lmplementing Staristical Criteria to Select Return Forecasting Models Table 3 (continued) Country SSC dy r-1 -2 Ik.-1 F6-2 P.-1 Yk.r-1 -1 Y-1 per-1 Netherlands PIC 0 0 0 0 0 cont'd. PLS 0 0 D PLS-MDC 0 0 0 0 0 0 00 Norway 家 0 0 0 0 0 AIC 0 0 0 BIC 0 0 FIC 0 0 0 0 PIC 0 0 0 PLS 0 0 0 PLS-MDC 0 0 0 Spain R AIC BIC 0 FIC 0 PIC 0 PLS 0 1 0 0 0 PLS-MDC 0 0 0 Sweden AIC 0 0 0 BIC 0 0 FIC 0 0 PIC 0 0 0 PLS 0 0 0 PLS-MDC 0 1 Switzerland R AIC 0 BIC 0 0 0 FIC 0 0 0 0 PIC 0 1 PLS 0 0 0 0 PLS-MDC 0 0 0 0 0 UK AIC 0 BIC 0 0 0 0 0 0 FIC 0 0 0 0 PIC 0 0 PLS 0 0 0 0 0 PLS-MDC 0 1 0 0 0 0 0 0 0 US R 0 AIC 0 0 D 0 0 1 0 BIC 0 0 0 0 0 0 FIC 0 0 0 0 0 PIC 0 0 0 0 PLS 0 0 0 0 0 0 0 0 PLS-MDC 0 0 0 0 0 0 0 0 d,denotes the January dummy;P,-1 is the stock price level at the end of month t-1."SSC"is short for statistical selection criterion.Results for the following SSC are displayed:R(adjusted R),AIC (Akaike information criterion),BIC (Schwarz's information criterion).FIC (Fisher information criterion),PIC (posterior criterion).PLS (predictive least squares criterion).and PLS-MDC (predictive least squares criterion adjusted with Markov dimension criterion).See text for additional information.The entry"0" indicates that the corresponding predictor was dropped by the selection criterion.The entry"1"indicates that the predictor was retained. 415