With SPY up $2.00 this afternoon, we’re opening a second August/September double-calendar, as follows:
Day limit order
Buy to open 2 SPY Sep 138 calls
Sell to open 2 SPY Aug 138 calls
Buy to open 2 SPY Sep 135 calls
Sell to open 2 SPY Aug 135 calls
for a net debit of $1.78† or better.
Note that we’re using the calls for both strikes in this double-calendar. And, as usual, 2 contracts is our base position for double-calendars. Trading whole-number multiples of the base-position size ensures that adjustments will not result in unbalanced positions. In addition, in order to come as close as possible to matching our Model Portfolio risk profile, it’s important to allocate an equal dollar amount to each initial opening trade in a cycle.
Analysis: Today’s big rally was sufficient to bring SPY in line with our trade-entry criteria on all fronts—strike diversification, delta and implied volatility. We had to keep the strikes narrow to achieve our target risk profile, but the results speak for themselves:
In short, we’re raising our upside breakeven and adjustment threshold significantly, while encroaching on downside risk only minimally. Daily time decay increases (from 0.74% of total capital at risk to 0.80%) with only a slight bump in gamma risk, mostly a result of moving the center of our risk curve closer to the current underlying share price. In addition, we’ve restored our positive delta bias almost precisely to our ideal target, and our probability of profit at expiration has actually increased.
So why a call/call double-calendar? Volatility skew. The ratio of front- to back-month implied volatility in the 135 puts is well under 0.9; in the calls, it’s over 0.94. This bit of esoteric detail doesn’t always prove useful in the long-run, but at any given point in time, trade decisions have to be based on what appears to give us the greatest advantage.
I’m still double-checking July’s net returns—at the point where we closed the last calendar spreads, the software risk profile was showing a much greater percentage gain than I’m getting from the raw numbers. Obviously, one has to be wrong, and I’ll run backtests tomorrow to determine where the error lies.
And, with any luck, we’ll be safe from having to adjust at least through tomorrow’s close.
†Average fill price.