three. The NYSE and the AMEX are the only floor exchanges; the others are electronic. Floor brokers nowhandle only 10% of NYSE-traded volume; all the rest iselectronic.Until just two years ago, the NYSE maintained itslong-standing market share of about 80% of trading inNYSE-listed stocks. Since then, dark pools and crossingnetworks have proliferated, causing the NYSE’s share of trading in its own listed stocks to drop to less than 50%.This extraordinary change in trading practices is oneof the driving forces behind the increase in liquidity frag-mentation, particularly over the past few years. In thisenvironment, the science of liquidity management hasbecome increasingly complex, quantitatively oriented,and computer driven. Intelligent, focused traders, how-ever, remain the critical factor in a successful liquiditymanagement equation.
LIQUIDITY MANAGEMENT: THE NEWBEST EXECUTION PARADIGM
Today, best execution depends on the ability to accessdeep sources of liquidity while adding performance,value,and cost efficiency to transactions. A keycomponent for achieving this new paradigm is algorithmic trading, whichcurrently is used to generate one-third of all trading activityin the U.S.and 40% of trades on the London StockExchange. The most advanced algorithmic strategiesdynamically adapt to market trends, enabling the trader to be more productive, have more control over execu-tion choices, and connect with other liquidity manage-ment technologies, such as dark pools and crossingengines, to optimize performance.Today’ssophisticated algorithms can be used to slicelarge orders into smaller sizes, often with the intention of hiding the size of these large orders from other marketparticipants—a technique known as “iceberging.” For momentum investors, “participating” algorithms can beused to ensure that a certain percentage of trading volumein a particular stock is captured. “Benchmark” algorithmscan be used to achieve specific benchmarks such as thevolume weighted average price (VWAP) or time weightedaverage price (TWAP) over a certain time period. Analgorithm may trigger a buy order on a certain percentageupward movement in a share price. Trades can be drivenwhen the spread between two stocks exceeds a certainamount, and they can even be generated by news events.Quantifiable events such as a company’s sales, net profit,or earnings per share can be e-mailed automatically inXML (extensible mark-up language) with tags aroundparticular figures so those figures can be read by algorithmsand instantaneously translated into stock trading orders.As the use of algorithms grows, trading is becomingfaster and more frequent. Daily trading volume hasincreased exponentially, and constant stock-price volatilityhas become a way of life. Product development continuesto accelerate as algorithms are developed to accommodateawide array of long and short strategies, multi-asset invest-ment portfolios, and the use of interest rate, currency,commodity, and credit derivatives. The next wave of algorithms will prove to be even more sophisticated andtrader focused, with such advanced features as embeddeddark pool seeking technology, true global reach, and moreprecise specialization for individual execution strategies.
DARK POOLS: UPPING THE LIQUIDITYMANAGEMENT ANTE
In many ways, algorithms are the answer to the ques-tion, “How can a trader do institutional size in markets opti-mized for retail trading?” With the evolution of the markets,often the best wayto work a one million shareorder is bydoing a few hundred shares at a time. Big computers andtrading algorithms make this possible. However, it is usefultoremember that, ultimately,this is a little likedigging aswimming pool one shovel full at a time—it is easy to arriveat the thought that there has got to be a better way. For thisreason dark pools were born. The dream is to match naturalorder with a natural contra order in the dark, without infor-mation leakage, gaming, or market impact. Not surpris-ingly, the reality is somewhat more complicated.
WHO NEEDS DARK LIQUIDITY—AND WHEN
Manytrading situations do not call for dark liquidity.Others virtually demand it. As we can see in Exhibit 1,whether dark liquidity is needed depends on whether thetrader is dealing with large cap, medium cap, or small capstocks (vertical axis), and the portion of average daily tradingvolume (ADV) that the trader is trying to buy or sell (hor-izontal axis). In the upper left, where the trader is dealingwith large cap stocks and not buying or selling more than5% of ADV, the open markets are perfectly satisfactory. Darkpools are not needed. But at the opposite end of the spec-trum, where investors are trading small and medium cap
22
B
EYOND
E
XECUTION
:T
HE
C
HANGING
R
OLEOFTHE
T
RADERINA
L
IQUIDITY
M
ANAGEMENT
E
NVIRONMENT
F
ALL
2007
LG_Ardell.qxp 12/13/07 9:37 AM Page 22
Leave a Comment