The theory works, until you factor in trading costs.
An impressive amount of academic evidence has documented the momentum phenomenon, which stubbornly defies explanation if you're a believer in efficient markets. The strategy gained great prominence in 1993 when a landmark study conducted by Sheridan Titman at the University of Texas and Narasimha Jegadeesh, who was then at the University of Illinois, suggested that maintaining a long position in past strong performers and shorting past weak performers could earn investors abnormally large returns over a six- to 12-month horizon. Since then, other research has produced similar conclusions prompting various academics to suggest that pursuing a relative-strength strategy can produce returns anywhere from 0.5% to 1.5% a month.
In theory the strategy sounds great, but now naysayers are questioning whether practical ways exist to exploit the momentum effect. In two new studies researchers, while not disputing the momentum anomaly, suggest that costs are too prohibitive to make the approach profitable for most institutional investors. Their findings conclude that any performance advantage gained by following the strategy would be swallowed, in most cases, by sizable trading costs.
One of the papers, The Illusory Nature of Momentum Profits, was written by David A. Lesmond of Tulane University, Michael J. Schill of the University of Virginia and Chunsheng Zhou of Peking University. It will be published in an upcoming issue of the Journal of Financial Economics. Donald B. Keim, a finance professor at the Wharton School at the University of Pennsylvania, authored the other, The Cost of Trend Chasing and The Illusion of Momentum Profits.
While previous studies didn't ignore transaction prices, the researchers of these latest papers suggest that transaction costs have traditionally been understated. Costs, they argue, have been downplayed in a variety of ways, including using mean commissions for New York Stock Exchange trades even though the momentum strategy is dominated by smaller, high-beta Nasdaq stocks.
What's more, a momentum strategy can require higher trading frequencies as well as shorting stocks, which is an expensive process that isn't always taken into account. In addition, stocks used in relative-strength strategies disproportionately trigger larger trading costs since momentum investors often clamor after the same equities at the same time, which reduces their liquidity and increases prices. Simultaneously, momentum traders tend to dump stocks at the very time when few investors want the dogs. Further, researchers contend it was misleading for previous academics to use average stocks, trading in average quantities, that are trending neither up nor down and that trade in markets with average liquidity.
A momentum strategy, when transaction costs aren't considered, may appear profitable, but when you take into account the total cost of trading, the momentum profits seems to be eclipsed in large part, says Lesmond, who is an assistant professor of accounting and finance.
Relying upon what they classified as "conservative" estimates of trading costs, Lesmond and his co-authors found little evidence that trading expenses generated by momentum strategies can dip below 1.5% per trade. In some cases, depending upon the momentum approach used and the size of stocks traded, the expenses were significantly higher than that.
Using a sampling period that extended from 1980 to 1998, Lesmond and his colleagues examined hypothetical trading costs derived from three relative-strength strategies chosen because of their popularity. Two of the strategies were lifted from the landmark 1993 study and Jegadeesh and Titman's later update, which documented similar findings. The third approach was borrowed from research conducted by Harrison Hong, Terence Lim and Jeremy C. Stein in 2000 that was published in the Journal of Finance. Using monthly return data from the Center for Research in Security Prices (CRSP), Lesmond and the other researchers divided the portfolios, based on gross returns, into poor, moderate and strong performers.
The study's findings illustrated that the costs of momentum trading were most prohibitive for small-cap stocks. The researchers, who used two different trading cost measures, examined the consequences of shorting the smallest small-cap losers and going long on the same pool's winners and then closing out these positions in a six-month period. The mean total costs for those four transactions would be either 12.13% or 12.92%. In contrast, the total performance return during that period was just 3.91%-a dramatic difference. When the largest and most liquid stocks were examined, the gap between costs and returns narrowed considerably, but the investor still would have lost money after costs. In this scenario costs were either 2.32% or 1.94%, which swallowed up the 1% return.
Discussing the study, Schill of the University of Virginia said the findings illustrated the futility of seeking a way to practice momentum investing while dodging prohibitive trading bills. "We suggest that if you eliminate high-trading stock costs, you eliminate the momentum trading effect," he says.