Abstract: We study the estimation, the dynamics, and the predictability of option-implied risk-neutral moments (variance, skewness, and kurtosis) for individual stocks from various perspectives. We first show that it is in the estimation of the higher moments essential to use an interpolation with a narrow grid as well as a wide interval. We show that implied moments are well explained cross-sectionally by a number of firm characteristics. We use the characteristics that have been shown to exhibit correlation with expected returns (like size and the market-to-book ratio of equity). In a next step, we investigate the joint dynamics of the three moments in a vector autoregressive model. We find that the moments are significantly linked to each other over time. Finally, adding exogenous variables to the vector autoregression improves the explanatory power of our model even further. Granger causality tests show significant differences between the three implied moments.