Abstract. Given the sustained attention to COVID-19, we explore if a reference dependent version of the CAPM has a good explanatory power. It views the CAPM with the prospect theory references - certainty effect, reflection effect and isolation effect. It firstly follows a linear prospect theory version of the CAPM, then, it extends that with the autoregressive distributed lag (ARDL) model, and finally, it augments them in the generalized autoregressive conditional heteroskedastic (GARCH-X) setup. It views the risk-free rate and market rate of returns as the certainty effect and reflection effect respectively while the lagged endogenous returns and the ARCH and GARCH effects as the isolation effects. With the NSE listed pharma-stocks’ data during COVID-19, pre- COVID-19 and both periods together, the prospect theory references depict that investors can build up different implications of the CAPM. With certainty effect, the prospect theory version of CAPM has less explanatory power while with reflection effect, the same has good explanatory power with the sample stocks over the data sets but it does not explain the isolation effects at all. Investors may re-look into that the CAPM if calibrated with the prospect theory references at ARDL and GARCHX augmentations, it provides better explanatory powers.

Keywords: CAPM; the prospect theory; reference-dependence; behavioral finance; ARDL and GARCH-X augmentations.