I have asked this question here, but did not get any answer. I was wondering if anybody knows a method of deriving skewness and kurtosis measures from different quantiles, mean, and/or variance. I do not have access to the full distribution of returns, but can get any quantile, along with mean and variance.
Thanks and let me know if you have additional questions.
