Abstract: In this paper, whether downside risk and signed jump risk have effects on pricing industry portfolios is examined. Assets pricing models with market risk premium, downside risk and signed jump risk are proposed firstly. Then, the new models are applied to study the contemporaneous/intertemporal pricing problem of industry portfolios. The results indicate that the contemporaneous market risk premium, downside risk and signed jump risk factors perform important interpretative functions for the excess return of industry portfolios. And the functions for cyclical industries are stronger than that of non-cyclical industries. However, the first-lagged market risk premium, downside risk and signed jump risk are limited in forecasting the contemporaneous excess return of industry portfolios. Furthermore, the first-order autoregressive model (AR(1) model), first-order autoregressive model with leverages (LAR(1) model), third-order autoregressive model (AR(1) model), first-order autoregressive model with leverages (LAR(3) model), heterogeneous autoregressive model (HAR model) and heterogeneous autoregressive model with leverages (LHAR model) are employed to obtain the predictive values of all risk factors, and intertemporal assets pricing models are constructed. It is found that new models show good pricing power for industry portfolios. Among them, HAR and LHAR models outperform other models, and their performances are particularly prominent for pricing the Shanghai material and public industry portfolios. The above results mean that the effects of downside risk and signed jump risk should not be ignored when pricing industry portfolios in shock market.

Key words: downside risk, signed jump, realized semivariance, asset pricing