http://clevelandfed.org/research/workpaper
Best available copy
Once size is accounted for, a significant price effect still exists during the
TOY and post-yearend periods. The SIZE portfolios do not exhibit significant
size-related effects in any of the subsamples, while the PRICE and MVPR
portfolios exhibit a significant price-related effect during the TOY and
post-yearend periods. This result is consistent with the LPSH, which argues
that size proxies for price during the TOY period.
Tables 1 and 2 show that mean daily returns are higher for all portfolios
during the pre-yearend period, post-yearend period and, therefore, the TOY
period. Although the pre-yearend mean daily returns do not exhibit any
factor-related bias, they are roughly 10 times larger than the sample period
returns for all of the portfolios. This would indicate that the adjustment in
stock prices from their tax-depressed lows occurs before the end of the calen-
dar year. In fact, Roll [23] finds the anomalously high returns at the turn
of the year begin the last trading day of December. Note that the anomalously
high returns for all the portfolios (except MVlO, PRIO, and MVPRS during the
post-yearend period) during the 10 trading days centered on the end of the
calendar year can be explained by recorded-price errors. An investigation of
the absolute price movements during the TOY period supports this conclusion.
Thomson [29] shows that the change in prices during the 10 days surrounding
the end of the year is within the bounds predicted by the LPSH. This is
further evidence that the TOY effect is a price-related effect and not a
size- related effect.
An alternative explanation for the anomalous TOY returns is that syste-
matic risk increases during the TOY period. If systematic risk increases,
then returns should increase to compensate market participants for the addi-
tional risk-bearing services provided. In other words, the abnormally high
TOY returns are not anomalous if risk-adjusted returns are no higher during
the TOY period than during the rest of the year. If systematic risk increases