While risk reversals are often sold to the buy-side as costless in terms of the premiums, their real effective price is the skew. Banks simply fiddle with the vol spread and the stikes prices to equate the premium: the premium paid nets off against the premium received. However, the real cost of the trade is the skew - the price in vol terms, of out-of-the-money calls vs. the price of out of the money puts. It is surprising to what extent this is not fully appreciated. Of course, there are times when the skew is large and in your favour when you can set the stikes at equal deltas and get paid positive $$ to do the trade. For example, if you were a natural seller of yen calls and buyer of yen puts or a natural seller of MXN puts and buyer of MXN calls you would very likely be able to do the trade and realize the net difference in the final price of the option premiums. Or, you can vary the strikes in such a way as to widen the spread between the strikes in your favour. As we pointed out in earlier weeklies the decline in the one-year MXN risk reversal may well be a sign that hedgers are getting less worried about a sharp down move in the peso. While the bid in the premiums for CAD puts vs CAD calls both in USDCAD and EURCAD is a sign of persistent buy-side concern with the potential CAD downside risk. Often the risk reversal is bid over for the lower interest rate FX pair. For example, JPY is bid over in USDJPY, the USD is bid over in USDMXN, and CHF is bid over in EURCHF. In those cases when it does not apply it can sometimes be a sign of a market that is overbought. For example on those rare occasions when Yen puts trade at a vol premium to Yen call or CAD call trade at a premium to CAD puts. But as a directional indicator, it is not flawless for sure, and it should be taken as one indicator that should be considered in relation to other systematic tools.