This month we learned the hard way just how damaging a sharp drop in implied volatility—a.k.a. “vol crush”—can be to calendar spreads, and that straying from our short-list of relatively well-behaved stocks and ETFs is a risky way of trying to avoid it. Back in December, we tried another approach for lowering volatility risk, using a diagonal spread instead of a straight horizontal calendar, and we found that starting out close to delta-neutral with a small strike spread provided little protection against a 10-point drop in IV. In our closing analysis, we mentioned that there’s a way to get the lower vega risk of a wider, directionally biased diagonal spread in a non-directional trade—the double-diagonal.
Much like an iron condor combines two directional trades into a neutral one, a double-diagonal marries a bull put diagonal and a bear call diagonal to form a hybrid between an iron condor and a calendar spread. To a condor trader, it’s an iron condor with the long strikes at a later expiration; to a calendar trader, it’s a double-calendar with the back-month options farther out of the money. The result looks a lot like a double-calendar, only with a fraction of the vega risk. Let’s compare the two:
The SPY double-calendar above is positioned to be direction-neutral—IV is skewed toward the puts, so we chose a lower strike that’s farther out of the money than the upper strike in order to reduce the initial delta. The risk graph below shows a double-diagonal with the same short strikes, but with the long strikes $5 farther out of the money. The position sizes (total capital at risk) are equivalent.
The profit/loss curves are a little different at expiration, but that difference doesn’t really come into play until the last week or so. For the most part, the characteristics of the two positions are nearly identical: gamma and theta are about the same across both time and underlying price. The precise delta at any given price differs between the double-calendar and the double-diagonal shown, due to vertical volatility skew in the back month, but the two risk curves have pretty much the same overall shape.
What the graph doesn’t show is the big difference in vega between the two positions. In the table below the graph, we see that the vega of the calendar spread is in the 130 range—which means we lose about $130 (2.77%) for every one-point drop in implied volatility. The double-diagonal, on the other hand, has a vega of less than 30, so we’re down only about $30, or 0.64%, for every point lost in IV.
Another advantage of incorporating a vertical-spread component is that we can use wider strikes compared to a straight double-calendar, because the delta of each spread is skewed toward the center. Theta drops a little as we move the strikes farther out, but vega also goes down significantly. Wider strikes mean less chance we’ll have to make an adjustment, and if we do, a higher probability that it won’t be necessary until we’ve had a chance to benefit from some time decay.
For more about double-diagonals and what they have to do with Calendar Options, see Part II, Putting the Double-Diagonal to Work.