While the positive alphas generated from the short portfolios shown in Table 2 are unexpected, as mentioned earlier, they are not without precedent in the bond literature. This result could be due to omitted risk factors given the lack of consensus on the correct expected return model for bonds. If this is the case, examining the long-short portfolio alphas is valuable because the portion of abnormal returns explained by the omitted risk factors is likely to be similar in the long and short portfolios and to get netted out in the long-short portfolio calculation. The longshort portfolio alphas then may become the best indication of which strategies are likely to be the most successful in practice. The