This is an archived article.

June 1, 2004

Momentum investing: UW research zeroes in on transaction costs

Financial wizards have known for some time now that sifting through accounts, worrying about the competence of a company’s management, or studying price-earnings ratios aren’t always the best ways to pick winning stocks.

About the only information momentum investors need to assemble a portfolio that will beat the market is knowledge of the best-performing shares during the previous six months. Buy them. Sell them six months later and, on average, an investor will beat the market.

But there’s always a but.

Only the investors who pay close attention to the price impacts of their trades, otherwise known as hidden transaction costs, can consistently earn a net profit, according to a new study by a University of Washington Business School researcher.

“Momentum investing is a strategy based on relative strength. You buy the winners and short the losers,” said Ronnie Sadka, assistant professor of finance, and co-author of the paper that appears in this month’s edition of the Journal of Finance. “This would also work when the market doesn’t perform well because the profitability of the long-short strategy does not exhibit strong correlation to overall market movements.”

Experts generally agree that this type of momentum investing is a very simple — and proven — strategy. Even equity analysts, known for their meticulous analysis, occasionally use momentum as a method of making stock selections. But Sadka and his research partner, Robert Korajczyk of Northwestern University, found that investment strategies relying on momentum alone often fail simply because of the cost of doing business.

Selling a stock is analogous to selling a house, explained Sadka. Say a house that was purchased for $100,000 a few years ago is today valued at $150,000. It appears that the homeowner has earned 50 percent on this investment. However, obviously to realize these gains, the house must be sold. The buyer will likely try to negotiate a deal for a lower price, and the actual selling price will be, say, $140,000. Thus the homeowner’s actual return after considering the hidden transaction cost of $10,000 is only 40 percent, and this is without considering direct transaction costs such as brokerage fees and closing costs.

Or in the case of stocks, a firm that puts in a large-scale buy order for a stock priced at $50 will likely end up paying more. It’s simple supply and demand, Sadka said.

“The seller is thinking, well maybe they know something that I don’t. If they want to buy it that badly, then it must be worth more.”

So the price ends up being $51, for example. And since maintaining and updating momentum-trading strategies requires relatively high stock turnover, one should be particularly concerned about the effect of hidden transaction costs on net profitability.

Korajczyk and Sadka devised an optimal strategy that took into consideration the hidden transaction costs investors would incur. They further winnowed the best performing stocks into a subset of best performing stocks with the lowest expected hidden transaction costs. The resulting group of stocks may change the investor’s return and are likely to increase risk because it’s a smaller pool of investments. However, their liquidity-conscious method turned out to do better than other momentum strategies because they weren’t hit as hard by the hidden transaction costs.

The paper is titled “Are Momentum Profits Robust to Trading Costs?”

###

For more information, contact Sadka at (206) 221-5383 or at rsadka@u.washington.edu