Passive income through option writing: Part 3

Back in 2016, I wrote a few posts on trading derivatives, especially options, to generate (mostly) passive income:

I’m still running that same strategy but it definitely evolved quite a bit over time. This might be a good time to write a quick update on what I’m doing and what I’ve changed since then. And for everyone who’s wondering what’s the use of this: I’m planning a future post on how selling options may help with Sequence Risk, so this is all very, very relevant even for folks in the FIRE crowd!

So, let’s take a look…

A quick recap

Selling put options exposes me to the worst possible return profile: I have (almost) unlimited downside risk, i.e., if the price of the underlying drops below the strike price then I lose the difference between the underlying and the strike price. But I have only limited upside potential; even if the underlying goes up by 100% (which is unlikely with an index over a short time frame, though) I only make the option premium, see the chart below:

putwriting-payoff-diagram2
From my earlier option post: A short put option gives you a cushion when the market drops but it limits your upside potential. If the premium is high enough this can be a highly profitable and low-risk “investment!”

Most people think that the return profile sounds unappetizing. It’s exactly the opposite of what most investors desire: unlimited upside and insurance against dramatic losses (positive skewness). But here lies the strength of this put selling strategy: Since it’s the opposite of what everyone desires the option premiums tend to be higher than the “fair” market price. It’s like a casino: people who frequent the casino pay a small fee to play and have a small probability to win a big payday (=buyer of an option) but the truly profitable business is being the casino (=selling insurance) not gambling in the casino! Same logic with the lottery. Or buying an extended warranty. Or buying insurance. It’s profitable for the sellers of those services!

So, show me the money! How would this strategy have performed in the past? Selling puts on the S&P 500 would have performed pretty well over the last few decades. You would have even outperformed (!!!) the index since 2000, see the chart below. Well, the recovery post-2009 would have been better with the S&P 500 index, but over the longer horizon, two bear markets and two bull markets, the put selling did phenomenally well!

PutW chart01
Cumulative (nominal) returns of the S&P500 (total return) vs. writing at-the-money put options once every month. Source: CBOE, see here and here for the source data.

Now, I don’t want to stress the outperformance of the put writing strategy too much. That’s because even performing in line with the S&P 500 or even slightly underperforming the S&P would have still been pretty impressive. What is truly amazing about the put writing strategy is that you generate equity-like returns but you do so with:

  1. roughly one third less volatility (~10% annualized vol, compared to about 15% in the S&P 500)
  2. significantly smaller drawdowns (i.e., drops below the all-time-high up to that point), and
  3. shorter-lived drawdowns, see the chart below:
PutW chart02
Drawdowns are much more shallow and shorter-lived when writing puts. This is much preferred from a sequence risk perspective!

That’s a huge advantage from a Sequence Risk perspective! That’s why I think selling options is a great strategy, especially for (early) retirees! If you’re still very early in the accumulation phase, sure go ahead and go crazy with equities. As I showed a few weeks ago, even a (temporary) drop in the stock market can be beneficial in that situation. But if you’re in retirement or close to retirement you’re much more concerned about Sequence Risk! So, in a previous post (Part 2), I detailed what exactly I am actually doing in my portfolio, which is slightly different from the simple short put strategy benchmark published by the CBOE:

  1. I sell puts that are “out of the money,” i.e., with a strike price a bit below today’s underlying value. The premium is a bit lower than for the at-the-money options but so is the volatility.
  2. I use some leverage to overcome the lower premium revenue.
  3. I use shorter-dated options.
  4. I invest the margin cash in higher-yielding bonds and also more tax-efficiently (Muni bonds). The CBOE PUT index assumes you invest the margin cash at a short-term (e.g. money market) rate.

But since I first wrote about this, here are some additional updates:

1: The account size is much larger!

When I first wrote the option trading posts in 2016, I already ran this with a six-figure account size. Not exactly “play money!” But that account has now grown substantially, due to both capital gains and additional contributions. Now the options trading account is about 35% of our total financial net worth, and about half of our financial net worth outside of retirement accounts. In other words, assuming that we don’t touch our retirement accounts until I turn 59.5, the put writing strategy now has to carry half the weight of generating income in retirement. And just in case you wonder, no, this blog does not in any material way contribute to our retirement budget. So, this is the real deal, not some academic exercise! This is what we actually use to finance our early retirement!

2: The leverage is lower, the premium target & option Delta is lower, and the “loss allowance” is larger

Retirement, especially early retirement, plays tricks with your mind! Suddenly, you become much less comfortable taking risks. So, when you run this strategy with some serious pile of real money and also without a day job to make up for potential losses you get a lot more cautious:

Leverage: I used to run this with roughly 3 to 3.5x leverage. Now I’m down to around 2 to 2.5x leverage. And again, there was some confusion about what exactly I mean by leverage. It’s very straight-forward to calculate your leverage ratio when you’re buying an asset. You have $100, you get a $40 loan and purchase $140 worth of assets. Your leverage is 1.4x. But how do you do this when you’re shorting a derivative? My preferred method is as follows. If I sell a put option with a 2,800 strike and a multiplier of 100 then the notional value is $280,000, also equal to the potential loss if the market were to drop to zero. If I have $125,000 per short option in my portfolio then the leverage is $280,000/$125,000=2.24. Notice that the $125,000 is about 5 times larger than the minimum margin requirement mandated by the exchange (roughly $25k). It’s always a good idea to keep way more margin cash as a cushion to avoid becoming a victim of margin calls, which is what wiped out the hapless OptionSellers.com investors.

Premium Target: The further out-of-the-money you write your put options the lower the premium. But the risk of losing money is also lower. So, I currently target an option time value of around 5-5.5% p.a., a little bit lower than the 7% I quoted in my posts from 2016. And again, the way I calculate this percentage is the annualized premium divided by the notional value. So, if I can make a $300 option premium per week and the notional value of the option (multiplier times strike) is $280k, then the time value yield is about 52×300/280000=5.57% (for an unleveraged position). The gross return is obviously larger once we apply leverage!

Also, different option traders use different methods to pin down their strikes. Delta or how many standard deviations (sigmas) you want to be out of the money. My strategy of targeting a certain yield is very close to targeting a 5 Delta or around 1.5-2.0 standard deviations below the current index level. Not every single transaction but over long-term averages.

Loss Allowance: I used to budget a loss allowance (i.e., how much of my gross premium revenue I lose – on average – when options go into the money) of around 50-55%. I’ve increased that to 60%. So in other words, out of every $100 of gross option revenue I budget I’ll keep $40. 2018 was my worst year so far, but I still managed to stay close to the 40% profit margin even after that horrendous February volatility spike. The overall average since 2012 was indeed 55% but, again, out of an abundance of caution I want to budget only 40% profits.

PutW chart03
Actual profit margins from selling puts 2012-2019. Note that in 2019 I haven’t suffered any loss (yet)! I wonder how long that’s going to last!? 🙂

So, what’s my expected option trading return? Simple Math. Let’s assume 2.25x leverage, 5.25% option yield and 60% loss allowance. I’d generate a net profit of 2.25×5.25%x(1-0.6)=4.725% annualized return. What? All this effort for such a measly return? Aren’t we supposed to earn 12-18% from investing in stocks? Hold your horses. First, your expected return from stocks isn’t 12-18%. Consult Dave Ramsey to get out of debt but fire him as soon as you get to a net worth of zero and listen to people who actually know finance to grow your assets! Second, the option revenue is only part of the equation. The beauty of the option writing strategy is that this is all done on margin! So, in other words, for every short Put option you also have another, say, $125k lying around to “play” with, i.e., to generate extra income. Theoretically, you could even invest that $125k in your standard 60% Stock, 40% Bond portfolio and use the short puts to make an extra 4.7% p.a. in addition to your 60/40 portfolio! For my taste, though, that would be loading up on equity risk a little too much. Especially considering that I already have tons of equity holdings in our other accounts. So, I keep my margin cash in a more stable portfolio. But since I first wrote this post I’ve certainly gotten a little bit more adventurous with my margin cash, which brings me to the next item:

3: Taking more risk with the margin cash

When I first wrote about this topic, I held most of the margin cash in Muni Bonds mutual funds. I have since transitioned over to a slightly more adventurous (=riskier) allocation

  • Muni Bond Funds (e.g. ABHYX): 30% of the portfolio. The yield is pretty decent! About 3.5% p.a. and that’s all tax-free!
  • Muni Closed-End Funds (e.g. NZF, BAF, etc.): 40% of the portfolio. These pay just above 5% in (tax-free!) interest. This comes at a cost, though: They are much more volatile (due to leverage!) than the lower-yielding plain-vanilla muni bond mutual funds.
  • Preferred Shares (e.g. ALLY-PA, GS-PK, MS-PI, STT-PG, etc.): 25% of the portfolio. All my preferred shares are floating-rate (or at least currently fixed, then transitioning over to floaters at a future date) to hedge against the risk of eventual interest rate hikes. Currently, the weighted average yield is just about 5.75%. Most of this is treated as (qualified) dividend income, though some also pay ordinary interest. The issuers may be very solid companies but make no mistake, Preferred Stocks are quite a bit riskier than your typical bond. They are called preferred stocks, not preferred bonds!
  • Cash balance: 5% of the portfolio. Interactive Brokers pays around 1.50% interest on unused cash balances and this is obviously ordinary income.

The weighted yield on all of the above is just about 4.6% p.a. Not too shabby! So let’s call that a 9% total return for the whole shebang; options plus margin cash returns. Subtract 2% inflation expectations and we’re at 7%. Not that bad! It’s in the same ballpark, even a little bit higher than long-term equity returns (6.7% real return) and much more than what I would expect conditional on being 10 years into a bull market. And the option strategy has lower volatility and better-looking drawdowns than equities! Case in point, October through December 2018 when I actually made money with this!

Also, I’m fully aware of the irony here. Over the last few weeks, I just finished my mammoth project, analyzing (mostly dismantling) the “Yield Shield” strategy. In case you’re not familiar with this, some other bloggers proposed the Yield Shield to eliminate Sequence Risk by investing in higher-yielding assets (spoiler alert: it doesn’t work!!!). But please note the big distinction here: I increase the yield in the fixed income portfolio knowing very well that this also increases Sequence Risk. I have no illusion (delusion?) that this would ever help with Sequence Risk. But I believe that the option-writing strategy actually has lower sequence risk than a plain equity portfolio so I can afford a little bit of extra sequence risk from my preferred shares.

4: Trading SPX options instead of S&P500 E-mini futures options

Over the years, people asked me why I would trade the futures options and not simply the SPX index options. Simple answer: while working in my finance job, that was the only asset class I was allowed to trade without preclearance from the compliance department. And by the way, my former employer was incredibly generous because I know a lot of friends at other banks that were strictly prohibited from trading any derivatives products (index options, futures, futures options). So, in any case, after leaving my job I eventually transitioned over to trading SPX options and never looked back because there are numerous advantages:

Lower commissions

One ES contract (50x) costs $1.42 in commission. That’s $2.84 for the equivalent of a 100x SPX contract. I pay a commission of around $1.24 per SPX contract. That doesn’t sound like a big difference but trading three times a week, this adds up to around $250 a year. It may not sound like a huge deal but considering that I budget around $13k to $14k in gross revenue per year from one single short put option and about $5,000 in net revenue (after paying for the occasional losses when the short puts expire in the money), then an additional $250 in fees per year is just too much!

Better margin efficiency

Both instruments, ES puts and SPX puts require around the same amount of margin. Roughly $12,000 for the ES put and $25,000 for the SPX at twice the size. But the SPX options are more margin efficient in the following sense:

I can use my Muni bonds funds, ETFs and Preferred shares as collateral for my SPX options. I’d keep a pretty slim cash cushion (maybe around 5%, or $6k out of a $125k of margin cash per option) ready to deal with a 60-point in the money loss. But I can keep the overwhelming majority ($119k) of my margin cash in income-producing assets, see top panel in the table below.

Not so if I held short ES futures puts. Interactive Brokers maintains two subaccounts, one for “securities” and one for “commodities.” If you hold short ES Futures options, IB will move over the margin requirement into the commodities subaccount. This could create a negative cash balance in your securities subaccount (see middle panel) if you hold too much of your margin cash in ETFs/MFs. That raises two issues: 1) you’ll start paying margin interest and 2) the IRS rules that your income from ETFs and MFs is no longer tax-advantaged (either tax-free or ordinary dividends) but “payments in lieu of dividends” which are considered “ordinary income” for income tax purposes! I guess the IRS doesn’t like it when people buy tax-advantaged assets on margin.

The only way to avoid the negative cash balance is to hold less money in ETFs and mutual funds, see the bottom panel. But that also lowers your income potential from the margin cash! One of the reasons why I prefer SPX options over the ES put options!

SPXoptionsMarginEfficiency
Because SPX index options are domiciled in the “US Securities” subaccount, the same as my Mutual Funds, ETFs, etc., I need very little idle cash sitting around to satisfy my margins. ES Futures options require more idle cash to satisfy margin requirements!

The convenience of cash-settled options

If the ES options end up in the money I will then end up with a long ES futures position in my account (=physical delivery). The only exception is the third Friday of March/June/September/December when the puts expire exactly on the same date and the same time as the underlying. So, with futures options, it’s my duty to sell the ES Future if short options get exercised. In contrast, with the SPX options, I simply see a debit for any option that ends up in the money. There are two advantages to cash-settlement. 1) I avoid the cost of getting rid of the long ES future and 2) I don’t have to sit front of my screen at exactly 1pm Pacific time and make sure I don’t end up ES futures (at 2.5x leverage!) right around market close. What if I miss that and the ES futures keep going down in after-hours trading? What if I have a medical emergency and I end up in the hospital and I can’t get rid of the leveraged futures position for a few weeks? The cash-settled SPX options are much easier to handle!

The only disadvantage: shorter trading hours

The only disadvantage of SPX index options is that they trade only during market hours (9:30am-4:00pm Eastern time on weekdays) while the ES future options trade even outside of market hours. You can even start your trades on a Sunday afternoon (Pacific time). But I’d never do so anyway because there’s normally very little liquidity outside of NYSE trading hours. Even when trading ES futures options I’d do so during normal trading hours 99% of the time. So, for me personally, it’s not a real disadvantage at all!

5: Trading three times a week (and sometimes four!)

When I initially wrote about this strategy I would sell options every Friday with an expiration the following Friday. Currently, I write options three times a week for the expirations on Monday/Wednesday/Friday. In other words, every Friday, I sell options expiring on Monday. Then every Monday I sell options expiring on Wednesday and – you guessed it – every Wednesday I sell options expiring on Friday. And if the last trading day of the month falls on a Tuesday or Thursday you get another expiration that week for a total of four trades that week!

There are (at least) two advantages to writing shorter-dated options. First, you have a lower probability of getting dinged by a sequence of bad days. Some of my worst losses with this strategy came when the S&P dropped four or five days in a row (think August 2015, January 2016, December 2018). With shorter options, you have the potential to avoid the big losses because you constantly “roll” your options and sell at new strikes. If the market is in a rut and keeps going down, you also move down your strikes over time. This helped me to actually make money in Q4 of 2018 because the S&P 500 fell “slowly enough” to almost never breach my strikes!

HitByMomentumChart
Writing shorter options is a hedge against a sequence of bad-luck shocks. In the top panel, we wrote an option with 4 trading days to expiration. The market drops three out of four days and we get to $13 in the money after day 4. In the bottom panel, the index stays above the strike of the option after two days. But due to the drop (and the likely rise in implied volatility), we can sell the next option with a far lower strike and avoid getting dinged from a continued fall in the index! The index dropped but we made money with both options. Sweet!!!

The second advantage is that more independent bets imply that the Central Limit Theorem has more power to work! Huh, what does that mean? As I previously wrote, selling put options generates a very (negatively-)skewed return profile: limited upside and unlimited downside, i.e., the kind of return profile that nobody wants because it’s so much at odds with most investors’ preferences that are biased toward positively-skewed returns! But even the most badly-behaved statistical distribution becomes more and more Gaussian-Normal (that nice, symmetric bell-shaped distribution) if you average over a sufficiently large number. Here’s a little numerical example:

Imagine you earn an option premium of $1 and with probabilities, 95%, 4%, 1% the option value at expiration is $0, $5 and $30, respectively. So, you earn $1 with 95% probability and you lose $4 and $29 with probabilities 4% and 1%, respectively. That’s a very negatively skewed distribution. If we simulate 20,000 samples of the average returns over 1, 10, 50 and 500 draws then the distribution of average returns over those 1, 10, 50 and 500 draws becomes more and more Gaussian-Normal, see below:

CentralLimitTheoremChart
Even a skewed distribution looks more and more Gaussian-Normal when you average over enough independent observations! 20,000

So, even a very unattractive and negatively-skewed distribution becomes better-behaved if you diversify over time. The more independent bets you can take the less scary the average distribution will look like! That’s why I try to take as many independent bets as possible, i.e., I use the shortest possible time to expiration to maximize the number of tries every year!

10_DM_Serie4_Vorderseite
Don’t you miss the old 10 Deutsche Mark note? It had a picture of Carl Fridrich Gauss and a small figure with the Normal distribution named after him! Source: Wikimedia

Of course, there are also (at least) two disadvantages of trading more frequently. First, it’s more work. Well, is it really? I’ve never timed exactly how long it takes me to do my trades every Monday/Wednesday/Friday. But last week we went skiing and I did my option trades on my Android phone while sitting in the chairlift for a few minutes. It’s not a huge time commitment! Of course, being a total finance geek I spend way more time in front of the screen looking at finance charts. Some people follow the Kardashians. I follow the S&P500 and VIX! But it’s not for everyone!

MVIMG_20190320_132851
Sitting in the chairlift and enjoying the view (Mt. Hood, Oregon’s tallest peak in the background). And doing a few SPX option trades on my Android phone, too!

The second disadvantage is that you have more chances to get “whipsawed.” It’s basically the opposite of the bad, bad, bad weeks mentioned above. Imagine the index goes down on Tuesday and Wednesday and breaks through your strike price. You lock in your loss on Wednesday only to see the index recover on Thursday and Friday. But when the index recovers you only recover the option premium, which may be way lower than the loss on Wednesday. I consider it the cost of doing business and I much prefer it over sitting through the scary negative momentum events like February 2018 or December 2018 with long-dated options!

HitByMeanReversionChart
Hit by Mean-Reversion: Sometimes you lock in the loss with the shorter-dated options. If the market reverts quickly you would have been better off selling the longer-dated options!

Side note: I’m not saying that this is the only way of doing this. A reader of this blog recently put together a nice all-in-one how-to post on how he intends to implement the options trading strategy.

Conclusion

This might be an exotic topic for the fans of the Safe Withdrawal Math posts. But I’m not a one-trick pony, so please bear with me! Also, I’m a big fan of put writing mainly because I think it has the potential to alleviate Sequence Risk. Not eliminate it, just alleviate it! So, stay tuned for a future post where I’d plan to combine the crazy exotic options trading stuff with the more mainstream FIRE, “how-to-deal-with-Sequence-Risk” stuff!

Hope you enjoyed today’s post! Please leave your comments and suggestions below!

Picture credit: Pixabay

149 thoughts on “Passive income through option writing: Part 3

  1. Love the article Karsten. I think this is the easiest to understand out of all your Options articles.

    Your posts are always actionable and get me motivated to do something. I’m going to switch from ES to SPX tomorrow.

    Just to clarify regarding the margin for SPX. This is the initial and maintenance margin? It’s not that you’re trading in margin, right? Just because IB has removed margin accounts for us Aussies .

    So I can keep almost all my initial and maintenance margin as non cash instruments instead?

    1. Thanks!
      For the SPX short puts, it’s around $27k initial and about $25k maintenance. But to be honest, I don’t even pay that much attention because I’m so far above the minimum.
      Not sure, though how the margin works for non-US investors.
      Best of luck!

  2. Very much appreciate your derivative posts. I was wondered if you have considered taking advantage of the momentum factor? With your put strategy you are still taking on equity and bond risk premiums. While cross sectional momentum would take on the same risks, time series momentum has been uncorrelated to both equities and bonds historically. This is available in the managed future space with high fees, except maybe with AQR. But otherwise relatively high barriers to entry for the individual investor. In theory, adding the uncorrelated asset to a bond/equity allocation or as adjunct to this strategy would boost return and lower volatility further, and have a much bigger impact on sequence of return risk. I am consider trying to run this in my own asset allocation but don’t have a finance backround so have some knowledge barriers to implementation. Any thoughts on this strategy?
    Here is a white paper to clarify
    http://docs.lhpedersen.com/TimeSeriesMomentum.pdf

    Thank you again for all your engaging writing and continued posts.

    1. I asked Karsten this on twitter a few days ago. He said this was on his to-do list. Great minds thing alike….

    2. Thanks for the link. Yes, I’ve thought about this issue the equity momentum is clearly a nice hedge against the catastrophic losses during the big bear markets. It will help with Sequence Risk!
      The nice thing about momentum is that it requires very little finance background and essentially no data subscriptions at all. You can derive the signals all from pretty easily available, zero cost sources.
      For equities, I did the exersice here:
      https://earlyretirementnow.com/2018/04/25/market-timing-and-risk-management-part-2-momentum/

      Not sure if the average investor can pull this off for the large specrum of available assets they have in the AQR paper, though!

  3. Nice post. Strange but it’s when I had the least amount of free time (while working full time) that I found time “playing” with options. It was quite exciting at times (I took more risk that you). I stopped doing options when I FIREd to simplify my life (even though it didn’t actually get simpler because then I started investing in small businesses, which was even more stressful). Fully agree that in the long run it’s better to be the casino. I still do like playing craps in Vegas though!

    1. Good points! You’ll still have to reserve some mental bandwidth to the trading, which is not what all FIREd folks want to do. But the risk profile (lower drawdowns and more consistent income) is much more appropriate for retirees than working folks! 🙂

  4. Another great post. Just a few questions/comments:

    1) I assume you are not managing your trades early with either a stop loss or profit target?
    2) What about the tax complications of SPX vs. futures? I thought that was a big issue for you?
    3) Why can’t any of these put writing funds like PUTW outperform SPX? Even in high IV periods they still seem to lag when you would most expect them to shine.

    Thanks!

    1. Obv not the guy you asked. But futures and SPX are treated by IRS the same way as a 1256 contract. So they are equally tax advantaged compared to equity options like you would trade for spy for example. PUTW would be a little different of a strategy they trade monthly at the money options. Also I don’t think they are leveraged. So they would have a different return profile and be less responsive to rapid changes in Volatility in theory. Though I suppose that depends somewhat on how they manage implementation. I don’t trade this strategy so I haven’t really been watching but it seems to me that volatility (looking at VIX) has been relatively low on a historical level with relatively short term spikes. I suspect that recently they just felt the pain but didn’t get the gain due to this.

    2. Great questions:
      1: The time to expiration is short enough that I’;ll just let them expire, potentially in the money.
      2: Same tax treatment: Section 1256, i.e., 60% Long-term, 40% short-term gains/losses.
      3: It’s hard for the PUTW to outperform when the S&P is going up as strongly as it did during the last few years. In the next bear market the PUTW will “outperform” (by losing less) again.
      And even in the sideways moving markets, the PUTW can get whipsawed if the drop is too short-lived and the bounceback to fast (think December 2018 and Jan 2019). So, the PUTW can struggle to keep up with the S&P at times.
      That’s the reason I do my adjustments to the option selling strategy as outlined here on the blog.

  5. Karsten,
    Great post – keep these coming. Could you elaborate a bit more on your mechanics:

    1) If you have a trade on, for example, that expires on Wed (you entered it on Monday) – If it is in-the-money, do you always close out the trade or will you take the assignment of SPX? I assume you always close it / buy it back before close on Wed.

    2) Do you ever pause on the strategy if the IV, IV Rank and/or VIX are too low to warrant the risk of the trades?

    Regards,
    Dave

    1. Great questions:
      1: I never close the trades before expiration. The time to expiration is so short.
      2: I never completely pause. But if the IV is too low I’d be willing to sell at a lower option yield. I’d like to be at least 1-1.5% out of the money, which could mean you have to temporarily lower your yield target if the market is really, really calm.

  6. You’re going to get tired of reading my comments today but I promise after this to wait for answers from the others before commenting any further. 🙂 I have four questions/comments and I hope it holds my spacing:

    1. Do you have any links (or posts of your own) on how to invest in Munis [funds] and what to look for with regard to preferred stocks?

    2. What do you mean when you say the put writing is done “on margin?” It’s not as if you still have all that cash available to invest. For example, if I sell a naked put and incur PMR of $20,000, then that decreases my buying power by $20,000.
    It is that buying power I use to make other investments (like buy bonds, stock)… so what do you mean, exactly?

    3. I think you can do better than $1.24/contract on SPX. I pay just over half that (including exchange fees), and I would think you could get the same rate I do. Maybe you’re happy enough with the brokerage to pay the higher rate.

    4. Why not just roll losers down for even (or slight credit)? If the market keeps going down then you’ll get more and more negative but isn’t it either pay (the piper) now or pay (the piper) later? And should the market reverse, your unrealized loss may never get realized. As long as you keep contract size constant, you won’t fall prey to the leverage demon (e.g. Martingaling).

    Great blog, Ern!

      1. I was talking about rolling at expiration rather than closing and starting a new trade at expiration.

        The TT study definitely suggests an advantage to closing losers early. I would have liked to see numbers on largest drawdown and maybe distribution of drawdowns to get a sense how much pain one might feel (and whether Net Liq–also a function of leverage–would remain positive) in the extreme cases. 17 rolls for them would be roughly six weeks trading 3x/week like Ern does. I think Ern has said his largest drawdown had him digging out for a few months so this is somewhat in the ballpark.

        Managing losers is anathema to the TT guys, which makes it interesting that they end up favoring just that in this study. Thanks Spintwig!

        1. Agree. These TT studies are interesting but only tell have the story. Wish they would at least include a benchmark such as SPY in their result graphs plus had information about volatility, drawdown, and sharpe. Without that we have no idea which strategies actually do better on a risk adjusted basis..

          Probably a long shot but I am thinking about paying some grad student with access to OptionMetrics data some money to go in and re-create most of the better TT studies so I can just see the real data.

          1. Agree.

            They have other studies that depict the performance of selling premium vs SP500. Unfortunately, just like the above link, the data isn’t available for peer review / scrutiny (or if it is I’m not aware of it).

            In my cycling experience every “aero” wheel manufacturer has wind tunnel stats that “prove” their wheel experiences the least amount of drag and is superior to the competition. Obviously they all can’t be #1. The data is sliced and diced just right to demonstrate the outcome they desire.

            When I worked for a healtcare constancy we had clients that were hiring us to help them get their new drugs to market. One such activity included getting the drug on formulary. A tactic to make that happen was to prove it had a lower total cost of care vs incumbents and pitch the research to the health ins directors that determined which drugs made the list. We would slice and dice the data every which way till Sunday to find a way in which the drug would outperform.

            To the same extent, I sometimes wonder if TT is going through this entire effort of research, live shows, their own trading platform, etc. for the sole purpose of marketing active trading and collecting the trading fees and margin interest where applicable. They’re not sponsored and don’t have advertisers, so they’ve gotta be making their $ and incentivized somehow. Counter argument is anyone can open a Robinhood account and execute the same concepts 100% free. Either TT is onto something or they’re an excellent marketing machine.

            I’d go in on funding that.

            1. I agree 100% with that last paragraph, Spintwig. The potential conflict of interest is an elephant in the room. When it comes to trading one’s own money, I really don’t believe anyone should take “published studies” for what they say without understanding the details of how the data were generated (or looked over the process themselves). The level of shadiness in this industry is, unfortunately, quite high.

            2. spintwig.com.

              I would like to offer up a comment in response to your speculation that TT might be focused on increasing trading volume as a “money grab” business model and that their research might not be objective. Consider some facts that run counter to what you suggest. 1.) TT have ALWAYS been against trading in the weeklies. Their central theme is trade out around 56 days and roll with around 21 DTE. This creates an average trading frequency of 1/4th that which could be done if they ‘manufactured’ their research to support using ‘weeklies’ and incresing trade acitvity/commisions. 2.) It would be very easy for them to make a persuasive pitch for use of weeklies since several of the TT principals actually developed the first weekly options for the CBOE (Tom Sosnoff and Tom Preston at least). I have been trading options for over 20 years and have found the TT research to be a a valuable and objective deep quantification of what I have experienced in those decades. Finally, I doubt seriously that Tom Sosnoff and some of the other founders need the hassle of starting a tough business and working all day at it. The sale of Think Or Swim to TD Ameritrade years ago set them all up well. I think you can lean into the “on to something” train of thought.

              1. 1.) It is true this approach yields 1/4 of the possible trades when compared to weeklies. However, it could also be true there was no way to slice the data to demonstrate superiority of weeklies. Going back to the healthcare consultancy example, it was on very rare occasion that we would have to eat some hours because a client’s drug was simply not superior in efficacy no matter then lens or angle we used.

                Without having the data available/published for peer review and scrutiny, we will never know. TT just happens to be the example here, but in general I’m skeptical when it comes to unsubstantiated claims. In fairness there’s enough info revealed to portray validity, but I can’t dig deeper. In academia the phrase is “publish or die.” Tom S. often refers to TT as a think tank and mentions his PhDs on staff. Well, publish the data 🙂

                To John’s point, I agree with the interpretation-of-data concern. When I was researching to write my post I encountered a few odd takeaways. I don’t recall the specific reference off the top off my head but on more than one occasion a takeaway would have been “don’t trade a particular way as it’s not optimal or there’s no scenario where ‘x’ is useful”, which would have a consequence of lowering trading volume [on the platform]. I’ve noticed there’s an aversion to saying anything that could lower trade volume.

                2.) I don’t doubt their involvement in the creation of weekly options, but I’m skeptical it’s the trade mechanics that caused weeklies to exist. There’s a market to be made and they have always very adamant about being market makers in earlier videos (the small talk between their actual slides).

                I have no idea what’s involved to start an ETF, but I’ve been curious why such a vehicle hasn’t been created to implement the same mechanics they, and myself, trade. Just like there are target-date funds with 5-year increments, I’d expect to see target-delta funds with 5 or 10 delta increments.

                Stand it up mechanically as an ETF, demonstrate its effectiveness as an S&P500 outperformance tool, whether on absolute terms or on a risk-adjusted basis, which should be an easy feat with all the data they have, and collect a modest expense ratio. Granted, one of their principles is moving away from money managers and being self directed, but I’d argue there’s still a market for such an ETF. I mean, PUTW exists after all.

                With all this skepticism, I’d like to reiterate I regularly use their mechanics in my own trading, it’s profitable and repeatable, and from what I can observe thus far is optimal, too.

                Thoughts?

                1. I agree 100% with the second paragraph. I’m skeptical, too, because of all the statistics (step 1) they don’t report and the inconsistency (step 2) of what metrics are reported from one episode to another. And, FCOL, PhDs! While the ideas are great and many of the numbers sparkling, what they present is two steps down (as just explained) from what I would expect as PhD (i.e. peer-review) quality.

          2. I agree, BradF22. There’s always a number of statistics I wish they _would have_ included in order to get a sense of whether they curve-fit studies, biased them, etc. I couldn’t even replicate many of their studies because they don’t give enough information on the “methods” slide. I’ve questioned them on multiple occasions and gotten different answers about things.

            With regard to your grad student idea, I’ve considered (and even gotten one detailed quote thus far) having an automated backtester built. I’m also looking at a couple available platforms. If you’re seriously interested in this work then get me some contact info and let’s talk about it. I’d also be happy to have Ern (or others?) along for the ride.

            1. Hi Mark. I may have a few people interested and I work at a university that has access to these data sets so I may be able to put together something. Not sure if emails will come through on here but you can reach me at FACHBC AT Gmail Thanks.

            2. I built a back tester in matlab a year or so ago, but at the time didn’t pony up for real historical data (didn’t realize you could get it when I was writing the program). I just calculated option prices based on the vix (adjusted to try to approximate shorter term OTM options’ IV) and the price of the S&P500 using black-scholes. I since found that you can buy historical option data for SPX for ~$400 if I remember correctly, so the barrier to entry is not too high. In terms of programming difficulty, the hard part in my opinion is making something that allows many different possible strategies to be simulated. I just kept modifying my code as I came up with new ideas to try or new metrics to look at, but I also knew all the caveats and which methods would and wouldn’t work together since I’d written everything. But making something that just works for any combination of scenarios is much more work.

              My problem with Tasty Trade was they frequently interpreted their results incorrectly, or at least not the way I would. For example, they had a study where they ‘showed’ that selling 1 month puts outperformed 1 week puts, but that conclusion was only reached after they used 4x the notional leverage for the 1 month puts (their explanation – because you’re selling the weeklies 4x in a month, so you have to sell 4x the monthlies to make it even). Of course then you’re taking 4x the risk with the monthlies, so of course you make more in dollar terms.

              1. Thanks! I should probably look into getting some data. Is that optionmetrics or some other provider?
                Agree on the TT issue: Looks a bit like a black box to me. As they say, “Don’t trust any statistic you haven’t falsified yourself” 🙂

                1. I’ve looked into iVolatility.com a couple times for data. I think there are multiple potential data sources.

              2. I know very little about programming, John, but I have been interested in having an automated backtester built for some time. I wonder why different strategies create such havoc for the code. The backtester itself is basically a glorified adding machine (track differences in option prices from trade inception, multiply by 1 or -1 depending on long/short, and multiply by number of contracts). Selecting different strategies could direct the program to what legs are required and, perhaps, what restrictions should go with those legs (optional?).

                I agree with TT’s faulty interpretation of results from time to time. At least twice, I have caught them basing conclusions off total return when risk-adjusted (something they emphasis often with regard to volatility of returns) return says otherwise.

    1. Thanks! Great questions! All very relevant!

      1a: Not really a website. I use Nuveen, Blackrock and Invesco for the closed-end funds. American Century for the Muni mutual funds. You’d have to customize this to your own needs and check for the muni funds that are tax-free in your state (e.g. CA, NY have a lot of customized funds).
      1b: Preferreds: https://www.preferredstockchannel.com/ has all the pertinent info I like to see before buying any preferred share.

      2: I mean that I can use the account value (cash, mutual funds, ETFs, etc) as collateral. You don’t have to hold the margin cash in idle cash.

      3: What broker are you using? I always thought IB was the cheapest for ES/SPX puts.

      4: It goes back to the central limit theorem. I want to get a fresh new draw 3x a week. If I were to hold on to a large delta position after a loss I’d expose myself to time-varying market exposure. I don’t like the large Delta after a big drop and the risk for further losses if negative momentum prevails.

      1. I use TD and get charged 0.$77 per option trade which includes SPX, but I had to ask for that. With your volume and size of account it likely would not be a problem for you to get. It’s actually quite a bit cheaper than /ES for me – I had only been able to negotiate that down to $2.37 per trade (including exchange fees), but you need 2 /ES to equal 1 SPX put. When you take into account those fees plus the extra cash to use with SPX (I get 2% for my cash sweep), plus the slight disadvantage I usually see on the bid/ask spread with SPX vs /ES, I calculate I get a boost of ~0.4% in return per year using SPX for how I trade.

  7. Hey Karsten,

    If you don’t mind sharing, what’s your portfolio theta as a % of net liquidating value? For example, theta of 500 on a 1MM portfolio would be 0.05% (5pbs). I’m sitting around 1bps.

    -Mitch

  8. The funny thing about this is that your use of margin is what actually gives me the most concern rather than the put writing strategy. Some of these closed end funds have pretty wild swings and people have been talking about munis defaulting for years. I assume you have researched this all and still feel comfortable with your use of margin. Would love to see a post on just this. Sorry for making you work so hard in your retirement.

  9. Wow this is a very extensive post and I am grateful that there are people like you out there to share it.

    I might have to do some research but maybe this is something I can potentially look into here in Australia with our own indexes.

  10. Interesting strategy Karsten. Looking at your diagrams you’re selling puts with an 80 strike, so make money if it finishes above that (or even slightly below taking into account the put premium) but have unlimited downside beyond that.

    Have you thought about hedging by buying a 60 strike put, ie having a short put spread trade on? I’d be curious as to what impact it would have on profitability, presumably the implied vol is higher at the lower strike so it would cost a bit more on the vol side but it’s further OTM so still cheaper and provides that downside protection.

    1. I have considered using the vertical spread like you described for hedging the ultimate downside risk. Haven’t really pulled the trigger on it because the hedging cost over time would have been way more than I ever lost in Feb 2018.

      1. Doing it as a put spread would definitely add costs, it does cap the downside risk though. I guess the question would be which is more important, maximising income from the strategy or minimising the risk. I’d be happy to pay a bit more to cap the downside risk of just being short a put, but there’s certainly a case to be made that it’s pretty unlikely to be an issue.

        1. Good point. So, I am defintiely considering doing a long-dated long put far out of the money as a tail hedge. I’d also consider long VIX futures as a tail hedge. Would have worked very well in Feb 2018! 🙂

          1. The two main things I’d be thinking about with a long-dated long put position would be that a) ideally it would be an american option so you can’t get hit on your 80 strike short puts and not be able to get out via your 60 strike long put and b) it’s a good hedge for tail risk but because you’re paying a lot more for time value than you would be with the shorter dated put you’d be looking at with a vertical put spread it may be fairly expensive. On the other hand you’ll have less theta decay for a while at least so maybe it works out fine.

            I’m sure you’re on top of it already though!

            1. a: If I’m long the put option then I wouldn’t have to worry about AMerican vs. European, but not sure I understood you correctly.
              b: the long-dated puts are more expensive in avsolute terms but cheaper on a per day basis (theta), so the calendar spread is actually net positive income. But again, I’m not even doing this yet. So, I have no super strong opinions on this one! 🙂

          2. I’m going start implementing your strategy of writing short term naked SPX puts on Monday. I just got approved for options trading through Fidelity on Friday and they gave me 500 free trades, so I can try this out for a couple years before moving to a different broker. Monday will be my first option trade ever and I am a little nervous about down side risk with the CAPE Shiller over 31. I am really liking this idea of also doing a long-dated long put far out of the money (probably at least 10% OTM) to protect against a Black Monday type event that could crush even 2 day expiration puts. I would think I could buy it fairly cheaply with the VIX having fallen so much recently, but I guess I’ll find out on Monday.

              1. I decided to buy 1 put with a one week expiration and a strike price 10% OTM which costs $25 per week. The extra $1,300 per year will certainly eat into my profits, but I feel much safer knowing I will lose at most $25,000 on anyone trade. Losing an entire year of expected profit would be a setback but not total devastation.

                I have $190K in a municipal bond ETF earning about 4% per year plus another $10k of cash in my brokerage account. My plan initially was to just sell one SPX option 3 times per week which would keep my leverage below 2X. However, now that I am hedging the options with a long put, I think I can afford to double up and sell two SPX options with each trade. Obviously, I would then need to buy 2 puts which would double my hedging cost to $2,600 per year. My maximum loss on any one trade would increase to $50K, but I think I can afford to increase my leverage to about 3X since I have the hedge in place.

                I have an additional $550K with fully marginable equity ETFs in my brokerage account, but I didn’t include that in my leverage estimate since those equity ETFs would be the last thing I would want to liquidate if a Black Monday type event happened. Do you think hedging the SPX trade affords greater use of leverage?

                1. Sounds like a good idea. 3x leverage seems not so bad when you hedge the extreme downside!
                  Also agree: since you don’t want to liquidate your equities in the other account after a Black Monday (or Tuesday or Wednesday) event it’s best to not include that other account in the leverage calc.
                  Best of luck!!!

              2. I’ve been buying two 10% OTM puts on SPX every week for the past two weeks as a hedge on the put selling strategy and everything seems fine, but I want to run a different idea past you. I think I will get more bang for my buck buying calls on the VIX. Buying the SPX puts costs about $2600 per year and I would still lose 50K if the market dropped 10% or more over any two day period, so I could still get beat up pretty badly.using my current hedging method.

                But if the market fell 10% or more I would expect the VIX to go through the roof, so I think it might be a more effective hedge to just by some VIX calls. I’ve been considering using VIX calls with expirations 6 months out. With the VIX so low right now, it seems like a pretty inexpensive way to hedge the SPX put selling strategy (at least for now).

                If the VIX spikes any time in the next 6 months, my plan would be to sell the VIX calls for a likely profit and then stop hedging the SPX put selling strategy altogether until the VIX drops again. In a high VIX environment, the 5 to 10 delta SPX puts would have strikes so far out of the money that I wouldn’t be as worried about being completely naked on them.

                1. I like the VIX approach. I think it’s much cheaper tail risk insurance than buying puts. If anyone has experience doing this for an extended time period (ideally spanning a few years) please share your experience! 🙂

              3. By the way, I have a new margin account with Fidelity who recently approved me for options trading, but they don’t offer futures contacts trading. Fidelity gave me 500 free trades when I opened the margin account, so I don’t want to switch brokerages any time soon. I would prefer long VIX futures over VIX call options, but I don’t have that option right now.

                1. Trading options on VXXB might be a way around the VIX futures limitation.

                  Just to confirm, you’re trading SPX for free as part of the 500 free trades deal?

                2. Yes, I can trade SPX for free until my 500 trades run out and they don’t expire for two years, I currently make only 4 options trades per week, so those free trades will last for a long time. After that, I certainly will have to find another broker unless Fidelity will give me a huge discount on their rate of $4.95 per trade. Ouch!

                  Thanks for the suggestion on VXXB call options as an alternative!

        2. I think this is a great research question. If total return is decreased but volatility of returns is decreased even more, then it might be worth doing. The spread might also increase duration of trade, though, which could make this analysis more difficult.

  11. This is a stupid question: Can SPX options only be exercises at expiration? Even for two day options, what happens if it becomes in the money on the second day, and then out of the money on the last day? Would option buyer exercises it on the second day?

    Thanks!

    1. Correct – SPX options are European and cannot be exercised before expiration. The alternative would be American, which CAN be exercised before expiration. SPY, for example, trades American-style options.

      1. Thanks for the reply!
        For short puts the Euro vs. American issue makes no difference. For calls on individual stocks there may be an advantage of exercising before expiration if the expiration is right after an ex-dividend date, but that’s not really applicable for the SPX index options. 🙂

    2. In the specific strategy being discussed here early exercise is an almost non-existent risk. If the short puts are OTM or ATM (or nearly so) the existience of extrinsic value makes it counter productive for the buyer (assignor). In the case of ITM options there is still extrinsic value to make early ex a distant threat and if the options were sold at a very low delta and they become ATM or worse ITM you can surely bet vol will have expanded greatly and as result extrinsic will have plumped up as well. L.S.S., not much to worry about with early exercise in any SP related puts. And in any event its a small admin. task to unwind the assignment which has not changed your risk exposure at all.

  12. I have trouble with your metaphors of being the “casino” or “insurance company” because the whole principle of those is that they are giant institutions that pool risk. An individual can’t literally sell home insurance because they would go out of business the first time someone’s house burned down. I can’t claim to fully understand the details of your strategy, but it seems to me that if this is actually a good strategy, then these must be misleading metaphors. Could you flesh out where the metaphors breaks down?

    1. 1) Casinos and insurance companies can sell contracts with negative probability-weighted expected value because the buyers of those contracts either receive other benefits (fun, comp drinks) or have no other options to reduce the risk of financial ruin or comply with external requirements (laws, mortgage terms). I’m not sure the options market works that way. You are always trading against a supercomputer that calculates it can buy your puts for a couple of pennies less than expected value.

      2) Casinos and insurance companies maintain thousands of bets at the same time. Statistically, this means it is extremely unlikely that a far-above-average number of insured homes burn down or that 1,000 people in a row win at blackjack. If an insurance company only insured, say, 3 houses, or if a casino only had one customer playing one round, they would be at risk from sequence of returns risk even if they sold their deals with positive (for them) expected values.

      So yes, the metaphor is problematic for an individual making only a few bets at a time, playing against a supercomputer and not comping its drinks. This is mainly a problem of scale.

      1. Actually in the case of casinos the main point is that they change the nature of their entire risk potential by way of house limits. This allows them to operate in the 1 to 2 sigma range with a positive expectation and never have to ‘pay the piper’ on the tail risk events (big winnings). As far as the insurance metaphor is concerned you can drill down into the ‘number of polices’ issue but at the end of the day it is a correct metaphor as the seller is literally accepting a premium to insure the option writer from some level of loss experience. The issue of bet volume is a detail that does not serve to diminish the appropriateness of the metaphor

      2. Very true! That’s why the scale is important and you only do this with proper risk management. I take 150+ plus “gambles” throughout the year and make sure that none of them individually is so catastrophic that I’ll never recover from it.

    2. Thanks! I disagree, though.
      Option selling is like selling insurance. Instead of selling insurance to 1,000 different households, I sell insurance over 1000 different trading windows. Statistically it’s the same. Even though the option selling is more sequential (more in line with the sequential process in the casino). So, I don’t see why the metaphors would break down here.

  13. Greetings, Ern. So what is the overall effect of the interest on margin loans for those instances that your puts go in the money? Is it negligible because, irrespective of interest rates, you pay it off promptly (say, the following business day)? And what about the cash to repay those margin loans; does this originates from, say, the sale of equities from an outside account?

    And, oh yeah, many thanks for this splendid write-up! You have a way of making the complex understandable without watering it down.

    Eric

    1. I never have to take margin loans in my current setup. I even make a little bit f interest income on the idle cash.
      One exception would be a sharp drop where the index drops by 60 or more points below the strike. In that case I’d very briefly have a negative balance and I’d sell some of the bond funds to raise cash.

  14. Seems the goal is minimizing sequence risk.have ya considered purchase of index options with the payout from the munis ? Currently costs about 5% for a year of SPY at the money. If ya get 4% from municipals, then max loss is 1% … no sequence risk. Some older simple spread sheet run from the 80s show ya need to replace about 75% of the market returns with a 0% loss.

    Thoughts?

    1. Very intriguing strategy, for sure. Not sure if the numbers add up.
      During calm periods (i.e. today) the time value of the ATM call is closer to 6%. Much higher when the crap hits the fan as late last year.
      Also, regarding the 1% max loss: you also have to finance your withdrawals and compensate for inflation. So your max loss isn’t 1%.
      Let’s look at the following scenario: you have $1,000,000, you keep $900,000 in munis, buy $60,000 of ATM calls ($1,000,000 notional) 1 year out and withdraw $40,000 for your expenses.
      After one year, if the SPX moves sideways you’ll have:
      $900,000 in munis
      $36,000 in interest
      =$936,000 portfolio value, which is $917,647 after 2% inflation. That’s -8.2%. If you go through 3 flat/slightly-down years in the S&P500 and 2% inflation, you’ll feel Sequence Risk even with this strategy.

  15. DONEAT53 you have made a most important point here. The put selling strategy that incorporates any leverage gearing actually has the potential to increase sequence risk as the D.S. profile becomes steeper at exactly the wrong moments. I always felt the 3x turns of leverage in ERN’s initial approach was exceedingly high and I still feel 2 to 2.5 turns may become an issue at some point but that is only my .02. More to the point I believe that amelioration of seq. risk is not nearly as meaningful in a put selling strategy (which, by the way, I do often so I am not opposed to this strategy) then it would be if one was to work on the long call side of the market where you are working for U.S. and truly attenuating D.S. in a meaningful way when it is most needed (large moves down.) The real problem is that being long calls (and call premium) is just another way of being in a married-put position and paying up for the put insurance which statistically is a large drag on performance. If one wants to truly reduce seq. risk and not have the drag they need to construct positions with greater than 1X upside potential and at the same time carve out a large hunk (but not all) of the initial D.S. It is not as easy as it seems but very possible especially if using LEAPS (long) on the call side and shorter duration further OTM puts (short) on the downside. In this way you have a very good chance of ‘financing’ the LEAP call inventory or even being Theta positive while removing some large slices of risk on the D.S. All that said I have for decades been a put seller (especially at elevated vol levels) and it’s one of my mane strategies.

    1. Disagree with the sequence risk. Even in Feb 2018 I only never lost more %-wise than the unleveraged S&P. That’s because I lose 0 until the strike price and only after going through the strike it becomes 3x (or 2x or whatever you use).
      The main issue of Sequence Risk, as the name suggests, is a long sequence of bad returns, i.e. negative momentum over several quarters, even years, so selling short-dated puts will alleviate SoRR even with moderate leverage because you reset the strikes so frequently. I always thought that long-duration short puts are much more scary and risky than my short-dated puts with some leverage.
      But for the faint-hearted I certainly recommend selling the ATM puts with a short duration and 1x leverage.

      I agree that long calls are expensive: put-call parity means you can’t win from long calls when you lose from protective puts. I doubt that LEAPS are a panacea either. If they were everybody would be doing it.

  16. Didn’t say the relief from Seq. risk was not in effect just that it has very real limits at which point it actually becomes a greater problem (vis a vie the gearing). As regards LEAPS, I will clarify my point by saying the financing costs in a ‘combo-ish’ position of long LEAP calls and short slightly OTM shorter duration puts (in smaller quantities) will give you a very high probability of fully financing your position with less D.S. exposure. You can be short fewer puts relative to the long calls and have small positive theta. In this case your net disposition is longer on the U.S. then you are on the D.S. (both the long calls and short puts are long ∆ positions). This gives you a toe-hold in a convex risk exposure and some time to work the position to increase that favorable relationship and constantly reduce risk. August 2011 allowed for one of the best experiences with this structure as it began a fairly long period of elevated volatility in the near term (selling puts rich) and not nearly as much of an elevated vol in the LEAPS (term structure of volatility at work). I ran this very trade from 8/11 to 3/14 and I am sure you would find the results very good especially as regards the nature of the net risk during the trade’s lifetime. Just another .02 for whatever it’s worth. btw… the SWR series is some of the best stuff out there and I recommend it often and highly. Great work there.

      1. ERN,

        Not sure you want me to post an excruciatingly long comment/reply so I’ll be kinda brief. That said happy to share any data or answer follow-ups. First to answer your questions on ∆ targets, yes there is targeting but the strikes are less important than the net ∆ disposition they leave you with (both contemporaneously and projected out in time). Also, the trade structure was born from my desire to hold a large long-term SP500 position but with a better risk profile-especially on the D.S. The objectives are 1.) Have an U.S. profile that is at least (if not better) than a vanilla SPX/SPY holding (long with the ability to get longer). 2.) Attenuate/truncate D.S. especially in large moves (this can only be done with a smaller (short put) contract count than what would be held in the equivalent vanilla long position. 3.) Since LEAP calls are being purchased, at a minimum the extrinsic (ATM and OTM it’s all extrinsic) MUST BE financed through selling other legs. Sold options can be both OTM calls and OTM puts, but again strikes are less important than the net ∆ disposition you have going forward. The biggest factor I consider is what the daily cost of the LEAPS are and that becomes my financing goal in the short calls and short puts, again keeping an eye on net ∆s. When I launched this trade I was replacing a 4000∆ long position (via Vanguard SP500) and I wound up long 120 OTM LEAP calls (about 4% OTM at that time). This gave me just a bit more long ∆ than the legacy ETF position giving me room to sell some OTM calls and hold on to required long ∆ (especially if the market moves up). Put sells were always much, much smaller than the equivalent 40 contracts (full notional). This takes management but can be made very mechanical. Hope this helps.

          1. I’d be interested in hearing more about this as well. It sounds like you (Joe) are trying to get leveraged upside potential (spending the entire account to buy LEAPS leverages by a factor of current market price divided by price of the option) with an eventual downside breakeven of the current market price (similar to buying stock)? You would still lose if the market goes down though, right?

            1. Kinda sorta 😉 It is true that in this strategy I become levered to the upside if the market climbs as that is what long option ∆ does. But, the core priority in this strategy is to carry less ∆ on the D.S. This is achieved through at least two factors. 1.) the long calls become less long in down moves. 2.) The short puts ( not always open and not static in position size) NEVER give you more maximum long ∆ on the D.S. then the size of the vanilla SP500 position. So, for example, in the description I supplied above the original ETF position equated to being long 4000∆. This means that the combined ∆ of the new position (which can be comprised of call spreads-with many fewer short calls than long and short puts.) The desired position should leave you long 4000∆ (at launch and longer on the U.S.) and less long on the D.S. It’s about having a convex relationship to the market movement. I also make a high priority of 1). Structuring the position so that initially it has a great attenuation of risk in a fast/big move down and 2.) Being long theta so that I can chip away (sometime aggressively) at the original risk basis at trade launch (finance the position). I would point out that a hard/fast rule is that there is never any potential to lose more than the vanilla SP position, so in response to your comment of ‘spending entire account’ that is not true here…never be levered to the D.S. My 2011-2014 trade started with about 40% capital outlay and was reduced to nearly 14% after the first year (aggressive call selling and sustained high volatility). While this degree of risk reduction is not common (unless you have those same volatility regime) it is an achievable goal to be constantly removing originating risk from the trade. And finally, yes p/l experiences loss behavior in D.S. moves but less so and at a decreasing rate if the current short put leg is small enough… Ain’t no free lunch.;)

      2. TD Amereitrade’s ‘Think Or Swim’ platform has look-back function for historical option pricing and volatility/IV levels. The only fly in the ointment is that it is E.O.D. data and there are, from time to time, intra-day moves that allow for much improved management trades. The differences may not be great but there is still a bit left on the table when back testing withe E.O.D. only data.

        1. Interesting. I don’t think that intra-day data for simulations is that easily managable for long-term simulations, so I’m happy with that. And totally agree: in real-life you could probably do a little bit better by managing positions throughout the trading day.

          1. I’ll play devil’s advocate on this point. I think any trade guideline you use to manage positions intraday for larger winners could also be whipsawed for larger losers should the market recover by EOD. Unless there’s reason to think the former would outweigh the latter, why unnecessarily complicate the backtest by involving intraday data?

            1. I should probably give a bit more color on some of the beneficial intra-day trades that can sometimes pop up. The big one has to do with very short term (minutes/hours) volatility spikes that give you better opening trades than the EOD data would indicate. Another one that happens from time to time is in strategies that are managed over a longer period of time (like the one I described at the start) you will often have to adjust positions ad there are times where intra-day moves present vastly improved environments for making those adjustmets.
              As I said it’s not a frenetic day-trading approach but rather an opportunity that shows up from time to time that is not reveal in EOD pricing. That said, I would be satisfied with EOD pricing for back testing and chalk those intra-day opportunities as a small margin of safety/tailwind. I never suggested NOT using EOD data I only pointed out there was some potentiality left unobserved in it.

              1. I don’t know the details of the particular approach you outlined, Joe, but theoretically speaking the tradeoff I would pose is as follows. Monitoring intraday would allow you to catch more spikes (similar to probability of touch being greater than probability of expiring). However, waiting until EOD gives you a better opportunity to catch the largest volatility spikes in the case of volatility trend days. Is there reason to think one is better than the other? I think this also applies to profit (or loss) management guidelines or adjustment guidelines that are implemented intraday (i.e. via alerts) vs. EOD (i.e. checked near the close).

                1. Nailed it on the P/L management guidelines. A position could breach a threshold intraday, triggering action to lock in profits, only the swing to the downside later and potentially be a loss due to a VIX spike.

                  I’d love the ability to set trade-duration-contingent profit takers in a reading platform.

                  For example: if a trade achieves 10% max profit over the span of 1 trading day, take profits, else if a trade achieves 20% in 2 trading days, take profits, else if a trade achieves 30% in 4 trading days, take profits, etc.

                  I do a poor-mans version of this currently, but it’s manual, cumbersome and frankly its too nice outside to be looking at a screen for extended periods of time.

                  Intraday backtesting would support the mechanic of profit takers.

                2. I think too much is being made out of my reference to EOD masking some occasional unobserved intra-day opportunities. I am with SPINTWIG below with “its too nice outside to be looking at a screen for extended periods of time.” Thanks god I can get real-time info and trade on my iPhone. It happens occasionally but like I said it IS NOT a day-trade approach. As far as being good at identifying volatility trends I wish I had that ability but I have to go with the markets as they are presented to me at the moment and trade accordingly. Over the past many years I would say those occasional intra-day opportunities have been a net positive especially since there are often 3 celebrate legs to the position that can be throttled up or back. Before too much more time is spent on this, again I will say it is not a huge factor only one I believed worth a quick mention. I think it comes as no surprise that success requires the market’s accommodation and if you don’t have that it really doesn’t matter. 😉

            2. Just for clarification, what I do intraday:
              I do nothing at all unless there’s an option expiration date.
              On the option expiration date, if the current options that expire at EOD that day are way out of the money (delta=0, time value=0.05) then I already sell the new options.
              If there’s a reasonable chance they they can still go in the money, I wait until right before the expiration time to roll.

              1. Like you, Ern, I used to sell a new option when the value of the sold option’s premium approached $0.01 and its delta and theta values approached zero. But like you, I trade 3x per week, and I noticed that market-moving news (earnings reports, employment and GDP figures, etc.) often occurs just after the closing bell or shortly before the opening bell of the market session. So in order to better move with the market (and reduce the odds of being “whipsawed” by breaking economic news), I always let the option expire and trade at the opening bell of the session that follows the prior sold option’s expiration.

                1. Understand. But I think that the market-making news (earnings after the bell and macro news before the bell) are why people are willing to pay for protection. So, I’m more than comfortable having delta exposure during the volatile times! 🙂

  17. Do you ever cycle through your holdings in the various funds/preferred shares listed? If so, how do you select which ones to own at any moment, and what’s the net effect on your portfolio from the capital gains/losses that come from selling your holdings?

    1. Yes, yes yes! Tax loss harvesting: I keep a list of ~5 extra funds on my watchlist and I would sell the existing ones if they were to drop into loss territory, especially short-term loss!!! Then sell the losers and buy the (very similar but slightly different) other funds to avoid wash sales.

      I did that in December 2018 and cycled all faunds at or close to their lows and realized a ton of short-term losses to offset some of my Section 1256 option trading income.

  18. Hi ERN

    Great post!

    What do you think of ETFs like Wisdom Tree’s $PUTW to achieve something similar for those of us in countries like the UK where there is next to no retail access to the derivative market?

    Thanks very much!

    James

    1. I don’t like the PUTW fund. The expense ratio is too high. And it doesn’t to a good job with the margin cash. The DIY route is much better.
      But again: the simple PUTW strategy has some nice return stats and is a good starting point, but I think we can do so much better!

      1. Thanks ERN that makes sense.

        Just out of interest, have you looked into selling covered calls to try and reduce sequence of returns risk?

        James

  19. THANK YOU THANK YOU THANK YOU.

    I traded a lot in my 20s, because I like trading. Now that I actually have a real nest egg, I like feeling like I am being productive with my play money. Your options position is what I intuitively believed and had been doing for a while using SPY/QQQ. Now, I’m a freaking pro on the Schwab futures desk, with my 500 free trades. Can’t believe I was paying all those taxes and commissions.

    Your series on options is exactly what I believe, and you just implemented far better than I could have thought to. Now, I can execute this on a chair lift too.

    This series might have just taken a few months off my working stiff life, thanks.

      1. This has been a crazy couple weeks, and I thought about sitting it out, but I held the line. It paid off!!! I feel like I cracked the code or something. And trading this on QQQ/SPY was so dumb at my tax bracket. Thank you.

        I like to keep a lot of cash, it’s just my personal preference, so writing puts on that cash makes so much sense to me.

  20. But how wise is to trade Spy options when holding VTI? Doesn’t this mean that you actually doubling your leverage?
    Also, what happens if 6k cushion is not enough? Do they liquidate all VTI position or just portion to restore margin?

    Thanks

    1. In retirement I strongly advise against using VTI as collateral for this option strategy. You double up too much on equity risk. But in the accumulation stage, hey, why not?! It’s not a bad idea to have some leverage, especially early on,.
      WHen you get a margin call they’d liquidate the SPX options, not the underlying ETFs that you hold as the margin cash.

  21. Hi ERN, I notice on Fidelity’s website that as of 4/30/2019 S&P PutWrite Index has 3 year avg return 6.65% vs S&P 500 14.87%. PutWrite does have lower std deviation and this is before any leverage used with PutWrite. But wouldn’t this influence the decision to pursue a PutWrite program on S&P 500 or S&P futures when in retirement? Thanks. Andrew

    1. I’m not surprised that during a long bull market the PUTW underperforms.
      THere’s also a lot of risk or call it “luch of the draw” when you have those short and steep drops (Feb 2018, Oct 2018, Dec 2018), where the PUTW with its 4-5 week long to expiration can be really badly whipsawed.
      That’s why I changed the strategy as outlined here: shorter time to expiration, be more creative with the margin cash, take more leverage, use OTM options, etc.

  22. I find this really, really, fascinating and have read almost everything on your site, so thanks for all of the amazing info you put out. I really hope you get your hands on some historical option prices to backtest your specific strategy compared to the suboptimal PUTW. I would be curious to know how much of the performance and volatility are due to the options and how much are due to the margin bond portfolio. I’ve read Sprintwig’s recent posts about backtesting and from his most recent post it looks to be an almost insurmountable task. I am a finance and investing novice, but I know my way around Matlab, so I wondered if I could use the Black-Scholes pricing functions (blsprice(), blsdelta(), etc.) to simulate the option price chain based on historical SPY and VIX values. Obviously this would only give an estimate for the potential premium of a trade, but at least it should give us the correct behavior and determine if trades are winners or losers, plus it would be free. I haven’t been able to get those functions to give me reasonable strike prices and premiums yet, but I did notice something while I was importing whatever free historical data I could get my hands on. I may be showing my financial naïveté, but it seems like the high-yield municipal bond fund is an absolute superstar. You used ABHYX, and I looked at VWAHX (vanguard’s high-yield tax-free bond fund) and over the last 25 years it has a cagr of 5.3% and annualized volatility of only 3.5%! The SPY only yielded 9% over that time with a volatility of 18%. I have no idea if this is a fluke of the time period (lying with personal finance error #4), but it seems to me that if I can get 60% of the growth with 20% of the volatility then what I should be doing is finding a way to lever into the high-yield bond fund. I’ve got to be making a mistake, though, because everything I can find on leveraged bond funds seems to indicate they underperform horribly. Any advice on where I should be looking for my mistake?

    1. Great idea calculating the options prices in matlab. I looked into a similar approach using excel. The formulas (in excel) work great for ATM options but the calculated values begin to skew as the deltas get farther OTM. The farther OTM the greater the skew. The concept is known as the “volatility smile.”

      If your experience is the prices are not expected / reasonable, it could be the matlab functions doesn’t account for the smile in their formula.

      1. I made a pretty ham-fisted attempt at accounting for volatility skew using a plot of the current IV vs call-delta curve that I found, and I think that is the source of my trouble. The blsdelta() function in Matlab is probably similar to the one in excel. It takes as inputs: spot, strike, risk-free rate, dte, and volatility. The outputs are call-delta, put-delta, and IV. So I tried an iterative approach where for a given strike I would keep calculating new deltas until the change was below a small threshold. For each iteration I would use an updated volatility based on the output IV of the previous iteration and a correction based on the volatility smile. The goal was to be able to set a target delta and then adjust the strike price downward until the delta was below the target. Once I found that strike price I could use the blsprice() function to calculate the expected premium.

        My biggest problem is knowing exactly what volatility value to use. I don’t really understand how the VIX is determined, so it’s probably inappropriate to use that number as the IV of an ATM put, and the Matlab documentation calls for the use of historical volatility, but I’m not sure that’s appropriate either. It’s also probably not valid to assume a static volatility skew, but maybe that relationship could be inferred from VIX. And then you’re always left with the fact that all models are wrong, and only some are useful. It’s nice to hear that the idea isn’t entirely without merit though.

        1. The VIX is computed over all strikes. Very likely, the ATM IV is lower than the VIX and most far OTM options have a higher IV.
          Also the VIX is over a 1-month horizon, mich longer than my options.
          There is also the issue that during stress periodsthe shape of the smile changes. In February 2018 I sold puts with an IV close to 100% even though the VIX was “only” 30 at the time.
          And during calm periods I sometimes sell options with IV less than the VIX, even though they are OTM.

          It’s a messy data problem to simulate this!

    2. Very good points!
      I’d like to do this by trading day and split the % performance into options vs. fixed income. So far, I’ve done it Jan 2017 – current but I’d like to expand it going further back. Your calculation is certainly correct: You get fixed income with a decent return and very low vol (and that vol is uncorrelated or even negatively correlated with stock risk) and then you need much more option income and you beat the index at much lower risk. That’s why I like this approach.
      But again, it would take some work to simulate this further back. Would like to see the performance in 2008, obviously!

  23. I love the shoutout to the Central Limit Theorem! It has to be one of the most under-rated statistical results. And you found a really wonderful illustration with put selling that I had never heard of.

      1. Mr Ern,
        Would you advise non-$ users implementing this strategy? I see issue with the currency risk by holding portfolio bonds/margin in US instruments. Otherwise not many non-US bonds have positive yield. Going 100% ETF solves currency issue, but together with selling options that’s too much equity exposure.

        Thanks,
        Thomas

        1. YES! I would probably hold the margin cash in something denominated in your local currency. This raises the issue of low yields abroad, though.
          Now that said, when yields were very low in the U.S. years ago, I was able to get something like 2-3% yield when going a little bit more “risky” than gov bonds. Not sure what your options are abroad…

  24. Curious if there is a viable methodology to replicate the Sell Put strategy in an IRA; one with Future and Options trading permissions….? Thanks.

    1. Many brokerages allow cash-secured options trading in IRAs. That is, your margin is reduced by the notional exposure of your position as opposed to the typical 20% for naked positions. Trading spreads may ease capital requirements.

      As for futures, IB allows futures trading in an IRA but the margin requirement is 3x or more vs non-netirement account. For smaller accounts, MES is a viable replacement to /ES at 1/10th the size.

  25. Hello BigERN. I’ve been learning a great deal from your blog as I have recently left my full-time job and I’m currently in Southeast Asia dipping my toe into early retirement. I dont have the same financial background and experience as you, so these posts have been invaluable.

    Have you considered additionally opening positions on the call side either by selling naked calls or creating a call credit spread? Seems like this could add additional income with no additional margin required. The only drawbacks I can see are
    1. not enough premium on the call side to make it worthwhile
    2. additional work to open and manage those position(s)

    But maybe there are more downsides I do not see.

    1. I have certainly considered that. You can also directly trade the spreads in IB (with one single B/A spread) if you want to do this on the Call side.
      Lots of people do so, but I have stayed away from that for a number of reasons:
      1: as you said, the premiums are slim on the call side
      2: I can’t get myself to root for the market not to rally. I’d find that immoral/un-American. 🙂
      Additional work wouldn’t be an issue, though.

      Hope this helps! 🙂

      1. One way to consider spreads is to understand options are contracts. They are not the “market” per se’. In an options trade there is a winner and a loser so it’s a different kind of risk than the risk of investing, as you said it’s akin to gambling i.e. making a bet, entirely speculative. The bet on a 3 day “market move” has nothing to do with betting against “the market” since you are betting on volatility and the direction and velocity of volatility, not economic health. I’ve traded spreads in commodities to reduce volatility and simply considered it a risk management tool, not an investment in the underlying.

        This is a great article

        1. All true, but this call spread has a negative delta. And it’s not for just 3 days but I’m doing this constantly, so this feels like constantly betting against the market. I’m also betting on volatility as you say, but the one doesn’t rule out the other… 🙂

  26. Wow, I discovered the blog yesterday and read many articles for several hours. The information is extremely well prepared. I have a lot of fun.

    I am a buy and hold trader and use ETFs. The world of options is therefore new to me. Seems very interesting to me though.

    Last week I sold a first put on Microsoft. Also worked directly. And I earned a few dollars.

    Now to My Questions. I read another blog. The strategy is as follows.

    Look at big companies / market leaders from different sectors like MC Donalds, Facebook, Microsoft, JPMorgan… sell puts when stocks have fallen. For example 5% and they quote at the bottom of the past trading range for example the last 4 weeks. Buy a strike 5% below the current price.

    This approach has advantages:

    Stocks are less capital intensive for the margin.

    If the stock falls, the IV goes up. The premium becomes higher.

    If a stock falls above the standard deviation, a rebound is minimally more likely. I think TastyTrade also proved that.

    How do you rate this approach?

    1. Sounds like a great approach!

      “Stocks are less capital intensive for the margin.”
      Very good point! But I also like the low t-costs of my options: ~$1.25-1.35 per SPX put option (100×2850=$285,000 notional).

      “If the stock falls, the IV goes up. The premium becomes higher.”
      The same is true for SPX options!

      “If a stock falls above the standard deviation, a rebound is minimally more likely. ”
      Stocks have idiosyncratic risk above their market beta. It’s possible that this also mean-reverts. Yeah, possible!

      So, yes, I find this intriguing. Maybe I even try this for fun. I still prefer the index options for me personally. I like the Section 1256 treatment of my index options! 🙂

  27. Hi Ern,

    Once you have mentioned that you closed position early on turbulent day. What was your approach and why?

    Also do you try to sell at higher premiums when IV is high and delta low? If yes, any tips?

    Regards,
    Roger

    1. My approach was “losing my nerve” but right now I try to just let the options expire. So far it worked out pretty well.

      I do sell at slightly higher premiums, exactly in the situation you mention: IV is high (sometimes astronomical!)
      Another route: if the market rallies and options I sold earlier during the day made most of their profit already, I’ll just sell another set of options.

      1. I do find selling issue with higher premium despite high IV as it is high for reason.
        But why would sell options during the session instead of before closing? Not afraid of intraday movement?

        Rog

        1. I’m certainly afraid of intra-day movements. But I also get a little bit of extra yield (or equivalently, sell more OTM for same yield) when selling before the close. Think of this as keeping my market delta constant during the day. The old options are at zero already and I like to get some delta even before close.
          For the math/stats geeks: Spread your risk as equally as possible and your Central Limit Theorem converges faster! 🙂

Leave a Reply

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