In equity markets, a selloff doesn’t really count until it gets the term structure up.
For instance, here’s the ratio of SPY 3 month implied volatility to one year implied volatility since 2012:
The only two occasions when we got close to backwardation here were in May 2012 and the end of December 2012 (fiscal cliff). In 2013, the mean level has been about 0.83, which just means that 3 month SPY options were typically about 83% as expensive, in IV terms, as one year SPY options. We’re intentionally using further dated options to filter out some noise of the sort that you would get with e.g. a VIX study.
There were three occasions that got the 3m1y term structure to perk up this year, but it’s the most recent jog higher that has my attention. Instead of spiking higher, the term structure rose steadily from mid-May until it peaked around 0.90 on June 5th.
Keep in mind that this flattening of the term structure – presumably on talk of reduced Fed asset purchases? – required a decline of less than 70 S&P 500 points. The chart below zooms in on this period with SPY prices for comparison. This behavior seems consistent with the nervous fidgeting I described in late April.
After the big rally on June 7, the market has already made up half the distance to its recent closing high. Option implied volatility, though, has further to fall. If stock prices continue to stabilize and the 3m1y estimate returns to its recent floor, a time spread opened for a debit (short 3 month, long 1 year) should perform well. Given the preference these days for unrelenting marches higher, I wouldn’t care to be short upside gamma too close to the money, so a double calendar or double diagonal makes more sense.