Skip to content


  1. MG
    June 29, 2020 @ 4:57 pm

    So per prior comments from you here and on BigERN’s site, you mention you’ve just stuck to buy-and-hold the index as it’s so easy yet tough to beat. However, this study appears to be the holy grail of them all; something that isn’t tied to a single stock, yet beats the SPX both in absolute and risk-adjusted terms.

    I get your point about time selection and the upcoming study based on a more limited time period; but if that study comes out agreeing with this one, wouldn’t we all be stupid not to just adopt this strategy?

    Also, huge props on another great study, and on your site in general.


      June 30, 2020 @ 10:19 am

      Thats a good question. On paper, yes, if the data suggests short (7 DTE) and ultrashort (0-3 DTE) short puts outperform buy/hold SPY then we’re all better off adopting the strategy.

      Two caveats come to mind: tax implications and why these option strategies outperform.

      I intentionally don’t speak to tax drag since everyone’s scenario is unique and thus generalizations aren’t terribly helpful. For me personally, the ability to defer capital gains is substantial. Combine this with the ability to use margin loans as a form of tax arbitrage ( and there’s a material “intangible” advantage to the buy/hold approach.

      As for why short-duration strategies outperform, my first thought is that these options are simply miss-priced. There’s very little [public] data on these strategies. Now that this info has been publicized, any edge on the strategy may fade away as retail and institutional players do their own due diligence and implement.

      The counter argument is that market dynamics will work themselves out such that the strategy will almost always outperform, similar to how indexing almost always outperforms despite the massive shift to the strategy over recent years.

      A happy medium might be to allocate 50% of one’s portfolio to the put writing strategy the other 50% to buy/hold.

      Thanks! Happy to hear it’s helpful. Don’t hesitate to reach out with any ideas / suggestions for future studies.


  2. Wes
    July 3, 2020 @ 9:10 am

    Thanks for the post. One question Ive got is when you are measuring the return on buy/hold SPY, are you accounting for dividends re-invested? On the compounded annual growth rate bar chart based on the SPY bar it looks like the return is around 8%. If this hasnt been accounted for over this same timeframe, how does this change the comparison between buy/hold vs the 50D, 7DTE, hold to expiration?



      July 3, 2020 @ 10:18 am

      Welcome! Yes, the buy/hold SPY numbers assume all dividends are promptly reinvested (total return).


  3. TFJ
    July 27, 2020 @ 12:23 am

    I’m having a lot of trouble following your curves. The summary here states All 30D and 50D strategies as well as 16D hold-till-expiration outperformed SPY, but it looks like only the dark green and light blue (50D) beat SPY based on the last P/L curve chart. Am I misinterpreting? Thank you.


      July 29, 2020 @ 5:16 pm

      The summary is based on the data in the section titled “Results.” This includes the overall P/L chart.

      The P/L chart in the “Discussion” section is based on a crude review of changing the start date of the study so as to avoid timing luck.


  4. Edwin Jose Palathinkal
    November 24, 2020 @ 1:13 am

    When you calculate the losses from immediately selling assigned shares, do you use the opening price of the next day?

    This is because according to the The Options Clearing Corporation’s Characteristics and Risks of Standardized Options, Page 51, Section named “Assignment”:

    “Assignments are ordinarily made prior to the commencement of trading on the business day following receipt by OCC of the exercise instruction”.


    “It is possible that an option writer will not receive notification from its brokerage firm that an exercise has been assigned to him until one or more days following the date of the initial assignment to the Clearing Member by OCC”

    I haven’t been assigned, I just need to know if the backtest accounts for it.


      November 25, 2020 @ 8:30 pm

      Great question! Losses are calculated on the day in which an option position expires ITM, not the day in which OCC logs an assignment event. Since price snapshots are taken only once daily at 3:46pm ET, the value of the option at 3:46pm is what’s used to calculate P/L. Price action of the underlying after 3:46pm ET the day of expiration is not accounted for and is an acknowledged gap in the methodology.

      In other words, assignment events are calculated as a buy back of an ITM option 14 min before the option expires. If we ignore the 14 minutes of time value and subsequent price action of the underlying between 3:46pm and when OCC determines assignment price, the P/L is the same.

      Rule of thumb: one can plan to realize ~80% of a backtest’s results in practice (not just mine, but anyone’s).


  5. Dan
    March 4, 2021 @ 1:22 pm

    What do you mean that all strategies after March 23, 2020 did not recover? From the visual chart, it looks like all strategies bounced back. What did I miss here?


      March 8, 2021 @ 8:39 am

      If you’re referring to the observation that none of the strategies recovered from the max drawdown as of the end of the backtest, that’s indeed accurate. The max drawdown occurred on March 23 2020 and none of the strategies have reached or exceeded the high-water mark associated with the max drawdown.


Leave a Reply

Your email address will not be published. Required fields are marked *