Comments and Discussion

Willim Poole: Mislikin's paper provides a thorough examination of efficient-markets theory and many useful ideas on its implications for macro models and monetary policy. His empirical results replicate and extend previous work on efficient markets. His work is careful and thorough. As far as I can see from reading the paper, he has been extremely careful in his treatment of the data and in his statistical analysis. Mishkin's empirical work uncovers only one puzzle. In equation 29, when he regresses the return from holding common stocks on the threemonth treasury bill rate, he finds a negative coefficient rather than the positive one predicted by the theory. The efficient-markets model, of course, predicts that, except for differences in returns due to risk and liquidity premiums, returns should be equalized on all assets; but Mishkin's equation comes up with the result that when the treasury bill rate is high, the rate of return expected on stocks is low. To understand that equation, suppose that the bill rate is relatively high at 8 percent. The quarterly rate of return on bills in decimal form is then 0.02. A bill rate of 0.02 times the regression coefficient of about -5.0 is -0.10. Add to this figure the constant term in the equation of 0.08, and a per quarter expected return is obtained from holding common stocks of -0.02, or -8 percent per year. Similarly, when the bill rate is relatively low-say 4 percent per year or 0.01 per quarter-the expected return on common stock is 0.03 per quarter or 12 percent at an annual rate. The puzzle is how the expected return on common stocks can be negative when treasury bills earing a positive rate of return can always be held. A possible explanation for this result depends on the existence of transactions costs. The time series of changes in the bill rate displays