Abstract
Options markets reflect the expectations of market participants regarding the future movements of the underlying asset across different strikes on each observation date (the so-called-option cross section). In the first chapter we demonstrate how this forward-looking information in the form of risk-neutral densities can be utilised to address the question of how option prices anticipate the financial impact of a given event over time. The general consensus in the literature is, however, that risk-neutral densities are biased and/or inefficient estimators of the future return densities, since they do not reflect the fact that market participants are risk averse rather than risk-neutral. Therefore, in the second chapter we extend the existing framework for transforming the risk-neutral densities into physical return densities via the stochastic discount factors that provide a link to investor’s level of risk aversion by considering their functional forms that can model locally increasing parts, a phenomenon known as the pricing kernel puzzle. In the third and final chapter we propose a way to make the estimators of the physical return distributions that are based on past return data, namely kernel densities, more forward-looking by minimising their dissimilarities from the risk-neutral densities. Overall, we provide further evidence on the value of utilising the information embedded in the risk-neutral densities for purposes of risk and portfolio management.