Naked Put Study 2 (Part 1)

I recently completed a second extensive backtest on naked puts.

This study incorporated three main differences. First, I sold the first strike under delta 0.20. Second, I kept notional risk constant, which I previously discussed in some detail. Also as previously discussed, I used a large enough notional risk to minimize granularity issues. Finally, I recorded daily account equity with the intent of avoiding the drawdown minimization problem shown here.

As done previously, I will begin with some overall trade statistics. I will not do a side-by-side comparison because position sizing was different this time.

I backtested from January 2, 2001, through July 20, 2016: 3,906 trades total.

     –90.1% trades (3,520) won

          –Mean profit on winning trades: $48,016 [standard deviation (SD) $18,629]

          –Largest winning trade: $160,524

          –Smallest winning trade: $129

          –Mean days in winning trades: 35.8 days (SD 10.3 days)

     –9.9% trades (386) lost

          –Mean loss on 386 trades: $277,108 (SD $341,504)

          –Largest losing trade: -$1,866,403

          –Smallest losing trade: -$529

          –Mean days in losing trade: 43.6 days (SD 9.2 days)

     –Average trade: $15,886 (SD $145,696)

     –Mean days in trade: 36.5 days (SD 10.5 days)

     –Profit factor: 1.58

Mull over these numbers for a couple minutes. What do you see?

One difference I see from the previous naked puts study is that the average duration of losing trades exceeds the average duration of winning trades. I used no stop-losses here, though, which means all losing trades were held to expiration.

The next thing I see is a profit factor of 1.58, which tells me this has a chance of being a decent system. On over 3,900 trades, winning trades made $1.58 for every dollar lost on the losers. The average trade is positive, which is necessary, but the SD is about ten times the average trade.

That’s a tad eye-popping.

Looking closer, the SD for losing trades is about 19 times that of the winners. I would expect some pretty large losers to cause that and I do see the largest losing trade is over 11 times the largest winning trade. I also see the average losing trade is well over five times the average winner.

Now I question whether I could trade this system live.

I will continue the analysis next time.