The Review ofl Size-sorted portfolio results for a one-week base observation period Number Number q of base observations aggregated ng of base Time period A, Portfolio of firms with market values in smallest NYSE-AMEX quintile 620906851226 1,42 620906740501 (6.12) 740502851226 B. Portfolio of firms with market values in central NYSE-AMEX quintile 620906851226 1216 19 (8.37) 740502851226 608 C. Portfolio of hrms with market values in largest NYSE-AMEX quintil 40502851226 608 (19 September 6, 1962, to December 26, 1985, and subperiods. The for the sample period from e ratios 1+ M, (q)are test statistics have a standard normal distribution G Vpothesis, the value of the variance ratio is 1 and indicate that the corresponding variance ratios are statistically different from I at the 5 percent level of in Table 2 indicate that the portfolio returns for the smallest quintile have a 42 percent weekly autocorrelation over the entire sample period! More over, this autocorrelation reaches 49 percent in subperiod 2( May 2, 1974 to December 26, 1985). Although the serial correlation for the portfolio returns of the largest quintile is much smaller(14 percent for the entire sample period observation interval of four weeks, much of the evidence od), it is statistically significant Using a ba against the random walk for size-sorted portfolios disappears. Although the smallest-quintile portfolio still exhibits a serial correlation of 23 per- cent with a z*(2)statistic of 3.09, none of the variance ratios for the largest quintile portfolio is significantly different from 1. In the interest of brevity we do not report those results here but refer interested readers to Lo and MacKinlay (1987a) The results for size-based portfolios are generally consistent with those for the market indexes. The patterns of(1) the variance ratios increasing in qg and(2)the significance of rejections decreasing in q that we observed for the indexes also obtain for these portfolios. The evidence against the