1. Erick C
    April 23, 2019 @ 9:42 pm

    I enjoyed the historical references to market downturns. One could argue having unused capital is inefficient, but reducing risk is key to any long-term strategy. Investing is like running a business, the opportunity cost of being 100% allocated doesn’t leave a lot of room for when life happens.


    • spintwig.com
      April 24, 2019 @ 1:06 am

      Absolutely! Striking a balance with risk and having reserves or another form of runway available allows the randomness of life (and the markets) to play out. A great approach may look good on paper but in practice there must be allowances for varying degrees of variance. Thanks for stopping by 🙂


  2. Tomaz Korosec
    April 24, 2019 @ 5:05 am

    I wanted to do something similar, but downside risk is too much 🙂 What I did now on my 70.000 EUR account is I just bought 1x ES ATM call option expiring in 330 days (7.700 USD outlay), sold 2x 3200 option also expiring in 330 days (3.800 USD income).This way I participate in any SPY rising up to 3200 level which is quite high level to reach in less than a year and after such run up. Even if it goes to 3.300 level I still make 3.500 USD which is almost 5% return on my 70k capital.. On the other hand, I do not need to worry about downside, because max I can loose is about 4.000 USD. But those 4.000 USD I plan to “recover” by selling 2-7 days out far OTM call options. I can do that selling 5-10 delta options and recover all the cost. So at the end I am essentially hedged for all the downside for free and participate in the upside. As I only have 70k EUR in my portfolio this is also using about 1.8 leverage, because ES is about 145.000 USD…. I am to risk averse to be able to sleep selling naked puts where we saw what can happen in february, december, and that was far from the worst selloffs that can happen. And we are in 10 years bull market already.. I think a risk of huge sell off just increases with each passing month/year..
    What do you think?


    • spintwig.com
      April 24, 2019 @ 11:52 am

      Hi Tomaz!

      Buying an ATM /ES call is certainly a way to capture the upside. I’m assuming when you say “sold 2x 3200 option” you’re referring to the call option.

      The max loss on the call you bought is the 7700. The max loss on the calls sold is unlimited. If /ES goes to 3300 you’ll have hedged one of your call positions, earning max profit of 250 points (3200 strike on short call – 2950 ATM /ES long call) or $12,500. However, the other call is naked and you’ll have a loss of 100 points or 5k, bringing total option P/L to 7500 before factoring in premium decay. If ES goes higher, the loss is greater. In your scenario you’ve traded the downside risk associated with puts for upside risk associated with calls.

      If managing max losses is of chief concern, might I suggest doing spreads? To limit loss exposure you could either buy an accompanying call to “define” the naked one or simply close (read: buy back) one of the short calls. I stopped selling strangles and calls for 1) the psychological reason of disliking the concept of potentially losing money when the market rises and 2) the data-driven reason of not seeing scenarios where selling calls has been repeatably profitable to an extent worth the time required / risk incurred.

      Hope this helps!


  3. Tomaz
    April 24, 2019 @ 2:31 pm

    This is exactly what I wanted. A 10 % upside and very low downside risk. I only opened this strategy because we are in 10th year of bull market. This is for sure not appropriate to do a year after a sell-off etc. But my thinking was, ok we are already up 15 % this year. Incredible. I want to participate in some upside if any more happens, but I want to be hedged if sell-off happens.. Can ES go to 3.400? Sure, not impossible, but in this case I will just make 0 returns because my prognose was wrong or I can roll calls higher. I sold two calls just for this reason alone. To bring break-even point down, to be hedged more and to earn more if SPY does go max 10 % higher from this level. And as you said, if it goes even higher, I gradually start to give back the returns. But I am much more concerned with downside crash that I am with “upside crash”.. This is long term play, short term if markets rise 10 % in a month, I will show loss because my double short calls will gain more than I gain on long call, but I want to play this with year end level in mind. Even if ES rises big from here there will be some profit taking towards end of the year and tax selling etc.. It is inevitable that one picks direction one thinks market is headed and then chooses appropriate strategy to play it. Mine in this case is slightly bullish for next 9 months using 1.8 leverage and hedging downside risk.


    • spintwig.com
      April 25, 2019 @ 4:03 pm

      Keep us in the loop on how it goes! One of the things I’d like to do is setup public paper-trading (or allow hooks into real) accounts on this site so people can execute their strategy while others can see how it performs and get trade ideas for their own portfolio.


  4. MarcB
    April 25, 2019 @ 4:12 pm

    Great read and a great source of information, thanks for the effort. About the leverage:
    1) you mention: “This is why I’m okay with the occasional over allocation of capital via the 15% ladder approach”, what is the 15% ladder approach?
    2) You mention that The ERN portfolio aims to have up to 3-5 short SPX options open totalling 1-1.25M in exposure (one 2850 SPX PUT = 285k notionally) but shouldn’t this be an exposure of 2-2.5M (leverage of 2 to 2.5)
    3) If I am correct: for a 100.000 USD portfolio, you only hold 1 SPY option (100 times 290 =29.000)?


    • spintwig.com
      April 25, 2019 @ 7:54 pm

      You’re welcome!

      1) The 15% ladder approach is a mechanic that says: “I’m going to allocate 15% of my available capital in a trade today.” The practice is repeated each trading day. It’s essentially a way of placing small bets often / over time.

      2) Correct – he’s generating 1-1.25M of equity exposure via SPX options IN ADDITION to the 1M exposure of his portfolio collateral [which happens to be invested in bond-like securities]. If his portfolio collateral was cash – i.e. not exposed to the market – then yes, he would need to generate 2-2.25M in equity exposure via SPX to achieve the 2-2.25x leverage. Leverage in this context is calculated as: notional market exposure / net liquidation value of collateral.

      3) Correct. For each 100k in portfolio value I allocate 20-30k for put writing. Bear in mind writing a single naked 290 SPY put in a 100k portfolio utilizes all capital in a single trade as opposed to laddering 15% of the capital (~4.5k) per day. Spreads would be used to make SPY puts possible on smaller accounts (eg: sell 290 put and buy 245 put at the same time). I probably should have used a hypothetical 1M as opposed to 100k in my example. I’ll update the post later tonight 🙂


      • MarcB
        April 26, 2019 @ 8:59 am

        I see, thanks.


  5. Dan
    June 15, 2019 @ 8:58 pm


    Great write-up! I was wondering if you could offer rebuttals to ERN’s points in:
    “4: Trading SPX options instead of S&P500 E-mini futures options”

    And more specifically, how you manage the funds when they go into the “commodities” account. That seems like it could become problematic.

    Also, does anyone know when options for Micro E-mini will become available?



    • spintwig.com
      June 15, 2019 @ 10:22 pm

      Thanks Dan! Glad you enjoyed it and thanks for stopping by.

      Sure – I agree completely with his points on all fronts. SPX is indeed more commission efficient, margin efficient and is cash settled which, for my purposes, is preferred. My preference would have been for a micro SPX product to be released instead of a micro /ES product.

      When I was exploring trading options on /ES I incurred a few dollars of margin interest due to having a “negative” cash balance in the securities “account” to cover the cash requirements in the commodities account.

      At IB the securities and commodities accounts are logical (they may indeed be “physically” separate behind the scenes but IB does a great job of making everything seamless from the end-user’s perspective). Visually and functionally there is no discernible difference between buying / selling / holding SPX, /ES and their options. Account balance is marked-to-maket, the positions and account cash balance appear no differently — all the cash shuffling ERN highlighted is automated and behind the scenes. There’s an account setting that controls a sweep feature – the automation that manages the account balances between the securities and commodities accounts. I have mine set to sweep-nightly any excess in the commodities “account” to the securities “account”.

      Other brokers, such as Schwab, don’t have a seamless interface and require traders to have a functionally-separate commodities account in which one must manually manage the cash collateral.

      Haven’t heard anything on MES options. I second that question.


  6. Curro
    July 12, 2019 @ 5:14 am

    Hi Spintwig

    I recently discovered your blog and I just wanted to thank you for your hard and systematic work. There is a lot of useful information in your posts, very nice ideas to try and backtest.

    Several comments I’d like to contribute, maybe you mentioned in different posts
    1. You can get VIX info in google spreadsheets with this function =googlefinance(“VIX”, “price”)
    2. Also you can get an excel with all historical VIX prices, 1st and 2nd month futures in this link http://investing.kuchita.com/2012/06/28/xiv-data-and-pricing-model-since-vix-futures-available-2004/ . Free to download, you can also buy the excel with the formulas for a very cheap price (I did some year ago when I was playing with SVXY, VXX)

    3. I like your backtest methodology. I’m trying to subscribe to optionstack, but it seems there is low activity (few discussions in forums, nobody answering e-mails). Did you succeed subscribing or just used the free version?

    4. When trying to backtest options, there are very few automatic solutions. I’m aware of optionstack, CMLviz and ORATS. Only optionstack seems to allow more customized strategies and the 3 of them use end of day prices. You also have onenetexplorer, which have all historical data and intraday prices. It has 30-day free trial, it is worth a try. But the backtest is a bit more manual

    5. All this bullput spreads, or naked puts or even iron condors suffer when you enter the trade in low volatility environment. So, any filter you can add, using VIX or simple EMA50 or whatever to avoid entering the trade is a good improvement.

    Well, I’m totally with the systematic 45 DTE put verticals, in SPY or SPX, but I also like the concept that someone mentioned in the comments about revolving your capital and the idea of, depending on VIX level, select different duration range for the trade. I’d like to explore that.

    Long comment, that’s all, than you againg.
    Best regards


    • spintwig.com
      July 15, 2019 @ 12:13 pm

      Hey Curro – appreciate the kind words and glad to hear you’ve found the material useful.

      That’s a fantastic link / article you referenced! Thank you for sharing.

      I used the free version (all symbols, limited to 1yr of data offset by 180 days). Curious there’s limited activity and responses.

      CMLviz is a new one to me. Will have to look into that.

      Ramping up capital allocation during high VIX environments is definitely a strategy I wanted to check out. The low IV environments were still profitable, such as 2016-2017, but agree there’s more opportunity when VIX is elevated.

      What’s interesting is despite all the strategies and finding optimal mechanics, the data suggests buy-and-hold SPY still outperforms. This past weekend I’ve become convinced to cease allocating 20-30% of my portfolio to options and to instead allocate it to SPY (assuming I intend to continue using leverage / margin). Both the absolute and risk-adjusted returns will be higher vs using the capital for option strategies.


  7. Brian McKellar
    October 19, 2019 @ 2:02 am

    “My preference would have been for a micro SPX product…” — I often use XSP (at IB) for exactly this (even although volume low). It is effectively 1/10th SPX (same as SPY), but cash settled. It is good for experimenting. I am not sure if this is what you were looking for.

    PS: your texts are excellent and I read them all from beginning to end. Sometimes twice 🙂


    • spintwig.com
      October 19, 2019 @ 4:37 pm

      That likely would have been a safer option. I only learned of the XSP product about a week or two ago.

      What kind of spreads do you see on XSP on days when VIX spikes?

      And thank you! Happy to hear they’re helpful 🙂


  8. Cooper Roberts
    April 14, 2020 @ 5:13 pm

    Hi! In part 2 you say “don’t use leverage” but in this article you say you do use leverage. Why the discrepancy?

    And what are your favorite marginable securities? I believe Ern likes muni bonds? Ern says he cant trade /ES with marginable securities through IB so that’s one reason he prefers SPX. But I don’t like that it’s cash-settled cause if it goes ITM I’d prefer to own the underlying and hold it selling calls until it recovers. Do you know if other brokers allow /ES trading with securities as collateral?

    I’m curious about /ES because of its 1256 tax advantage. But I saw you comment elsewhere that that advantage gets negated because of slippage and fees and that you’re just as well off trading SPY—?


    • spintwig.com
      April 14, 2020 @ 8:44 pm

      If the portfolio assets are in cash and the notional exposure of the options is equal to or less than the cash in the account, there is no leverage. However, if the entire portfolio is held in BIL (index fund of 1-3 month US treasuries) which is essentially the same thing as cash, when someone write options against that collateral they’re technically leveraged since there is no literal cash in the account (other than a small buffer so one doesn’t have to keep selling BIL whenever there’s a loss). It’s leverage by definition but in practice it behaves essentially the same as a non-leveraged account.

      The miniseries is due for an update. I wrote those over a year ago and have learned lots and generated much research since then. Specifically, I updated the leverage section just now as a stop gap. The data suggests leverage can actually boost risk-adjusted returns in certain scenarios. Will do a comprehensive update soon. Great question!

      I suppose my favorite is BIL since it’s essentially immune to fluctuations in value (aside from the monthly interest that accrues). Muni bonds, preferred shares and longer-duration bond indexes such as AGG or IEF are alternatives but have the potential to experience correlation with SPY as well as experience material price decreases.

      As for ES, it’s a futures product and is subject to different regulation. It’s this regulation on futures that prohibits using something other than cash as collateral. Thus, the requirement will be the same at all brokers.

      With regard to slippage vs 1256 tax advantage, the 1256 instruments begin to outperform when scaled. The trade fee differential and slippage of the bigger tools will tend to be a drag relative to multiple contracts of SPY. However, there’s a crossover point where the tax advantage will outweigh the drag. One would need to run some calculations based on their unique tax scenario and forecasted returns to identify where that crossover point is.


  9. Coop
    April 14, 2020 @ 10:42 pm

    Thanks so much!!

    One thing I’m not understanding. I see the average 3 year return on BIL is 1.5%. Brokers charge much higher interest rates for margin. Let’s just say 5% for this example. So you’re earning 1.5% in BIL but spending 5% in margin, so netting 3.5% cost to use margin. Whereas if you liquidated all the BIL and used the cash collateral so you didn’t need to borrow expensive margin, then wouldn’t that be much cheaper? You wouldn’t be earning the 1.5% but you’d be saving the 5% margin cost. So BIL+Margin seems much more costly than just having cash and no margin. Does that make sense? Am I missing something?


    • spintwig.com
      April 15, 2020 @ 10:59 pm


      The margin used when opening short option positions is “earmarked.” It’s not actually debited from the account since there is no realized loss; no interest is assessed. In fact, the account balance will actually increase by the amount of the premium received when the position is opened. The broker does this earmarking in order to protect it, and to a lesser extent the investor, from runaway losses. The mechanics are federally regulated under “Regulation T.” This is where the term “Reg-T” comes from when describing margin accounts (as opposed to portfolio margin, a different way of earmarking for large accounts)

      The margin in your example is an explicit debit against an accont such as purchasing shares on margin or withdrawing cash against the account. Your logic is sound, but this is not how margin works in the context of option selling.

      Using Interactive Brokers as an example, the first 10k in cash collects no interest. Further, the interest rate on idle cash is prorated for accounts with a NetLiq of less than 100k. Even more, IBKR’s interest rate on cash (for non-lite accounts) is benchmark minus 1% – this is more expensive than the 0.14% expense ratio of BIL. By purchasing BIL, all collateral dollars earn a non-prorated return with no strings attached.


      • Coop
        April 17, 2020 @ 1:14 pm

        Thanks so much for clearing that up! Makes perfect sense.

        Unfortunately, it looks like RobinHood won’t let me use margin as collateral for put selling…am I understanding that correctly? https://robinhood.com/us/en/support/articles/360001227606/options-collateral/

        “We don’t use Gold Buying Power as collateral. We must hold cash.”


        • spintwig.com
          April 19, 2020 @ 3:49 am


          Over the last year or so since publishing the miniseries RH has been caught with its metaphorical pants down on several occasions.

          That quote is likely due to their most recent mishap: the infinite leverage glitch.

          Updating the recommended broker will be added to the list of upcoming revisions 🙂


  10. Coop
    May 1, 2020 @ 10:31 am


    Have you written anywhere about puts compared to selling put spreads? Spreads seem way way more margin efficient so I’m curious how that would perform in a backtest compared to just selling puts (with the same account size allocation).


    • spintwig.com
      May 3, 2020 @ 9:07 am

      I haven’t compared the two side by side, but the closest thing would be reviewing the SPY short put cash-secured and leveraged studies against the SPY short vertical put spread study.

      In order of increasing total return: SPY vertical put spread < SPY short put cash-secured < SPY short put leveraged.


      • Cooper Roberts
        May 3, 2020 @ 11:19 am

        Thanks Spint! But I see a big problem with the spread study: it assumes commissions. I use RobinHood and there’s zero commissions on spreads. What would the study’s conclusion look like without factoring in commissions?


        • spintwig.com
          May 3, 2020 @ 11:51 am

          At face value it would seem reasonable to simply remove commissions from the model and voila. Unfortunately it’s more complicated than that. Removing commissions won’t actually reflect actual performance of fee-free trading.

          While RH doesn’t charge fees it does participate in PFOF (payment for order flow). In fact, that was their primary source or revenue at the time of launch and for years after. I touched on the topic at https://spintwig.com/retail-broker-business-model/#Payment_for_Order_Flow after the mini-series was published. The takeaway being fees are waived in lieu of execution price being worse.

          If I remove the trading fee I would need to negatively adjust the fill rate to compensate. I don’t have any data on the frequency or severity of fill price drag associated with PFOF so it’s a shot in the dark.

          Given that:

          1) RH announced fee-free options trading in Dec 2017

          2) most large brokers charged 4.99-12.99 or more from 2007 through 2019 per trade

          3) the studies assume a $1 trade fee for the entire backtest duration

          The depicted option strategy performance is overly optimistic [for retail traders] as it underestimates commission drag by 80% or more on 84% of the study’s occurrences.


          • Cooper
            May 3, 2020 @ 12:14 pm

            But since most of the main brokers engage in PFOF the question is, how much worse is RobinHood’s execution in comparison? RobinHood was recently fined over a million for how they used pfof and have since been under more scrutiny…so perhaps their fill quality has caught up with other brokers? RH says this about it: https://robinhood.com/us/en/support/articles/stocks-order-routing-and-execution-quality/


            • spintwig.com
              May 5, 2020 @ 11:24 am

              That’s the million dollar question. It depends whether the purchasers of the trade data are fulfilling at “best price” vs “fastest execution” or some other approach.


          • Cooper Roberts
            May 3, 2020 @ 12:20 pm

            $1 per leg to open and close? So $4 per spread total to open and close? If RH has indeed improved their execution quality in comparison to the other PFOF brokers then the $1 fill rate in the study might possibly be way too high and screwing the results if one is using the strategy with RH—?


            • spintwig.com
              May 5, 2020 @ 11:52 am

              Correct – $4 per spread; $1 to open, $1 to close. If a position expires worthless then no fee.

              From Dec 2017 through present it’s a plausible argument the results are a tad more pessimistic than in reality, assuming RH and the PFOF providers have improved results. Unfortunately there’s no visibility into execution quality so it’s unknown how they compare to a broker that doesn’t participate in the practice. It could be RH went from “terrible” to “mediocre” while non PFOF brokers are “good” or “excellent”. At this point it’s speculation.

              Nevertheless, any potential skew from excessive trade fees from Dec 2017 onward is more than offset by the $5+ trades that were common for the 10+ years prior to Dec 2017.

              For napkin math purposes there are ~252 trading days in a year. All non-0-DTE studies open trades daily, so add $504 per leg (assumes no positions expire worthless) to the total P/L per year starting with 2018 onward and subtract $2,016 per leg from the total P/L per year from 2017 and earlier.

              ( $5 trading cost – $1assumed paid in backtest ) * 2 trades [open and close] * 252

              If you’re interested, I offer paid custom private backtests for a fee. Trading fee is one of the many customizations available. I’ll have more details on this service published on the site soon.


Leave a Reply

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

This site uses Akismet to reduce spam. Learn how your comment data is processed.