Read on Twitter
@OlympusQuant 1/5 thanks for asking me to review the below. I appreciate your hard work and willingness to share it. I found your work well thought, executed and informative. I am now following you to track your progress. Here are my comments. https://twitter.com/olympusquant/status/1351002799774052355
2/5 first off, whenever you analyze variables and their predictive function, I encourage to take a step back and check if it actually bears significance. In this case looking closely if skewness variation predicts future payoffs in FX pairs. There are many ways to do so.
3/5 Here is a suggestion: indexing skewness over a long enough rolling period and comparing to subsequent returns in a fwd pre-defined period of reinvested futures rolled at a set date ore-expiry. Someone like @M1tchRosenthal would be a great follow and guide.
4/5 From a pure theoretical point of view I need to alert you to the fact that skewness in FX is largely explained and influenced by carry. Carry naturally counterbalances a negative bias due to a structural fear of a $ rip for all short $ pairs @VolQuant is also great
5/5 i would also encourage you to integrate FX forward premiums into your analyzes as they also have a strong influence. Finally, continue and never be shy to share. @nope_its_lily and @sajidnizami are both great follows and guides on the how to. 
