| Abstract: | Monthly holding period returns for U.S. Treasury bills and notes of identical maturity indicate a significant coupon effect upon term premiums. Hotelling's T2 test of the vectors of mean term premiums indicates that term premiums are not statistically significant for notes but are significant for bills. Mean-variance and stochastic dominance criteria indicate an investment preference for bills over notes on a pretax basis. Because the data set is Treasury bills and notes, which are identical except for coupon level, these results are evidence of a coupon effect on term premiums. |
