Will the Smart Institutional Investor Always Drive Prices to Fundamental Value?

We study gains to momentum trading from 1946 through 2002. Past papers assume a zero-investment strategy where short sales of losers fund the purchase of winners. In practice, the broker holds the cash from the short sale as collateral, and the investor funds long positions with his own or borrowed capital. To estimate implementation costs, we assume that the investor borrows through securing other investments, pays margin costs on the borrowings, and earns the rebate rate on the cash from the short sale. We find that the difference between the call money and rebate rates is 1.4% per year over the full sample, but in more recent years, this cost approaches 2%. This zero-investment friction and trade costs restrict arbitrage and smart-investor trades. Also trade costs change over time and allow us to look at the effects of trade cost changes and momentum returns. Congress deregulated trade commissions on May 1, 1975. Commissions fell dramatically for large stocks but did not materially change for small stocks. The commission change sets up a natural experiment to test if momentum fell after commissions fell. We find that large-stock momentum fell but standard momentum strategies, which use a substantial number of smaller stocks, did not materially change.