACTIVE TRADER . the trailing stop was triggered at 49.6.00 44. Because of the uptick rule.80 42. it also is important that the time and sales window confirms large block trades are going our way and that most trades are being executed at the bid price (indicating selling pressure). without exceeding the 0. and you should not expect to capture more than 50 percent of the retracements following the bounce. However.20 43. This is not an exact science. Note that the stop is slightly tighter in this trade than in the first example.00 42. Because the stock already has traded at or close to this price in the pre-market. After the entry at 41. Had we sold immediately on the open without adjusting the entry price to take this price action into account. The next day (May 3) we therefore looked to go short if the market fell to 41. but never more than 20 minutes. It’s a good idea to tape that to your monitor. can last a little longer. The stock kicked off the official trading session with a two-minute rally.20 Previous day (compressed) 9:30 5/3/01 10:00 Current day 10:30 11:00 5/2/01 Source: Data Broadcasting Corp.20 40.20 44.20 45.20 42.4 points behind the current price level. should not last any longer than five to eight minutes. it’s a good idea to move the initial entry point farther away from the price action to avoid being caught on the FIGURE 2 SELL SIGNAL wrong side when the market opens.60 43. 0.40 44.80 44.60 40. and not mistakenly ADBE had already reached the pre-determined entry price in pre-market trading. yielding a net profit of 0.20 41. not the other way around. Figure 2 (left) is an example on the short side of the market. Sometimes you will jump into a trade too soon despite this step.80 41.4 points below the whole number (42) closest to the previous day’s low.3 points away from the entry price. Bottom line Successful trading is much more difficult than it first appears. we adjusted the entry to 41.00 41.4 and 45.4.40 43.4 on May 2. other times this precaution will save you from taking an unnecessary loss.3-point stop we’ve set for this trade.80 40. Adobe Systems Inc. this trade will be invalidated as soon as the market trades above this level. Check your trade confirmation window to make sure you are executing a single short trade.7 — 0. Don’t be afraid to hit the short button on your trading platform software several times (assuming you are using a directaccess broker) so you can get in on an uptick.40 40. It requires a long process of market watching and practicing chart pattern recognition. we would have been stopped out with a loss. Your trades should be at least 80 percent breakouts and no more than 20 percent bottom bounces. (ADBE).75 for a . Because of the support-resistance level created by the pre-market gap to 41. Adobe (ADBE) traded between 42. Previous day’s low: 42.00 40.60 44. Trading with the trend on breakouts using these criteria will help traders avoid overtrading and selectively trade the strongest and most powerful chart patterns. ADBE gapped down to 41. The only exceptions to trading breakouts of the previous day’s trading range are those rare occasions when a stock makes a rapid multi-point drop from the previous day’s high and bounces off the previous day’s low.6. two-minute 45.60 45. In this trade. along with the words. But this is a trade for experienced traders only. In fact.40 42.80 43. Trades entered after 10 a.We trailed a stop no more than 0.40 45. When this happens.80 11:30 12:00 entering multiple trades.65-point profit. we trailed a stop a few spreads behind the open trade.00 43.6 in pre-market trading.60 42.40 41. Planning ahead to trade breakouts should be done daily using the previous day’s high and low to set trade alerts.

By contrast.000 on Sept. 28. Swing trading has been increasingly popular ever since the Securities and Exchange Commission (SEC) raised the minimum margin requirement for pattern day traders (PDTs) to $25. you have to line up as many market factors as possible. professional traders and institutions favor breakout trading. BY KEN CALHOUN T 35 rying to outguess the market by picking bottoms and tops is usually unsuccessful. Swing trading. and training in directaccess trading methods as well. These 10-day “channels“ provide clear criteria for entering breakout trades once these price levels are triggered.” Active Trader.TRADING Strategies Swing trading 10-day CHANNEL BREAKOUTS To trade breakouts successfully. October 2001). 2001 (see “New rules for the intraday trader. Combining 10-day support and resistance lines with confirming signals such as volume breakouts and reversals is a practical approach to identifying swing trade opportunities. by contrast.activetradermag.com • March 2002 • ACTIVE TRADER . www. but in computer hardware and software. can be highly stressful and requires a significant initial investment — not just in trading funds. Swing trading is also an effective way to learn many of the “classic” technical indicators and limit risk with small-share or paper trades. Incorporating volume and momentum into your trading plan can put you on the inside track to breakout trades that won’t break apart. This makes swing trading a viable alternative for active traders who are unwilling to meet the new margin requirements and/or uncomfortable with the technology and capital demands of day trading. The following strategy uses simple volume and sector-strength filters to determine when to trade breakouts of 10-day price channels. in which trades often are entered and exited in a matter of seconds.000 can make no more than four intraday trades in a five-day period. Traders with less than the $25. Day trading. Why swing trading? Swing trading is a shorter-term trading style in which positions are held anywhere from one to 10 days. and more often than not results in a large numbers of whipsaw trades. Because it does not require a trader to watch the market all day. software or equipment. is generally less stressful and does not require as large an upfront investment in capital. swing trading can be done on a part-time basis using online discount brokers. Professional day trading requires a full-time commitment and a fast direct-access broker. those who exceed this limit must meet the new day trading margin requirements or face potential position liquidation or account closure.

Enter 50 to 60 cents above the nearest whole number above the highest high of the past 10 days. this would trigger a long trade. The TRIN measures the net buying pressure vs. 4. The rationale is prices with a “9” tend to look expensive and often meet resistance. The most conservative initial stop-loss is the previous day’s high.5. EST. Compare sector indices such as the SOX. Here are the rules: 1.m. selling pressure in the market at a given point in the trading day.com . or during late-afternoon rallies — between 2:30 and 3:30 p.5. Confirm entries using market indicators such as the Arms Index (TRIN) as well as the time of day. See Indicator Insight. versions of which are available for both NYSE and Nasdaq stocks.e. The following rules are given in terms of upside breakouts and long trades.5. avoid highs reached on lower-than-average volume or those reached by a stock in a weak sector that day. The TRIN. Go long 50 to 60 cents above the nearest whole number reverse for shorts): above the highest high of the past 10 days. 19. Confirmation: The best entries are those in which volume is higher than in the previous session and/or when the stock is in a strongly trading sector. accompanied by volume that is higher than the previous day’s volume at the same time. reverse the rules for short trades. 39. from 9:45 until 11 a..5 points below the current trading range. 29. The trigger for a 10-day breakout long trade would be 38. A TRIN reading of 1 means buying/selling volume and the number of advancers/decliners are equally matched.activetradermag. which would result in long trade triggers at 20. If the stock opened today at 37.. for example. as long as the volume in the current session is higher than it was at the same time in the previous session.g. choppy price action. December 2000. It’s usually best to enter 10-day channel long trades in the early morning. Compare volume bars on the current trading day to previous trading days. Avoid highs that are reached on lower-than-average volume or those reached by a stock in a weak sector on the entry day. or both near multiples of 10. 30. For a breakout swing trade. The only exception is when the entry price would contain a “9” — e. a TRIN under 1 is bullish. Look for volume breakouts on the 10-day chart.5. in such cases. Define the 10-day high and low for the stock using a 15-minute candlestick or bar chart. this strategy is better for long swing entries. trail the stop . 3.5. wait until the stock clears the nearest multiple of 10. Exit/risk control: The widest initial stop should be the closer of 1. the highest price a stock traded at was 37. and so on. in which case it’s necessary to take your profit before the market does. Certain cautionary indicators (“red flags”) can be used to eliminate poor trades. buy when a stock breaks out at least 50 cents over the whole number above the 10-day high. Once in a profitable trade.8. use a 15-minute chart encompassing the most recent 10 days (i. for more information on this indicator. and give preference to entries in the strongest sectors 36 Strategy snapshot Strategy: 10-day channel breakout Markets: Nasdaq or NYSE stocks trading between $5-$60. can help determine whether a trade is advisable by highlighting whether momentum is bullish or bearish at a given time. Be sure to include volume bars on the chart. etc.5 points or the previous day’s low. GHA and GSO to determine which are strongest.000 shares and average daily range of 1 to 4 points. NBI. 2.6 and traded up to 38. a TRIN reading above 1 is bearish. it will often reverse and fill the gap. assume that during the previous nine trading sessions plus today. including today. today and the previous nine trading sessions).5. It’s also helpful to enter at times of the day when the market is the strongest and most volatile.If a stock gaps The tools For this approach.5..000 shares. ACTIVE TRADER • March 2002 • www. The best entries are those for which volume is higher than in the previous session. For example. However.m. Active Trader. 40. which was set on the previous trading day. with average daily volume of at least 800. Entry (for longs.5. with average daily volume of at least 800. and average intraday trading ranges of 1 to 4 points. The rules The best types of stocks to trade with this approach are Nasdaq or NYSE stocks priced between $5 and $60. The formula is: {number of advancing issues/number of declining issues}/{volume of advancing issues/volume of declining issues} open more than 10 to 15 percent from its previous close. For long entries.

. trail a stop no more than 50 cents below the current pre-market trading range to lock in your profit.50 48. It is frustrating to panic out of a gap-down swing trade only to see the stock turn around and fill the gap in the first few minutes of the trading day.m. Again.. forming an ascending triangle toward the end of the period as it challenged the resistance level another time.5 points below entry (roughly 2 percent).50 49.e. A good initial stop-loss is the previous day’s low or 1. or use a time stop of no longer than 10 days (i.000 400.50 41. the position should be closed with a profit.00 39. 37 Mixed.50 43. use a stop-loss of no more than 50 cents below the current pre-market trading range. It’s best to enter swing trades on days where all sectors are convergent and the broad market has strength in a single direction.8. those up 1.00 51. Whenever one of the exit signals appears. Nvidia (NVDA).8. (INTU). Entry would occur at 55..50. a trailing stop is used to lock in profits. 5.00 54. If you trade on a shorter time frame. When a stock gaps open significantly above the previous day’s high (in your favor for a long swing trade).00 42.50 How to handle gaps Managing gap opens on swing trades is always a challenge. in which case it’s necessary to take your profit before the market does.50 55. Calmly give it a few minutes to establish a trend and see if it consolidates and Support Volume 39. Trail a stop to protect open profits at 2 percent (generally from .5 to 66. Source: eSignal FIGURE 2 AFTER THE BREAKOUT After the stock fulfills the entry requirements and breaks out above resistance. Also check to see if sectors are convergent (all green or red — i. Stop-loss values are determined by the previous day’s high and low. This is especially true if the stock gaps up above the previous day’s high.5.activetradermag.000 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 11/16/01 11/19/01 11/20/01 11/21/01 11/23/01 11/26/01 11/27/01 11/28/01 11/29/01 11/30/01 Source: eSignal — e.g. because down gaps frequently attract buyers who can bring the price back up.50 53. The initial maximum stop-loss for this trade would be at 66. if a stock gaps open more than 10 to 15 percent from its previous close. 15-minute Entry Resistance 44. Intuit Inc. For example.5 points.50 51. 15-minute Resistance 55. whichever is smaller.50 40. exit all remaining open positions after 10 days). Conversely.50 50.50 54.50 42.00 6 million 4 million 2 million 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 12:00 11/16/01 11/19/01 11/20/01 11/21/01 11/23/01 11/26/01 11/27/01 11/28/01 11/29/01 11/30/01 Ascending triangle Support Volume your risk tolerance. you might consider setting a tighter stop at the previous day’s high.5 points) below the current level of the open trade.5 to 1.e. we would enter EBAY on a breakout above 67.5 to 3 percent or so on the day at the time of the trade entry. it will frequently reverse and fill the gap.FIGURE 1 POISED TO BREAK OUT NVDA traded in a channel from 48 to 55 for 10 days.00 40. when a stock gaps down significantly against you.50 52.00 53.com • March 2002 • ACTIVE TRADER . If the previous day’s range was 65.8 and a high (resistance level) of 66.or five-minute charts. 6. It is sometimes helpful to wait until approximately 9:45 a.00 41. say three.00 600. Re-enter on subsequent breakouts after retracements have occurred.000 200.00 48. choppy days are poor days for swing trade entries. EST to see where the stock trades before exiting a position.00 50.00 52.00 49. 50 cents above the highest high of the past 10 days. it’s often best to wait until 15 to 20 minutes after the open to exit the position.00 43. depending on the intraday market trend and www. moving up or down) or divergent (mixed). let’s say the 10-day channel range for EBAY is bounded by a low (support level) of 61. 1. However. Either of these price points can provide you with an initial stop-loss value.

5).00 18.com . which was not triggered. Cup patterns. For example. The “time stop” is 10 days from Nov.5 (44-. the entry would be reset over the next number up.00 Resistance Breakout entry Exit 32. Using volume and price action filters will help you avoid false breakouts in choppy markets. ACTIVE TRADER • March 2002 • www.6 point profit on Dec.5 (because the entry price would have been 29.6-point profit. We continue to trail a stop .00 16. 2001. cup patterns also provide well-defined resistance levels for breakout trades. Using 10-day trading channels to identify entries on volume breakouts can help you better define support and resistance levels and provide techniques you can integrate with other technical indicators to develop a swing-trading plan. at 56. saucer-shaped retracements. Notice this stock forms an ascending triangle pattern following an initial downturn earlier in the 10-day channel.e. when price falls back from the right side of the cup instead of breaking out above the resistance level of the cup).00 24.FIGURE 3 OVERFLOWING CUP Like channels and ascending triangles. say.5.00 36.00 Volume 1 million 500. When adding to an initial position. to Nov.5).8 in pre-market trading. The test of sellers that occurs at a resistance level prior to a long cup-pattern breakout validates the entry and provides a support level after the trade. This keeps the average entry price toward the low end of the total position. 7. 38 Dynamic position sizing using cup breakouts Dynamic position sizing is the process of adding to an initial position once a stock has broken out and is continuing to attract buyers. 55. 3. 29. Using a trailing stop approach raises the stop to 43.00 Initial cup Second cup 22. a trader may buy 200 shares initially and add another 100 shares on a subsequent breakout as the stock continues to climb. 5 (forming a short-lived double top). and in such cases entry is made above the nearest multiple of 10). Invision Technologies (INVN).5 points is a tighter stop than the previous day’s low (see trade rule No. Based on the guideline to enter 50 cents over the nearest higher whole number above the highest high of the 10 days. 29. 7. this would provide additional confirmation for a long trade. pulled back from the resistance level of 28.activetradermag. Bottom line Swing trading provides traders with opportunities to manage multiple positions and entries at a more leisurely trading pace than is possible in the hectic world of day trading.5 points below the entry. The key to trading them is to apply volume and other filters to avoid false breakouts that turn out to be double tops (i. 1. Trade examples Figure 1 shows Nvidia (NVDA) trading in a 10-day channel (between 48 to 55) from Nov. On the afternoon of Nov. The stock gapped open higher on Dec. a long entry would be triggered at 55. 16. The initial stop would be placed at 54.00 26.00 38. at which point it consolidated.5).5 behind the current trading range until the stop is taken out. 30. every trader should research and experiment with different trading styles to help determine his or her preference and level of comfort. measured and specific strategy to trade breakouts continues to produce entries that are more consistent than intra-range or “bounce” trade approaches. 11 (when the trailing stop was hit) at 34.1 for a 3. 30. Cup-pattern breakouts. and is poised to break out to new highs if it clears the 55 resistance area. 2001. 11.00 28.000 12:00 12:00 12:00 11/29/01 11/30/01 12/3/01 12:00 12/4/01 12:00 12:00 12:00 12:00 12:00 12:00 12/5/01 12/6/01 12/7/01 12/10/01 12/11/01 12/12/01 Source: eSignal reverses this initial gap move.00 20. which are extended. formed another cup and finally broke out above the resistance on the afternoon of Dec. 15-minute 40. however. the stock gaps up to. not shown) is higher than it was at the same time on the previous day. (If.5-1. the stock continued to rally throughout the session to a high of 44.5 or higher). In this case the stock is up more than a point on an overnight hold. The stock broke out above a second cup pattern on Dec. A long trade was triggered at 30. Figure 3 shows a cup pattern that started on Dec.50.59 on Dec.. If the volume when the trade is entered (on Dec. it is sound risk management to “extend yourself” only on the strongest of patterns. 1.5. the stock cleared the 10-day high and a long trade was entered at 42. like ascending-triangle and consolidation breakouts. 10 and the trade was exited on Dec.00 30.00 34. and was stopped out with a 3. The initial stop loss was set at 41 (42. Using a comprehensive. 2001. On Nov. However.60. more conservative stop-loss level is the previous day’s high — in this case 54. because 1. Figure 2 provides an example of how to manage a profitable long swing trade. The key to using dynamic position sizing is to add no more than half the number of the initial trade size on subsequent 10-day high cup-pattern breakouts. The alternate. are much stronger than cases where a stock simply trades to a new high without penetrating any kind of resistance level. appear fairly frequently.

000 50. price must move below its 30-day moving average and break a lower Bollinger Band that is set Buy-and-hold stats: one standard deviation away (instead of the usual two). T).MSFT.activetradermag. which are lines typically plotted two 12/7/92 12/7/93 12/7/94 12/7/95 12/7/96 12/7/97 12/7/98 12/7/99 12/7/00 12/7/01 standard deviations above and below a moving average.00 46.000 2. avoid being stopped out prematurely with a loss.00 July www.00 Buy 48. there DJIA 175% 31. LX Buy Sell 52.000 trade if the market is already trending in one direction but quickly reverses to establish a new trend in the opposite 500. Test results: The system was originally tested on all 30 stocks in the DJIA.500.5% (current) 29 months (current) will be times when the system is out of the market completely. Exit with a profit or loss if 58. Bollinger Bands expand during high-volatility periods and contract during low-volatility periods.00 42. The number of shares to trade (ST) is calculated using the following formula: ST = AC * PR / R where AC = Available capital PR = Percent risked R = Distance between entry price and exit price (stop-loss).00 Reverse the rules for short trades. 39 February Source: Omega Research ProSuite March April May June Account balance ($) Volatility breakout system 2.000 market is about to burst out of a congestion area and potentially establish a new long-term trend. IBM.5% (current) 26 months (current) Markets: This system will be tested on stocks and also on futures (p.000 direction. 1. C. the system is designed to stay out of the (AXP. daily price moves below its 30-day moving average and penetrates LX = Long exit 56. the system will work to keep you in the trade to sions. Money management: 1.000 System logic: The volatility breakout system is a classic trading strategy based on identifying situations when a 1. though. KO. GM. breaks the upper Bollinger Band. Singling out these stocks. 2. JPM. HWP. Risk 4 percent of available equity per stock traded. MRK.000.00 Philip Morris (MO). CAT. market to avoid taking any unnecessary risk.000. Test data: Daily prices for 14 Dow Jones Industrial Average stocks When volatility is high. Because the Total Maximum Longest next entry cannot occur until price moves back above the lower Index return drawdown flat period band.EQUITY CURVE 3. Test period: November 1992 to June 2002. Go long tomorrow if price moves above its average price for the last 60 days and breaks the SAMPLE SIGNALS upper Bollinger Band.00 44. 70). INTC. but if an entry is trig. MO. It also can signal a 1. S&P 500 120% 40% (current) 26 months (current) Nasdaq 182% 80. To exit. This system uses Bollinger Bands (complemented by 0 moving averages).00 54. above its moving average and above the upper band. A long entry is triggered when price moves above its 60-day moving average and Starting equity: $1 million (nominal).com • October 2002 • ACTIVE TRADER .000. the 14 in this test were selected because they were the Rules: ones that showed a profit. with $10 deducted per trade for slippage and commisgered anyway.500. 2. DIS.00 LX a lower Bollinger Band set one standard deviation away.

57% -5.com.70% 10. the current drawdown has gone on for 42 months.88% -15./Year — trades per market per year • Tr. Start by e-mailing system logic (in TradeStation’s EasyLanguage or in an Excel spreadsheet) and a short description to editorial@activetradermag. raised by 1/n.72% 30.627(109.40% 83.78% 162. Doing otherwise would run the risk of using all the capital on only a few positions. Note: Each system must have a clearly defined stop-loss level and a suggested optimal amount to risk per trade. p.DRAWDOWN CURVE 12/7/92 0% -5% -10% -15% -20% -25% -30% 12/7/93 12/7/94 12/7/95 12/7/96 12/7/97 12/7/98 12/7/99 12/7/00 12/7/01 failed to create results as good as the ones produced by the test on currency futures (see Futures System Lab. If we were to test only the remaining 11 that showed a profit.8 TIM (%): Tr. The individual stocks either trend well. respectively • Lrg. This in turn will result in relatively small dollar gains despite large price swings. In this system.58% 37. and we’ll get back to you. (%) — average continuously compounded annual return • Profit factor — gross profit/gross loss • Avg.29% 12 months 16. DIT: 35.0 LEGEND: Cumulative returns — Most recent: most recent return from start to end of the respective periods • Average: the average of all cumulative returns from start to end of the respective periods • Best: the best of all cumulative returns from start to end of the respective periods • Worst: the worst of all cumulative returns from start to end of the respective periods • St. tied cap (%) — average percent of total available capital tied up in open positions • Win. where a group of stocks interacts within a single trading account.08% 17. months (%): 53 Drawdown Max. maximum 60 markets.8 41. trades (%) — percent winning trades • TIM (%) — amount of time there is at least one open position for entire portfolio.04% 24. win/loss ($): 391.500 trading days).172 Total return (%): 149 Avg./Mark. three stocks that were profitable when all 30 stocks were tested showed a loss when the field was pared to 14.: Annualized Most recent: Average: Best: Worst: St. dev: the standard deviation of all cumulative returns from start to end of the respective periods Annualized returns — The ending equity as a result of the cumulative returns.99% 12.29% 24 months 36 months 48 months 60 months -0. trades — number of trades • Avg. respectively • Win.activetradermag. though. In a dynamic portfolio such as this one.06% 12. DD (%) — maximum drop in equity • Longest flat — longest period.26% 24 months -0.30% STRATEGY SUMMARY Profitability End. Our system-testing technique lets all markets share the same account and is based on the interaction within the portfolio as a whole.61% 26.86% 15. We’ll test it on a portfolio of stocks or futures (for now.78% 36 48 months months -1.34 Avg. close to 75 percent of available capital is tied up./Month — trades per month for all markets Send Active Trader your systems If you have a trading system or idea you’d like tested.76% 58.33% -16.5 Trade statistics No. One thing to keep in mind is that portfolio composition is more complex than simply eliminating instruments that don’t perform well. and each market.com 40 .26% 3.22% 22.53% 22. annual ret.090 (13.70% 10.00 Profit factor: 1. win/loss ($) — largest wining and losing trade. which results in a large amount of whipsaw losing trades and rather severe drawdowns.270 Avg.19% 10.50% 4. send it to us at the Trading System Lab.81% 64. where n is the respective period in number of years LEGEND: End. win/loss ($) — average wining and losing trade./Month: 100 57.41% 42. the large distance between the entry and exit prices requires trading rather small positions. trades (%): 38. DD (%): Longest flat (m): 25.3 4. Past performance does not guarantee future results.61% 3. The way to overcome this is to diversify by trading as many stocks from different sectors and groups as possible.91% -5.29% 155.19% -19. win/loss ($): 31.57% 16.23% 27. It is not meant to recommend or promote any trading system or approach. This system is also a bit passive in its trade frequency. 44). Traders are advised to do their own research and testing to determine the validity of a trading idea. As proof of this. historical testing may not reflect a system’s behavior in real-time trading.08% 17. or they don’t. To make it more aggressive. Also. ACTIVE TRADER • October 2002 • www.10% -10.94% 20. spent between two equity highs • No.61% 13. in months. the lookback period can be shortened and/or the standard deviation boundaries tightened. using the last 2./Year: Tr. Even though the system trades only 14 stocks.26% 1. tied cap (%): 73 Win.12% 8. the ability of an individual stock to turn a profit or loss also depends on the behavior of all the other stocks. equity ($): 2.07% 60 months 4. using true portfolio analysis/optimization. This correlation also makes it extremely difficult to trade a longterm system on the stock market.76% 58.0 Avg. dev: 12 months 16.83% 9.437) Win.491. ROLLING TIME WINDOW RETURN ANALYSIS Cumulative Most recent: Average: Best: Worst: St.546) Lrg. trades: 456 Avg. and it is not very likely that a system like this will start producing a profit anytime soon. trade ($): 3. Most system-testing software only allows you to test one market at a time. dev.7 4.88% -15. it’s likely a few more would turn into losers. months (%) — percentage profitable months over test period • Max. annual ret.45% 11. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.43% 36. This is likely a reflection of the disappearance of the stock market’s pre-2000 trending characteristics. respectively • Tr.11% 7.77% 87. trade ($) — amount won or lost by the average trade • Avg.50% 42.95% 102. equity ($) — equity at the end of test period • Total return (%) — total percentage return over test period • Avg./Mark. DIT— average days in trade • Avg. (%): 10. almost all at once.

82% Worst: -28.54% LEGEND: Cumulative returns — Most recent: most recent return from start to end of the respective periods • Average: the average of all cumulative returns from start to end of the respective periods • Best: the best of all cumulative returns from start to end of the respective periods • Worst: the worst of all cumulative returns from start to end of the respective periods • St. the currency market doesn’t care all that much about the stock market.600 (131. 31.000 2. Finally. The reason is the smooth. Reverse the rules for short trades. (%) — average continuously compounded annual return • Profit factor — gross profit/gross loss • Avg.78% 24. dev.90% 15. months (%) — percentage profitable months over test period • Max.52% 6.500.000.86% 27. equity ($): 4.73% 114. as many as 40 to 50 different futures contracts could be traded before reaching the same level of margin fewer than 20 stocks would require. To make it more aggressive. win/loss ($) — average wining and losing trade.13 Profit factor: 1. DIT— average days in trade • Avg. and each market.75% 76. Theoretically.000 Futures volatility breakout system System logic: The system uses Bollinger Bands and a moving average to trade volatility breakouts. respectively • Tr. Rules: 1.33% 4. trades: 303 Avg. such as currencies.17% 7.91% 15. ROLLING TIME WINDOW RETURN ANALYSIS Cumulative 12 24 36 48 60 months months months months months Most recent: 34. long-term economical and political climate./Month — trades per month for all markets 41 www./Year — trades per market per year • Tr.31% 58. and even trade the system more aggressively.7 Trade statistics No. where n is the respective period in number of years STRATEGY SUMMARY Profitability End.77% St. Test results: If there ever was a system built for the currency markets.29% Annualized 12 24 36 48 60 months months months months months Most recent: 34. DD (%): 31. without extending ourselves too much.95% Average: 15.000 1./Year: 4. the lookback period can be shortened and/or the standard deviation boundaries tightened.activetradermag. tied cap (%) — average percent of total available capital tied up in open positions • Win.536) Lrg. Exit with a profit or loss if price moves below its 30-day moving average and penetrates a lower Bollinger Band set one standard deviation away (instead of the usual two).057 (29. Dollar index. Go long tomorrow if price moves above the 60-day moving average and breaks the upper Bollinger Band. respectively • Win.07% 15. trade ($) — amount won or lost by the average trade • Avg. the most recent drawdown of approximately 30 percent is the only one over the last 10 years of such magnitude. Risk the following percentages of available equity per market: 2 percent for Australian dollar.500. trades — number of trades • Avg. the drawdowns can be both deep and long. dev: the standard deviation of all cumulative returns from start to end of the respective periods Annualized returns — The ending equity as a result of the cumulative returns.12% 28.000 4. trades (%): 42. In other words.26% 23. Swiss franc. 56.68% 80.com • October 2002 • ACTIVE TRADER . (The yen and the Euro are traded with twice the risk because they are the most liquid currencies. 1999)/Euro (after Dec. which resulted in both a drawdown and flat time that would be deemed unacceptable by most professional money managers. long-term trends currencies sometimes exhibit.000 3.6 LEGEND: End.67% 1. however.717 Avg. This is because futures have much smaller margin requirements than stocks. trade ($): 11.82% 15.350) Win. spent between two equity highs • No. These numbers should stay under 30 percent and 18 months. equity ($) — equity at the end of test period • Total return (%) — total percentage return over test period • Avg. energies and interest rates. win/loss ($): 53. 2.27% 22. respectively.500. win/loss ($): 345. the stock market is very sensitive to all other markets.000 1. probably because they are mostly influenced by the global. months (%): 51 Drawdown Max. The major disadvantage of a trend-following system is most markets that work well with this type of a system are usually correlated. Australian dollar.FUTURES & OPTIONS System Lab EQUITY CURVE 5.94% Worst: -28. win/loss ($) — largest wining and losing trade.) 2.07% 34. and 4 percent for Japanese yen. Markets: Most trending futures markets. this is it. Test period: November 1992 to June 2002 Test data: Daily futures prices for eight currency futures: Japanese yen. and D-mark (until Dec. Canada dollar.85% 18. Account balance ($) 3.8 Avg. PV = Dollar value of a one-point move. The number of contracts to trade (CT) is calculated with the following formula: CT = (AC * PR) / (R * PV) where AC = Available capital PR = Percent risked R = Distance between entry price and exit price (stop-loss).20% 24.10% -0.82% 138. in months.22% 5.36% 6.66% 137.000.16% Best: 74. This means there is plenty of room to add markets. dev: 18. Dollar index and Swiss franc. The system also has a relatively low trade frequency.6 Longest flat (m): 19./Month: 2.550.77% St./Mark.000 0 11/25/92 11/25/93 11/25/94 11/25/95 11/25/96 11/25/97 11/25/98 11/25/99 11/25/00 11/25/01 The risk for this system was 2 or 4 percent per trade.1 TIM (%): 97 54.0 Tr.000 2.26% 6.: 18.74% 37. 31. Starting equity: $1 million (nominal).82% 7. The logic and tools for this system are described in the Trading System Lab on p. This is reflected in the erratic look of the system’s equity curve.39% 65. Therefore.14% Average: 15. Money management 1.1 Tr. they will either all work well or perform poorly simultaneously. DIT: 35. D-mark and Euro.000. Because of this.10% -1. The system ties up an average of 11 percent of capital.290 Total return (%): 355 Avg. tied cap (%): 11 Win. respectively • Lrg. while the stock market cares a great deal about the currency market. Most of the previous drawdowns bottomed between 10 to 20 percent.06% 18. (%): 17. and it usually requires a couple of very good trades to get the system profitable again.31% 25. the latest drawdown is quite possibly an anomaly. This system was also tested on stocks (see Trading System Lab). and your worst drawdown is always still to come. annual ret. annual ret.17% 16. Canadian dollar.000 4.15% 55.53% 109.75% 32.000. there are no guarantees in the market.34% 21.76% 38. 1999). British pound.32 Avg.90% Best: 74.50% 108. In contrast. Research has shown that a trend-following system will work best when traded on 15 to 20 select markets from various sectors of the economy. trades (%) — percent winning trades • TIM (%) — amount of time there is at least one open position for entire portfolio.000 500.500. raised by 1/n. British pound. That said.000.94% 202./Mark. DD (%) — maximum drop in equity • Longest flat — longest period.

we will add a noise filter. The noise channel breakout system shows how a filter can improve the performance of intraday breakout trading. The discussion is broken into two parts. Unfortunately. The rules for the resulting noise channel breakout system (NCBS) are: Buy rule: If price crosses above the highest high of the last n rice trends begin with a breakout of a previous high or previous low. (For an important point on testing stock trading strategies. many breakouts are random — mere market noise. which goes long on a move above the highest high of the last n bars and goes short on a move below the lowest low of the last n bars. • nlo.” p. covering 1) the system rules and data selection and 2) testing procedures. The strategy will be tested on International Business Machines (IBM). As a result. One example of a simple filter is to wait for consecutive closes above or below a breakout level. 21 to April 6. There are three system parameters to find: • nhi. Another example is waiting for price to penetrate a breakout level by x percent or points before acting on the signal. which is the number of bars in the lookback period used to determine the lowest low price (llp). and whipsaws. which is the number of bars in the lookback period used to determine the highest high price (hhp). . Intraday data has a high noise level. P The noise channel breakout system The basic system we will use here is a fairly simple and effective breakout system that has been in the public domain for many years: the channel breakout system. see “A note on price data and dividends. traders sometimes attempt to use filters to improve the odds of catching a successful breakout trend. to the basic channel breakout system.D. In the tests that illustrate this strategy. whipsaws) can completely overwhelm a trading system.TRADING Strategies Better breakout trading: THE NOISE CHANNEL SYSTEM All breakout traders have to deal with the reality of false moves BY DENNIS MEYERS. This will give you the necessary tools for performing similar research and tests on other markets. we’ll use five-minute bars of IBM from Feb. designated by the symbol f. which is the amount price must exceed the hhp or llp to trigger a buy or sell. The following discussion will analyze a variation on the simple channel breakout system that uses the latter type of filter to minimize whipsaws on an intraday basis. False moves and reversals can repeatedly whipsaw traders who act immediately on typical breakout signals. To help eliminate such random movement. Without some kind of filter. 75). the losses generated by the random price movement (that is. • f. PH. meaning it contains a great deal of random price movement that looks significant but turns out to be meaningless.

buy at market. stable — i.547 Max. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg. In other words. consec. Statistics based upon trading 1.613. Therefore. when short. losers: Avg.64 1. holidays. Again. you would go long when price moved 2 points above the highest high of the last 20 bars. contracts held: 1 All trades Open position P/L ($): Gross loss ($): Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg.840 -1. For example.000 shares of IBM. win/avg. it cannot be higher than the previously calculated hhp as previous highs are dropped out of the lookback window.060 -1. no.490 35. For more information on walkforward testing and how it was used for this strategy.activetradermag. In addition. 21 to March 23. a situation can occur where there is a higher hhp without the price filter f being hit.38 3 21 Testing the strategy The “walk-forward testing” approach will be used to test this strategy because of the volatile nature of intraday stock prices.370 54 22 -2. In addition.000 shares of IBM. It cannot go higher.40 223. no.260 48 26 5..e.191 3 26 Source: TradeStation by TradeStation Group Inc. no. it cannot be lower than the previous calculated llp as previous lows are dropped out of the lookback window. bars in winners: Max intraday drawdown ($): Profit factor: 13.. if n = 20 and f = 2 (points). the profits. events and trader sentiment. and when calculating the highest high price (hhp). winners: Avg.com • September 2001 • ACTIVE TRADER . bars in losers: 0 -25.419 Max. winners: Avg. five-minute bars from Feb. loss: Max. no. consec. when short the stock. Exit rule: Close the position five minutes before the NYSE close (no trades are carried overnight). bars in winners: Max. win/avg. trade(win & loss) ($): Max. contracts held: 1 All trades Open position P/L ($): Gross loss ($): Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg.890 39.500 47 22 5.613 3 39 -9. Also. In addition. tests performed on intraday data three months ago may no longer be representative of today’s intraday price action. when long the stock. Performance summary: Total net profit ($): Gross profit ($): Total no.010 47 25 -1. www. trade(win & loss) ($): Max. five-minute bars from Feb.940 1. bars in losers: 0 -25. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg.bars (nhi) by an amount greater than or equal to f. consec. wins and drawdowns should not change much as the parameters move by a small amount away from their optimum values. Otherwise. As a result. no. sell at market. the time of year — such as the season. when long and when calculating the lowest low price (llp).18 289. consec. the results should be 43 TABLE 2B TEST PERIOD 2 Performance summary for noise channel breakout system: IBM. It cannot go lower.000. — affects the character of intraday markets. to avoid the situation where a lower llp occurs without the price filter f being hit. the llp can only stay the same or go higher.510 1. Sell rule: If price crosses below the lowest low price of last n bars (nlo) minus an amount greater than or equal to f. no.940 1.” The best parameters will be defined as those values that generate the best net profits combined with the minimum drawdown and minimum largest losing trades. Intraday price dynamics are constantly changing because of economic surprises. Performance summary: Total net profit ($): Gross profit ($): Total no.153. see “Proper system testing. vacation time. etc. Statistics based upon trading 1. intraday drawdown ($): Profit factor: 10.660 1. 28 to March 30.309 4 39 -8. loss: Max. losers: Avg. the hhp can only stay the same or go lower. the system TABLE 1 OPTIMUM PARAMETER VALUES FOR TEST DATA Start date 2/21/01 2/28/01 End date 3/23/01 3/30/01 nhi 8 8 nlo 4 4 f 1 1 TABLE 2A TEST PERIOD 1 Performance summary for noise channel breakout system: IBM.470 1.

52 99.52% -0. March 26 to April 6) of the noise channel breakout system shows the strategy actually worked better on the short side than the long side.070 50 8 -1.420 950 390 800 220 350 2.900 380 3. This trade summary for the out-of-sample test (five-minute bars.com 44 .190) (840) (1.67 # bars in trade 67 68 75 60 10 41 34 15 60 71 49 25 75 27 8 40 P&L ($) P&L (%) Max.02% -0. Table 3 summarizes the combined performance of the two out-of-sample data segments from March 26 to April 6.890) (500) (10) (940) (940) (1.150) 20 (500) (50) (1.740 1.30 Exit date 3/26/01 3/27/01 3/28/01 3/29/01 3/29/01 3/30/01 3/30/01 4/2/01 4/2/01 4/3/01 4/4/01 4/4/01 4/5/01 4/6/01 4/6/01 4/6/01 Exit time 15:55 15:55 15:55 15:05 15:55 13:05 15:55 10:55 15:55 15:55 13:50 15:55 15:55 11:55 12:35 15:55 Exit price 94.05 94.performance using an nhi of 10 bars should be similar to that using nine bars or 11 bars.350) (1. from March 26 to April 6.20 96.160) (500) (1.59 97.38% 10:20 15:50 12:00 10:30 15:15 11:55 13:15 10:05 15:40 15:40 11:20 14:00 15:25 11:15 12:05 12:35 10:35 10:15 9:40 11:20 15:45 10:00 14:35 10:55 11:40 10:00 10:30 14:20 9:40 11:55 12:35 13:55 ACTIVE TRADER • September 2001 • www.40 93. Tables 2a and 2b show test results using these parameters. we considered only those results with four or less maximum consecutive losses.00 92. Table 4 is a trade-by-trade summary TABLE 4 TRADE-BY-TRADE SUMMARY Combined walk-forward out-of-sample performance summary for the noise channel breakout system: IBM five-minute bars from March 26 to April 6.92 96. contracts held: 1 All trades Open position P/L ($): Gross loss ($): Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg.50 94. the same nhi and nlo values tested without any filter resulted in a loss of $1.470 (940) (940) (370) -0.000 1.59 94. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg. The trades in this time period are the out-of-sample trades TABLE 3 OUT-OF-SAMPLE RESULTS generated from the optimized parameters from the two test sections of Feb 21 to Test results Table 1 shows the optimum parameter values for the test window described in “Proper system testing.38% 1. no. Statistics based upon trading 1.480 2.00% 0.30 97.660 300 Time Max.49% -1.” The nhi was eight bars.50 2.70 96.390 14.15 98.000 shares of IBM. By comparison. profit 0 4.620) 0 (380) (1. loss: Max. Performance summary: Total net profit ($): Gross profit ($): Total no.82% 4.75 96. the nlo was four bars and f was 1 point.24 97.20% 0.96% -0.40 94.25 96.960) Time 3/26/01 10:20 3/27/01 10:15 3/28/01 9:40 3/29/01 10:05 3/29/01 15:05 3/30/01 4/2/01 4/2/01 4/3/01 4/4/01 4/4/01 4/5/01 4/6/01 4/6/01 4/6/01 9:40 9:40 10:55 10:00 9:45 13:50 9:40 9:40 11:55 12:35 3/30/01 13:05 (770) 4. Entry Date Entry time Buy or sell Sell Buy Sell Buy Sell Buy Sell Buy Sell Sell Buy Sell Buy Sell Buy Sell Entry price 93.97% -0.300 3.000 3.00 95.activetradermag.69% 0. This performance represents what would have happened in real time if you used the system parameters found in the test section (not including slippage and commissions).38 3 37 Source: TradeStation by TradeStation Group Inc.350) 1.382 5 54 -4.30 98.50 92.18% 3. drawdown ($) (1. Also. consec.24 97. no.900 2. bars in winners: Max.85 98.420 (1.75 524.350 -758.68 97.500 0 150 2.20 97. trade(win & loss) ($): Max. no. win/avg.040 550 1.90 96.90 94. in choosing the “best” parameters.300) 0 (1.58% -0.97% 2.16% 2. losers: Avg.382 Max.460 16 8 4.50 90.807. consec.150.75 95. winners: Avg. intraday drawdown ($): Profit factor: 8.600 2. bars in losers: 0 -6.00 91.05% -1.88 96.00 92.

testing on price data the system parameters were not originally derived from — can determine if a system is robust and has a chance of performing well in real trading. If you were an owner of IBM. On the “Ex-dividend dates” (approximately one month before the payable date). However. using optimization. been cherry-picked for this particular data period (a process known as “curve-fitting”). no adjustment needs to be made. and again finding the parameters values through optimization on this new data. Breaking down the numbers With respect to average winning and losing trades. Thus. you would receive those dividends in cash.com • September 2001 • ACTIVE TRADER . when developing and testing a system using historical stock data. drawdowns and profit factor. Almost any period of historical prices can be curve fitted easily to give the false illusion of future profitability. 28 to March 30. the important point is how well it will perform on price data it has not been optimized on — that is.activetradermag. without out-of-sample testing. In short. An important (but unspoken) point in walk-forward testing is that if you cannot get good results in the out-of-sample data segments. Figure 1 is a five-minute chart with the noise channel superimposed. IBM has a small downward price bias equal to the amount of the yearly dividend. The only thing the statistics from the test section tell you is how well you have curve-fitted the data in the test section. However. if the intraday time period fell on an ex-dividend date.g. 45 www.. 28 to March 30. In other words. 10. many market pundits still make the unproven claim that statistics generated solely from optimized buy and sell trades in the test section (the initial period of price data)have value in predicting whether or not the system will perform well in the future. an adjustment would have to be made to avoid distortion. and with hindsight. A walk-forward testing procedure was applied to the noise channel breakout system as follows: Five-minute bars from a period of four weeks from the start of the test period — Feb. these performance measures in no way reflect how a system will perform on price data it has not been optimized on. to Feb. the only thing indicated by the optimum values in the test portion is that the data has been “curve-fitted” as best it can with this system. IBM pays dividends on a quarterly basis. As a matter of fact. Nothing could be further from the truth. the price of the stock is adjusted down by the value of the dividend.A note on price data and dividends A n overlooked aspect of testing a stock trading strategy is the effect of dividends. June 10. 10 and Dec. However. out-of-sample data. Sept. over the course of a year. real-time system performance will be random. Because no dividends were paid in the data sample used for the test in this article. W These parameter values were then applied to an out-ofsample data period following the test segment (March 26 to March 30). it’s almost impossible not to get an excellent fit with great statistical results. it’s nothing more than a hope and a prayer to believe that system performance in the future will be anywhere near the optimized performance. number of days in lookback period. or parameters. but it does indicate that four weeks of test data was enough to capture the intraday price dynamics of this stock. the out-of-sample performance (Table 3) was better than the test sample performance (Tables 2a and 2b) The better performance of the out-of-sample section could have been coincidental. Proper system testing hen testing any trading strategy. and are unlikely to perform as well on other historical test periods. or in actual trading in the future. prices are not adjusted for dividend payments. making up for the small downward bias. Without further testing on out-of-sample data. these optimized parameter values have. For example. These optimized parameter values are then applied to the next out-of-sample five-minute intraday data window (April 2 to April 6). This creates a small distortion in parameter selection and forward-adjusted results. However. Despite these facts. noise filter value) were determined by testing a range of values for each. as well as some of the buy and sell signals from the Table 4 trade-by-trade summary. This walk-forward process was repeated by moving the test data window forward one week. the last three years (using daily price data). there is no way to tell if the system will work in the future. it’s possible to take a trading strategy with four independent variables. in essence. the “best-performing” system parameters (e. find the values for each of them that give the best (optimized) results on a specific historical period — say. usually on the “dividend payable dates” of March 10. Only out-of-sample testing — that is. 21 to March 23 — were chosen and system parameter values were found through optimization on this intraday data segment. March 23 and Feb. For example. At this stage of system development.

you should continue to perform walk-forward testing to determine if there is a shift in the system’s effectiveness and whether better parameters have emerged. but without taking the time to adequately research a technique. Overall. five-minute 101 100 -1 99 98 97 96 95 1 94 93 3/26 11:20 12:15 1:10 2:05 3:00 3/27 11:15 12:10 1:05 2:00 2:55 3/28 11:10 12:05 1:00 1:55 2:50 Source: TradeStation by TradeStation Group ACTIVE TRADER • September 2001 • www. Also. There were no big winners or big losers. Every market has subtle differences because the participants vary from market to market. at least 10 additional test and out-of-sample windows should be examined to ensure the performance summarized here was not the result of chance. this could indicate a “negative” bias for the system. On one hand. International Business Machine (IBM). Also. Any trading method should be tested before you risk capital on the technique. the system constraint of not carrying positions overnight eliminated many negative opening surprises. indicating steady returns. Building on the results To use this system in real-time trading.com 46 . The system successfully captured intraday trends. the chances of success are low.activetradermag. there is a considerable amount of work involved. On the other hand.4 times average losses in the out-of-sample section. Figure 1 shows how the system was able to efficiently capture intraday trends in IBM. given the current bear market environment. the ability to cash in on the short side has value. FIGURE 1 NOISE CHANNEL AND TRADE SIGNALS A few of the buy and sell signals generated by the noise channel breakout system are shown. the NCBS did a good job of minimizing the whipsaw losses prevalent in breakout trading systems and maximizing the profits from major intraday trend moves. Average wins were 2. Granted. Consequently. To determine if this approach can be used on other stocks or markets it would be necessary to follow the same proce- dures and determine the appropriate parameters for each.The out-of-sample trade summary (Table 4) shows the system did better on short trades than it did on long trades. market activity can change over time. traded on IBM.

To help eliminate www. again applied to IBM fiveminute price bars. 21. but quickly retraces.TRADING Strategies The long and short of it: The Noise Channel Breakout SYSTEM 2 BY DENNIS MEYERS. The NCBS is an intraday breakout system based on five-minute bars. we’ll review the basics of breakout systems in general and the NCBS in particular. A lthough price breakouts are the basis for many trading approaches.activetradermag. respectively. Here we will use the NCBS. delaying trade entry until the initial breakout has been confirmed by a price move of a certain size or duration in the direction of the breakout. whether the filter should be different for long and short trades. First. traders often apply filters to breakout systems. to April 6. 70) showed how a simple channel breakout system. thus avoiding false breakouts. a 40-day channel breakout goes long when price moves above the highest high of the last 40 days and goes short when price falls below the lowest low of the last 40 days. breakout systems are plagued by false signals — when price initially breaks out. since uptrends and downtrends have different characteristics. For example. Breakout systems can be used on intraday price data. as well as daily or weekly data. One aspect the original noise channel breakout system (NCBS) did not address is whether to treat the long and short sides of the market the same — that is. to see if some improvement can be made by using different filters for long and short trades. 2001. 2001. Because intraday price action can be very volatile. the following tests will use the same five-minute bar prices of IBM from Feb. “Better breakout trading: The noise channel breakout system“ (Active Trader.D. The noise filter delays taking a breakout signal until the market provides some confirmation the breakout is sustainable. To compare the new version of the system to the previous one.com • October 2001 • ACTIVE TRADER 47 . September 2001. triggering a buy or sell. without some kind of filter the losses generated by the random price movement (that is. resulting in a losing trade. p. could be developed to trade IBM on an intraday basis using five-minute bars. whipsaws) can completely overwhelm a trading system. PH. with an additional noise filter to minimize whipsaws. NCBS refresher The basic channel breakout system goes long on a move above the highest high of the last n bars and goes short on a move below the lowest low of the last n bars. To combat this problem.

and with hindsight. As a result. the system parameters used for the sample data should be applied to different “out-of-sample” price data (say. to the basic channel breakout system. daily bars from 1993 up to 1998). real-time system performance will be random. • nlo. in essence. there are four system parameters for the NCBS-2: • nhi. as would occur in actual trading. However. the important point is how well it will perform on data on which it has not been optimized — that is. which is the number of bars in the lookback period used to determine the lowest low price (llp). or in actual trading in the future.com 48 . • f. which is the number of bars in the lookback period used to determine the highest high price (hhp). This “walk-forward process” simulates the application of a system to future data. The three system parameters for the NCBS are: • nhi. out-of-sample data. which is the number of bars in the lookback period used to determine the lowest low price (llp). which is the number of bars in the lookback period used to determine the highest high price (hhp). all you know is how well your system parameters performed during this particular period. An important (but unspoken) point in walk-forward testing is that if you cannot get good results in the out-ofsample data segments.” designated by the symbol f. W such random movement.activetradermag. WALK-FORWARD: Proper system testing hen testing any trading strategy. the last three years (using daily price data).Traders often use additional rules or filters to prevent being whipsawed by breakout trading strategies. ACTIVE TRADER • October 2001 • www. or parameters. and are unlikely to perform as well on other historical test periods. these optimized parameter values have been. In short. cherry-picked for this particular data period (a process known as “curve-fitting”). the system’s performance results have no implication outside this sample data set. it’s possible to take a trading strategy with four independent variables. which is the amount price must exceed the hhp or llp to trigger a buy or sell. The Noise Channel Breakout System 2 The Noise Channel Breakout System 2 (NCBS-2) uses different filter values (f. it’s nothing more than hope to believe that system performance in the future will be anywhere near the optimized performance. find the values for each of them that give the best (optimized) results on a specific historical period — say. from the original system) for the long and short sides of the market. To get an idea of how the system might actually perform. the NCBS adds a “noise filter. daily bars from 1998 to the present). without out-of-sample testing. • nlo. For example. Find out if using different filters for long and short trades improves the performance of an intraday breakout strategy. If a certain system is first tested on a “sample” piece of historical price data (say.

The same data will also be used so we can judge whether this new modification can improve performance. then buy at the market. in this case March 26 to March 30.350 1. consec.640 34. contracts held: 1 Open position P/L ($): Gross loss ($): Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg. it cannot go lower.510 1. Without further testing on out-of-sample data.18 289. Performance summary: All trades Total net profit ($): Gross profit ($): Total no. contracts held: 1 Gross loss ($): -25.25 Sell rule: If price crosses below the lowest low price of last nlo days by an amount of greater than or equal to xoD. there is no way to tell if the system will work in the future. only stay the same or go higher than its Statistics based upon buying and selling 1. when long and when calculating the lowest low price (llp).030 1. See “Walkforward: Proper system testing” for addi- FIGURE 1B TEST PERIOD 2 Performance summary for NCBS-2. The NCBS-2 rules are: Buy rule: If price crosses above the highest high price of the Start date End date nhi xoU nlo xoD last nhi bars by an amount greater then or equal to xoU. bars in winners: Max intraday drawdown ($): Profit factor: Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg.com • October 2001 • ACTIVE TRADER . which is the amount price must exceed the hhp to trigger a buy signal. bars in losers: 0 -24. the llp can Performance summary for NCBS-2. 44 www. losers: Avg. no. This process was repeated by moving the test data window forward one week to Feb.820 52.68 2 28 Source: TradeStation by TradeStation Group Inc. consec. when short. bars.38 3 21 Total no. IBM five-min. bars. winners: Avg. Exit rule: Close any position five minPerformance summary: All trades utes before the New York Stock Exchange Total net profit ($): 13. • xoD. Feb. The parameter values found were then applied to an out-of-sample period.460 38 20 5.400 -1. 21 to March 23. Testing the strategy As in last month’s article. the only thing indicated by the optimum values that are found in the test portion is that the data has been curve fitted as best it can with this system.547 Max. no. most recent value. loss: Max.63 18 -3. The walk-forward testing procedure was applied as follows: A four-week period from the start of the IBM five-minute bar data from Feb. Gross profit ($): 39.activetradermag. bars in losers: • xoU.260 48 26 5. In FIGURE 1A TEST PERIOD 1 addition. 21 through March 23 was chosen and system parameter values were found through optimization on this intraday data segment.470 1. 28 through March 30 and again finding the parameters values through optimization on this new data test window. In addition. which is the amount price must fall below the llp to trigger a sell signal. trade(win & loss) ($): Max. The logic behind modifying the filter values is because market behavior associated with buys and sells is quite different. TABLE 1 OPTIMUM PARAMETER VALUES FOR TEST DATA the noise channels associated with buys and sells should be independent of each other.940 1. win/avg. which in this case was April 2 to April 6.153.89 253. “walk-forward” optimization will be used here. loss: Max.) It should be noted that in this stage of system development. the hhp can only stay the same or go lower than its most recent value. Feb.Think of the symbols xoU and xoD as the “crossover Up” and “crossover Down” values.060 -1. no. bars in winners: Max. no. it cannot go higher.000 shares of IBM. winners: Avg. consec. consec. losers: Avg. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg. win/avg. IBM five-min. (Optimization refers to the search for the parameter values that give the best results. Statistics based upon buying and selling 1. intraday drawdown ($): Profit factor: 9.890 Open position P/L ($): 0 close (no trades are carried overnight).370 54 22 -2. no.000 shares of IBM.25 3 48 -10.378. no. then sell at the market.39 Max. trade(win & loss) ($): Max.25 12 0. The parameter values found were then applied to the next out-of-sample data set. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg. 28 to March 30. and when calculat2/21/01 3/23/01 8 1 4 1 ing the highest high price (hhp).309 4 39 -8. 2/28/01 3/30/01 18 1.723 1.

In addition. Figures 1a and 1b) show the performance summary of the test windows using the optimum parameters shown in Table 1.50 2.47 8 -1. Figure 3 shows a specialized percentage trade-by-trade summary from March 26 to April 6.tional information on optimization and walk-forward testing. IBM five-min. loss: Max. Statistics based upon buying and selling 1. and the out-of-sample performance summary of Figure 2a.000 shares of IBM. 21 to March 23 and Feb.740 0. Figure 5 is a five-minute chart of IBM with the NCBS-2 channels superimposed and some of the buy and sell signals from the Figure 3 trade-by trade summary indicated on the charts. consec. win/avg.480 2. winners: Avg. show the out-of-sample performance was better than the test sample performance with respect to average winning and losing trades. Statistics based upon buying and selling 1.660 2. no. win/avg. nlo.000 shares of IBM. contracts held: 1 All trades Open position P/L ($): Gross loss ($): Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg.730 -842.activetradermag. no. 28 to March 30. for comparison with Figure 3.198.070 50 8 -1.50 576. The in-sample trades are. IBM five-min. The lookback periods refer to number of bars and the filters values are given as dollar amounts.) Also included at the bottom of the chart is the bar-by-bar profit or loss of each trade. Improved performance? The optimum parameters in Table 1 show the first test data section produced the same optimum parameters as the original NCBS. wins and drawdowns should not change by much as the parameters move by a small amount away from their optimum values. bars in losers: 0 -6. March 26 to April 6.460 16 8 4.61 2 57 -5. no. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg. we considered only those parameter sets with maximum consecutive losses of four or less. the “best” parameters are defined as those values that give the best net profits and best total winning bars/total losing bars ratio with the minimum drawdown and minimum largest losing trades.57 2. Note that the trades from March 26 to April 6 are the out-of-sample trades generated from the optimized parameters from the two test sections of Feb. right) shows the combined performance summary of the two out-ofsample data segments from March 26 to April 6 for the original NCBS. trade(win & loss) ($): Max. curve-fit and are not of interest here. ACTIVE TRADER • October 2001 • www. Slippage and commissions are not included. Figure 2b (bottom. contracts held: 1 All trades Open position P/L ($): Gross loss ($): Percent profitable (%): Number losing trades: Largest losing trade ($): Average losing trade ($): Avg.com 50 . by definition.390 15 7 4.807.650 15. consec.g. Also. FIGURE 2A TEST PERIOD 1 Combined walk-forward out-of-sample performance summary for NCBS-2. no. This can been seen by observing that both xoU and xoD are exactly the same and are equal to f of the original NCBS. trade(win & loss) ($): Max. the profits.activetradermag.390 14. This improved performance in the out-of-sample section could have been due to Results Table 1 shows the optimum parameters for the IBM five-minute data series. March 26 to April 6. drawdowns and profit factor. Performance summary: Total net profit ($): Gross profit ($): Total no. consec.375 3 37 Source: TradeStation by TradeStation Group Inc.75 524. consec.28 Max. Figure 2a shows the combined performance summary of the two out-ofsample data segments from March 26 to April 6 for NCBS-2. the results should be stable — e. in choosing the best parameters. no. This performance represents what would have happened in real time if the parameters found in the test sections (Table 1) were used. intraday drawdown ($): Profit factor: 8.382 Max.000 1. of trades: Number winning trades: Largest winning trade ($): Average winning trade ($): Ratio avg.000 2. (All the signals. losers: Avg. loss: Max. bars in losers: 0 -6. The sample performance summaries in Figures 1a and 1b. as well as expanded performance statistics. Figure 4 contains the specialized trade-by-trade summary from the original NCBS for the same out-of-sample dates. no.382 5 54 -4. can be found at www. For comparison. Of the four system parameters to find (nhi.350 -758. Performance summary: Total net profit ($): Gross profit ($): Total no. xoU and xoD). winners: Avg.com. bars.67 3 38 FIGURE 2B TEST PERIOD 2 Combined walk-forward out-of-sample performance summary for NCBS. bars in winners: Max. bars. losers: Avg. bars in winners: Max intraday drawdown ($): Profit factor: 8. In addition.

96) (0.02 (0.650 3.000 shares.000 3.40 93.25 96.05 94.890) 10:30 11:55 Buy 12:35 Sell (1.59 94.52 99.50 90.30 98.58 (1.300) 11:40 0 (500) (10) (940) (940) Average (911) 10:00 14:20 9:40 11:55 12:35 (1. Trade size = 1.620) 10:35 10:15 9:40 15:45 14:35 (1.99 92.97 2.040 550 1.05 94.040 550 1.00 92.15 98.420 20 (500) (50) (1.75) (0.960) 13:55 Total Average Average 8.18 3.68 97.25 94.02 (0.52) -0.070 910 930 3.02 97.88 96.20) (1.40 94.55 91.16 2.20) 0.250) 11:40 (1.75 97.90 96. NCBS.420 950 390 800 220 350 2.650 0.730) (650) Trade % P&L (0.15 98.300 3.50 90.20 96.20 96.97) (0.. NCBS-2.55 95.69 0.830 (440) 700 2.68 97.49 (1.190) 10:00 (1.49) (1.75 95.150) 20 (500) (50) 1.75 95.150) (1.90 94.740 1.150 20 Time Trade Max $DD 0 (380) (500) (840) (670) (660) (520) (10) Time 3/26/01 10:20 3/27/01 10:15 3/28/01 9:40 3/29/01 10:05 3/29/01 15:05 3/30/01 4/2/01 4/3/01 4/4/01 4/4/01 4/5/01 4/6/01 4/6/01 4/6/01 9:40 10:55 9:40 10:55 12:55 9:40 9:40 12:00 12:35 3/30/01 13:05 10:20 15:50 12:00 10:30 15:15 11:55 13:15 15:40 15:40 11:20 14:00 15:25 11:15 12:05 12:35 (1.com • October 2001 • ACTIVE TRADER .FIGURE 3 OUT-OF-SAMPLE TRADE BY TRADE SUMMARY IBM five-min.30 98.730) 12:35 (2.660 300 Time Trade Max $DD 0 (380) (500) (840) Time 10:20 15:50 12:00 10:30 15:15 11:55 13:15 10:05 15:40 15:40 11:20 14:00 15:25 11:15 12:05 12:35 (1.92 96.500 0 150 2.67) Trade Max $Pft 0 4.38) 1.20 96.58 (0.67 67 68 75 60 10 41 34 15 60 71 49 25 75 27 8 40 (770) 4.24 97.50 94.75 97.61% 1.02 3.20 97.70 96.05) -1.350) 1.240) 13:55 Average (948) Total Average Average 8..350) 10:55 (1.90 94.02 Exit date 3/26/01 3/27/01 3/28/01 3/29/01 3/29/01 3/30/01 3/30/01 4/2/01 4/3/01 4/4/01 4/4/01 4/5/01 4/6/01 4/6/01 4/6/01 Exit time 15:55 15:55 15:55 15:05 15:55 13:05 15:55 15:55 15:55 12:55 15:55 15:55 12:00 12:35 15:55 Exit Bars Trade price in trade $ P&L 94.190) 10:00 (1. March 26 to April 6 Entry date 3/26/01 3/27/01 3/28/01 3/29/01 3/29/01 3/30/01 3/30/01 4/2/01 4/2/01 4/3/01 4/4/01 4/4/01 4/5/01 4/6/01 4/6/01 4/6/01 Entry time 10:20 Sell 10:15 Buy 9:40 Sell 10:05 Buy 15:05 Sell 9:40 9:40 Buy Buy 13:05 Sell 10:55 Sell 10:00 Sell 9:45 9:40 9:40 Buy Buy Sell 13:50 Sell Entry price 93.82) 4.475 Source: Meyers Analytics 51 www.00 92.49 0.950 2.420 950 390 800 220 2.38) Trade Max $Pft 0 4.160) 11:20 (1.160) 11:20 (1.30 97.85 98.90 96.03 (0.05) 2. Trade size = 1.493 Source: Meyers Analytics FIGURE 4 OUT-OF-SAMPLE TRADE BY TRADE SUMMARY IBM five-min.75 96. March 26 to April 6 Entry date Entry time Sell Buy Sell Buy Sell Buy Sell Sell Sell Buy Sell Buy Sell Buy Sell Entry price 93.76 2.18 3.45 93.000 3.50 94.85 98.30 Exit date 3/26/01 3/27/01 3/28/01 3/29/01 3/29/01 3/30/01 3/30/01 4/2/01 4/2/01 4/3/01 4/4/01 4/4/01 4/5/01 4/6/01 4/6/01 4/6/01 Exit time 15:55 15:55 15:55 15:05 15:55 13:05 15:55 10:55 15:55 15:55 13:50 15:55 15:55 11:55 12:35 15:55 Exit Bars Trade price in trade $ P&L 94.450) (1.55% 1.52 99.59 97.450) 12:00 (1.24 97.00 95.620) 10:35 10:15 9:40 15:45 14:35 9:45 11:00 13:20 9:40 (1.88 96.70 96.50 92.470 (1.67 67 68 75 60 10 41 34 60 75 24 36 75 28 7 40 (770) 4.59 94.92 96.47) 0.activetradermag.900 380 3.420 (1.52) (0.600 2.900 2.390 0.470 (940) (940) (370) Trade % P&L (0.300 3.50 92.59 97.82) 4.33 92.00 0.00 91.000 shares.

the results from at least 10 to 20 more tests and out-of-sample periods would have to be examined to make sure that the results above were not due to pure chance. given the current bear market. The NCBS-2 had one less trade and slightly better numbers. The less-complicated NCBS. this could be normal. The charts show the system constraint of not carrying positions overnight eliminated many negative opening surprises. The out-of-sample trade-by-trade summary of Figure 3 shows the system did better on short trades than on long trades. average trade drawdowns were -$948 and the average trade net profit was $576. There were no big winners or big losers. The performance summaries in Figures 2a and 2b show there is very little difference between the NCBS and NCBS-2. indicating steady returns. the difference wasn’t enough to claim any superiority or to justify the added complication of another optimization parameter. has a smaller drawdown and a smaller largest losing trade. Whatever the reason.500 1.000 2.com determine if the more complicated NCBS-2 offers any advantage in the trade-by-trade figures.493.6 times average losses in the out-ofsample section. There seems to be little advantage: Both systems’ totals and averages are nearly the same. Comparison of Figures 2a and 2b favors the simpler NCBS. 52 . The blue and red lines are the long and short filter levels. Average trade run-ups were $1. or perhaps. This could indicate a negative bias for the system.000 500 9:55 10:50 11:45 12:40 1:35 2:30 3/27 10:50 11:45 12:40 1:35 2:30 3/28 10:50 11:45 12:40 1:35 2:30 chance but does indicate that four weeks of test data were enough to capture the intraday price dynamics of IBM. Average wins were 2. while having a slightly lower net profit and average win/average loss ratio. five-minute 0 100 101 -1 99 98 97 96 -1 1 0 0 95 94 93 4. respectively. International Business Machine (IBM). this bias warrants further investigation.activetradermag. the system did a good job in minimizing the losses resulting from the inevitable whipsaws that will occur in any trading system and maximizing the profits from the major intraday trend moves of IBM.FIGURE 5 NCBS-2 SIGNALS Trade signals for the NCBS-2 are shown on a five-minute chart of IBM. To use NCBS-2 in real time trading. It’s also instructive to compare Figure 3 with Figure 4 to ACTIVE TRADER • October 2001 • www. However. Overall. The NCBS-2 did very well in catching every major intraday trend of IBM.

ADVANCED Strategies The multibar range BREAKOUT SYSTEM Breakouts of price channels can be profitable — if the volatility is there and you’re on the right side of the trade. llv1=l.xs(c). The second way is to determine the range of each bar (high minus low) and add that range (or a percentage of it) to. There are two simple ways to define support and resistance levels for price channels.hhv2(h). the more emotional the market becomes. FIGURE 1 TRADESTATION CODE FOR THE MULTIBAR RANGE BREAKOUT SYSTEM {Strategy: #MultiBarRangeBO} Input: n(45).xb(c). and enter trades when price breaks out of either the up side or down side of a channel. vars: hhv1(h). if l[ii]<llv2 then llv2=l[ii]. the upper and lower boundaries represent the price channel. One advantage to the second method is it better reflects the volatility of the market — it will expand and contract as the volatility changes. a trade will become profitable only if it continues to move in the direction of the breakout.llv2(l). end. if marketposition>=0 then Sell Short Next Bar xs stop. In the following system.45).llv1(l). if time<XTime then begin if marketposition<=0 then Buy Next Bar xb stop.ii(0). This is often reflected by the fact markets fall faster than they rise. traders have to cope with more risk. In both cases. value1=hhv1-llv1.m(15). hhv1=h. A breakout above or below the channel’s resistance or support creates buy or sell signals. Breakout strategies require the market to be in a high-volatility period. PH. the current close. www. The first way is to use the highest high and the lowest low of the lookback period. if XTime<>0 then SetExitOnClose. if l[ii]<llv1 then llv1=l[ii].XTime(1515). This stop-and-reverse system tries to capture intraday trends in the S&P E-Mini contract by recognizing differences in the characteristics of up moves and down moves.D. llv2=l. Volatility and emotion go hand in hand. the channel is determined by using the range of the price bars in the lookback period. In either case. xs= c . it is first necessary to define a lookback period. As volatility increases.45). value2=hhv2-llv2.sx(0. end.com • January 2004 • ACTIVE TRADER 53 . hence. for ii=1 to m-1 begin if h[ii]>hhv2 then hhv2=h[ii].activetradermag. BY DENNIS MEYERS. end. Such systems typically identify support and resistance levels when price has been moving in a range or channel. xb= c + (Value1 * bx). or subtract it from.bx(0. for ii=1 to n-1 begin if h[ii]>hhv1 then hhv1=h[ii]. B reakout systems are popular when markets are volatile. hhv2=h.(Value2 * sx).

2.50 ($125.22 ($862.95% 27. losing Largest winning trade Largest losing trade Largest winner as % of gross profit Largest loser as % of gross loss Net profit as % of largest loss Max.50 $173.00) 3.00% 11 9 0 $72. Sell rule: Sell the next bar at sellCh.85 $0.03% 1.sx*SRange where n = The number of lookback bars (including the current bar) for buy signals. sx = the percentage multiplier of the SRange for sell signals. 1.00% 21 13 1 $98.50 $6. bars in winning trades Avg.93 $225. 3.00 ($97.00) Short trades $3.160. this system avoids the opening gap whipsaw problem — trades being triggered because of large gap openings ACTIVE TRADER • January 2004 • www.00) ($362. drawdown (intraday peak to valley) Max.95 ($887. consecutive winning trades Max. Intraday bar exit rule: Exit the position on the close (no overnight trades). TABLE 1 MULTIBAR RANGE BREAKOUT SYSTEM PERFORMANCE SUMMARY.69 ($975.50 ($462. The sell price is determined by subtracting a percentage (a different percentage than the buy percentage) of the range of the last m bars from the current close.96 166.41) 2. the system issues a buy signal.32 $700.activetradermag. meaning it is always in the market: When a sell signal occurs. the parameters for the buy signals will be different than those for the sell signals.00 ($300.39% 1.18 17. bars in losing trades Max. but produced profits on long trades. consecutive losing trades Avg.38 $362. SRange = the price range over the last m bars. c = the current price.com Avg.00) Long trades $1.74 $0.55% 17.50) Multibar Channel Breakout rules This is a stop-and-reverse system. The trade rules are simple: 1. the lookback periods the system references for buys (n) and sells (m) are also different.262.450.12) 2. stop. These are the system’s parameters: ES = E-Mini price.00) 14. m = The number of lookback bars (including the current bar) for sell signals. losing trade Ratio avg.42 ($78. 2003 The system triggered more short trades during the test.912. If the next bar’s price exceeds the buy price.18% 32 22 1 $89. short trades are exited and a long trade is entered. long trades are exited and a short trade is entered. Buy rule: Buy the next bar at buyCh. All trades Total net profit Gross profit Gross loss Profit factor Open position P/L Total number of trades Percent profitable Winning trades Losing trades Even trades $4.05 143.76% 1.154.88 91.637. buyCh = c + bx*BRange. trade drawdown Source: TradeStation 54 .637.50 $4.912.00 20 55. The resulting Multibar Channel Breakout system will trade the S&P 500 E-Mini futures on an intraday basis using one-minute bars. JULY 7 TO AUG.00) ($400. when a buy signal occurs.00 ($1. If the next bar’s price falls below the sell price.725.50% 5 5 134.86 ($51. winning/ avg.8 69.725. bars in total trades Avg. the system issues a sell signal.00) ($475.00) 18.32 $207. Notice that not only are the percentage multipliers for long (bx) and short trades (sx) different. The TradeStation Code is shown in Figure 1 (opposite page).14 $0. winning trade Avg. drawdown (trade close to trade close) Max. trade net profit Avg. as well.50) ($175.65 $700.91 218.50) 4.462.00 $1.However.50 ($1.00 35 60. Although it may not be immediately obvious.17% 4 3 185.00) 10.50) 3.00% 5 2 45. sellCh = c . bx = the percentage multiplier of the BRange for buy signals.81% 23.00 ($300.00) ($475. BRange = the price range over the last n bars.39) 3.00 55 58. stop. The range for the last x bars will be defined as the highest high of the last x bars (including the current bar) minus the lowest low of the last x bars (including the current bar). because of the propensity for markets to fall faster than they rise. The buy price is determined by adding a percentage of the range of the last n bars to the current close — the previously described volatility-adjusted technique.50) ($300.

75 204 994.00 $387.50) ($50.00) $0.50 996.004.002.75 1.00 $87.25 240 996.50 $87.000. LLC 55 www.50 $187.50 $25.50 $100.50 47 986.00 Trade max$Pft $175.00 ($50.50 17 1.50 ($300.50 $575.50 113 979.50) $137.25 20 993.00 $300.00 989.75 370 1.75 93 992.00) ($12.50) ($275.000.00 $175.25 15 988.50 $150.75 111 1.00 $100.00) ($62.50) ($50.00) ($37.50) ($62.com • January 2004 • ACTIVE TRADER .00 $75.00) Time 10:36 12:38 13:01 13:08 13:47 10:19 14:11 9:33 9:04 11:23 10:01 9:46 8:48 12:18 15:10 9:04 11:02 11:06 9:02 11:27 12:03 13:01 14:41 8:32 10:01 10:18 14:02 14:49 8:35 12:55 13:39 14:35 10:27 12:45 13:11 8:45 12:33 13:25 15:12 10:29 12:37 14:04 9:01 10:34 12:33 9:56 11:52 12:01 13:04 14:21 9:08 12:14 13:16 13:22 8:51 Source: Meyers Analytics.75 317 1.50 996.50) ($37.50 $25.002. 1. 58.50 1.50) ($12.50) $687.50 993.50 ($12.75 1.50 94 990.50) $50.00 $87.50 253 985.00 $125.50) ($25.25 988.00) ($25.00 $37.004.50 ($62.25 18 989.25 258 1.75 9 1.50) ($12.50) ($12.00 342 988.activetradermag.25 1.00 $487.75 49 987.00) ($62.50 $50.00) ($112.002.75 978.002.00 227 989.50 $187.00 $700.50) ($12.00 2 986.002.75 377 997.00 $0.75 986.00) ($37.50) $362.00 $100.00 127 997.00 $0.00 $412.00) ($87.25 983.50) ($250.50 993.50 $187.00) $475.75 38 984.003.50) ($112.50) $162.75 995.25 403 978.25 173 986.50 ($87.00 $337.50) ($25.001.75 66 987.25 982.50 985.00 $0.75 1.50 $225.00 $0.00 18 988.007.50 ($50.50 119 1. Entry date 7/7/03 7/7/03 7/7/03 7/7/03 7/7/03 7/8/03 7/8/03 7/9/03 7/10/03 7/11/03 7/14/03 7/15/03 7/16/03 7/16/03 7/16/03 7/17/03 7/17/03 7/17/03 7/18/03 7/18/03 7/18/03 7/18/03 7/18/03 7/21/03 7/22/03 7/22/03 7/22/03 7/22/03 7/23/03 7/23/03 7/23/03 7/23/03 7/24/03 7/24/03 7/24/03 7/25/03 7/25/03 7/25/03 7/25/03 7/28/03 7/28/03 7/28/03 7/29/03 7/29/03 7/29/03 7/30/03 7/30/03 7/30/03 7/30/03 7/30/03 7/31/03 7/31/03 7/31/03 7/31/03 8/1/03 Entry time 10:36 Sell 12:35 Buy 12:52 Sell 13:01 Buy 13:24 Sell 9:51 Sell 14:09 Buy 9:33 Sell 8:58 Sell 9:05 Sell 9:54 Sell 9:05 Sell 8:48 Sell 12:10 Buy 12:22 Sell 9:04 Sell 11:01 Buy 11:02 Sell 8:49 Sell 11:27 Buy 12:02 Sell 13:01 Buy 14:35 Sell 8:32 Sell 9:20 Sell 10:01 Buy 11:37 Sell 14:37 Buy 8:35 Sell 12:51 Buy 13:38 Sell 14:27 Buy 9:21 Sell 12:45 Buy 13:10 Sell 8:43 Buy 9:01 Sell 13:01 Buy 15:08 Sell 8:33 Sell 12:37 Buy 12:57 Sell 9:01 Sell 10:34 Buy 11:28 Sell 8:34 Sell 11:38 Buy 11:56 Sell 13:02 Buy 13:04 Sell 9:00 Buy 11:20 Sell 13:07 Buy 13:22 Sell 8:46 Sell Entry price ($) 1.00) ($112.75 986.00 ($37.25 25 980.75 992.50 $12.50) ($125.50 $762.50 $275.50) ($87.001.75 1 980.00 988.25 986.50 $525.50 $62.50 $62.50 $0.00 $37.50) ($50.00 $62.50) ($75.50 Exit date 7/7/03 7/7/03 7/7/03 7/7/03 7/7/03 7/8/03 7/8/03 7/9/03 7/10/03 7/11/03 7/14/03 7/15/03 7/16/03 7/16/03 7/16/03 7/17/03 7/17/03 7/17/03 7/18/03 7/18/03 7/18/03 7/18/03 7/18/03 7/21/03 7/22/03 7/22/03 7/22/03 7/22/03 7/23/03 7/23/03 7/23/03 7/23/03 7/24/03 7/24/03 7/24/03 7/25/03 7/25/03 7/25/03 7/25/03 7/28/03 7/28/03 7/28/03 7/29/03 7/29/03 7/29/03 7/30/03 7/30/03 7/30/03 7/30/03 7/30/03 7/31/03 7/31/03 7/31/03 7/31/03 8/1/03 Exit time 12:35 12:52 13:01 13:24 15:15 14:09 15:15 15:15 15:15 15:15 15:15 15:15 12:10 12:22 15:15 11:01 11:02 15:15 11:27 12:02 13:01 14:35 15:15 15:15 10:01 11:37 14:37 15:15 12:51 13:38 14:27 15:15 12:45 13:10 15:15 9:01 13:01 15:08 15:15 12:37 12:57 15:15 10:34 11:28 15:15 11:38 11:56 13:02 13:04 15:15 11:20 13:07 13:22 15:15 15:15 Exit Bars price ($) in trade 1.50 35 985.00 1.25 184 988.00 990.50) ($112.50 23 1.25 991.00 $262.50 ($75.50) ($50.00) ($25.25 984.50) ($87.50) ($37.00 1.00) ($12.007.25 202 992.002.00 995.00 $375.00 $300.75 370 993.50 $450.00 $275.25 131 1.50 59 991.50 $150.50) ($337.00 $112.50 $37.50 983.50 $62.00) ($150.25 996.50 244 996.00 Time 12:06 12:35 12:55 13:23 14:22 11:21 14:48 11:03 13:48 9:08 14:46 14:05 10:12 12:10 14:30 9:39 11:01 14:01 9:20 11:31 12:19 14:11 14:53 14:10 9:50 11:10 12:37 14:38 11:16 13:35 14:14 15:13 9:50 12:57 14:57 9:00 9:52 14:54 15:08 8:52 12:55 14:42 9:34 10:58 14:17 10:32 11:38 12:19 13:02 13:14 10:14 11:40 13:19 14:56 9:35 Trade max$DD ($12.50 $62.00 ($125.50 1.002.00) ($12.00 ($125.50) ($50.00 7 996.50 $0.50 ($225.50 993.00 $362.50) ($25.50 $200.50 985.002.003.50 41 985.75 48 993.75 987.50 138 986.25 988.50) ($175.003.002.25 140 1.25 992.50 $725.50 $162.00) ($150.50 ($50.00 ($87.50 ($50.50) ($62.25 12 995.50) ($187.75 256 986.00 $187.50) $0.50) $0.00 $412.00) $100.50 $500.00 $187.50) $0.00 ($25.50 $275.00) $75.00 $362.00 $512.00 $237.00 ($87.25 986.00) ($400.00) $0.50 1.00) $0.50 $625.50 66 1.00 $37.50 $0.50 988.00) ($50.75 992.TABLE 2 TRADE-BY-TRADE SUMMARY: JULY 7 TO AUG.50) $37.00 $62.75 96 987.75 987.25 985.50 $337.25 1.002.25 983.00) ($87.25 989.18 percent of trades were profitable.00 $75.25 1.50) ($87.50 389 Trade $P&L $62.75 994.25 125 983.50 $37.00 985.00) $337.50) ($87. Overall.002.50 991.00 ($12.25 158 985.00) ($25.00) $162.50 $12.001.00) ($150.001.50) ($337.00 $700.25 1.00 $25.00 54 989.002.25 986.00 $450.25 117 983.00 ($62.012.00) $0.00) ($62.001.25 180 986.00 $375.00 976. 2003 This list contains each trade in the test period.00 995.00 107 1.00 40 978.00) ($62.50 $287.00 $300.75 1.50 986.

the bottoms of Figures 2 and 3 56 Testing The system was tested from July 7 through Aug. Table 2 (opposite page) is a trade-by-trade summary of all the trades. To avoid curve fitting. if there is a gap on the opening bar. Figures 2 and 3 are one-minute charts of the S&P E-Mini that span July 31 to Aug. sx = 0. we eliminated all cases that had more than five losing trades in a row. Table 1 (p.FIGURE 2 RIDING THE TREND During this period. These losses are small compared to the total net profit of $4. one-minute 1. we chose the one that had the highest total net profit and the lowest drawdown. and all the buy and sell signals are marked. difficult to sustain more than a handful of consecutive losses.activetradermag. since such performance was unlikely to be duplicated in the future. 2003. September 2003 S&P E-Mini futures (ESU03). Also.45. Breaking the expanded ranges takes time and avoids the opening gap whipsaw. The parameter ranges tested for the initial optimization test were: n =10 to 50 in steps of 5. the buy and sell ranges are expanded and no trades are made until the buy and sell ranges contract or the price breaks the expanded ranges.05. using September 2003 E-Mini futures (ESU03) one-minute bars. we had to choose one set of parameters that produced the most realistic results. m = 10 to 50 in steps of 5.0. 1. With this system.4 to 1 in steps of 0. we eliminated all results that had profit factors (gross profit divided by gross loss) greater than 4.000 995 990 Buy End of day exit End of day exit Short 985 980 600 400 200 0 -200 7/30 9:11 9:33 9:55 10:17 10:39 11:01 11:23 11:45 12:07 12:29 12:51 13:13 13:35 13:57 14:19 14:41 8/1 Source: TradeStation that quickly reverse and stop out the position. The Multibar Range Breakout channels are superimposed on the price series. Finally. and the biggest intraday drawdown was $887.05. Of the remaining test results. and produced small losses on two signals when the market was flat. one intraday downtrend. 43) shows the performance summary for the fourweek test period (slippage and commissions not included). m = 15. because it is ACTIVE TRADER • January 2004 • www. 1.com . bx = 0. the system caught one intraday uptrend.912. After the initial test. A wide range of parameter values was tested to find the optimal ones for the system. sx = 0. The average net profit per trade was $89 — well above slippage and commissions for a typical S&P E-Mini trade. The largest losing trade was $300. bx = 0.45.010 Short Short 1.005 Buy 1.4 to 1 in steps of 0. The optimization procedure produced the following system parameters: n = 45.

Figure 4 is a daily chart of the S&P EMini futures from July 7 to Aug. The Multibar Channel Breakout system’s positive performance warrants further investigation.com • January 2004 • ACTIVE TRADER . never had more than three consecutive losses.015 1.com/ purchase_articles. Individual articles can be purchased and downloaded from www. FIGURE 4 DAILY PERSPECTIVES The daily chart of the test period shows the system was able to profit on both sides of the market when conditions shifted from uptrend to downtrend to consolidation. the system will stay in the same direction the entire day. September 2003 S&P E-Mini futures (ESU03). The system was able to produce profits on both the long and short sides of the market. September 2003 S&P E-Mini futures (ESU03). If the numbers begin to deviate. 1. one-minute 1.FIGURE 3 ONE DAY.000 995 990 985 980 975 7 Source: TradeStation 14 21 28 57 www. daily 1. and shows the market moved up. All trades are exited at the close — no positions are held overnight.activetradermag. If you consider following this system in real-time. and aside from a streak of five losing trades near the outset of the test period.005 1. down and sideways during this period.010 1.activetradermag. another review of the system parameters are in order.002 1.htm. pay close attention to how the real-time statistics compare to the hypothetical numbers shown here.000 998 996 994 992 990 Sell End of day exit 988 986 984 982 980 978 End of 976 day exit 600 400 200 0 14:19 14:41 8/1 9:02 9:24 9:46 10:08 10:30 10:52 11:14 11:36 11:58 12:20 12:42 13:04 13:26 13:48 14:10 14:32 14:54 Source: TradeStation include the bar-by-bar profit or loss of each trade. ONE TRADE If no signal in the opposite direction is triggered.

The TRADING Systems Lab
EQUITY CURVE

DeMark variation
Markets: Stocks, stock index futures, index stocks
(SPDRs, DIAs, QQQs), futures and currencies

600,000

500,000

System logic:

Sell

This system is based on a simple pattern, named TD 300,000 Carrie, described by Tom DeMark in his book New Market Timing Techniques (John Wiley & Sons, 1997). It trades a move above or below the true high (the 200,000 highest of one bar’s high and the previous bar’s close) or the true low (the lowest of one bar’s low and the previous bar’s close) of the bar four days 100,000 prior to the current (active) bar. However, for the breakout to be valid it must be qualified by a few criteria. (The following rules are described in terms of 0 11/12/91 11/12/92 11/12/93 11/12/94 11/12/95 11/12/96 11/12/97 11/12/98 11/12/99 11/12/00 a long trade; reverse for short trades.) First, to identify a strongly trending market, the true high of four days ago must be higher than the high five days ago. requiring the breakout to take place intraday, we enter into an If this requirement is not met, it’s still possible to get an entry signal if orderly market instead of a highly volatile one. the market has made a correction counter to the direction of an eventual trade (i.e., a downward correction in the case of a long trade). In Rules: an uptrend this correction is identified by the highs of either two or 1. Prepare to go long today if a. the true high from four days ago is higher than the high from three days ago being lower than the true high of four days ago. either two, three or five days ago, and Second, the close of the bar prior to the anticipated breakout b. yesterday’s close is lower than the close two days ago, and needs to be lower then the previous bar’s close. This is to ensure that c. today’s open is lower than the high four days ago. most traders still have a short-term bearish outlook prior to the upside breakout. That will increase the force of the up move as the 2. Prepare to go short today if a. the true low from four days ago is lower than the low from traders are caught on the wrong side of the market and scramble to either two, three or five days ago, and get out of the market. b. yesterday’s close is higher than the close two days ago, and Finally, the breakout must take place intraday and exceed the true c. today’s open is higher than the low four days ago high of four days ago by a sufficient amount. That the trade needs to take place intraday means, for a valid upside breakout, the open- 3. Go long today with a stop order at the true high of four days ing price of the day for the breakout must be lower than the true ago, plus 0.1 percent. high of four days ago. This is to avoid entering into too strong an 4. Go short today with a stop order at the true low of four days opening, which often marks the end of the current trend. Also, by ago, minus 0.1 percent. 5. Risk 2 percent of available equity per trade. 6. Exit all trades with a loss if the market moves SAMPLE TRADES against the position by 4 percent or more. 7. Exit all trades with a profit if the market Amgen (AMGN), daily 71.00 Sell moves in favor of the position by 12 percent or LX#3 LX 69.00 more. 8. Exit all trades after five days, counting the Buy 67.00 day for the entry as day one, and no matter how late in the day the trade was made (i.e., a trade 65.00 LX#3 executed at 2:50 p.m. on Monday would be exitLX#3 Buy ed Friday the same week). 63.00
62.00
Sell Buy

Account balance ($)

400,000

60.00 58.00

Test period: November 1991 to June 2001 Test data: Daily stock prices for the 30 highest capitalized stocks in the Nasdaq 100 (excluding Intel and Microsoft, which are also part of the Dow Jones Industrial Average). $10 commission deducted per trade. Starting equity: $100,000 (nominal) Buy-and-hold stats:
DJIA: Total return – 254 percent; Max DD – 22.5

SX Buy SX#3

Buy

56.00 55.00 53.00 51.00

2 9 16 23 30 April May Source: TradeStation by TradeStation Group Inc.

7

14

21

28

4 June

11

58

www.activetradermag.com • September 2001 • ACTIVE TRADER

DRAWDOWN CURVE
11/12/91 0.00% -5.00% -10.00% -15.00% -20.00% -25.00% -30.00% -35.00% -40.00% 11/12/92 11/12/93 11/12/94 11/12/95 11/12/96 11/12/97 11/12/98 11/12/99 11/12/00

percent (current); Longest flat – 18 months (current). S&P 500: Total return – 216 percent; Max DD – 30.4 percent (current); Longest flat – 15 months (current). Nasdaq: Total return – 519 percent; Max DD – 72 percent (current); Longest flat – 15 months (current).

go with the entry strategies. We therefore arbitrarily attached a 4-percent stop-loss and a 12-percent profit-exit target, plus a time-based stop that exits all trades after five days, no matter what. (All these stops are completely un-optimized, which means a little optimization should increase performance considerably.) Also, note the system operates with no trend filter, such as a long-term moving average. Such filters, which allow only those trades that are in the direction of the underlying trend, also improve performance. Finally, note that many of the stocks traded in this example weren’t tradable until a few years ago, which explains the initial large drawdown and exceptionally long flat period. Had we been able to test the same 30 stocks throughout the entire period, it’s highly likely performance would have improved considerably.

ROLLING TIME WINDOW RETURN ANALYSIS
Cumulative Most recent: Average: Best: Worst: St. dev.: Annualized Most recent: Average: Best: Worst: St. dev: 12 months 24 months 36 months 48 months 60 months

System analysis
In DeMark’s original work, the amount by which the price had to clear the breakout level was set to the smallest price increment for the market in question. In this version, this is changed to one-tenth of a percent to make the system consistent across all markets. This means that for a stock that trades around $50, this amount is about five cents; for a stock that trades around $100, it comes out to approximately 10 cents. DeMark did not suggest any exit strategies or stop-loss levels to

4.27% 21.73% 87.23% -26.11% 29.26%
12 months

34.67% 81.32% 157.51% 59.85% 115.57% 185.15% 206.04% 294.97% 393.04% -25.61% -21.38% -15.55% 61.15% 95.27% 125.03%
24 months 36 48 months months

336.03% 254.23% 494.23% 12.15% 137.24%
60 months

STRATEGY SUMMARY
Profitability End. equity ($): 415,573 Total return (%): 316 Avg. annual ret. (%): 16.03 Profit factor: 1.13 Avg. tied cap (%): 58 Win. months (%): 53 Drawdown Max DD (%): Longest flat (m): Trade statistics No. trades: 3,529 Avg. trade ($): 158 Avg. DIT: 3.0 Avg. win/loss ($): 1,150 (1,393) Lrg. win/loss ($): 12,674 (8,131) Win. trades (%): 39.4 TIM (%): Tr./Mark./Year: Tr./Month: 97 /15.1 12.3 30.7

4.27% 16.05% 21.73% 26.43% 87.23% 74.94% -26.11% -13.75% 28.26% 26.94%

21.94% 29.18% 58.07% -7.70% 24.99%

26.68% 29.95% 49.01% -4.14% 22.48%

34.25% 28.78% 42.82% 2.32% 18.86%

LEGEND: Cumulative returns — Most recent: most recent return from start to end of the respective periods • Average: the average of all cumulative returns from start to end of the respective periods • Best: the best of all cumulative returns from start to end of the respective periods • Worst: the worst of all cumulative returns from start to end of the respective periods • St. dev: the standard deviation of all cumulative returns from start to end of the respective periods Annualized returns — The ending equity as a result of the cumulative returns, raised by 1/n, where n is the respective period in number of years

37.5 57.2

Send Active Trader your systems
If you have a trading system or idea you’d like to see tested, send it to us at the Trading System Lab. We’ll test it on a portfolio of stocks or futures (for now, maximum 30 markets, using daily data starting Jan. 1, 1990), using true portfolio analysis/optimization. Most system-testing software only allows you to test one market at a time. Our system-testing technique lets all markets share the same account and is based on the interaction within the portfolio as a whole. Start by e-mailing system logic (in TradeStation’s EasyLanguage or in an Excel spreadsheet) and a short description to editorial@activetradermag.com, and we’ll get back to you. Note: Each system must have a clearly defined stop-loss level and a suggested optimal amount to risk per trade.

LEGEND: End. equity ($) — equity at the end of test period • Total return (%) — total percentage return over test period • Avg. annual ret. (%) — average continuously compounded annual return • Profit factor — gross profit/gross loss • Avg. tied cap (%) — average percent of total available capital tied up in open positions • Win. months (%) — percentage profitable months over test period • Max DD (%) — maximum drop in equity • Longest flat — longest period, in months, spent between two equity highs • No. trades — number of trades • Avg. trade ($) — amount won or lost by the average trade • Avg. DIT— average days in trade • Avg. win/loss ($) — average wining and losing trade, respectively • Lrg. win/loss ($) — largest wining and losing trade, respectively • Win. trades (%) — percent winning trades • TIM (%) — amount of time there is at least one open position for entire portfolio, and each market, respectively • Tr./Mark./Year — trades per market per year • Tr./Month — trades per month for all markets

Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts. It is not meant to recommend or promote any trading system or approach. Traders are advised to do their own research and testing to determine the validity of a trading idea. Past performance does not guarantee future results; historical testing may not reflect a system’s behavior in real-time trading.

ACTIVE TRADER • September 2001 • www.activetradermag.com

59

EQUITY CURVE

Dynamic breakout system
Account balance ($)

$300,000

$250,000

Markets: Any market with a propensity to trend.

$200,000

$150,000

System logic: This system enters a long (short) trade if the last closing price is above (below) the $100,000 highest high (lowest low) of the lookback period. The word “dynamic” refers to the fact that $50,000 the lookback period will change based on the volatility of the market. This is similar to the system tested in the $0 3/26/93 January Trading System Lab (p. 50), when Bollinger Bands were used to trigger trades. Since the Bollinger Bands moved farther away from price as the volatility of the market increased, the higher the volatility, the more difficult it was to enter and exit trades. This month’s system functions in a related manner. The higher the volatility, the longer the lookback period will be. Because buys and sells will be based on price making a new high or low for the lookback period, the longer the lookback period, the more difficult it will be to enter or exit a trade. In this case, the lookback period can range from 20 to 60 days and will change daily depending on the level of volatility. (Volatility reading is the daily standard deviation of the closing price over the last 30 days.) You can read more about the logic of the system in the Futures System Lab (p. 70), where we have tested it on 15 different commodity futures markets. Rules: 1. Go long on the open if yesterday’s close is higher than the highest high of the lookback period. 2. Exit by reversing the position. Reverse the rules for short trades. Money management: 1. Risk 2 percent of available equity per market traded. 2. The number of shares to trade was calculated with the following formula: ST = AC * PR / Dist where ST = Shares to trade AC = Available capital PR = Percent risked Dist = Distance between the entry price and the exit price on the day of entry. Test period: April 1993 to October 2002.
May June Source: Omega Research ProSuite

3/26/94

3/26/95

3/26/96

3/26/97

3/26/98

3/26/99

3/26/00

3/26/01

3/26/02

Test data: Daily prices for 30 of the most widely traded Nasdaq 100 stocks. $10 per trade deducted for slippage and commission. Starting equity: $100,000 (nominal). Buy-and-hold stats: Total Index return DJIA S&P 500 Nasdaq 146% 100% 180%

Maximum drawdown 39% (current) 51% (current) 83% (current)

Longest flat period 33 months (current) 30 months (current) 30 months (current)

Test results: Although the system has fared no better than buy-and-hold over the life of the test period, it has fared much SAMPLE TRADES
Cisco (CSCO), daily
Buy Sell Buy 14.00 13.00 12.00 17.00 16.00 15.00

Sell

11.00

10.00

9.00

July

August

September

October

60

www.activetradermag.com • February 2003 • ACTIVE TRADER

trades: Avg. trades — number of trades • Avg.: Lrg.activetradermag. win/loss ($) — largest winning and losing trade. DIT— average days in trade • Avg.04% better than a Nasdaq 100 buy-and-hold strategy over the last 30 months (a 43 percent drawdown compared to the Nasdaq’s 83 percent).27% -38.com.96% -27.053 Win.484 (2.49% Trade statistics No.85% 10. the lookback period for the short side would be shorter than that for the long side. send it to us at the Trading System Lab. trades (%) — percent winning trades • TIM (%) — amount of time there is at least one open position for entire portfolio.69% -14.1 31. annual ret.1 Worst: St. Overall. tied cap (%): 74 Win. It is not meant to recommend or promote any trading system or approach.209) TIM (%): Tr.78% 4.24% -5.06% 24 months -19. (%): 5.3 104./Month — trades per month for all markets If you have a trading system or idea you’d like tested.: Annualized Most recent: Average: Best: 12 months 8. Start by e-mailing system logic (in TradeStation’s EasyLanguage or in an Excel spreadsheet) and a short description to editorial@activetradermag. We’ll test it on a portfolio of stocks or futures (for now.03% 186. DIT: Avg. As a result./Year — trades per market per year • Tr.18% 20.41% 36 48 months months 4.67% 43.22% 46.500 trading days).48% 60 months 26.44% 32.25% 33. DD (%): Longest flat (m): 43.54% -21. raised by 1/n. Another way to improve this system could be to use different lookback periods for long and short trades. Traders are advised to do their own research and testing to determine the validity of a trading idea.77% 48 months 63. win/loss ($): 1. Most system-testing software only allows you to test one market at a time.94% 28. and we’ll get back to you.69% 8.3 (435) 34. historical testing may not reflect a system’s behavior in real-time trading.77% -27. using true portfolio analysis/optimization. dev. win/loss ($) — average winning and losing trade.22% 36 months 15.69% 8.23% 58.93% 27.15 Avg. you should be able to increase profits considerably by experimenting with different settings for the indicators and adding a few ideas of your own. Our system-testing technique lets all markets share the same account and is based on the interaction within the portfolio as a whole. this would translate into more profitable months. equity ($) — equity at the end of test period • Total return (%) — total percentage return over test period • Avg.77% 104.: the standard deviation of all cumulative returns from start to end of the respective periods Annualized returns — The ending equity as a result of the cumulative returns. trade ($) — amount won or lost by the average trade • Avg.77% 9. It is quite difficult to succeed over the long term with a system that sells short because of the inherent upside bias of the stock market.51% 60 months 4. using the last 2.10% 9. annual ret. months (%) — percentage profitable months over test period • Max. dev.06% 12 months 8. dev.86% 7. win/loss ($): 37. trades (%): 1.40% 40.49% 1. Most likely.96% 106. and each market. ACTIVE TRADER • February 2003 • www. in months. (%) — average continuously compounded annual return • Profit factor — gross profit/gross loss • Avg.849 Total return (%): 65 Avg.53% 14. Note: Each system must have a clearly defined stop-loss level and a suggested optimal amount to risk per trade./Year: Tr./Mark. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.03% 131.28% 34.24% 148. The best way to improve the results for this system would probably be to reverse the logic for the exit so that it would be easier to exit during times of high volatility. respectively • Win.133 57 63.25% 8.36 Profit factor: 1.8 3. As with all systems tested in the Trading Systems Lab.com 61 . This would most likely result in more and smaller losing trades. spent between two equity highs • No.42% 0./Mark. and the high volatility associated with bear markets. DD (%) — maximum drop in equity • Longest flat — longest period. maximum 60 markets. equity ($): 164. trade ($): Avg. For this system. where n is the respective period in number of years Send Active Trader your systems LEGEND: End. a smoother equity curve and a higher average annual return.9 9. but also in larger winning trades because the system would exit closer to the STRATEGY SUMMARY Profitability End.79% 13.29% 7.DRAWDOWN CURVE 3/26/93 0% -5% -10% -15% -20% -25% -30% -35% -40% -45% -50% 3/26/94 3/26/95 3/26/96 3/26/97 3/26/98 3/26/99 3/26/00 3/26/01 3/26/02 maximum open profit of each trade. the high volatility makes it difficult to enter potential winning short trades and exit losing trades. months (%): 50 Drawdown Max.9 LEGEND: Cumulative returns — Most recent: most recent return from start to end of the respective periods • Average: the average of all cumulative returns from start to end of the respective periods • Best: the best of all cumulative returns from start to end of the respective periods • Worst: the worst of all cumulative returns from start to end of the respective periods • St. Past performance does not guarantee future results.03% 10. respectively • Lrg.46% 24 months -10. ROLLING TIME WINDOW RETURN ANALYSIS Cumulative Most recent: Average: Best: Worst: St.93% 27.01% 25.69% 23./Month: 100 92.45% 5. this one is totally unoptimized. respectively • Tr. tied cap (%) — average percent of total available capital tied up in open positions • Win.42% 30.

the 1980s.00 190. The more popular it became. soy oil.000 Nasdaq stocks (p. Test results: It is fair to say this system did not work very well from 1996 to 1999. $50 deducted for slippage and commission per contract traded.00 140. Exit by reversing the position. Risk 2 percent of available equity per trade. Starting equity: $100. soybeans. where you can read more about the system’s logic).00 May June July August September October Source: Omega Research ProSuite 62 www. the more other traders These markets should be able to produce profits large enough modified the system by varying the lookback periods.000 $140. Test period: April 1993 to October 2002. as it is highly likely that at following formula: SAMPLE TRADES Oats (O). 3/26/93 3/26/94 3/26/95 3/26/96 3/26/97 3/26/98 3/26/99 3/26/00 3/26/01 3/26/02 The Donchian breakout system was invented by Richard Donchian in the 1970s and refined by Richard Dennis in least some of the markets traded should be in strong trends. soy meal. 2. Applying the system to many different futures markets should result in a steady profit.00 160. particularly in the commodity futures market.00 Account balance ($) Dynamic breakout system CT = AC * PR / Dist where CT = Contracts to trade AC = Available capital PR = Percent risked Dist = Distance between the entry price and the exit price on the day of entry. Trading System Lab are likely the most commonly used strategies of all time. The idea is that capturing a strong trending move should more than make up for a large number of small Money management: 1. 60.000 above or below the highest or lowest price of the last $0 four weeks (approximately 20 days).00 Buy 120. Reverse the rules for short trades.FUTURES & OPTIONS Trading System Lab EQUITY CURVE $160. oats. However.00 130.000 $120. feeder cattle. lumber. System logic: This is the same system tested on 30 $60.000 (nominal). rules. pork bellies.00 180.000 $100.000 The system is based on the Donchian breakout system. cotton.com • February 2003 • ACTIVE TRADER . since 1999 it 110. daily 210. orange juice. coffee. applying to make up for the whipsaw losses produced in the other marother types of filters and attaching various money management kets plus enough additional profits to make trading worthwhile. losing trades produced during times of consolidation and stag2. This is a result of commodity futures markets’ historical tendency to trend.000 $80. which enters a trade as soon as the market trades $20. The dynamic breakout system is a modified version of the Donchian system that alters the lookback period between 20 Rules: 1.000 Markets: Any markets with a propensity to trend. This system and the volatility breakout system used in the January the highest high of the lookback period. Go long on the open if yesterday’s close is higher than and 60 days depending on the volatility of the market.activetradermag. Test data: Daily prices for 15 commodity futures markets: cocoa.00 150. sugar and wheat.00 170. The number of contracts to trade was calculated with the nant prices.00 Sell 200. rough rice. corn. $40.

2 1. they argued.38% -19.163 (2. and the return is close to 30 percent for the last six months./Month — trades per month for all markets If you have a trading system or idea you’d like tested. Most system-testing software only allows you to test one market at a time. in months.67% 21.91% 48 months 0. For example. The reason. months (%): 52 Drawdown Max.18% 49. win/loss ($) — average winning and losing trade.2 TIM (%): Tr.97 percent. the best way to optimize it for the futures markets is to trade it in many markets.34% -28.03 Avg.25% 2. respectively • Tr.832) Win. dev. spent between two equity highs • No. However.57% 0. (%) — average continuously compounded annual return • Profit factor — gross profit/gross loss • Avg. equity ($): 123.6 However. things often take a turn for the better.: the standard deviation of all cumulative returns from start to end of the respective periods Annualized returns — The ending equity as a result of the cumulative returns.36% 24 months 17. DIT— average days in trade • Avg.7 LEGEND: Cumulative returns — Most recent: most recent return from start to end of the respective periods • Average: the average of all cumulative returns from start to end of the respective periods • Best: the best of all cumulative returns from start to end of the respective periods • Worst: the worst of all cumulative returns from start to end of the respective periods • St.221 Total return (%): 23 Avg.75% 60 months 2. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts. annual ret.43% 6.045) Lrg. this system is potentially poised to launch itself into a new period of prosperity.71% 21. Start by e-mailing system logic (in TradeStation’s EasyLanguage or in an Excel spreadsheet) and a short description to editorial@activetradermag. Our system-testing technique lets all markets share the same account and is based on the interaction within the portfolio as a whole. the return for the past 12 months has been 9.20 Profit factor: 1. trade ($): 117 Avg. Trading a large number of markets is possible thanks to the relatively low margin requirements of the futures markets (often only 5 to10 percent of the total contract value). We’ll test it on a portfolio of stocks or futures (for now.23% 60 months 0. Note: Each system must have a clearly defined stop-loss level and a suggested optimal amount to risk per trade.1 Avg.com 63 . tied cap (%) — average percent of total available capital tied up in open positions • Win. there still is a long way to go before the system can find its way out of a drawdown that has lasted for almost seven years./Year: Tr. It is not meant to recommend or promote any trading system or approach.activetradermag. maximum 60 markets.17% 2. using true portfolio analysis/optimization./Year — trades per market per year • Tr.DRAWDOWN CURVE 3/26/93 0% -5% -10% -15% -20% -25% -30% -35% 3/26/94 3/26/95 3/26/96 3/26/97 3/26/98 3/26/99 3/26/00 3/26/01 3/26/02 has gained ground little by little.40% -5.53% 44.11% 11. DD (%) — maximum drop in equity • Longest flat — longest period. slowly increasing its average annual return. respectively • Lrg. win/loss ($): 2. tied cap (%): 40 Win. annual ret.53% 44. (%): 2. dev. dev. equity ($) — equity at the end of test period • Total return (%) — total percentage return over test period • Avg.95% -9. and each market. win/loss ($): 21. win/loss ($) — largest winning and losing trade.73% -8./Month: 100 51. historical testing may not reflect a system’s behavior in real-time trading.00% 4./Mark.64% 47. Traders are advised to do their own research and testing to determine the validity of a trading idea. STRATEGY SUMMARY Profitability End. ACTIVE TRADER • February 2003 • www. In the late 1990s.51% 3. as the performance summary shows.61% -20.62% 22.73% 10.85% 24 months 8.4 1.4 percent.93% 36 48 months months 6. DD (%): Longest flat (m): 32.500 trading days). and we’ll get back to you.89% 2.68% -24. trades (%): 32.97% 2. This shows the market works in cycles.48% 8.24% 47.56% 36 months 20./Mark. trades (%) — percent winning trades • TIM (%) — amount of time there is at least one open position for entire portfolio. raised by 1/n.30% 21. trades: 199 Avg.11% 11.85% 17. Aside from the various ways of improving this strategy suggested on p.61% -20.26% Trade statistics No. 60.08% 5.83% 0. months (%) — percentage profitable months over test period • Max.9 82.36% 12 months 9. trades — number of trades • Avg. respectively • Win. and just when you’re about to throw in the towel. ROLLING TIME WINDOW RETURN ANALYSIS Cumulative Most recent: Average: Best: Worst: St. the average return for the last three years has been 6.: Annualized Most recent: Average: Best: Worst: St. Past performance does not guarantee future results.: 12 months 9.46% 4.40% 1. where n is the respective period in number of years Send Active Trader your systems LEGEND: End. was that the markets had become so sophisticated over the last several years that whatever inefficiencies made the strategy profitable during the 1970s and 1980s had been eliminated. send it to us at the Trading System Lab.06% 0.40% 13. but the recent upward trend confirms our findings from last month that the long-term trend-following strategy is ready to stage a comeback as a profitable trading system. Granted.com.40% -10.43% 8.028 (1. many analysts claimed that long-term trend-following systems would no longer make money. using the last 2.94% 50.97% 2. trade ($) — amount won or lost by the average trade • Avg. DIT: 90.95% -25.

00 plied by 10 percent.00 840.000 400.000 500.00 truth is most likely somewhere in the middle and.00 stop level.00 the recent low used as the initial reference point.000 250.com) tiply this result by 12 (days) and. as a result. is selected. we December 1997 January 1998 Feb. This amount is then added to 560. multiplied by 13.000 800. The blue line tracks the price action more closely.000 Account balance ($) FIGURE 3 EQUITY CURVE (SIMPLE EXIT RULE) The simple 55-day low exit technique actually outperformed the modified stop. This is a long-only system. greater than the highest high of the last 55 days. 660.000 150.00 hinges on money management and diversification.wealth-lab.1 (10 percent). all components of a trading system must be ade760. 860.000 950.activetradermag.000 450.000 50.000 600.000 900. and then multiplied by the number of 580.00 Buy lowest low of the past 55 days. We will compare the results of this stop with using the lowest 2. The magenta line represents the 55-day low stop and the actual average true range for each day in the market.00 of the trend as possible by tightening the stop relative to the number of days the trade has been open.00 This system’s entry technique is simply a breakout above 720.000 250. Figure 1 shows the difference between the the lowest low of the last 55 days plus 10 percent of the two stops.000 350. 1998 would multiply the 20-period ATR by 0. Nasdaq 100 index (ND).000 550.000 500. mulSource for all figures: Wealth-Lab Inc.00 880. 650.000 0 3/22/94 5/1/95 Equity 6/3/96 7/3/97 8/3/98 9/1/99 Cash Linear Reg.000 850.00 After initiating a trade. 11/1/00 1/2/02 2/3/03 Long Short 64 www.00 days the trade has been open.000 0 3/22/94 5/1/95 6/3/96 Equity Cash 350. The next day. (www.00 trailing stop technique that is designed to capture as much 680. 1.000 50.000 300. The 780. For example.00 Experimenting with exits Market: Futures. added to the lowest 1.com • June 2004 • ACTIVE TRADER .000 200.00 System concept: Some traders believe exit signals are more Sell 820. Enter long on the next bar’s open if today’s high is low of the past 55 days. add the result to the lowest low of the past 55 days. the 20-day ATR Rules: would be multiplied by 0.000 100.000 650. 740.1.000 700.000 600.000 450. such as the 640.000 150. and so on. finally. 11/1/00 1/2/02 2/3/03 Long Short Account balance ($) 7/3/97 8/4/98 9/3/99 Linear Reg.000 400. To determine the 620.000. The more important part of the strategy is a 700.000 100. while others believe success 800.000 750.000 200. if a trade has been open for 12 days.FUTURES Trading System Lab FIGURE 1 COMPARING THE STOPS Notice how the modified exit rule approaches price at an accelerated rate as time passes.00 important than entry signals. blue line is the modified exit strategy.00 quately developed and tested.000 550. FIGURE 2 EQUITY CURVE (MODIFIED EXIT RULE) The modified exit rule produced modest profits over the test period.000 300.00 a 55-day high. a recent low point. daily 900. Exit on the next day’s open if the lowest low is less than low of the past 55 days. the 20-day average true range (ATR) is multi600.

Japanese yen (JY).74% -9.31 percent over approximately 10 years and 9.com. euro forex (FX).00% Drawdown -8.56 -28.87 6/9 — Volker Knapp of Wealth Lab PERIODIC RETURNS Avg. The test used ratio adjusted data from Pinnacle and how long the trade has been open. consec. The system produced 351 trades with an average simple stop approach suggests the modified version exited trades profit of $963. gold (GC). consec. hold time — The average holding period for all trades • Avg.204 compared to $338.19% 9 8 Monthly 0. return ratio return return profitable consec. LEGEND: Net profit — Profit at end of test period.00% -22. consec.00% Test data: The system was tested on the Active -26. hold time (losers): Max.95% 52. profitable — The largest number of consecutive profitable periods • Max. lean hogs (LH). periods profitable unprofitable Weekly 0.97 58. If you have a system you’d like to see tested.63 40. German bund (DT).activetradermag. return — The average percentage for the period • Sharpe ratio — Average return divided by standard deviation of returns (annualized) • Best return — Best return for the period • Worst return — Worst return for the period • Percentage profitable periods — The percentage of periods that were profitable • Max.7 percent).282).000 (nominal).77 per trade. DD (%) — Largest percentage decline in equity • Longest flat days — Longest period. Test period: March 1994 to August 2003. which contains the following 19 futures: DAX30 (AX).00% -10. uses the simple 55-day low exit rule. drawdown • Max. historical testing may not reflect a system’s behavior in real-time trading. Net profit ($): 338. copper (HG). with the worst year being a loss of 8.00% -2.16% 55. The single largest annual drawdown of 19. It is not meant to recommend or promote any trading system or approach. the concept behind this exit idea is worthy of experimentation and could prove to be a more effective exit strategy Profitability Trade statistics when combined with other trading methods. trades — Number of trades generated by the system • Win/Loss (%) — The percentage of trades that were profitable • Avg. win/loss — The maximum number of consecutive winning and losing trades Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts. Max. win/loss: 42. gain (winners) — The average profit for winning trades • Avg. DD (%): Longest flat days: 135.99 percent reduce drawdown. increased to $3.78% -10. The system returned markedly higher (36. please send the trading and money-management rules to editorial@activetradermag.03 7.65 percent compared to 28.88% 0. Deduct $20 slippage/commission per round-turn trade.00% -16. Of course.00% -24. Starting equity: $250. consec.88 1. hold time (winners): Avg. less commission • Exposure — The area of the equity curve exposed to long or short positions.32 1.30% 70.41% 4 3 Annually 9. 3/22/94 5/1/95 6/3/96 7/1/97 8/3/98 9/1/99 11/1/00 1/2/02 2/3/03 euro dollar (ED). unprofitable — The largest number of consecutive unprofitable periods Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc. Setting the “maximum risk percent” to 3 means at any given time we should not lose more than 3 percent of the total account equity.00% 7 1 LEGEND: Avg. gain — The average profit for all trades • Avg. loss (losers) %: Avg. consec.57% -8.70 760 Win/loss (%): Avg. hold time (winners) — The average holding time for winning trades • Avg.FIGURE 4 DRAWDOWN Money management: Risk a maximum 3 percent of total account equity per trade (“stopbased risk %”).69 26.70 2.00% -14. ACTIVE TRADER • June 2004 • www. Traders are advised to do their own research and testing to determine the validity of a trading idea. in days.073 and the average holding time increased from 40 days to 95 days.17 0. 0. profit (winners) %: Avg. coffee (KC). as opposed to cash • Profit factor — Gross profit divided by gross loss • Payoff ratio — Average profit of winning trades divided by average loss of losing trades • Recovery factor — Net profit divided by max. sugar (SB). Comparison to the occurred in 1996. although its drawdown was also percent of the total portfolio equity per trade.’s testing platform.00% -18. This version had fewer trades live cattle (LC). hold time: Avg.63% 6 4 Quarterly 2.00% -4.60% 0. The simple system’s net profit was much higher System results: Figure 2 shows the equity curve when risking 3 ($639. but it also reduced profits. hold time (losers) — The average holding time for losing trades • Max.65 40.54 percent annually. natural gas (208) than the modified system because the stop maintains a con(NG). The average profit per trade Data Corp.00% -20.282.57 22.95% -9. crude oil (CL). the system is between two equity highs • No.59 16.00% Trader Standard Futures Portfolio. trades: 351 Net profit (%): Exposure (%): Profit factor: Payoff ratio: Recovery factor: Drawdown Max.86% 0. STRATEGY SUMMARY However.21% 0.31 33.91% 56. silver (SI). -28. gain/loss (%): Avg. Nasdaq 100 (ND). S&P 500 (SP) and Tstant distance from price action regardless of volatility changes Notes 10 year (TA). loss (losers) — The average loss for losing trades • Avg.09 -4. Sharpe Best Worst Percentage Max. too quickly and did not give the system enough time to ride the Figure 3 shows the equity curve of the comparison system that trend.00% -6.com 65 . soybeans (S). a total profit of 135.00% corn (C). Past performance does not guarantee future results. The modified exit rule succeeded in reducing the system’s risk (28 percent compared to 36 percent).00% -12. overnight gaps and limit down moves could lead to higher losses.00 No.3 perBottom line: The trailing exit strategy enabled the system to cent in 1999.59 11.

For the 10 years covered in the test. the higher the number of contract. through Dec. long-only monthly breakout system on a single market.000 0 1/2/85 12/2/85 1/2/87 1/4/88 1/3/89 1/2/90 1/2/91 1/2/92 1/4/93 1/3/94 1/3/95 Total profit Buy & hold Linear reg Risk control: The system does not use a fixed dollar. would be to buy at the highest high of the past month and sell at the lowest low of the past two months. but it had a much lower risk level.com) y-month low is used to indicate the market is no longer in an uptrend.000 expect the performance to be worse than the in-sam35. Test data: Daily continuous T-Bond futures (US). 31. Whenever performing out-of-sample testing. 2 PROFIT CURVE The optimized monthly breakout system slightly underperformed buy-and-hold. System concept: Some traders believe designing a robust trading system requires dividing your data into two sets. Instead.000 10. Rules: 1. This is not a reflection of the quality of the 30. No commissions or slippage were deducted. 1. 1985. 1. we test a basic.wealth-lab. We will explore that concept here. 56 for the significance of this statistic). Monthly breakout Market: Futures. “unseen” price action). The first half (referred to as “in-sample” data) is used to develop trading rules and optimize their parameters. Second “out-of-sample” period: Jan. Test periods: Initial “in-sample” period: Jan.000 system. To illustrate the principles in a straightforward fashion. the best parameters win or lose.” which is the net profit divided by the maximum drawdown. 20. Test results — In-sample: All parameter combinations were profitable. 2002. 2.FUTURES Trading System Lab Optimization parameters: The x and y parameters will be optimized from one to eight months. Account balance ($) resents new. to Dec.com • March 2004 • ACTIVE TRADER . the optimal settings did not produce many trades (19) over the test period. we will use end-of-day data to take opening gaps into consideration. point or percentage stop.activetradermag. The process is referred to as “walk-forward testing” because the system can be progressively applied to new data to see if it continues to perform. It is simply because the system was devel25.000 5. the system is considered to have a better chance of working in real trading.” which is the absolute value of the system’s net profit divided by its maximum drawdown (see the stock Trading System Lab on p.000 ple tests. Money management: Each position will consist of one T-bond the shorter the breakout length. trades. 31. 40. the second (“out-of-sample” data) is used to simulate trading the system and see if the results are both favorable and consistent with the initial test. 1995. (www.000 15. the T-Bond. at which point all positions are liquidated. 1996. 66 www. However. Figure 1 (above) shows the top combinations sorted by “Recovery Factor. a reversal below the Source for all figures: Wealth-Lab Inc. Entry: Buy at the highest high of the last x months. Exit: Sell at the lowest low of the last y months.000 oped and optimized on a different data set. As expected. FIGURE Even though the monthly highs and lows define the entry points. If the system performs well on the second data set (which repFIGURE 1 OPTIMIZATION RESULTS FROM IN-SAMPLE TEST The results for different parameter combinations are sorted by “recovery factor.

If you have a system you’d like to see tested.428). please send the trading and money-management rules to editorial@activetradermag. The best parameter combination from the in-sample test (buy above the one-month high and sell below the two-month low) became the third worst combination in the out-of-sample test.886) than the system’s drawdown ($6. Figure 4 (above) shows the equity curve for the optimized parameters from the in-sample data period tested on the outof-sample period.00 99.006) over the 10-year — Dion Kurczek and Volker Knapp of Wealth-Lab Inc.50 99.com. all the 15.’s testing platform.50 105. it is more important to look for parameter stability. suggesting the basic trading approach is sound. A -5.00 95.00 104. ACTIVE TRADER • March 2004 • www.00 96. in-sample period was higher than the system’s profit ($40. the most part follows the underlying market curve (blue line). T-bonds (US).00 100. The optimized parameters from the in-sample test performed much worse on the out-of-sample data.50 101.50 Sell Buy Volume August 2001 September 2001 October 2001 November 2001 Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.00 103. a sample trade of which is shown in Figure 5 (right).com 67 .000 parameter combinations were profitable in both peri10. Because of this.00 101.50 100.000 ods. the buy-and-hold drawdown was more than 50 percent higher FIGURE 5 SAMPLE TRADE ($10.000 broad range of parameter values should be profitable 1/2/96 7/29/96 3/31/97 12/1/97 7/28/98 3/30/99 12/1/99 7/27/00 3/28/01 12/3/01 8/1/02 and deliver consistent results. It is not meant to recommend or promote any trading system or approach.000 parameter set of the first 10 years became the third worst in the following seven years. but it should be used to confirm parameter stability rather than to try to find the parameter combination with the highest net The equity curve (dark green) in Figure 2 (opposite page) for profit. Even though the buy-and-hold profit ($44. However. The best parameter combination from the in-sample test was the third worst in the out-of-sample test. FIGURE 4 PROFIT CURVE Bottom line: The testing process illustrated here shows how the performance of even fundamentally sound trading systems can vary over time. which indicates the system rules have some merit.00 97. Two combinations from the former top 11 are now in the bottom eleven. Even though the performance was positive.000 The best-performing parameters in one period will 0 almost never be the best in future data. Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc. all the parameter combinations have remained profitable. it was quite volatile compared to the smooth ride of the optimal parameter combination (buying above the one-month high and selling below the eight-month low) for this test period. Traders are advised to do their own research and testing to determine the validity of a trading idea.50 98. Account balance ($) Test results — Out-of-sample data: Testing the parameters on the out-of-sample data set produced significantly different results.50 96.00 105. Past performance does not guarantee future results.012).50 102.FIGURE 3 OUT-OF-SAMPLE TEST RESULTS The system behaved much differently on the out-of-sample data. 5.50 97. The system performed best in the out-of-sample test when it bought above the one-month high and sold below the eight-month low.00 98.50 103. On the other hand.50 104.activetradermag. Total profit Buy & hold Linear reg Optimizing is a very useful tool.00 102. historical testing may not reflect a system’s behavior in real-time trading. daily 106. Figure 3 (top) sorts the 11 worst parameter combinations by recovery factor. The best 20.

FIGURE 1 EQUITY CURVE We tested the portfolio of 18 stocks on 30-minute bars (more than 3.000 140. All open positions are exited at the close of the day. International Paper (IP). AT&T (T) and Wal-Mart (WMT).576 in commission charges.com. if the closing price of the current 30-minute bar is above the high of the range. System concept: This intraday system takes a trade when price breaks out of the trading range established in the first hour of the session. Hewlett Packard (HPQ). this is one of the more active systems tested here. Test data: The system was tested on 30 minute bars of the Active Trader Standard Stock Portfolio. 0. Long trades were slightly profitable.00 -20.546 trades in the test period. There will be only one trade per day. General Motors (GM).000 40.60% 0.000 120.com • January 2004 • ACTIVE TRADER .000 (nominal).20% -80.40% 0.30% 0. A short position is established if the closing price is below the range. a breakout of the early trading range is sometimes caused by a specific news item that will cause the trend to continue.000 60.60-minute breakout system Markets: Stocks.000 100. and a move above or below the 60-minute range might signal a trend in that direction. Early in the morning the market often is trying to establish its direction.000 0 10/10/02 Equity 11/25/02 Cash 1/14/03 2/28/03 Linear reg 4/14/03 Long 5/29/03 7/14/03 Short 8/26/03 10/8/03 Buy & hold Entry rules: Long trades: Buy if the closing price of the third 30-minute bar is above the high of the first 60 minutes of the day.000 10. generating a 4. International Business Machines (IBM).00 -40. there are no trades in the first 60 minutes. Starting equity: $100. A signal in the opposite direction is used to exit the current position. After the first 60 minutes have ended.00 0. Test period: October 2002 through October 2003. Cisco (CSCO).000 50. Also. However. Initial test results: The system’s performance was disappointing — an overall loss of -3. which contains the following 18 stocks: Apple Computer (AAPL).50% 0.01 per share slippage and commissions. Starbucks (SBUX). Increasing the FIGURE 2 ADDING A FILTER percentage would require dropping trades that The bars show the average per-trade profit that would have been exceeded the available cash limit. Intel (INTC).80% 0.000 70. JPMorgan Chase (JPM). Short trades: Sell short if the closing price of the third 30minute bar is below the low of the first 60 minutes of the day.70% 0.00 40.22-percent profit during this period. Boeing (BA).qcharts.00 68 www. even setting the commission at one cent per share (advisable for a system that generates this many trades) still resulted in $13. we will go long on the next open. Deduct $0.000 80. Coke (KO). Microsoft (MSFT). Data from www. This will allow all trades the system is generating to be executed.000 90.500 total trades).000 20. Exit: Exit all positions on signals in the opposite direction or at the end of the day.00 80. With 3.00 60. Figure 1 shows long trades were slightly profitable. Citibank (C). Caterpillar (CAT).000 30. Sears (S). Each signal used five percent of the portfolio equity value.000 110. Disney (DIS).25 percent.000 130.activetradermag. captured by adding the CMO as a trade filter. 160.000 150.00 CMOlevel 20.00 -60. Because the highest volatility of the day often occurs in the first trading hour. Money management: Each trade is sized at five percent of the current account equity.

17% 38. consec.22 18/23 Weekly Monthly — Volker Knapp of Wealth-Lab Inc. for more information on this indicator).46 0. Subsequent black bars have closing prices within the first-hour range.g. profit (winners) (%): Avg. hold time (losers) — The average holding time for losing trades • Max. indicating bullish momentum.activetradermag.247. consec. return ratio return return profitable consec. The filter consisted of taking trades only when the CMO reading for the previous day’s bar was above 60.00 20. hold time (winners): Avg. This illustrates how testing a strategy on a single instrument or very limited portfolio can lead to the wrong conclusions about a system’s value. the total portfolio itself did not.71 -0.90 Short 240 @20.56 -0.87 Cover 240 @20.98 0. as opposed to cash • Profit factor — Gross profit divided by gross loss • Payoff ratio — Average profit of winning trades divided by average loss of losing trades • Recovery factor — Net profit divided by max. though — such as trying other filters.11% -2. Another alternative is to use shorter bars (e.80 20. profitable — The largest number of consecutive profitable periods • Max.com 69 .30 20. Figure 2 (opposite page) shows waiting for a CMO reading above 60 before taking trades would have increased profitability.Adding a filter: To see if we could reduce the number of trades and improve performance. hold time (losers): 3. periods profitable unprofitable -0. gain/loss (%): Avg. it did not produce satisfactory results.34 9/24/03 9/25/03 Buy 246 @20. There is plenty of room for further experimentation.90 20. p.57% -4. Cisco Systems (CSCO).’s testing platform.45 Avg.00 19. consec.60 1. The number of trades decreased to 251 from 3. five minutes) to get into positions faster. historical testing may not reflect a system’s behavior in real-time trading.70 20.96 0.85 6.60 20. DD (%): Longest flat days: -3. please send the trading and money-management rules to editorial@activetradermag. the system is between two equity highs • No. 30-minute 21.99% -2. hold time — The average holding peri od for all trades • Avg. It is not meant to recommend or promote any trading system or approach.48% 1. hold time: Avg. Past performance does not guarantee future results. Bottom line: Although the strategy concept sounds reasonable. drawdown • Max. we added a filter — a 14-bar Chande Momentum Oscillator (CMO) to the basic system (see the December 2003 Trading System Lab. FIGURE 3 SAMPLE TRADE The green bars are the first trading hours of the days shown here. unprofitable — The largest number of consecutive unprofitable peri ods Trading System Lab strategies are tested on a portfolio basis (unless otherwise noted) using Wealth-Lab Inc. in days.51 LEGEND: Avg..87 6. DD (%) — Largest percentage decline in equity • Longest flat days — Longest period.00% 4 2 1 6 5 1 -7.47% 47. including those that would differentiate between long and short trades. the average profit was 0.70% 2. consec.20 21. 48.40 20.06% -0. trades — Number of trades generated by the sys tem • Win/Loss (%) — the percentage of trades that were profitable • Avg. Max.26 0. hold time (winners) — The average holding time for winning trades • Avg.48 percent.55 -0.34 STRATEGY SUMMARY Profitability Net profit ($): Net profit (%): Exposure (%): Profit factor: Payoff ratio: Recovery factor: Drawdown Max. 3.20 20. ACTIVE TRADER • January 2004 • www.12 percent and total profit was 1. red bars indicate the closing price is below the range and blue bars show bars with closing prices above the range.50 20.41 -3.46% 60.10 20.com.40 21. trades: Win/loss (%): Avg. loss (losers) — The average loss for losing trades • Avg.24% -0. Traders are advised to do their own research and testing to determine the validity of a trading idea. return — The average percentage for the period • Sharpe ratio — Average return divided by standard deviation of returns (annualized) • Best return — Best return for the period • Worst return — Worst return for the period • Percentage profitable periods — The percentage of periods that were profitable • Max.02 6. loss (losers) (%): Although some individual stocks produced very good equity curves. profit (winners) — The average profit for winning trades • Avg.63% -0. PERIODIC RETURNS Avg. win/loss — The maximum number of consecutive winning and losing trades Quarterly -0.10 21. less commission • Exposure — The area of the equity curve exposed to long or short positions. profit — The average profit for all trades • Avg. Disclaimer: The Trading System Lab is intended for educational purposes only to provide a perspective on different market concepts.078 Max. Sharpe Best Worst Percentage Max.546 49.30 21. If you have a system you’d like to see tested. win/loss: LEGEND: Net profit — Profit at end of test period. The filter reduced the total number of trades and eliminated short trades.546.42 Trade statistics No. Testing our portfolio with the additional filter turned the losing system into a winner — although not a spectacular one.25 -3.50 21. consec.

Sign up to vote on this title
UsefulNot useful