The stock market, always an object of fascination for researchers, shows predictable patterns of return over time.
Over the short term - say, a week or a month - returns are likely to reverse, while over a medium time frame - three months to a year - they are likely to regain momentum in their original direction. Over the long term - three to five years- stock returns tend to show reversals once again.
In a hypothetical world of zero-cost portfolio trading, then, an investor holding long positions of winning stocks and short positions of underperformers would achieve superior returns over an intermediate horizon, researchers have shown.
But what impact do trading costs have on these strategies? And does this observation hold true for portfolios regardless of their size? Kellogg Professor Robert A. Korajczyk and his co-researcher Ronnie Sadka, PhD ‘03, decided to find out.
“There’s a long literature that seems to indicate there are opportunities for investors following momentum-based strategies. But none of these have taken into account the costs of implementing those strategies,” says Korajczyk, the Harry G. Guthmann Distinguished Professor of Finance and Director of the Zell Center for Risk Research.
Korajczyk and Sadka, a professor at the University of Washington, looked at how trading costs affect the profitability of investment strategies that rely on these patterns of momentum. They focused in particular on how large a momentum-based fund would have to be before the cost of trading drives profits to zero. Their findings were published in a June 2004 article in the Journal of Finance.
The trading of large institutional portfolios leads to price impact, with the purchase of large amounts of stock pushing up the price and the sale of large amounts pushing down the price. When transaction costs are not figured in, the size of the profits can be misleading. “If you think you’re going to get 10 percent on the trade, but your trading pushes the price up by 5 percent, then you’re only really getting a 5 percent return,” Korajczyk says.
Trading costs can have an even larger effect on smaller, less liquid stocks. These costs can wipe out a small-capitalization-focused fund’s profits when a momentum-based strategy is applied.
“It was surprising how rapidly the profits went away when they were based on traditional strategies,” Korajczyk says. “You get to a $10 million to $20 million fund and they just disappear because of the adverse effect of price impact. They were touted as being very profitable but were not, after accounting for the cost of trading. However, we found with a tweak of the strategy, we could trade much larger portfolios before profits were driven away.”
Korajczyk and Sadka discovered alternative strategies that provide greater profits: in particular, one that tilts portfolios in favor of stocks that are more liquid, and away from those that are less liquid.
“When a stock is illiquid, there are more frictions to trading,” Korajczyk says. “You can’t just buy or sell an indefinite quantity at some fixed price. In the case of an illiquid stock, it could be that there aren’t many shares being traded, so your trading has a large impact on the price. It’s not like going into a currency market, where you could buy tens of millions in a currency and it wouldn’t change the exchange rate very much.”
The findings are likely to be of particular interest to professional money managers trading larger portfolios, Korajczyk says.
“A naively implemented momentum-based strategy will not be successful at very large investment sizes,” he says. “Thinking more carefully about how the difficulty of trading assets intermingles with trading strategies can help you implement the strategy at a more significant investment size.”