Put Diagonal Backtest (Part 1)

Because the put diagonal fits together in a portfolio with this type of strategy, I spent a couple hours backtesting it. Today I begin discussion of related comments and impressions.

I will start with the strategy guidelines:

1. Buy one ATM LEAPS LP in Dec of following year.
2. Divide cost by total number of weeks to get TEV.
3. Sell ITM put in shortest-term expiration available (M/W/F) at highest strike whose EV > TEV.
4. At expiration, roll SP to the higher of strike determined via (3) or strike just closed.
5. If the market has rallied 10% from LP purchase then then roll up and out to ATM Dec closest to two years thence.
6. If one year has passed since LP purchase then roll LP out to ATM Dec closest to two years thence.
7. Execute with midprices and charge $18/contract commission.

My usual approach for dealing with slippage is described in the fifth paragraph here.

In live trading, I regularly deal in options priced $15-$35. I sometimes have to let the trade work for a few minutes, but I usually cave no more than $0.10 for two legs.

The current backtest should factor in more slippage because it:

• Trades options up to $200-$400 in price
• Trades options sufficiently DITM as to have little open interest and virtually no volume
• Includes 2020 crash conditions with fast-moving markets and wider bid/ask spreads to offset price uncertainty

$18/contract, which includes $2 commission and $0.16 slippage, is significantly more than my average realized slippage of under $0.05/contract. Furthermore, I close out each option in the backtest because OptionNet Explorer (ONE) does not have an “Expire Options” button (another issue I have with the software). I am effectively charging $0.32/contract for each short option because in live trading, I could take assignment of ITM options at expiration with no transaction fee whatsoever. I would therefore be saving the $0.16 and, on the rare occasions when options are OTM, the remaining premium as well (ONE usually displays OTM options to be valued at $0.025 just before expiration).

Overall, I think I have done a decent job simulating slippage. I’m a bit light on the LP, but I believe most of these trades would not incur anything close to $0.32/contract. To satisfy the extreme naysayers, I could certainly increase slippage a bit more (e.g. to $0.20) and say “peace!”

I will start to discover the reality when I actually go live. For now, let’s move on.

Despite the guideline defined by (3), I was somewhat discretionary when choosing what strike to sell in the backtest. I sometimes took the fat premiums when IV kicked up and I could still remain seemingly DITM (e.g. 1.5 SD or more).

Why?

I will continue with this dilemma next time.