Ferrin Vol Fund sells one-month at-the-money index straddles every month at 16% implied volatility, while realised volatility averages 12%. Roughly what does it earn in a normal month, and what happens in a month when the index moves 12%?

Ferrin collects about 3.69% of notional a month and pays out about 2.76% on average, so it earns about 0.92% in an average month and about 1.3% in a typical one. A month in which the index moves 12% costs about 8.3%, roughly 9 average months of gains. The volatility premium pays steadily until one large move takes back most of a year.

Price the straddle with the shortcut every desk uses. An at-the-money straddle is worth about 0.8 times volatility times the square root of time: 0.8 x 16% x 0.289, about 3.69% of notional. At expiry Ferrin pays out the size of the index's move, in either direction. If the index really moves with 12% volatility, the average absolute move in a month is about 0.8 x 12% x 0.289, 2.76%. The difference, about 0.92% a month, is the volatility risk premiumThe tendency of option prices to imply more volatility than later turns up, which is the pay option sellers receive for taking on crash risk., worth about 11% a year on notional equal to capital.

Mostly quiet gains. It works like selling flood insurance in a town where it rarely floods: most years every premium is profit. With monthly index volatility of about 3.5%, about 71% of months move less than the premium collected, and the median month earns about 1.3%. That is why the strategy's track record looks smooth: many small gains, few losses, and a Sharpe ratio that looks excellent for years. The shape is the whole story. The gain is capped at the 3.69% premium; the loss is not capped at all.

Ferrin owes the whole move and keeps only its premium. 12% less 3.69% is a loss of about 8.3% of capital, which erases about 9 months of average profit in one month. A normal model says such a month, more than three and a half standard deviations, comes about once in 157 years. Equity indices have produced moves of that size in several crises within living memory, far more often than the bell curve allows, which is the whole reason the premium exists. Sold on twice the notional, the same month costs 17%.

Close with what a risk manager would do about it. Size the notional so the worst plausible month is survivable, not so the average month looks attractive. Buy some protection further out, such as cheap out-of-the-money puts, which gives up part of the premium to cap the tail. And judge the strategy on its worst months, not its Sharpe ratio, which flatters any strategy whose losses have not happened yet.