June 10, 2020
Welcome back to another post dealing with an investing strategy that’s central to our own retirement strategy here in the ERN household. Just a bit of background, about 35% of our financial net worth is currently invested in this strategy. But it accounts for more than 50% of our taxable assets, so for our early retirement cash flow planning, this is really serious business. This puts food on the table in the ERN household!
If you’re not familiar with this strategy, I’ve written about the topic of option writing to generate (retirement) income in general and my personal approach here:
- Trading derivatives on the path to Financial Independence and Early Retirement
- Part 1 – Intro
- Part 2 – Extended Intro
- Part “2.5” – Trading like an Escape Artist: October 2018 update
- Part 3 – Strategy details as of 2019
- Part 4 – Surviving the 2020 Bear Market!
- Part 5 – A 2018-2020 backtest: Guest Post by “Spintwig” (plus a quick update on last week’s volatility)
- Part 6 – A 2018-2021 backtest with different contract sizes: Guest Post by “Spintwig”
- Part 7 – Careful when shorting long-dated options!
- Part 8 – A 2021 Update
- Part 9 – A 2016-2021 backtest: Guest Post by “Spintwig”
The first three links are more about the general philosophy and the last link, Part 3, is about how I’ve been running the strategy most recently. The strategy involves writing (=selling/shorting) put options on the S&P 500 index with a little bit of leverage. And one can also keep the majority of the account in income-producing assets (bond funds, preferred stocks) to generate additional cash flow. Sweet!
In light of the recent market volatility, of course, it would be a good time to do an update on my strategy because I’ve gotten a lot of questions on how that strategy has been holding up during the bear market. Did it blow up? You are all a bunch of rubbernecks, aren’t you? 🙂
Long story short, my strategy did pretty well so far this year. Not just despite but even because of the volatility spike. Let’s take a look…
OK, before I even get into the weeds of the 2020 Bear market, let’s start with a few more general option topics that I always wanted to write about. They will also be a good rationale for some of the trading decisions that I made during the bad bear market of 2020.
One question I get frequently is this one:
What is the “Option Delta” and why is it so important to monitor?
In a nutshell, the option delta can be interpreted many different ways:
- Mathematically, it’s the derivative (slope) of the option price with respect to the price in the underlying, ceteris paribus (all else equal), i.e., while holding all the other parameters constant. In other words, the in the chart below, Delta is the green line and it’s first derivative of the blue line.
- Simply speaking it’s the sensitivity of the price of the option to changes in the underlying price, all else equal (i.e., implied volatility, risk-free interest rate, dividend rate, etc. all stay the same).
Side note: The Delta is also a rough estimate of the probability that the underlying will fall below the strike price (for put options) or rise above the strike price (for call options). For the math geeks, I included the derivation below:
As always, there are limitations:
- For very large market moves you can’t use the slope approximation anymore because the delta itself will change with the underlying (see the green line in the chart above). The rate at which the delta changes is measured by another option greek called gamma.
- Also, for very large market moves it’s not just the underlying price that will change but also the implied volatility. In other words, if the S&P were to fall by 100 points over a short period, your short put option position will lose a lot more than 100 times your portfolio delta, thanks to another other option greek: vega, the change in the option price due to changes in the implied volatility. And, yes, I know, “vega” isn’t even a Greek letter.
- But again, for small enough market moves, the Delta is a rough estimate of how much market risk I’m exposed to.
So, why is the Delta so important? I like to keep my account risk as constant as possible. I don’t want to have essentially zero exposure to the market on Tuesday and then bet the house on Wednesday. Monitoring my Delta is a great help in that task. And I can express that on the individual option level where the option Delta is normally expressed as a value between 0 and 1 (or -1 and 0 for puts). But I can also monitor the overall portfolio Delta exposure, aggregating over all options and also factoring the option multiplier (100x for the SPX index options and 50x for the e-mini futures options). In that case, it’s expressed in $ per SPX index points, i.e., if I have a portfolio Delta of 30 it means, my portfolio moves $30 for every SPX index point. Interactive Brokers provides that figure on the main portfolio overview page on their mobile app.
More info on option, derivatives, etc.: Since a lot of folks keep asking that same question about what option pricing reference book I use, here it is: Robert Whaley’s Book “Derivatives”. It comes with a CD-ROM that has an Option-Pricing Excel Plugin. The book is quite expensive (slightly over $100) but the Excel plugin alone is worth the money. Robert Whaley, by the way, is the researcher credited with developing the VIX index!
Most importantly, the Delta also helps me picking my strikes, which brings me to the next item:
What strikes do I target?
That’s probably the #1 question I get. Again, there’s some science, but also some “art” and “gut feeling” and maybe a little bit of luck involved in this process. There are (at least) 5 different methods to target the strikes based on science:
- Target a fixed yield. I try to shoot for an annualized yield of around 5%. I detailed the math behind that number in Part 3 of this series: with a leverage of about 2.5x and an annual yield of 5%, I’d generate maybe about 12% gross annual yield before losses and probably somewhere around 5% after budgeting for the occasional put option that goes in the money. Together with the yield on the fixed income portion of the portfolio, I’m targeting around 9% annualized return. In any case, if the index is at around 3,000, a 5% put option yield means your options should fetch around $150 a year (per index multiple). That’s just under $3.00 per week, so about a $1.00 premium Monday to Wednesday and Wednesday to Friday and probably a bit lower than $1 over the weekend. You may have three calendar days but only one trading day between the Friday and the Monday close, so you have to adjust for that!
- Target a fixed Delta. I like to stay at an option delta of around 5 or below. (For the purists, this would be a -0.05 delta for the put).
- Alternatively, I can also target a fixed Delta times SPX volatility. So, I would reduce the Delta whenever stocks become more volatile, so as to keep my portfolio risk, expressed as % or $, constant. If you recall, back in March, we had triple-digit moves in the S&P almost every day (and 1,000+ point moves in the Dow Jones) and a VIX above 80. So, during the crazy-high volatility period in March, I was writing puts with a delta wayyyy below 0.05, probably closer to 0.01 to 0.02 to limit the daily volatility.
- Target a fixed multiple of the most recent memorable daily loss. For example, if the most recent worst daily drop in the S&P was 80 points, I’d probably want to be at least 160 points out of the money for a two-day option.
- Target a fixed percentage below the current index level. For example, no matter how low the VIX may be, I’d never want to write a put within one percent of the current market value, at least not when I’m running this with 2.4x leverage.
In practice, I’d use a combination of all five criteria. And sometimes not all criteria can be satisfied all at the same time and then you’d have to make some compromises. That’s where you have to go beyond the science and rely on experience and gut feeling. And a little bit of luck! 🙂
If you want to see an example of my recent trades, here’s a screenshot of my Monday morning (June 8, 2020) trading screen. I had already traded one option with a strike of 3,060, while the index was slightly above 3,200. The Delta is only -0.023 and the time value about 5.1%.
When do I sell the new puts?
That’s probably the second most frequent question I’m getting. For example, when I get up on a Wednesday morning and I have a bunch of contracts expiring that day at 1:00 pm (Pacific Time Zone!), then when do I sell the Friday options? Do I write them at 6:30am that morning nice and fresh at the market open right after I get up? Or do I wait until 12:59:59pm to write the new contracts to make 100% sure.
The answer: It depends. Of course! If my current options are still at risk of going into the money before market close, I’d be crazy to double my risk. And if there’s very little risk of crashing through my put strikes then I have much more leeway in writing the new puts already earlier that morning. And how do I measure how risky my current options are? You guessed it: through the option Delta! So, there’s another application of this option Greek.
Moving from the general options trading philosophy to how this whole process worked in 2020 so far:
How did the options trading perform during the 2020 Bear Market perform so far?
Long story short, the strategy performed as I would have hoped for. In fact, even a bit better than I had hoped. Aside from two very minor losses earlier in January, I suffered a significant loss on February 24 when my puts went about 40 points into the money. Ouch! Considering that I plan to make about $3.00 in premium per week I had just wiped out 3 months’ worth of option premia. Not the worst loss ever but nothing to be happy about!
Amazingly, after the Feb 24 loss, all my puts earned the maximum premium. And the reason is exactly as I’ve detailed in this series before; after a major loss, implied volatility and thus put option prices skyrocket. In other words, …
… insurance becomes more expensive when the house is already on fire!
Which in turn means I can sell my options far enough out of the money (strikes below the current index level) that even with the crazy market gyrations in March 2020 the index never went below my put strikes again. Sweet!
But 2020 was even better than that! Not only did I suffer no more losses, but I recovered the loss within a few weeks by mid-March already. No need to wait for 3 months! And I am now up for the year by about 4.5%. Pretty sweet! Of course, it was still a wild ride. Even though I eventually earned all my put option premia there were a few scary moments in between. For example, on Thursday, March 12, the S&P 500 dropped all the way to 2,480, -9.5% in one day. I had a bunch of options expiring on March 13 (Friday the 13th, oh, my!!!) with strikes between 2,250 and 2,450. But since the S&P stayed above those strikes on the expiration date I easily made all the money back on March 13. Whew!
So, how did I manage to boost my option writing revenue so significantly in March? It’s a combination of (at least) four factors:
- Sell at higher yields: after yields went through the roof I could have simply sold options with my proven and tested target yield. But if I can sell options 20% out-of-the-money over the weekend and make three times my normal premium, then, hey, I’ll take that money! I sold some contracts with a 1,700 strike in March! Crazy times!!!
- Sell more contracts: I strongly advise against jacking up leverage willy-nilly, especially during high-volatility periods. You’d potentially set yourself up for an OptionSellers-like wipeout. Never, ever raise your leverage when things are going against you. I occasionally raise my leverage when things are going in my favor, though. Here’s an example: I sold a batch of Monday options on Friday morning. The market goes up on Friday afternoon and I essentially made 80% of the premium already. I might as well sell a few more Monday options and also still keep the almost worthless close-to-zero-Delta puts. It looks like a risky trade, but I consider it manageable if the overall Delta is low enough.
- Off-day option writing: Suppose you wrote your Wednesday options on Monday. Tuesday rolls around and the market moved sideways or even up and I made most of the premium already on Tuesday. The same logic as in the previous bullet point works here again: with the Delta so low now, I would occasionally sell a few more options for the Wednesday expiration. I had a quite few occasions like that in March 2020 and made pretty good extra income with that.
- Same-Day option writing: Now, let’s go even more exotic! Imagine on a Monday morning, my options expiring that day are already essentially worthless. Instead of selling the Wednesday options on Monday morning, I would then first sell options that expire that same day. And then hold off selling my Wednesday options until later on Monday afternoon when all of my Monday options are “secure.” Extra income! Hey, I’ll take it, especially after the painful loss in February!
Related to the question of how to best recover from a substantial loss, I always get questions like this one:
Why not just keep selling puts at the last strike at which you lost money?
The idea here is that if you simply keep selling options with the last strike that went sour, you’ll eventually recover the loss. You just have to wait for the market to recover again, which it normally does.
But this is a really bad idea! Let’s look at an example where this approach would have backfired catastrophically. And I have to look back no further than February 24, 2020, when my put options with strikes around 3,260 went in the money by about 40 points. Had I kept selling puts at that 3,260 strike, I would have had a wild ride when the index went all the way down to below 2,200 intra-day on March 23. A loss of about 1,100 index points! You take a 40 index point loss, about 3 months worth of premia, and turn it into a loss worth multiple years worth of option premia. Ouch! Had you done that with any degree of leverage it would have likely wiped you out along the way and you never could have recovered after all. The classical Gambler’s Ruin Problem!
Selling at that same strike only works when you experience a quick turnaround. If the market keeps dropping as in 2020 or during any other Bear Market you’re not going to enjoy the ride! And the rationale is – you guessed it – tied to the option Delta again. When the index falls far enough below that 3,260 strike, then the Delta of that put option is essentially -1. That’s 20-times the maximum risk I normally want to take on. Looks like a bad idea to me!
How did the rest of the options trading account perform?
Now for the (slightly) bad news. As I detailed in Part 3 of this series, I’ve shifted over the portion of the portfolio that holds the margin cash to slightly riskier investments. More yield but also more risk. Well, that portion certainly underperformed this year. It’s still down by about 4% year-to-date which drags down the nice positive return from the put selling back to +0.06% as of June 9.
Well, I could have just stuffed the money under the mattress. But I also made a lot more money in prior years thanks to holding the higher-risk/higher-return assets as margin. With the overall portfolio (put writing plus bond funds and preferreds) I significantly beat the S&P 500 index return in 2014, 2015, 2018 and closely matched the index return in 2016, 2017, 2019 and 2020 year-to-date, all at about half the volatility. So, I’m not going to complain too loudly about getting hit in my preferred share portfolio this year. In the big picture, this is still a great strategy to generate income in retirement: great returns and much-reduced volatility relative to an all-stock portfolio!
So, to conclude, before 2020, my strategy had only been exposed to relatively mild volatility and the occasional correction but I’m happy to report that the pure option writing strategy also did very well in 2020 during its first bona fide bear market. Not despite but exactly because of the high volatility! Good to know because there will be more bear markets to come during a multi-decade retirement!
113 thoughts on “Passive income through option writing: Part 4 – Surviving a Bear Market!”
To what extent does trading derivatives affect your SWR? In other words, you’re giving up equity upside for a captured yield.
But isn’t the upside what will create a bigger safety buffer for future bear markets?
Or perhaps it works either way. I have a poor understanding of the strategy but I guess it can also make money when the index goes down as long as it doesn’t fall below your set price. So perhaps it’s improving your target SWR.
Also as a software engineer, I’m always thinking that if the gut-feeling can be translated to maths (and code) then an automated algorithm should be the next step.
Good point. So far I’ve managed to keep up with the equity market even in 2019 when the inde gained 31%+ and I was only about 0.1% behind.
But again: I’m not worried about missing out on winning the lottery. The SWR is determined by the short/medium-term drawdowns. And having lower and shorter-lasting drawdowns allows you a higher SWR. Not I am banking on it but it’s still nice to have it as a reserve.
Yeah, higher SWR definitely makes sense; part 3 of this series thoroughly convinced me on that. The conclusion was just a qualitative “higher SWR” though. I would love to see an analysis of what sort of historical SWR boost would come from moving a portion of equities over into your put writing strategy – perhaps if you’re someday in the mood to write a post, but can’t decide between the SWR series and options!
Yeah, good idea. I have some historical data for the CBOE call writing and put writing strategies, but they are only since the late 90s and only for the monthly ATM option writing. Not exactly the best way of implementing this.
As another SWE, I feel exactly the same! I left another comment on this article with the approach I came up with; check it out (but keep in mind that I’m an amateur learning as I go with this option stuff, and might be spouting nonsense).
On the operational side of things, I will warn you: Interactive Brokers, which seems to be the only reasonable choice, is utter dog garbage in every way other than the financial market backend stuff. Everything about them – their website, their software, their security practices (worst example: when logging in with your username+password through their desktop software you literally have the option to toggle SSL off for some nightmarish reason), their API, their example code – every step of the way, I’m constantly running into something else broken, and having to work around it. Prepare for a serious time investment, which will probably be a bit less fun than your typical side coding project.
Also, from a conversation I had with a support person, it sounds like 2FA is soon going to be fully mandatory. (In typically confusing fashion, the current status is that your account starts with SMS-based 2FA, which you can disable, but not until you first add a mobile device to be used with their app-based 2FA). Why is this a problem? Because they follow the finance industry standard of “our code is too broken to run indefinitely, good thing markets close.” Your automated thing will be logged out weekly, requiring a retyping – yes typing, you have to start up their whole GUI gateway thing again – of your password. Assuming you’re willing to set up automation (that knows the password for one of your financial accounts) to take care of that, this login will then also require a 2FA dance. Since it sounds like 2FA will become mandatory, I just bit the bullet and wrote an Android app to watch for the SMS and forward it to my setup.
So… it’s doable, but I would recommend not starting until you’re sure you’re willing to commit serious time. Sorry for the long post, I think part of me has been desperately waiting for a chance to really vent on IBKR lol.
Oh wow I didn’t realize the gateway was going to require actually typing in the password once a week in the future, that’s going to be a pain indeed.
Yeah, IB has some issues. Customer service is a bit slow/incompetent sometimes. 🙂
Karsten, I can’t think of any questions right now but I’m here to chime in and let you know that I’m your number 1 fan when it comes to these option articles. 🙂
Please get in line and form an orderly queue behind me!
Fully agree they are great articles.
Haha, thanks! 🙂
Thanks, Bob. Always glad to create content for my “two #1 fans,” i.e., you and PJ. 🙂
I hope I can at least crack the top ten!
Sure, there’s space for about 15 more people in the top-10! 🙂
Do you only trade S&P or the individual stocks? I’ve had more gains (and losses) in individual names around earnings.
Of course now with Dave and his Robinhood followers jumping in; Ive noticed the premiums climbing steadily for individual names. Might soon stop my options trading for the year and pick up closer to elections.
Never touched individual stocks. I don’t like the tax treatment and the record-keeping requirement.
Great breakdown of the concepts and decision making process!
On those ~2-Delta SPX positions, do you recall how much slippage exists?
I nromally face a $0.10 B/A spread and set my limit in the middle. Most of the time I get filled at that limit.
Very similar experience here:
1. Somehow did not have any loss in January. I guess maybe luck or that I was trading just 1 ES contract selling EoD (I assume you got whacked for early morning sells or so).
2. Feb 24 got a lesson.
3. Feb 27 was “sour” with crazy closing. Was still trading ES options and was sure my option expired worthless, but actually got assigned close to closing price. Luckily did not close trading app to see few minutes later 1 contract with +10 points. (had opposite experience in last October, when was first time waiting for assignment to see it -10 points).
Now moved to SPX and also doing some off-day or intraday sells. What’s your target here, similar to weekend sells?
If I do same day, I try to wait at least 15min to skip initial market overreaction. Maybe it’s me being stupid, but find it hard to do same day sells, when volatility is low and markets opens +1.5% and you can sell option with strike at +0.5% only.
I also hate that SPX margin is not released straight after expiring, opposite to ES. Any ideas when it is released, midnight CT?
Also, if I have 100k account and did Monday/Tuesday sells for same Wednesday with all expiring worthless, then doing sell for Friday would raise margin 3x to ±96k until margin is released. Makes me wonder what would happen if there is sudden drop until margin is released?
Thank you Ern for all the info you have provided for this process!
Thanks for sharing! My Jan losses were some quirks that others might not have suffered. I think I just had some bad intra-day luck. 🙂
I have never felt a constraint from margin not being released right away. I need $25k-30k margin per short put and I have $100k+ in assets per contract.
But: when trading puts on ES futures you will face the margin headache and potentially margin interest from the cash accounting at IB, as described in part 3, item 4, where I detail the “better margin efficieny” of the SPX puts.
I had some cash that I was unable to invest for tax reasons for a period so dipped my toe in the put writing water after reading your articles. Been very successful so far. So many thanks for your comprehensive articles on this subject.
That cash is now free to invest and I am considering investing a proportion of it into PUTW rather than straight into a S&P index. Having reached financial independence I am less interested in big upside gains and so PUTW appears to offer an attractive alternative.
So I would have a mix of S&P, PUTW and your put writing strategy for my equity based holdings. Do you have any view on PUTW as an investment?
I would stay away from PUTW. Fees are too high and the margin cash is invested in money market at 0% now.
I also don’t like the approach of the ATM puts with such a long expiration date. Got very badly whipsawed this year!
Great site and plenty of good work by Karsten. Thank you. New to the site, based in Switzerland and managing a reasonably sized account I have been doing this successfully for quite a while.
Interesting. In what currency/assets do you keep margin? Btw, did you have to deal with tax authority declaring you activities yet? I guess you know what I mean 🙂
Swiss-based here too. Tom you have triggered a crucial point indeed. I am wondering we can still be considered non-professional traders and if these premia gains would be considered as capital gains as we start following Big ERN’s strategy.
This being said, I think I will give it a go anyway and worse comes to worst only this strategy would be taxed as income, my ETF investments would being unrealized gains..
In the US, being consdered a professional trader actually has tax advantages. Might be different in Switzerland…
Yeah it’s actually the exact opposite over here in Switzerland.
0% taxes on capital gains for non-professional traders and income rates tax for capital gains made by so-called “professional traders”. The tax authority judges whether or not ones qualifies as professional based on a set of principles/rules.
I plan on going ahead and implementing the strategy, and straight up point at PUTW if the tax guys doubt my non-professional status, stating that I am just replicating an ETF (well, sort of..), how would I be more of a pro simply doing things manually?..
Sounds like a plan! Good luck and Tschuess/Servus/Ade!
Nice! Switzerland also has some good tax treatment for capital gains, so this should be ideal for Swiss investors! 🙂
Hello Walter, I was wondering how option writing has been treated for you tax wise? do income taxes apply to these or have they been treated as capital gains? I am also based in Switzerland (permit b so tax declarations yet, just preparing for the future)
Hi Karsten – Thank you very much for exposing to options trading for me. I have approached this with great caution because I am knew so my delta’s are super low and most of the time I am trying to be two standard deviations away. In few hundred transactions, I will ramp up once I will get the “art of it” down. Basically, train my brain to not panic.
Very nice. Good approach. It takes some time to gear your risk and see how you’d react during some of the crazy trading days. It took me a while before I put any serious money into this! 🙂
This answered some lingering questions, thanks! I’ve been doing this for a few weeks with a $30k account to get a feel for it, and using SPY instead of SPX, so 2 puts is almost exactly 2x leverage. I sold 2x JUN10 306 at .1 so it was validating to see you post your almost identical 3060 for .9 on SPX.
Question: I haven’t invested any of the margin cash as I’m trying to get a handle on the margin requirements. I noticed on expiry days where I set up the next round of sells but still have an almost-worthless pair waiting to expire, the margin requirement is large as a percentage of the account, 20k for the 4 open puts, so I’m not sure what the appropriate amount of margin cash to invest would be. With Reg-T 50% overnight requirement, I think I could put up to 20k into something (30k cash deposit, -20k max options margin = 10k excess x2 = 20k investable), but that seems to be right on the ledge of a margin call, no? Am I missing something?
I’m not a tax/regulatory expert. I try to keep over $100k in margin per contract. But you are allowed to hold than in other assets (bond funds) and those assets will count as margin.
So, with over $100k in margin and an initial/maintenance margin of around $28k/$25k I’m pretty safe.
For smaller accounts, I recommend the XSP options (10x) or ES puts (50x)
I second using XSP and am using it for my small account. From my research it has all the benefits of SPX (cash settled, no risk of share assignment, better tax treatment) and is functionally exactly the same size as SPY. It also has the advantage of tracking the index exactly instead of only indirectly (which matters if there are large shifts in price).
It has a higher commission/premium ratio than SPX, but you’re probably already suffering that problem with SPY.
Quick (dumb) question for the group that I can’t seem to figure out in IBKR:
How do you get the calls off the page so I can see just the puts like Karsten does?
Thanks! And great work, Karsten. Really appreciate the great articles. Best in the business!
I usually “dump” all contracts for one expiration date into my trading screen. Then delete the rows of the calls and the puts with strikes that I don’t need.
Hi ERN, given that you stated your objective is to earn 3.00 per week, I’m assuming $300 per week $15000 per year pre-tax. If that’s not correct please let me know. Given your academic credentials and work experience would it be easier, less stressful to simply consult for a month, make the $15k, have expense deductions for tax savings, and be done for the year? Or would that violate the principle of “early retirement?”
$3 is the revenue per one index per week. One contract is 100x, so you get the roughly $15k in annual revenue for ONE contract. Multiply by 10 because I’m usually writing 10 contracts at a time and you get around $150k gross revenue. Subtract the loss provision but add the revenue of a $1m+ portfolio of bond/preferreds and you should get to the high-5-figures, even 6-figures.
We can live well off the 35% of our portfolio and even leave the remaining 65% mostly untouched most years.
Thanks. That’s helpful. If you took the 35% portion of your portfolio that you devote to the SPX put selling, and instead invested it in IVV using an equity to margin debt ratio of say 100 / 20, i.e. for each $120 of IVV bought, use $100 of your money and $20 of IB’s margin loan, you’ll surely have a higher std. dev. But I wonder how the long term returns would have compared? 📈
That would mean I’d have essentially a 107% equity, 0% bond and -7% Cash portfolio overall. Looking at my SWR simulations that seems way too “sequence risky”
Whats your highest Strike for tomorrow? I am at 2980 – now i m addicted to the charts. It is running against me arrrrg
I had been doing my research for a while and sold my first SPX put (expiring Jun 12) at 3070 yesterday 10 minutes before close when the underlying was at 3198. What a rough first day so far! Hopefully we get a quick bounce back or I’ll be working out of the hole for a while to recover losses on my first non-paper trade.
At least vol will be up haha! Definitely going to test my mettle but that’s why we do the research first.
Yeah, bad timing. I also started my option trading at the worst possible time in 2011, right before the US bond downgrade.
It will average out! 🙂
Wow, how did you get 2980? I sold my puts for the 6/12 expiration between 3010 and 3055 on Wednesday.
It was a wild ride! But thanks to the late rally on Friday I could limit the loss to only a few thousand dollars total. Could have been a lot worse!
Long time reader, first time commenter. What do you think of selling one put every day instead of two puts on Monday, Wednesday and Friday? Then three of the puts would expire in one trading day and two would expire in two days. Do you think the risk/reward would be better or worse?
One put per day would violate my “keep the delta as constant as possibe” rule. On average, you’d have twice the delta on Tue->Wed compared to Mon->Tue.
Hi ERN, first time commenting so I have to start by saying: thank you so much for your writings! You saved me from the 4% mistake, which in my case happens to have been extremely lucky, since I had saved enough that I may very well have quit my job at the start of this lovely year! I especially like your CAPE-based SWR, your look at how “I want to save $X” retirements will cluster towards the end of bull markets, and especially your super thorough takedowns of all conceivable flexibility arguments. And, of course, spurring me to even be aware of option trading.
I don’t have the patience to get manually involved in weekly trading. Probably more importantly, I don’t trust myself to stay the course and not get super anxious about price movements. Being a software engineer, and having had positive experiences with automating a similarly exotic financial side project over at LendingClub, I figured I would try to develop an algorithm I could feel comfortable letting run eventually unsupervised. Since the approach you describe has a bit (however slight) of intuitive human touch to it, I had to go my own way a bit. Where I’ve ended up: I use a GARCH method to estimate volatility, plug that into Black-Scholes, look at the gap between my computed prices and the market prices, and pick the largest gap. To manage risk, rather than doing anything with the greeks that B-S could give me, I apply a handful of cutoffs of the form “given the volatility we’re assuming and the amount of puts we’re considering writing, the probability of losing 5% of the backing cash must be <.01" (weekly), for a collection of values that outline roughly what I'm willing to tolerate. I also buy an equal number of puts down at ~92% of the current underlying, both to put a known hard limit on max loss, and because discarding historical weekly returns below there as outliers makes the past few decades' data look much closer to a true Gaussian, making me way more confident in trusting the B-S math.
Does that sound basically reasonable? I feel vaguely uncomfortable to be approaching it so differently from a pro, but thinking it over repeatedly I'm convinced that my risk model is directly evaluating what I need to care about – probabilities of various levels of uncomfortable loss – and keeping them within sane bounds. Probably important to mention: as part of keeping myself uninvolved, the idea is that everything will always go to expiration; no bailing out early. So, how the price of the option evolves is not relevant to me (other than margin call; I keep my max loss smaller than the assets I hold in the account).
Finally, fun fact: my first tentative still-mostly-manual live fire test had a Feb28 expiry, and my fourth is tomorrow (June12). So uh, horrible luck, haha. I'm mostly undeterred – although the fact that I stepped back while everything was crazy shows that I clearly can't trust myself to have an active hand in the day-to-day investment decisions! Fortunately this is all still just a single XSP at a time. Can't automate without IBKR Pro, Pro isn't worth it without writing at least a full SPXW, certainly not going to write that much until I'm super sure of everything.
Wow, that’s a pretty fancy methodology!
How much variation in your delta do you get? How do you keep the risk roughly equal over time?
Do you use IB with their API?
Also, you comment got cut off in the second paragraph, so if there’s more to your approach, please share. Overall this looks like a great way of implementing the put writing strategy.
Hmm, are “the B-S math” not the last three words you see in that second paragraph? Reading back over it here on the site, everything looks intact. Briefly: I also buy puts at ~92% of current underlying, and manage risk by calculating probability of experiencing %X loss in a given week. I hold everything to expiration.
Regarding delta, well this, uh, might worry you (which is why I’m slightly hesitant)! I don’t directly pay attention to delta. I manage risk by keeping Pr[losing >X% of portfolio this week] < Y, for a handful of values of X and Y I think I am emotionally comfortable with. I compute that probability from the volatility I'm estimating for the upcoming week – I'm assuming the week's log return will be normally distributed with the stdev I estimated. In order to make "normally distributed" a reasonable assumption, for every put I write I buy one down around the "long tail" of historical weekly returns, which I found to start around -8% simple return. That's based on playing around with the Shapiro-Wilk test and discarding outliers.
Regarding risk over time, I'm hoping that describing risk in terms of X% of current portfolio is reasonable, both at higher and lower portfolio values. In particular, it gives you the opposite of the really dangerous "oops we're underwater, write a couple more at higher strikes this week to catch up" behavior.
Yeah, IB. My automation is still just paper trading, since I won't pay for IB Pro until I'm confident enough to do a full SPX every week. I'm basically out of the weeds, but there was a stretch where really only sunk cost fallacy was keeping me going, haha.
Nice! Thanks for the info.
I think, your explicit formula is close to my delta*SPXrisk criterion.
Awesome, that’s encouraging to hear, thanks. Good luck to you as well 🙂
I’ve done a lot of work on this over the past year. On the off chance anyone ever uses what I wrote here as a starting point, I thought I should update it to avoid leading them astray. I’ve dropped that GARCH+Black-Scholes approach; in fact, I now no longer (directly) use volatility at all. Instead, I build – from whatever subset of historical data I consider “close” to current conditions by a certain couple of metrics – a CDF of what log returns might look like the next couple of days. That gets me estimates of expected value of net income (for picking the best), and the probability of losing more than a given amount (for keeping risk at a level I’m comfortable with). Good luck to anyone else giving it a try!
Interesting. Though, in the comparison today vs. historical, I don’t see how you get around comparing vol or some variant of it. You allude to using it “indirectly” but it’s not clear how,
Correct, it certainly is related. I use the log ratio between the recent (10 day) SPX high and low, together with the log return from SPX 10 days ago to SPX at the time of writing. It’s crude, but that actually makes me feel better – I had gotten nervous that I couldn’t really fully justify the sophisticated stuff to myself.
OK, makes sense! Thanks for the clarification.
Please provide us with details about how you handled the sell off on June 11 related to the put selling strategy. I had one contract on and I ended up buying back the original 7 delta put when the SPX was down about 90 points. Still a hit but nothing like it could have been by end of day.
Why did you buy it back if it was only 7∆?
When I bought it back it was 45-50 delta due to the swift market decline. After I bought it back, the market dropped 70-80 points below my original strike price. Ugly situation but could have been much worse. I was assuming that there was no way the market will recover above my strike by expiration on June 12.
Well yes your marks would look really bad and even worse so because of the vega explosion and the fact that you were pretty much ATM when you bough it back. I think there are a couple things here and maybe some cross purposes. In one instance having trading rules to close at a certain los level is fine but it is not part of the strategy ERN issuing (or PUTW) etc etc. In another instance you bought back nearly 100% premium (if you bought back when it was a 50∆ option) and that is a lot of juice that would go away. I won’t comment on what you should have dome or not done but you should get clear on which of the two (both reasonable) approaches you want to use.
Thanks Joe – totally agree with your comments. Overall, I like the strategy of selling low delta puts in SPX – 2 day expiration, tax benefits, cash settled, etc. I knew I was buying back alot of premium but wanted to limit losses – calculating my loss (even with the premium paid) vs. the loss I would have taken if I had waited until expiration at close on June 12, I would have taken a ~$900 additional loss if I held to expiration. I am open to recommendations on how best to manage the risk in these situations – I guess there are probably only 2 options – set a max loss level and buy back OR ride it out in hopes of expiration above strike price (aka – Karsten’s approach).
Hi, just learning, so if the contract was down 90 points when you bought it back, does that equal a $9,000 loss on it? Thanks.
If it goes $90 below the strike, then it’s indeed a $9,000 loss. (but you still net against the premium you collect upfront).
But remember: the index first has to fall all the way to the strike and then some more dor you to lose serious $$$! 🙂
I dealt with it like always. Hang tight and let the options expire.
I had strikes at 3010, 3025, 3030, 3050, 3055. I sold the lower strikes between 3010 and 3030 in the morning (my time) on Wednesday and the 3050 and 3055 strikes a little bit before market close.
Luckily, we saw a late Friday market rally and I ended up with only half my contracts in the money and even those only slightly ITM ($8 and $13, respectively).
Was it a bit scary? Yeah, but put this in perspective: even if the market had closed at 3000, my % loss would have been less than the unleveraged S&P500 index loss.
Thanks Karsten – I clearly had my strike too high as I was at a 3075 which, at the time of sale on Wednesday (6/10) near market close, was a 7 delta. Maybe lowering my strike closer to a 3 delta and trading a wide put credit spread would be better as I am not sure I have the stomach for letting it ride on these really down days!
No – I am going off of memory at the moment but the loss was approx $2,500 – it ballooned up to about ~$7K+ if I had waited until later in the day to buy it back. As it is, I waited too long as my 1 hour workout in the am cost me $1k :). Price to buy back ballooned from $1,500 to $2,500 in about an hour. Karsten says he never buys back the put but in these extreme moves, I am not sure there is a better alternative to manage risk / losses.
I’m in at 299 (SPY), close was 300 yesterday, staying in. We’ll see what happens today!
Looks like it worked out! Almost a point landing! 🙂
One way would be to trade credit spreads. Sell a put but buy a put with the same expiration date at a lower strike. You have a maximum $ amount you could lose.
Thanks for your fantastic articles, ERN!
Would you have any hesitation in holding index funds in your margin put writing account, and holding the muni bonds/pref shares in a separate tax-sheltered account?
I live in Canada where it would be much more tax-efficient for me to structure the accounts in this way. In theory I would hold the exact same investments, just in different accounts. The only potential pitfall I can think of is margin requirements. Would you decrease your put writing leverage if you were using index funds as collateral? Is there anything else I’m missing?
Not to barge in but….@JR watch it. If the index funds are volatile assets (especially equity funds) you must count that net ∆ into you leverage gearing. Say for example you held $100 in an Spooz index fund and another $100 notional risk in a Spooz short put strategy you would already levered 2X.
Thanks. Correct me if I’m wrong – but wouldn’t this be true of ERN’s approach as well since he holds an S&P index fund in another account? I’m simply talking about holding the bonds/pref shares in a separate account from my put writing. On the whole, wouldn’t my risk remain the same, save for the risk of a margin call (i.e. because the collateral in my margin account would have a higher correlation to the short puts)?
If so, how much less leverage, if any, should I be using to mitigate this margin call risk?
Yes it would be true in any situation where the notional value of your risk is larger than the assets supporting that risk. Trading on margin is not always a leveraged proposition, margin is only a measure of the brokerage’s (via the exchanges’) requirement for capital to support your risk therefore you need to consider your balance sheet in the aggregate to derive if there is any real leverage gearing. ERN mentions about 2.5X turns of leverage in his put writing program and that would apply in a situation where the account was simply the short puts and cash or CEs (like T bills). But, when the cash is used to buy other (albeit less) risky assets then the gearing increases… maybe not by full turns but by something more than the simple short puts/cash combo. At a higher level you can look at the correlations to tweak your leverage (subjective) determination but this can change at any time. You mention holding bonds and preferreds against the short put portfolio and for sure that is going to be less gearing than holding all equity risk instead but it is not the same as cash or CEs. As far as holding them in another account you are correct that beyond whatever the final leverage gearing there is the entirely separate matter of what can be seen by the brokerage. You could have a large excess of capital to support your positions but if it’s not in the actual account the brokerage won’t (actual they cannot) consider it. In the case of a margin call, proof of assets held elsewhere may buy you a very short bit of additional time to meet the margin call but I would not rely on it. The best bet (account size permitting) is to apply for portfolio margin (PM) as this (when used properly and prudently!!!) will give you much more cushion in the margin calculation. I stress that the additional risk permitted in a PM account (versus Reg-T) should NOT be used for slathing on bigger size (more risk) but just to give you much more room before you have to go through the actual hassle of moving funds/assets form one brokerage to another. If your strategy is selling very low delta puts it is difficult to reduce the leverage without cutting deeply into the revenue potential. I would only add that the only true risk management happens at order entry via position size, after that all other so-called ‘risk management’ decision are merely trades in the other direction, be they are closing transactions or putting on ‘adjusting’ trades. As far as how much less leverage… I repeat your best bet is to apply for PM to create more room but al also repeat (shouting) use the extra risk permitted ONLY to lower the margin requirement and NOT to increase the aggregate risk you would take. Size kills.
I hear you, thank you very much for your replies Joe. FYI for others – my quick Google search suggests that Canadians are not eligible for portfolio margin due to restrictions in place from the regulatory body here.
Yeah, that’s my concern. But again, money is fungible.I do hold that much equity in other oaccounts.
Yeah, money is fungible. I could put the equity index funds I currently hold in my retirement accounts and hold them in the taxable IB account and – conversely – hold the income-generating assets in the tax-preferred accounts.
I might have gotten cold feet though on March 23 when the SPX was down 34% and this would have hindered my recovery during March. But I see the point. I’m mulling over this! 🙂
Would you mind elaborating on how your recovery would have been hindered if you had switched the money around and held equities instead of income generating assets? Would it be due to not being able to take as many contracts due to bigger losses on the equity portion of the portfolio?
That’s exactly the reason. I’d be afraid that the drop in the margin cash would reduce the # of contracts I can short.
Karsten, Thank you for the amazing write-ups to help us mere mortals make sense of these instruments! Can you recommend any books or resources to further study these techniques?
I use the book by Whaley:
This is just a reference book. It has a CD ROM with an option-math plugin for Excel.
Since I live in Australia, it’s only really feasible for me to wake up at close to market close to trade due to time difference hence my strikes were 3055 and 3060 so I got whacked a bit.
It hurt psychologically at first but I felt better after doing sum quick sums. I am still in profit since the the big March drop so the trading method is just doing what it’s supposed to do.
Yeah, my lower strikes were from the earlier trades on Wednesday. But sometimes the early trades can go against you and it turns out it would have been better to only trade right before the close. Win some, lose some! 🙂
I am new to investing and receiving dividends. I study all possible ways of investing. Thank you for sharing this information and your experience!
You bet! Good luck!
Always start with “practice money” for a few years until you’ve racked up a few losses. 🙂
I am not too deep into options. But from Nassim Taleb I learned that this kind of seemingly easy “steady income” from writing options works until it doesn’t, usually with its blow-up sooner or later, when its long left tail risk hits you hard. In finance, the emotionally hard way of allocating short left tail risks, which are unpopular and thus underpriced due to constant uncertainty and small losses, is usually the much better antifragile way for high long-term risk-adjusted returns.
The recent article https://blogs.cfainstitute.org/investor/2020/08/03/creating-anti-fragile-portfolios/ explains very well, why the question, if a portfolio is long or short volatility, is becoming more and more critical for your long-term financial health, particularly when quite a different market regime will prevail for decades. But even if not, the ride will be much smoother with high returns.
For this, long volatility trend-following CTA and recently outright long volatility are the most suitable options and futures trading strategies if understood well enough to select respective mutual alternative funds for quality to buy&hold them conveniently for good. Only combined with equities, these two non- to anti-correlated assets can provide for high equity-like returns but with almost negligible WDD and SoRR even without bonds.
I’ve gone through many (dozens) of losses. Inever claimed that this is “steady income”
Also, what kind of tail risk are we talking about? 2008? Or a global pandemic? Put selling would have done well through both episodes.
Also, I think that specific long-vol strategies are actually quite attractive. Long VIX futures or options on VIX futures are a good idea:
I just don’t think that plain long put options will do very well. That’s why I take the short side on that trade.
Couple of questions about the margin cash portfolio. What do you think about using preferred stock ETFs vs picking individual preferred stocks? And, how do you think about the embedded leverage inside your muni closed end funds as it relates to your 2-2.5x leverage target for the short puts?
I used the PFF (iShares) before but was turned off by the high expense ratio: 0.46%. You can do better holding your own individual Pref shares.
I also like to have control about the shares I hold. I like (“prefer” so to say) pref shares with adjustable interest rates to hedge against future rate hikes.
Some of the margin cash is in the CEFs with some built-in leverage. But keep in mind that I also like to keep a little bit of cash in the IB account.
So what is the plan in a 1987 styled -22% one day drop? The 9.5% drop this year was pretty crazy, so interesting to see your options were still pretty far out of the money and it didn’t do much damage. This strategy seems like a modified volatility swap strategy that risks a bigger black swan wiping out multiple years (all?) of profits though. Very short gamma!
Large drops like that normally occur when the IV is already elevated. During March 2020, I sold puts so far OTM that even the big drops didn’t even reach the strike, much less wipe out 3x equity.
But don’t get me wrong; a large drop indeed wipes out multiple weeks of put income. Maybe 2-3 months. Not really sure how you would wipe out multiple years.
Can you post an update of the chart SP500 vs Put strikes? Thank you.
Added two more charts at the end!
ERN, how concerned are you about a 1987 type, one day 20% decline in the market? My broker has a risk tool that showed that i would suffer a 60% decline in my account if that occurred today(I’m using a similar strategy to you with about 3x to 4x margin).
Seeing that type of potential loss has me wanting to dial back the margin or increase my DTE so I can sell lower strikes.
Nobody I know would run a 3x leverage strategy with at-the-money options and risk a 60% loss.
I can’t really replicate what would have been the option quotes back then, but it’s likely that on the Friday before the Black Monday, implied volatility would have been extremely elevated and you probably sold options way out the money. You might not have lost any money at all. You might have made money if you had sold options 25% out of the money! 🙂
Case in point, March 16, 2020. How much money did I lose when the market dropped 12% in one day? Did I lose 36% with my 3x strategy?
No, I MADE money on those options. I sold options 20% out of the money and the market never fell below my strike.
Amazing display of knowledge! Thank you for sharing also for (spy) users of this strategy like me how wide would you suggest the strikes and can this be used on qqq or iwm since they now have 3 expirations a week now?
Yes, good point. I still prefer the index options (large notional + tax efficiency due to S. 1256). But some other indexes/ETFs should work just as well for other investors.
Have been going through this again – loved the entire series and putting it in practice.
Got a couple of questions:
– Is there any risk of writing ‘too low’? (i.e. where the delta is too low and you are collecting too little for the inevitable drawdowns)? Have you modelled this before?
– the delta of -0.04 fluctuates depending on what time of the day you write and where the index seems to be trading. Any tips around this? Can make a huge difference on the premiums collected and strikes.
– you mentioned credit spreads in a scenario of a free fall. Do you mind elaborating please?
– how do you work out annualised yield? Would it just be premium/number of days * 365?
Thank you so much.
1: yes, there’s a risk of generating too little income. You can write puts for a premium of $0.05 and hundreds of points OTM. But it’s not worth the effort. I got bills to pay.
2: Yes, intra-day timing matters. Right before market close you get less premium than early in the monring. You have to weigh how much overall delta you got still live to determine if you want to write new options for T+1 before your date T options expire.
3: I don’t usually do credit spreads. The risk is lower but the income is also greatly reduced. Especially for just 1-day options it’s best to do naked puts.
Longer horizon plays, yes, it would be best to hedge the extreme downside.
4: Annualized yield is premium/strike*365/days2expiration. Correct.
Thanks for your responses.
1) Was not referring to that low. But more like -0.03 or even -0.02 delta in the event of low volatility.
2) What do you mean by overall delta? Is there like a ‘total’ delta guideline you stick to?
3) Would you ever intervene in trades to reduce losses where the put is so far in the money and it’s just a question of how much? Perhaps by buying a put that’s ITM?
1: I currently sell at around -0.02 Delta. So we’re likely on the same page.
2: sum of the delta over all contracts. How much does your *portfolio* move in response to a 1pt move in the index?
3: I do sometimes. You win some you lose some. It probably didn’t make a difference averaging over all interventions
Thanks so much. Just wondering if you considered selling a same-day call option that’s also around the same delta to increase returns (aka a short strangle)? If not, do you mind why not?
I have. But I don’t like the smaller premiums (due to vol smirk). And I don’t like betting against the market going up. Feels unpatriotic.
But I know people use both sides of the vol premium successfully.
You mentioned that you sell more contracts when conditions are favourable quickly.
“Here’s an example: I sold a batch of Monday options on Friday morning. The market goes up on Friday afternoon and I essentially made 80% of the premium already. I might as well sell a few more Monday options and also still keep the almost worthless close-to-zero-Delta puts. It looks like a risky trade, but I consider it manageable if the overall Delta is low enough.”
Question. For the Monday options, do you sell at the same strike or move up strike (and keep the same delta) now that the index has run up?
Also, can you confirm if your strategy is to alway roll an ITM position before expiration (and is it always for a credit) or just hold the ITM position to expiration (take the loss) and open a new position?
I noticed that if I tried to roll the ITM position down for a credit, I won’t be able to get to the desired (<-0.05 delta), than if I just take the loss and open a brand new position.
Which do you prefer?
Thanks in advance and great article as always!
I sell at the strikes and premium that looks attractive at that point. That will likely be at a higher strike if the market moves favourably.
I have no set rules. Most of the time, I would let the puts expire ITM. But I have shifted to cutting losses when the price of the put gets to about 10x the premium target. That would normally be before they get to ITM.
Hi ERN, do you sell the puts at market close?
Yes. At or close to the market close.
Wow! Thanks Earn. I see this process is always evolving and constantly improving! 🙂
Do you take into account any macro factors e.g. fed announcements etc? or you just systematically place orders and let the probabilities play out over time?
I found it was helpful to hold out until AFTER the Fed announcement at Jackson Hole when the market tanked. I got a better price and further OTM.
I’m pretty stoic and systematic. Obviously, I notice the high-vol periods (Fed meetings, CPI releases, payroll employment releases, etc.) but they will simply show up as higher IV and thus further OTM strikes.