Call Me Crazy (Part 5)

I want to continue discussing the long call backtest by understanding what it means to trade with insurance.

Trading with insurance has a few different interpretations. The long call is synthetically equivalent to underlying shares plus a long put (also known as a “married put”). Puts are commonly recognized as insurance, which few people purchase. The long call represents insurance because it controls stock shares for a fraction of the cost. These are two sides of the same coin.

Deleveraging limits loss. In the backtest (see table here), long calls return almost as much as the underlying stock shares for a much lower cost. The capital used to buy the call is the only portion of the account I can lose so long as the call is in play.

Deleveraging complicates performance comparison. Depending on reference, the long-term average stock return is about 8% (1957 – 2018 for SPX). I think many people come to believe they need 8% annually to keep pace with the market. This benchmark, however, implies a 100% stock portfolio. Who does that? A blended (deleveraged) portfolio with 60% stock returning 10% and 40% bonds returning 1.4% (average 3-month T-bill APY over last 20 years) generates an overall return of (10 * 0.6) + (1.4 * 0.4) = 6.56%. I believe many people would be unhappy, thinking this falls short of the 8% benchmark.

Given that mentality, beating the market is an incredibly difficult task. Stocks in the blended portfolio need to return 12.4% annualized to match the headline average stock return! Rumor has it most self-directed and active investors fail over the long-term. An apples-to-oranges comparison of a blended portfolio with a pure benchmark may be one reason why: people investing with greater risk in search of better returns ending up suffering outsized loss.

Those who can’t at least match the benchmarks are told to “dump it all into index funds” or “leave it to the professionals.” Regardless of what assets are included in the portfolio, the appropriate weighted average benchmark should always be used when evaluating returns. Maybe with reasonable expectations, self-directed investors would fare better.

I think significant deleveraging coupled with long-call implementation should somehow make its way into performance metrics. The call is most expensive in Jan 2009 at less than 17% of the underlying index. In 2008, the call loses no more than its then-maximum limit of 20% for the year. The long shares lose much more and could bankrupt the account on any given day. This limited exposure allows investors to sleep well at night. One way to account for this apparent safety is what I called in Part 2 “RAR by MPDD:” CAGR divided by % exposure. This is how I came up with 6.6 vs. 1.1 in favor of long calls.

Might this safety be an illusion? I will discuss that next time.