“…For example, instead of using historical skewness based on daily data, Conrad et al (2013) use option price data to estimate skewness, and Amaya et al (2015) use intraday data; they seem to find supporting evidence for unconditional models, although their samples are limited by the availability of data. In contrast, our new asymmetry measures do not suffer from these shortcomings and are consistent with theoretical models such as those of Barberis and Huang (2008), Goulding (2017), and Han et al (2018), which predict that high upside asymmetry implies a lower expected return. Table 8 reports the time-series average of the slope coefficients and their t -values (given in parentheses) from Fama-MacBeth regressions of stock excess returns on ISKEW, IEϕ, or ISϕ, as well as other variables representing firm characteristics (see column 1), from Aug. 1963 to Dec. 2015 in high-and low-ALIQ (aggregate stock market liquidity) periods.…”