Professional Documents
Culture Documents
TG - MR Swing NAAIM 2010 PDF
TG - MR Swing NAAIM 2010 PDF
Aquantitativesystemformeanreversionand
swingtradinginmarketregimes
March2010
DavidAbrams ScottWalker
QuantitativeSystemDesigner PortfolioManager
www.daveab.com www.walkertradingpartners.com
dave@daveab.com swalker@walkertradingpartners.com
1. Abstract
MR Swing is a quantitative system that employs daily meanreversion and swing trading in different
marketregimestoproduceenhancedabsoluteandriskadjustedreturns(e.g.inoneconfiguration23%
CAGR, 54% risk adjusted CAGR, 1.37 Sharpe, 13% max drawdown). The system uses four key design
androbustnesstoregimewhipsaws.AnextensiveanalysisofoutofsampleETFsandmanagedfutures
demonstratestherobustperformanceofthesystemovertenyears.Finally,MRSwingisincorporated
asacomponentinadiversifiedportfolioofETFs(modeledafteruniversityendowments)andisshownto
significantlyincreasetheportfoliosreturnwhilereducingthemaximumdrawdown.
This paper describes a quantitative method for trading equity indexes (ETFs and futures) that can
increasereturnandreducevolatilitywhenusedinadiversifiedportfolio.Thesystemisdesignedtobe
adaptivetochangesinmarketvolatilityovertheshortterm,andalsodeploystechniquesalignedtothe
dominant market regime. After recapping the robust performance metrics across various ETFs and
futures,wedemonstratehowMRSwingimprovestheperformanceandriskprofileasacomponentina
portfoliowithassetclassallocationsinspiredbylargeuniversityendowments:TheIvyPortfolio(Faber
andRichardson2009).
TheMRSwingsystemisbasedonfourcoreprinciples:(1)Marketregimeswitching,(2)Nonsymmetrical
trading algorithms, (3) Volatility adaptive metrics, and (4) Robustness to regime whipsaws. The first
quantifythecurrentmarketenvironmentorregime,andthenadoptatradingapproachbestsuitedtoit.
An example of regime switching (Kestner 2003) uses a trendfollowing method (the 40day/20day
channelbreakout)intrendingmarkets,andcountertrendtrading(14dayrelativestrengthindex(RSI))
in nontrending markets. The average directional index (ADX) is used to define the market regime as
trendingornontrending.
Table1:ExampleofRegimeSwitching(Kestner2003)
ADX<20 trendtobeginsoon 40day/20daychannelbreakout
In MR Swing we simply quantify two regimes: bull and bear. One could have more complex regimes
based on price volatility, option implied volatility, and other factors. We use a shortterm mean
reversionmethodinthebearregime,andswingtradingtechniqueinthebullregime.
Thesecondprincipleisnonsymmetricaltradingalgorithms.Mosttradingstrategiessimplyreversethe
rules for bull and bear. For example, in Tactical Equity Allocation Model (T.E.A.M.) the system (Lent
2009) buys shortterm meanreversion on a daily timefame, when the weekly percent trend channel
hashadanupsidebreakout.Thesystemreversestheserulestoshortarallyinthedailytimeframeafter
aweeklychannelbreakdown.Mostcountertrendsystemspickoneoverbought/oversoldoscillatorand
then symmetrically buy oversold and short overbought. These common techniques assume that
markets behave in a symmetrical manner in different market regimes. We show that entirely unique
techniques are more responsive to regime conditions rather than using symmetrical techniques. The
swing trading component uses different algorithms for entries and exits. In addition, we employ
differentmethodsbasedontheregimetomakeanonsymmetricalalgorithm.
Ourthirdprincipleofvolatilityadaptivemetricsrequiresthateachindividualcomponentofthesystem
mustbecapableofrobustlyhandlingchangesinmarketvolatility.Forexample,whenwequantifythe
dominant market regime using a trendfollowing method, we employ a channel based on price to
reducewhipsawsandgetthebenefitsofhysteresis.Next,everymetricweuseformeanreversionand
swing trading entries/exits was carefully chosen to include volatility in the algorithm. In the mean
reversion component, the overbought/oversold levels are calculated using an adaptive technique that
normalizesthedatausingnonparametricstatistics(i.e.wedonotassumethatpricesexhibitaGaussian
distribution).
The final principle is robustness to regime whipsaws. Regardless of the method we use to define the
current market regime, false alarms or whipsaws will occur. Classic moving average trendfollowing
systemsareverypronetolossesinasidewaysmarket.Insteadoftryingtoeliminatewhipsawsinthe
regimemodel,weinsteadstructureeachcomponenttobeabletowithstandwhipsaws.Thatmeans,an
entry in the swing trading regime must be able to handle a change to the meanreversion regime
without causing the system to become unstable. MR Swing is designed to address these cases and
ensurerobustnesstochangesinregime,includingwhipsaws.
2. MeanReversioninEquityIndexes
Themajorequitymarketindicesexhibiteddailyfollowthroughlastcentury,buthavebeenverymuchin
ameanrevertingenvironmentsinceabout2000.Stokesshowed(Stokes06/09/2009and08/10/2009)
howdailyfollowthroughworkedwell1950to2000,butthenitreversed.Wesuggestthatthischange
tomeanreversionisduetotheincreasedspeedatwhichnewsanddataispricedintothemarket.In
particular,computerizedtrading,discountbrokerswithlowcostcommissions,andtheworldwideweb
startedtoreachcriticalmassaround2000.Thesefactorsmayhavestructurallychangedhowtheindices
behaveandarereflectedinoursystemdesign.
In the study below, we compare daily followthrough to daily meanreversion in the S&P 500 equity
index.Therulesforfollowthrougharesimple:iftodayscloseishigherthanyesterdays,thengolong.If
itislowerthanyesterdaythenreverseandgoshort.Dailymeanreversionistheopposite.
Algorithm1:DailyFollowThroughRules Algorithm2:DailyMeanReversion(MR)Rules
if(C>C1)thenbuy if(C>C1)thensellshort
if(C<C1)thensellshort if(C<C1)thenbuy
Theresultsofthesetwosystemsoverthelasttenyearsareshownbelow.
Table2:ComparingDailyFollowThroughtoDailyMeanReversion
$100,000portfoliofrom10/01/2000to02/01/2010
System CAGR Sharpe Portfolio Max Max Profit Percent MaxDD
Value Wins Losers Factor Profitable
DailyFollow 16.20% 1.01 $19,097 7 9 0.82 33.67% 87.30%
Through
DailyMean 15.42% 0.78 $383,899 9 7 1.25 66.17% 26.13%
Reversion
Asyoucansee,dailymeanreversionhasbeenarespectablesystemoverthelasttenyearswith15.42%
compound annual growth rate (CAGR) and 66% of trades profitable with a reasonable drawdown of
26%.Thechartbelowcomparesthegrowthofa$100,000portfoliotradingthetwosystemsbetween
2000and2010.
Figure1:DailyMeanReversionvs.DailyFollowThrough
$500,000
$450,000
$400,000
$350,000
$300,000
$250,000
$200,000
$150,000
$100,000
$50,000
$0
DailyFollowThrough DailyMeanReversion
We will delve deeper into meanreversion and identify ways to improve on the base system. The
benchmarkforoursystemisthedailymeanreversionsystembecauseofitssimplicityandperformance
metrics.Thesystemthatwedesignmustimproveonthesimpledailymeanreversionshownabovein
ordertojustifytheaddedcomplexity.
2.1. MeanReversioninMarketRegimes
Wedefinemarketregimesbasedona200daysimplemovingaverage.Wechoosea200daymoving
averagebasedonresearchdoneby(Siegel1998)thatshowedoverthelastcenturya200daymoving
averagefilterwouldhavereducedvolatilityinalongtermstockportfolio.
where N = 200 ,1 , 1 (Algorithm 3)
1 1
Belowshowstheresultsofmeanreversioninbullandbearmarketregimes.
Table3:DailyMeanReversion(MR)inMarketRegimes
$100,000portfoliofrom01/05/1999and02/12/2010
System CAGR CAGR Sharpe Portfolio Timein Max Max Profit Percent MaxDD
riskadj Value Market Wins Losers Factor Profitable
MRin 0.69% 1.30% 0.09 $108,346 53% 11 7 1.03 65.36% 21.62%
Bull
MRin 11.89% 28.80% 0.72 $348,498 43% 12 5 1.40 66.02% 26.14%
Bear
Daily meanreversion is a more dominant force during a bear market (11.89% CAGR, 0.72 Sharpe)
comparedtobullmarkets.Meanreversionthrivesontheheightenedvolatilityofbearmarkets.
Figure2:MeanReversioninMarketRegimes
$400,000.00
$350,000.00
$300,000.00
$250,000.00
$200,000.00
$150,000.00
$100,000.00
$50,000.00
$0.00
MRinBullRegime MRinBearRegime
TheMRSwingsystemwillbedesignedtotakeadvantageoftheoutperformanceofmeanreversionin
thebearregime,andwewilluseadifferentstrategyinbullregimes.
3. QuantitativeTradingSystem
Whileweuseddataanalysisintheprevioussectionstounderstandmarketstructureforoursystem,itis
important to avoid curve fitting when designing a quantitative trading system (Bryant 2006). In
particular,westartwiththeseguidelines:
Default parameter settings: all indicators and system components should use default values,
insteadofoptimizingthemforthehighestnetprofit.
Avoidoptimization:optimizationwillonlybeusedtotestifasystemcomponentaddsvalue,not
tochoosespecificsettings.
LongTimeframe:wechoosethelasttenyearsasourtimeframe.Thisgivesusenoughdatafora
robustanalysis,andalsokeepsustradingthecurrentmarket,andnotlastcenturysmarket.
ManyTrades:trendfollowingsystemthatonlytradeafewtimesperyearmayhavedrastically
differentresultsifyoumissasmallsampleoftrades.Wepreferasystemthatgeneratesmany
trades (e.g. MR Swing generates ~400 trades in the current analysis) because this adds to our
confidencethatsystemresultsarenotoverlyinfluencedbyasmallsetoftrades.
OutofSampleTesting:weoriginallydesignedthesystemfortheSPY,andwillshowtheoutof
sampleresultsontheQQQQ,EEM,EWM,VTIaswellasthefutures@ESand@NQ.
characteristicsofmarketsbyusingamarketregimeswitchingmethodtotakeadvantageofshortterm
meanreversioninthebearregime,anddeployswingtradinginthebullregime.Next,wewillemploy
(2) non symmetrical trading algorithms for entries, exits and the regime specific trading algorithms.
Everysystemcomponentmustbebasedon(3)volatilityadaptivemetricssothatitcanhandlechanges
involatilityoveralongtimespan.Finally,werecognizethatnoregimeswitchingmodelwillbeableto
eliminate all false signals and each core system component must exhibit (4) robustness to regime
whipsaws.
3.1.QuantifyingMarketRegimes
There are many ways to define the market regime. A complete study of different market regimes is
beyond the scope of this paper. MR Swing identifies market regimes based on a 200day moving
averagechannelofthehighsandlows.Whenthesystemisinthebullregime,itonlyallowsthechannel
toincrease,andviceversainbear.Thereasonwechooseachanneloverjustamovingaverageistoget
thebenefitsofhysteresis(Alves2009)andreducewhipsawsbyusingarangeofvaluesbeforeswitching
frombulltobearregime.
Hysteresis(Hysteresis2010)isanaturalphenomenonthatappearsinmagnetism,elastics,cellmitosis
and control theory (e.g. thermostats). These systems exhibit path dependence in which the current
statedependsonthepathtakentoachieveit.Thesystemhasmemoryandtheeffectsofthecurrent
inputareonlyfeltafteradelayorrangethresholdisexceeded.Webelievethatmarketsalsoneedtime
torespondtonewinformationand thatresponsedoestakeintoconsiderationrecentmarkethistory.
Our market regime filter requires price to close below the low of the channel before switching into a
bearregime.
Algorithm4:MovingAverageTrendChannel
1, MAX , , ,
, 1, 1
default N= 200
Simple moving average channel of the bar highs and lows. A bar close below the trailing channel
switches it to a down trend ( = 1). In an uptrend ( = 1) we only allow to increase, such that in
sidewaysmarketorretracementdoesnotchange.
ThemarketregimemodelshownbelowinFigure3isfrom2005to2010onadailychartoftheSPY.
Figure3:MarketRegimeTrendChannelonSPYfrom2005to2010
200daySMAchannelofbarhighsandlows(incrementonlyindirectionofthetrend).
Thistechniquecapturesthemajormarkettrendswithfewerwhipsawsthanusingthetraditional200
daySMA.Insteadofattemptingtodesignalongtermmarketregimefilterthatremovesallwhipsaws,
werequireeachcomponenttoexhibit(4)robustnesstoregimewhipsawsinthecoretradingalgorithm.
Akeydesigninthealgorithmisnottoexitapositionjustbecausethedominantmarketregimechanged.
Instead,weholdontotheposition,andsimplychangetheentry/exitrulesbasedonthenewregime.
Themeanreversionandtheswingcomponentsmustbeabletohandlethisbehaviorwithoutaserious
drawdown. This is one reason the system structure is designed for major equity market indexes as
opposedtoindividual(especiallyilliquid)equities,whichmayhaveapriceshockduetoastockspecific
news announcement. MR Swings patience in waiting for a high probability exit to the trade after a
regimechangehasbeenextensivelytestedandshowntogreatlyimproveintheequitycurve.
3.2.BearRegime:ShorttermMeanReversion(MR)
Wecanimproveonthedailymeanreversionsysteminsection2byemployinganadaptiveshortterm
meanreversionalgorithm.Somequantitativesystemshaveusedatwoperiodrelativestrengthindex
RSI(2)fordailymeanreversion.AlthoughthiswouldworkfineinMRSwing,wechoosetheDVO(David
VaradiOscillator,2009)becauseitmeetsourprincipleofa(3)volatilityadaptivemetric.ThePercent
Rankfunctionisusedtonormalizethedailydata,andthisnonparametrictechniquerobustlyadaptsto
marketvolatilitychanges.
Algorithm5:GeneralizedDVOandDV2
, ,
1 1
, , , , , ,
5 252
0.5 0.5 0 0 , 0.5 0.5 0 0 0 , 5, 252
TheDavidVaradiOscillator(DVO)wasdesigned(Varadi07/29/2009)tobeashorttermoscillatorusingbarhighs
andlows(maximumsmoothingperiodN=5).TheDV2isonespecificsettingoriginallydesignedfortheSPY(the
weightingperiodwas50/50overthelasttwodays).TheDVOcanbeusedtocreateuniqueweightingschemes
thatfunctionbestforeachclassofsecurityanddefinealocalweightdensity( fortheETForfuturescontract.
Itcapturesdifferentcyclelengths,amplitudes,andreturndistributionsofthesecurity.
TheDVOgivesustheflexibilitytocustomizethesystemtothedifferentcyclelengths,amplitudesand
return distribution of the particular security. Although we use the default values of the DV2 shown
above,MRSwingsperformancecouldbeimprovedusingthisreadilyavailablefeatureoftheDVO.The
chart below shows the DVO in the meanreversion component of MR Swing between 02/2008 and
08/2008.ThesystembuysthenextdayopenwhenDVOislessthan40%andshortsthenextdayopen
whenDVOisabove70%.Themagentalineaboveshowsthemarkettrendchannel.
10
Figure4:MeanReversion(MR)TradesinBearRegime
UsingDVOonSPYfrom02/2008to08/2008
MRSwingactualtradesontheSPYin2008duringthebearregimetodomeanreversionusingtheDVOwithdefaultsettingsof
theDV2.
3.3.BullRegime:SwingTrading
Therearemanymethodsforswingtradinginbullmarkets.Onesystematicwayofdiscretionaryswing
tradingusingmultipletimeframesistheTripleScreensystem(Elder1993).First,thelongtermtrendis
determined by requiring that the security must show positive momentum on the weekly timeframe
(moving average convergence divergence (MACD) histogram for this week is higher than last week).
Next,themarketshorttermtrendonadailychartisusedtoidentifyapullbackusinganoscillator(e.g.
EldersForceIndex).Thethirdandfinalscreenistoputinatrailingbuystopatthehighsofthebars
untilfilled.Theinitialstoplossisatthelowofthepreviousbar.
11
WetakeourinspirationforMRSwingsquantitativealgorithmfromtheTripleScreensystem.First,the
dominantmarketregimemustbeabullswingtradingenvironment(Section3.1).Second,welookfora
pullback entry (Section 3.3.1). Third, we patiently use a limit order entry technique (Section 3.3.2).
Exitsemployanintermediatetermexhaustionindicator(Section3.3.3).
Figure5:SwingTradesinBullRegime
SPYfrom06/2009to02/2010
ValueChartoversoldisusedforentries.SVAPOoverboughtsignalsexhaustionpointsfortheexits.Noticetheregimechanged
from bear to bull in June 2009, switching from shortterm MR to swing trading. Also the system handled the false alarm
whipsawsinJune/July.
3.3.1.SwingEntryinBullRegime
Wehavefoundthroughexperiencethatbullmarketstendtohavequick,sharppullbacksandthenslow
grindingmoveshigher.Thisobservationleadsustochoosinganonsymmetricaltradingalgorithmfor
swingtrading.Entriesuseavolatilityadaptivemetricthatissuitedtofindingfastpullbacks,andexits
use an intermediateterm exhaustion method that includes price and volume. We choose the Value
12
Chart(HelwegandStendahl,2002)forentriesbecauseitdynamicallyidentifiesoverboughtandoversold
levelsevenwhenpricevolatilitychangesovertime.Firstarelativepricechartiscreated,similartoade
trendedpriceoscillator.Then,itisadjustedbyadynamicvolatilityunit.
Algorithm6:ValueChart
| | , | |,
2
0.16, 5
Relative price chart that adapts to changes in market volatility. The value chart of the close is the
distance between the close and the floating axis of relative price value (mean of the median price)
dividedbythedynamicvolatilityunit(meanofthebarrange).
WeusethedefaultN=5barperiodfortheValueChart.Anoversoldlevel(7.5)isbasedonthelowsof
thebars,soweareonlyusing inthealgorithmtosignalapotentialpullbackentrypoint.
3.3.2.LimitOrderEntryTechnique
MRSwingusesextremepatienceinentriesandexits.Limitordersareusedinswingtradingtoattempt
togetanentryclosetothelowofarecentbar.
Algorithm7:LimitOrderEntries
, ,
Eachbarthelimitordertargetpriceisrecalculatedtodynamicallyadaptto
recentpricechangesandsearchforanentry.MRSwingpatientlywaitsfor
thetargetpriceofthelowestlowofthelasttwobars.
13
Sometimesanentryismissedbecauseofthistechnique,butMRSwingislookingforlowriskentriesat
theexpenseofpotentiallymissingatrade.
3.3.3.SwingExitinBullRegime
Asaquantitativesystem,MRSwingneedsaprecisedefinitionofmarketexhaustionwhenswingtrading
thebullregime.Wechooseavolumeandpriceoscillator(Vervoort2007)thatmeasuresexhaustionby
combiningtheshorttermpricetrendwiththeslopeofthevolumeovertime.
Algorithm8:VolumeandPriceOscillator(SVAPO)
SVAPO(Period=8, Cutoff=1, StDevH=1.5, StDevL=1.3, StDevPeriod=100)
//basic volume trend is triple exponential moving average of linear reg slope of volume
VTrend = TEMA(LINEARREGSLOPE(VolumeValue, Period),Period);
//Day Trend=UP if smoothed Heiken-Ashi and volume trend higher than previous bar
Up = haC>haC[1] * (1 +(0.001*Cutoff)) and COUNTIF(VTrend >= VTrend[1],2)=2;
Dn = haC<haC[1] * (1 -(0.001*Cutoff)) and COUNTIF(VTrend <= VTrend[1],2)=2;
Using the default settings, the SVAPO tends to find excellent exhaustion points in the bull regime by
combining shortterm price trend and the slope of the volume trend. Note that we do not wait for
confirmationofanexhaustionpointlikethetraditionalSVAPOsystem,insteadwestartlookingforan
exitusingourlimitordertechnique(Section3.3.4)whenSVAPOcrossesabovethetopband.
14
3.3.4.LimitOrderExitTechnique
SimilartoentriesMRSwingusesextremepatiencelookingforahighprobabilityexit.Limitordersare
usedinswingtradingtoactivelysearchforanexitclosetothehighofarecentbar.
Algorithm9:LimitOrderExits
, ,
Eachbarthelimitordertargetpriceisrecalculatedtodynamicallyadaptto
recentpricechangesandsearchforanexit.MRSwingpatientlywaitsfor
thetargetpriceofthehighesthighofthelasttwobars.
MRSwingkeepssearchingforalimitorderexitoneachbar.Wealsotestedapplyingatrailingstopset
at the simple moving average of the lows of the last 2 bars. This combined with the limit order exit,
providesabalancedapproachtoexitsthatmosttimeshitthelimitorder,andusesthetrailingstopto
protectprofits.
4. CompositeStrategyandPerformanceResults
Nextwewillcombineeachcomponentofthealgorithmintoanendtoendautomatedtradingstrategy.
Thecomponentswillbetestedinthreedifferentvariations,shownbelow:
Table4:Strategy Rules
Takeallcountertrendtradesinallregimes.Alwaysinthemarket
(longorshort)unlessstoppedoutforthattrade.
MRSwing(counter)
Aggressive BearRegime:golongandshortMRrules
BullRegime:golongandshortswingtraderules
TakeallcountertrendtradesinbearregimeusingMR.Swingtrade
longonlyinbullregime,donotshortabullmarket.
MRSwing(swinglong,MRcounter)
BearRegime:golongandshortMRrules
BullRegime:swingtradeloneonly(nevershort)
Conservativesystemthatonlytradesinthedirectionofthedominant
marketregime.Nevercountertrendtrade.
MRSwing(nocounter)
Conservative BearRegime:sellshortusingMRrules(neverlong)
BullRegime:golongswingtrades(nevershort)
15
Wecompareanaggressivesystemthattakesallcountertrendtradesinbothbullandbearregimes.A
countertrendtradeintheswingtradingsystemissimplytosellshortinsteadofexitingwhenthelimit
orderpriceisfilled(Section3.3.4).ThenexttableshowstheperformanceofMRSwingontheSPY,an
ETFthatfollowstheStandardandPoors500IndexoflargecapitalizationUSstocks.
Table5:MRSwingonSPY
$100,000portfoliofrom02/01/2000to02/12/2010
System CAGR CAGR Sharpe Portfolio Timein Max Max Profit Percent Max
riskadj Value Market Wins Losers Factor Profitable DD
MRSwing 37.90% 38.34% 1.58 $2,484,277 98.81% 17 6 2.81 69.47% 23.65%
(counter)
MRSwing 35.16% 51.25% 1.54 $2,068,616 72.79% 17 6 2.75 69.43% 21.68%
(swinglong,
MRcounter)
MRSwing 23.07% 54.06% 1.37 $801,735 48.01% 15 3 3.25 74.23% 13.09%
(nocounter)
Thefullcountertrendsystemisinthemarket98.8%ofthetime(usinga20%pertradestoplosscauses
it to be out of the market 1.2% of the time). It shows a solid 37.9% compound annual growth rate
(CAGR) and Sharpe of 1.58 with a max daily equity drawdown of 23.65%. Note that the drawdown
figuresaremarkedtomarketeverytradingday,sothatyoucanseethetruedrawdown(asopposedto
monthly figures or a tradebytrade figure). The system reinvests the entire portfolio on every trade.
Commissionsandtaxesareexcluded.Slippageisexcluded,butwastestedasinsignificantduetoentries
andexitswithlimitorders.
TheconservativeversionofMRSwingtakesnocountertrendtrades.Itisonlyinthemarket48%ofthe
time, and it still delivers 23% CAGR and the max drawdown is greatly reduced to 13%. Taking trades
onlywiththetrendalsoreducesthenumberofconsecutivelosingtradestojustthreeoverthelastten
years. This is a key pragmatic issue for any manager or trader following an automated system, since
thereisawellknowntendencyforinvestorstowithdrawfundsandtraderstoturnoffthesystemifthe
16
number of consecutive losses leads to a lack confidence in the system, often just before the system
recovers (How many consecutive losing trades will you suffer before turning off the system?). We
believethatthisisoneofthekeystrengthstoemployingMRSwinginlivetrading.
The next chart shows the performance of MR Swing, markedtomarket daily, over the last ten years.
Theconservativesystemshowsmuchlowerdrawdown,butbetween2004and2006performancewas
flat. The system compounds portfolio equity by reinvesting profits, which accelerated 2008 and 2009
returns.
Figure6:MRSwing SPY
$3,200,000
$1,600,000
$800,000
$400,000
$200,000
$100,000
MRSwing(counter)37.90%CAGR
MRSwing(swinglong,MRcounter)35.15%CAGR
MRSwing(nocounter)23.07%CAGR
17
5. MRSwing:ExchangeTradedFunds(ETFs)
MRSwingwasoriginallydevelopedfortheSPY.WetestedtheoutofsampleresultsonotherETFsand
foundarobustsystemperformanceacrossdifferentgeographicalandmarketsizeindexbasedETFs.
Table6:MRSwing(counter)ExchangeTradedFunds07/2000to02/2010
Sym CAGR CAGR Sharpe Portfolio Max Max Profit Percent MaxDD
riskadj Value Wins Losers Factor Profitable
Emerging EEM 24.76% 25.18% 0.69 $365,399 10 4 1.46 64.46% 37.63%
Markets
Russell2000 IWM 14.60% 14.69% 0.60 $322,973 13 4 1.61 65.89% 45.07%
SmallCaps
Nasdaq100 QQQQ 36.88% 37.27% 1.12 $2,083,961 10 5 1.74 67.23% 23.16%
LargeCap
Vanguard VTI 28.56% 28.95% 1.33 $697,403 10 6 2.05 65.51% 19.93%
TotalMarket
Nextweconsidertheconservativeversionthattakesnocountertrendtrades.
Table7:MRSwing(nocounter)ExchangeTradedFunds07/2000to02/2010
Exchange Sym CAGR CAGR Sharpe Portfolio Max Max Profit Percent MaxDD
TradedFund riskadj Value Wins Losers Factor Profitable
Emerging EEM 4.59% 8.82% 0.19 $130,031 8 5 1.13 60.00% 41.83%
Markets
Russell2000 IWM 13.70% 27.10% 0.78 $301,865 10 4 2.07 68.15% 27.68%
SmallCaps
Nasdaq100 QQQQ 29.25% 66.90% 1.28 $1,196,429 9 4 2.29 72.28% 15.93%
LargeCap
Vanguard VTI 16.40% 34.84% 1.07 $323,548 8 4 2.27 66.43% 13.57%
TotalStock
Market
NoticeagaintheoutofsampleETFresultsofthenocountersystemshowmuchlowermaxdrawdown,
and still a solid rate of return. Max consecutive losing trades of both aggressive and conservative
systems are between 4 and 6, making this a tradable system. In particular, QQQQ shows a 29%
compoundannualgrowthrate(CAGR),1.28Sharpe,77%profitabletrades,and16%maxdrawdown.
18
6. MRSwing:ManagedFutures
WetestedMRSwingonoutofsampledataonthe@ESand@NQfuturescontracts.Futuresprovide
the benefits of excellent liquidity, low cost commissions, flexible use of leverage, and tax advantages
overETFs(e.g.marktomarketresults60%longtermcapitalgainsand40%shortterm).Oneimportant
differenceintradingfuturesisourlimitordersforentriesandexitsmaygetfilledinovernighttrading.
We have found that overnight order fills to be another positive for trading MR Swing in a managed
futureaccount.
Fixedfractional(FF)moneymanagementisusedat$50,000percontract,andalso$25,000percontract.
managementsystemtradesaggressivelyforasmallaccountthengetsmoreconservativeastheaccount
sizegrows.Forexample,$10,000getsonecontract.Thenitneeds20Kforthenextcontract.At30Kit
istrading2contracts,andneeds30Kmoreinprofitforthethirdcontract,etc.
Table8:MRSwing(counter)$100,000portfoliofrom07/19/2000to02/22/2010
Sym Money CAGR CAGRrisk Sharpe Portfolio Max Max Profit Percent MaxDD
Mgt adj Value Wins Losers Factor Profitable
ES Fixed 34.76% 35.41% 1.14 $1,839,589 14 4 2.56 67.75% 46.43%
Ratio
ES FF 28.96% 29.49% 1.26 $1,197,509 14 4 3.18 67.75% 22.24%
50K/ctr
ES FF 70.12% 71.59% 1.28 $17,892,862 14 4 4.60 67.75% 46.59%
25K/ctr
NQ Fixed 28.01% 28.26% 0.89 $1,091,448 9 4 1.68 68.41% 33.84%
Ratio
NQ FF 19.90% 20.07% 0.95 $579,387 9 4 1.72 68.41% 21.37%
50K/ctr
NQ FF 43.26% 43.66% 0.98 $3,243,852 9 4 1.74 68.41% 42.09%
25K/ctr
The aggressive version of MR Swing on the ES and NQ contracts with fixed fractional and fixed ration
moneymanagementisshownbelow.Weincluded$2.30percontractforcommissions.
19
Figure7:MRSwing(counter)ManagedFutures
withDifferentMoneyManagement
$3,200,000
$1,600,000
$800,000
$400,000
$200,000
$100,000
$50,000
Jul00
Dec00
May01
Oct01
Mar02
Aug02
Jan03
Jun03
Nov03
Apr04
Sep04
Feb05
Jul05
Dec05
May06
Oct06
Mar07
Aug07
Jan08
Jun08
Nov08
Apr09
Sep09
Feb10
NQ50K/contract NQ25K/contract NQFixedRatio
ES50K/contract ESFixedRatio
One would expect the 25K/contract results to be approximately twice the 50K/contract system. Our
testsshowed70.12%CAGRvs.28.96%CAGR.Sincetwicethe50K/contractresultswouldbe60%,where
did the extra 10% come from? This is due to the interval. A $100,000 portfolio has to accumulate
another $50,000 before trading a 3rd contract. The 25K/contract system more quickly adds contracts.
$100,000 portfolio into $17.8M over ten years. We leave this as an option for the aggressive traders
interestedinMRSwing.
20
Figure8:MRSwing(counter)ManagedFuturesES
withFixedFractional$25,000/contract
$25,600,000
$12,800,000
$6,400,000
$3,200,000
$1,600,000
$800,000
$400,000
$200,000
$100,000
$50,000
MRSwing(counter)ES25K/contract
Inthenextsection,wewillshowhowMRSwingcanbeusedtomakeamuchmoreconservativesystem
thatstillgivessolidreturns.
Table9:MRSwing(nocounter)$100,000portfoliofrom07/19/2000to02/22/2010
Sym Money CAGR CAGRrisk Sharpe Portfolio Max Max Profit Percent MaxDD
Mgt adj Value Wins Losers Factor Profitable
ES Fixed 26.66% 57.22% 1.21 $1,004,598 18 4 3.03 70.59% 23.20%
Ratio
ES FF 18.52% 38.45% 1.11 $525,439 18 4 3.50 70.59% 10.71%
50K/ctr
ES FF 45.65% 105.41% 1.23 $3,927,298 18 4 4.48 70.59% 21.86%
25K/ctr
NQ Fixed 24.46% 50.61% 0.98 $831,042 11 4 2.44 74.77% 29.06%
Ratio
NQ FF 16.41% 32.89% 1.04 $435,128 11 4 2.51 74.77% 16.38%
50K/ctr
NQ FF 37.45% 81.38% 1.12 $2,173,202 11 4 2.57 74.77% 34.23%
25K/ctr
21
The conservative version of MR Swing in a managed futures account shows exciting results. In
particular,wecanpickalowdrawdownversionof50K/contract(10.71%forES,16.38%forNQ)andstill
getexcellentannualreturns(18.52%CAGRforES,16.41%CAGRforNQ).Further,ifweuseleverageon
theconservativesystem,25K/contract,wedoubleourmaxdrawdown,andgetevenmorethan2Xthe
performance(45.65%CAGRforES,37.45%CAGRforNQ).
Figure9:MRSwing(nocounter)ManagedFutures
withDifferentMoneyManagement
$3,200,000
$800,000
$200,000
$50,000
Oct01
Aug02
Oct06
Aug07
Dec00
May01
Mar02
Jan03
Jun03
Nov03
Apr04
Sep04
Feb05
Dec05
May06
Mar07
Jan08
Jun08
Nov08
Apr09
Sep09
Feb10
Jul00
Jul05
7. CreatingPortfolioSynergy
Modern Portfolio Theory asserts that assembling diversified assets with low or negative correlations
createsaportfolioofcollectivelylowerriskwithoutsacrificingreturn;effectivelyafreelunch.TheIvy
Portfolio (Faber and Richardson 2009) illustrated a meanvariance portfolio that uses noncorrelated
indexestoproduceamuchbetterabsoluteandriskadjustedreturn.Theassetallocationsaremodeled
22
aftertheHarvardandYaleendowments,whichinvestmoreheavilyinrealassets(commoditiesandreal
estate),absolutereturnsfunds,andprivateequitythantraditionalstockheavyportfolios.Additionally,
The Ivy Portfolio showed that combining low correlated assets with long term timing rules both
increasedreturnanddecreasedvolatilityoverabuyandholdstrategyofthesameassets.
Lintner(1983)wrote:thecombinedportfoliosofstocks(orstocksandbonds)afterincludingjudicious
investmentsinleveragedmanagedfuturesaccountsshowsubstantiallylessriskateverypossiblelevel
of expected return than portfolios of stocks (or stocks and bonds) alone. We demonstrate how
combiningandemployingasetofnoncorrelatedstrategiesachievessubstantialimprovedreturnsand
lower drawdowns. To achieve this we substitute a 20% portion of portfolio assets to our MR Swing
tradingsystemfortheabsolutereturnportionoftheendowmentsassetmix.Integratedintothebase
allocation of the Ivy Portfolio, this raises the CAGR by a factor of 1.7 while cutting the maximum
drawdownby3/5thscomparedtotheIvyETFswithTimingreturns.
7.1IvyPortfolio
InordertocomparetheIvyPortfoliowithtimingtoMRSwing andablendedapproach,wefirstalign
dataandtimeperiods.TheIvyPortfolioisbasedonabasketofequallyweightedindexes:
1. UnitedStatesequitymarket(S&P500)
2. MorganStanleyCapitalInternationalEAFEIndex(MSCIEAFE),
3. GoldmanSachsCommodityIndex(GSCI),
4. NationalAssociationofRealEstateInvestmentTrustsIndex(NAREIT),
5. UnitedStatesgovernment10yearTreasurybonds
To facilitate backtesting comparisons with our models on a comparable basis, we test the concepts
against tradable market instruments. To that end, we identified the following instruments for our
versionoftheIvyPortfolio.
23
Table10:IvyPortfolioInstruments
1. Vanguard500IndexInvestor(VFINX)
2. VanguardTotalInternationalStockIndex(VGTSX)
3. ThomsonReuters/JefferiesCRBIndex,acommoditypriceindex1
4. VanguardREITIndex(VGSIX)
2
5. VanguardIntermediateTermBondIndex(VBIIX)
TradableETFsusedtoconstructtheIvyPortfolio.Thetestingperiod
covers8/1/2000through1/29/2010.3
WemodeledanETFversionofTheIvyPortfolioforbuyandhold,withanequalallocationoffundsper
asset and a yearly rebalance period. We compared that to the Ivy Portfolio using basic timing rules:
Theassetsportionoftheportfolioisfullyinvestedintheassetwhentheclosingpriceisabovethe10
monthmovingaverage;itgoestocashwhenitisbelow.Cashismodeledatthe30dayTBillaverageof
approximately2.2%andweincludenormalcommissionsforadirectaccessbroker($.005/share).
TheIvyPortfoliodiversificationaddssignificantvaluevs.abuyandholdoftheS&P500inbothreturn
and drawdown. After implementing the timing rules with The Ivy Portfolio it increases performance
versus Buy & Hold, adding 3.4 to the annualized return each year while dropping the maximum
drawdownfrom30tounder12%4.
Table11:IvyBuy&HoldversusTimingResults
8/1/2000to12/31/2009
CAGR Sharpe MaxDD
SPY(S&P500)Buy&Hold 2.48% 0.16 56.49%
IvyBuy&Hold 5.05% 0.26 29.83%
Ivyw/Timing 8.45% 0.99 11.65%
1
TheCRBistheonlyexceptionofnotbeingdirectlytradableinthisform.CommoditybasedETFs/ETNsdonothavesufficient
historyforourpurposes,butDBCcouldbeusedtoday(althoughtheproblemofrollyieldfromcontango/backwardationisnot
solved).
2
VanguardIntermediateTermBondIndexchosenasTreasuryETFwithsufficienthistorynotavailable.
3
8/1/2000isthefirstperiodwhendataisavailableforallinstrumentsinoursystems.
4
BoththeIvyBuy&HoldandIvyw/TimingusetheinstrumentsfromTable10.
24
7.2IntegratingResultsofManagedFuturesintoIvyPortfolio
Measuring the correlations using the monthly delta of the equity curve values, several of the
subcomponents of the Ivy Portfolio have low cross correlations, particularly the commodities and the
bonds. MR Swing futures components, NQ and ES, balance each other nicely with only an r= 0.39
correlation. Of greater importance, MR Swing components have low and often negative correlations
withtheIvyPortfoliocomponents.Wewillseehowtheselow/negativecorrelationsenhancedtherisk
basedmetrics(Sharpeanddrawdown)asweputthepiecestogethertoenhancetheportfoliosstability.
Table14:CrossCorrelationofIvyPortfolioComponentsandMRSwingFutures
MR Swing trading the Futures with ES and NQ work well together since they respond differently to
marketmoves,withtheSharpe,maximumdrawdownandvolatilityallimprovingonthecombinedbasis.
Table13:CombingFuturesMRSwing(nocounter)FixedRatio
8/1/2000to1/28/20105
Sym CAGR Sharpe MaxDD Volatility
ES 28.00% 1.25 15.86% 19.2%
NQ 26.42% 1.05 14.65% 20.9%
CombinedNQ+ES 27.26% 1.29 11.56% 17.7%
Next,weincludetheconservativeversionofMRSwing(nocounter)managedfutures,withasmall10%
allocationeachtoESandNQ,usingfixedratiomoneymanagementintothecombinedportfolio.
5
InTables11to14thesystemsaremarkedtomarketusingmonthlydata(Sharpe,MaxDD,andVolatility)tomatchtherestof
theIvyPortfolio.
25
Figure10:IvyPortfolio+MRSwing
MRSwing
@ES
Bonds(VBIIX)
10%
16% MRSwing
@NQ
10%
RealEstate(VGSIX)
16% DomesticStocks
(VFINX)
16%
Commodities(CRB)
16% InternationalStocks
(VGTSX)
16%
Werebalancebacktothe10%allocationtoESandNQatthebeginningofeachyear.Includinga1/5th
allocationtoMRSwingintothecombinedportfoliodrivesanimpressiveincreaseof5.5pointsofCAGR,
0.7Sharpe,cutsthedrawdownby3/5th,whileonlyincreasingvolatilityby0.2%!
Table14:CombingIvyw/Timing&MRSwing
8/1/2000to1/28/2010
Portfolio CAGR Sharpe MaxDD Volatility
Ivyw/Timing 7.92% 0.89 11.65% 6.6%
Ivy+NQ+ESMRSwing
w/Rebalancing 13.41% 1.60 4.48% 6.8%
ChangeDuetoMRSwing +5.49 +0.70 +7.17 +0.2
26
Figure11:IvyPortfoliow/Timingvs.CombinedIvyMRSwing
$400,000
$350,000
$300,000
$250,000
$200,000
$150,000
$100,000
$50,000
$0
Aug00
Nov00
Mar01
Jul01
Nov01
Mar02
Jul02
Nov02
Mar03
Jul03
Nov03
Mar04
Jul04
Nov04
Mar05
Jul05
Nov05
Mar06
Jul06
Nov06
Mar07
Jul07
Nov07
Mar08
Jul08
Nov08
Mar09
Jul09
Nov09
RebalIvy+NQ+ES
Ivyw/Timing
We are not using the Futures to hedge the portfolio, but rather as a source of negatively correlated
positive returns. Blending a base of lowcorrelated assets (The Ivy Portfolio) with a noncorrelated,
marketregimeswitchingandvolatilityadaptivetradingsystem(MRSwing)createsapowerfulsynergy.
Theresultissuperiorandsaferreturnsthananyofitsindividualparts.
27
8. Conclusion
MR Swing is a quantitative system that employs daily meanreversion and swing trading in different
marketregimes.Overthelasttenyears,itshowedarobustequitycurve(37.90%CAGR,1.58Sharpe,
23.65% max drawdown) on the SPY. Removing countertrend trades cut the maximum drawdown in
halfandproducedexcellentriskadjustedreturns(23%CAGR,54%riskadjustedCAGR,1.37Sharpe,13%
max drawdown). In addition, MR Swing was tested using managed futures in various money
management configurations. The conservative nocounter setup with leverage (25K/contract fixed
fractional) produced solid results on the @NQ and @ES (45% CAGR, 105% risk adjusted CAGR, 1.23
Sharpe,21.86%maxdrawdown).MRSwingalsotestedwellonawiderangeofETFs.
Thesystemisbasedonfourcoreprinciples:(1)amarketregimeswitchingmethodtoexploitdifferent
characteristics of markets by using shortterm meanreversion in the bear regime, and deploy swing
tradinginthebullregime.Next,ituses(2)nonsymmetricaltradingalgorithmsforentries,exitsandthe
regimespecifictradingalgorithms.Everysystemcomponentisbasedon(3)volatilityadaptivemetrics
thatitcanhandlechangesinvolatilityoveralongtimespan.Finally,sincenoregimeswitchingmodel
willbeabletoeliminateallfalsesignals,eachcoresystemcomponentexhibits(4)robustnesstoregime
whipsaws.
Finally,weincludedMRSwingasa20%componentintoadiversifiedportfoliooflowcorrelationETFs
basedontheIvyPortfolio,modeledafteruniversityendowments.MRSwingincreasedthecombined
portfolioCAGRbyafactorof1.5whilereducingthemaxdrawdowntoonly4%.Thisdemonstratesthat
smoothandprotecttheirportfolioperformance.
28
9. Acknowledgements
Thanks to our excellent reviewers: Beth Borchers (editor), David Varadi (quant), Hardin Abrams
(engineer), Lee Girer (engineer/finance), and Ted Schnur (professional trader). TradeStation 8.7 was
usedtogeneratethetradecharts.AmiBrokerwasusedtosimulatetheIvyPortfoliofromYahoodata
source(includingdividends).MicrosoftExcel2007wasusedtogeneratemostcharts.
10.Glossary
CAGR Compoundannualgrowthrateisthesmoothedyearoveryeargrowthrateofthesystem.Itcovers
theentiretimeperiod.
CAGRriskadj. Riskadjustedcompoundannualgrowthrateistheyearoveryeargrowthrateofthesystem,butit
onlycountstimeinvestedinthemarket.
DVO DavidVaradiOscillator(DVO)measuresshorttermmarketmovements(Varadi2009)
FixedFractional(FF) Fixedfractionalmoneymanagementisamethodofreinvestingprofitsinafuturestradingaccount
(numcontracts=Floor(portfoliovalue/dollarspercontract)
FixedRatio Fixedratiomoneymanagementisamethodofreinvestingprofitsinafuturestradingaccount.The
moretheportfoliogrowsthemoreconservativeitbecomesinaddingnewcontracts.
MaxDD Maximumpercentagedrawdownofthesystemfrompeaktotroughoverthetimeperiod.Inour
analysistheequitycurveismarkedtomarketdailyinsteadofjustatthestart/endoftradesinorderto
showthetruemaximumdrawdownpotential.Onlysection7ontheportfoliousesmonthlyfigures
becausetheIvyPortfoliousesmonthlydata.
MaxWins Maximumconsecutivewinningtradesinarow.
MaxLosers Maximumconsecutivelosingtradesinarow.
PercentProfitable Numberofwinningtradesdividedbytotalnumberoftrades.
PortfolioValue Valueofa$100,000initialinvestmentafterrunningthesystemfortheperiodoftime.
ProfitFactor Grossprofitdividedbygrossloss.Apositiveprofitfactordenotesanetprofitpertradeonaverage.
SVAPO Vervoortsvolumeandpriceoscillatorisusedtofindexhaustionpointsinthebullregime.
Sharpe Sharperatiomeasurestheriskadjustedperformanceofthesystemusingmonthlydata.Theriskfree
rateofreturnis2%inourcalculations.
TimeinMarket Thepercentageoftotaltimeinthemarket(eitherlongorshort),whichisnotincludingflatperiods
betweentrades.
Volatility HistoricvolatilitybasedonBlackScholesusingmonthlydataintheportfoliosection.
29
10. References
Alves,Marco(2009)."JoinTheBand:ApplyingHysteresistoMovingAverages,TechnicalAnalysisof
Stocks&CommoditiesVol.27,No.1.
Bryant,Michael(2006)."TheMythofOptimization,"
http://www.breakoutfutures.com/Newsletters/Newsletter0306.htm
Elder,Dr.Alexander(1993).TradingforaLiving:Psychology,TradingTactics,MoneyManagement.
Wiley&SonsInc.,pp235243.
Faber,MebaneandRichardson,Eric(2009).TheIvyPortfolio:HowtoInvestLiketheTopEndowments
andAvoidBearMarkets.Wiley&SonsInc.
Helweg,MarkandStendahl,David(2002).Dynamictradingindicators:winningwithvaluechartsand
priceactionprofile.Wiley&SonsInc.,pp2546.
Hysteresis.(2010,February24).InWikipedia,thefreeencyclopedia.RetrievedFebruary24,2004,from
http://en.wikipedia.org/wiki/Hysteresis
Kestner,Lars(2003).QuantitativeTradingStrategies:HarnessingthePowerofQuantitativeTechniques
toCreateaWinningTradingProgram.McGrawHill,pp118121.
Lent,Justin(2009).TacticalEquityAllocationModel(T.E.A.M.).NationalAssociationofActive
InvestmentManagers(NAAIM)Conference2009.
Lintner,John(1983).ThePotentialRoleofManagedCommodityFinancialFuturesAccounts(and/or
Funds)inPortfoliosofStocksandBonds,AnnualConferenceofFinancialAnalystsFederation,May
1983.
Siegel,Jeremy(1998).StocksfortheLongRun,2ndEdition:TheDefinitiveGuidetoFinancialMarket
Returns&LongTermInvestmentStrategies.McGrawHill,pp246251.
Stokes,Michael(06/09/2009).StockMarketFollowThroughonSmallDays,
http://marketsci.wordpress.com/2009/06/10/stockmarketfollowthroughonsmalldays/
Stokes,Michael(08/10/2009).TheStateofShortTermMeanReversion,
http://marketsci.wordpress.com/2009/08/10/thestateofshorttermmeanreversionjuly2009/
Varadi,David(07/29/2009).TheDVO,http://cssanalytics.wordpress.com/2009/07/29/thedvo/
Vervoort,Sylvain(2007)."ShortTermVolumeandPriceOscillator",TechnicalAnalysisofStocks&
CommoditiesVol.26,No.11.
30