From Skews to a Skewed-t

In this paper we present a new methodology to infer the implied risk-neutral distribution function from European-style options. We introduce a skewed version of the Student-t distribution, whose main advantage is that its shape depends on only four parameters, of which two directly control for the levels of skewness and kurtosis. We can thus easily vary parameters to compare different distributions and use the parameters as inputs to price other options. We explain the method, provide some empirical results and compare them with the results of alternative models. The results indicate that our model provides a better fit to market prices of options than the Shimko or implied tree models, and has a lower computation time than most other models. We conclude that the skewed Student-t method provides a good alternative for European-style options.