All parts of this series:
- 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”
* * *
Last week we made the case for generating passive income through option writing. A quick recap of last week: buying puts to secure the downside of your equity investment is a bit like casino gambling: pay a wager (put option premium) for the prospect of winning a big prize (unlimited equity upside potential). Unfortunately, the average expected returns are also quite poor, just like when you gamble in the casino or buy lottery tickets.
Since we can’t beat the casino, let’s be the casino!
Being the casino means we act as the seller of put options. Let’s see how we implement this:
- Brokerage account: we use Interactive Brokers. It seems to be the cheapest provider in terms of per contract trade fees, but they also nickel-and-dime us with all sorts of other small fees. Given the large volume of contracts we trade every year, we currently bite the bullet and pay those nuisance fees knowing that we save a lot on commissions. But if anybody has experience with other providers who charge lower fees (IB: currently $1.41 commission per futures option contract) please let us know.
- Underlying: we want to sell a put option, but on what? Currently, we sell put options exclusively on S&P500 futures contracts, specifically, the e-mini contract (ticker “ES”). It sounds really scary: we sell a derivative on a derivative. A derivative squared! But economically, this is almost indistinguishable from selling put options on, say, an S&P500 index ETF. But: There are a number of advantages when we implement the put writing strategy with futures options rather than options on equities or ETFs:
- We can run a tighter ship with our margin cash. Selling a put, you face the possibility of having to buy the underlying at the put option strike price. If you are forced to buy a large chunk of equities or ETFs it’s better to have that cash available and ready. Otherwise, you face paying margin interest if your cash balance drops below zero. But we don’t like a lot of idle cash sitting around; remember our theory about emergency funds?! Everything is much easier with futures contracts because they don’t require any cash outlay (except for a small cash margin cushion) and they are extremely liquid and cheap to trade. We are able to keep around 70% of the margin cash in a municipal bond fund to generate extra (tax-free!) interest income. We also found that the bond fund serves as a diversifier: When equities go down, bonds do well. And with a 70% bond weight, it’s actually enough of a bond weight to make a difference in terms of diversification!
- Tax season is a breeze: we trade about 16-20 contracts a week, or 800-1,000 per calendar year. If we had to keep track of all trades and itemize them all on our tax forms it would be taxmageddon every April 15. But since options on index futures are considered Section 1256 contracts for U.S. tax purposes we have to report only one single number on tax form 6781, line 1: the net profit/loss of all such contracts combined. Everything is already net of transaction cost. I spend more time documenting my 2-3 mutual fund trades on our capital gains tax forms than our 800 to 1,000 option trades! Moreover, all profits from our option trading are automatically considered 60% long-term capital gains and 40% short-term gains, irrespective of holding period. Sweet!
- Account size: we currently run this strategy in an account worth around $500-600k. I personally started with $30k of play money just to learn my way around but then eventually grew the account once I felt more comfortable. We like to short one put option for each $30-35k of account size. So that $30-35k would be the minimum recommended account size.
What option(s) do we short?
Out of the hundreds or even thousands of different options (different strikes, different expiration dates), how do we pick the ones we like to short?
1: Picking an expiration date
We pick the shortest possible time to expiration. That means every Friday we sell a new set of put options expiring in exactly 7 days. Then, next Friday we sell the next round.
Update (September 2017): For most of the year 2017 we’ve shifted to even shorter-dated options. There are three expirations every week (Monday-Wednesday-Friday). So, now we write options on Friday that expire on Monday, then on Monday, we write options that expire on Wednesday and every Wednesday we write options that expire on Friday. The premiums for the shorter-dated options seem “richer,” i.e., we get the most premium per unit of risk we take on.
We like to keep the maximum number of independent bets because that’s how casinos make money; when the house has an advantage the more people play and the longer they play the more certain it becomes that the house wins!
“In the casino, the cardinal rule is to keep them playing and keep them coming back. The longer they play, the more they lose. In the end, we get it all.” Sam Rothstein (Robert DeNiro) in the 1995 movie Casino
2: Picking a strike price
Even though we initially introduced the put writing strategy as selling at-the-money puts, what we do in practice is slightly different. Here’s a snapshot I took last week on Wednesday (about half-way through the trading day). The ES future was sitting right at 2150.00. You could sell an at-the-money put for $15.75. For the roughly 9.5 days to the expiration that would mean a whopping 28.15% annualized yield. Remember from last week: hedging out the downside gives you “only” about 20% p.a. extra return! So, option premiums are quite rich, especially at weekly frequency! But we also include the puts that are out of the money. You can find strike prices in steps of 5 points, but we list only the strikes 2,075-2,100 in steps of 25 to save space.
The option with strike 2,100 looks most attractive to us (note we didn’t actually trade any of those on Sep 28 because we still have the options expiring on Sep 30, so this is just a theoretical exercise):
- It offers a decent yield of 7.33% p.a. (= option premium as % of underlying index, annualized). Note that this is the gross revenue if the option expires worthless. Every once in a while you lose money on the trade and our long-term average experience has been that we keep about half of the option premium as profit and pay out the other half to the option buyers. During the very calm periods (parts of 2012, all of 2013, the first half of 2014) we actually kept close to 90% of the gross option premium. But that’s not typical; during the last two years, with volatility and significant drawdowns, half of the gross revenue is all we got. We will have to use leverage to get to our desired expected return level.
- It’s far enough out of the money that we have enough of a cushion against losses; the S&P can drop by around 2.5% and we still wouldn’t lose money. We like that kind of wiggle room. When you sell put options at the money the premium is higher, but even the first dollar of a decline will already eat into your profit!
- The implied volatility is about 14.5%, higher than the prevailing market-implied volatility (VIX) level of 13.2%. We like that! Is has been documented that, in average, the VIX is higher than realized volatility (see page 3 sidebar table in this paper), and if our option has higher implied vol than the VIX we have two cushions:
Option Implied Vol > VIX > Realized Volatility
3: How much leverage?
We use leverage for two reasons:
- To compensate for the impact of marginal taxes on investment income we like to scale up by at least a factor of 1/(1-tax rate).
- For the same reason leverage is normally used: boost an attractive return. Shorting only one single put option per notional value of the ES future ($2,150 x 50=$107,500) would create too little return and very little risk as well. At the inception date, the short 2,100 strike option had a delta of 0.15, so it is only 0.15 times as volatile as the underlying index future. We can “safely” scale that up to 3x leverage and still maintain less volatility than the underlying, most of the time.
How much more risk comes from leverage? Last week we pointed out that with the simple short put option without leverage you would never lose more than the underlying. That changes once you introduce leverage. In the chart below we plot the payoff diagram of the 3x short put option:
- In region 1 we lose more than the index. But it’s still not 3x the index loss. Even if the index were to drop all the way to 2,000 (-7%) we lose about just over 13%, not 21%. That’s because the 3x only starts after we drop below the strike price. Because of this cushion, our strategy will actually look less volatile than the index, most of the time. Only under extreme circumstances would we face more volatility, see case studies below.
- In regions 2, 3, and 4 we beat the index. Sweet!
- In region 5 we make money but less than the index. Again, as pointed out last week, we are not too concerned about this scenario because we have plenty of other equity investments, so our FOMO (fear of missing out) is not too pronounced.
Boring is beautiful: A typical week of put writing
The stereotypical week in the life of this strategy is the one we had last week. Here’s the path of P&L for the 3x leverage Short Put vs. the simple index investment (through the S&P500 index future):
- September 23 midday: sell 3 put options, strike 2,110, while the underlying was at 2158.00.
- September 23, market close: the ES future closed at 2,158. Our P&L was about very slightly down. (Volatility went up a little bit before market close, hence the increase in price, despite an unchanged underlying!)
- September 26: the ES contract closes down at 2,139.75, almost 1%. Bummer, that’s a bad start to the week. Our position lost money but less than the underlying. We were still far enough away from the strike, so nothing to worry about (yet!).
- September 26-29: the ES contract recovered from its “bad case of the Mondays” and bounced around between 2,148.50 and 2,163.25. Our short options roughly mimic the path of the equity future P&L.
- September 30: the ES contract closes at 2,160.50. That’s above the strike price and we made the maximum premium. We are slightly ahead of the equity index for the week!
With the exception of a small scare on Monday, this was a very uneventful week. We earned the maximum option premium, while equities bounced around quite a bit. Despite the equity volatility throughout the week and our 3x leverage, it was a smooth ride. We had less volatility than the underlying index and made more money. Making money the boring way, one week at a time!
Murphy’s Law: when this strategy goes horribly wrong
OK, for full disclosure: put option writing with leverage is not for the faint-hearted. Sometimes things can go wrong and when they do one can lose a lot of money in a short time. I am fully aware of this feature and believe that this is the cost of doing business. To use the casino analogy again, sometimes a slot machine pays out a big prize. If it didn’t, nobody would want to play it.
Let’s look at what happened in the first week of January 2016. We had initiated a bunch of short puts on December 31 (Thursday because Friday was a holiday). Even between selling the option and the closing that day, the index future dropped, though not by much. For the first few days in the new year, the index kept going down and we mimicked that path, though our losses were actually muted despite the 3x leverage. That’s because if you are still far away from the option strike price then you still have that “cushion” and the volatility even in the 3x leverage short put portfolio is muted. (For finance nerds: The option Delta is still far below 1!)
But then came Thursday, January 7: The index dropped by 2.4% and our P&L went below the index. Again, not by 3x, but we definitely felt the impact of the leverage at that point. On the Friday exit, the index dropped further, though we had the wisdom of pulling the plug and closing the position while the ES Future stood at 1,928.50. Wow, what a ride! Instead of making $405 with the three short puts (2.70 x 3 x 50=$405.00), we lost over $7,000. The portfolio lost a lot more because we had a total of 20 short puts (some at better strike prices with lower losses, though), but the damage was done. We had the worst start to a new year ever! It would take until mid-March this year to just get back to zero return, and even that was aided by the excellent returns in the Muni bond fund. Considering only the short put strategy it took 18 weeks to dig out of the hole!
When this strategy goes “horribly right” – Yhprum’s Law
Meet Yhprum (a second cousin of Murphy) and his law applies when, for a change, everything that can go wrong actually goes right. I have had a few instances of Yhprum, most recently around the Brexit mess in June 2016. Let’s look at the week of June 17-24.
There was a pretty bad drop on June 24, but we still made money. That’s because the uncertainty about the Brexit was already reflected in the option prices on June 17. Thus, we were able to sell put options with strike prices so far out of the money that even the steep decline on June 24 never got even close to causing any losses. In the P&L chart below, note how between 6/17 and 6/23, the P&L of our strategy and the index have a positive correlation, but our movements are very much muted. Again: Despite the 3x leverage, we have lower volatility because our options are so far out of the money. Then comes Friday 6/24. The S&P index drops by 3.6%. The ES Future goes all the way into the low 2,000s. That wouldn’t harm us because our option strike was at 1945. We actually made a small profit that day. Despite 3x leverage! And in case you wondered: the post-Brexit week was also profitable: on 6/24 we were able to sell puts with strikes in the 1,800s, because everybody got so scared on Friday. On Monday the market dropped again, but then recovered swiftly and we earned the full option premium that week as well.
Returns over the last two years
Case studies are fun, but what was the average performance over the last year or two?
The last two years have been a tough environment for equity investors. The second half of 2014 was volatile, and 2015 saw the mess with the Chinese devaluation and a Federal Reserve rate hike. January and February of 2016 were pretty awful, but we did reach new all-time highs in August. But the path was very bumpy (did I mention the Brexit?) and the average equity return was 6.8% from September 30, 2014, to September 30, 2016, with dividend reinvested. That’s still a decent return but less than the long-term average.
Our option writing strategy performed significantly better, see chart below. We got an average annualized return of above 15%, more than twice the equity return. In contrast, we had a volatility of only 6.2%, about half of the index ETF volatility.
For full disclosure: our returns include the additional returns from investing in the Muni bond fund, which had excellent returns over this 2-year window, not just interest but also price appreciation. But we want to stress this issue again: The bond returns are part of the strategy because we have $600,000 of cash sitting around, which we should put to use. Also, the returns are net of all the fees and commissions. We don’t expect 15% returns to last forever but 12% before tax and 8.5% after tax with 6-8% annualized risk would be our target return profile.
Our strategy has major drawdowns around the same time as the S&P500. But the advantage of our strategy has been that if additional drawdowns occur after the initial event (September 2015, February 2016), we actually make money. That’s because investors were in panic mode and drove the option premiums up so high that we sold puts at strike prices far out of the money: none of our short puts lost money even when the market dropped further in consequent weeks. Thus, despite our 3x leverage, we had a pretty smooth ride after the initial drop. And, as mentioned above, the Brexit didn’t harm us either!
Last week we introduced the option writing strategy for passive income generation. The run-of-the-mill strategy would be to sell a cash-secured put, at the money. It’s so popular, Wisdomtree even made an ETF out of it. We take this well-known strategy and make four adjustments: 1) leverage, 2) sell out of the money puts, 3) use weekly options instead of monthly, and 4) hold margin cash in longer-duration bonds (not just low-interest cash) to boost returns. So far we have fared pretty well with this strategy, easily beating the S&P500 benchmark. We do have drawdowns, about in line with the large S&P500 weekly losses during the recent stress periods. But the overall volatility is much lower than the S&P500.
439 thoughts on “Passive income through option writing: Part 2”
Can you share whats your cut loss strategy is?
I am trying out the “cut loss when i lose the whole premium” i sold for that expiry, but i would think if i do a very short to expiry like what ERN does, i would be cutting loss ever so often as the premium is little, and as you mentioned, an increase in implied volitlity , even if the price is stagnant would make me cut loss.
Im still trying to find my comfort zone between ERN ways ( weekly or lesser expiry) and optionsellers.com ways ( 6 months away with futher OTM)
I don’t really have a strategy around losses other than keeping track of my leverage (keeping it where I want it, not letting it get too high) and increasing leverage when volatility is higher / decreasing when vol is lower. I’ve seen studies from Tastytrade around closing a loss around 2x the credit received, but they aren’t super compelling and they say they don’t usually do that in their own trading. I think they’ve also looked at closing positions when the option goes in the money. I think 1x the credit would be too tight of a limit. A lot of times the position will come back as you allow time to pass, and if you get out at the lowest point you’re giving up the opportunity for that to occur. In the past I have extended that idea in time by ‘rolling’ positions forward. That means if I have a losing option, I’ll close it and sell an option in the next expiration cycle at the same strike for more money or possibly for the same money at a lower strike. This has pros and cons – you now have a position with lower probability of success as the strike is closer to the underlying’s price / the delta is larger, but if the underlying goes sideways for long enough or if it reverses back up, you can make back some or all of what you lost. Alternatively, you can close the position and move to the next expiration and start again selling the same delta put you always do. Statistically this should produce more winners over time but it takes longer to dig out of the hole of a large loss. I am still digging out from the February drop after using this method where as if I’d rolled out in time for a small credit I’d probably be about back to even by now. But if the market had continued dropping, I would have been better off from doing what I did.
Mostly I focus on closing winning positions a bit earlier to try to prevent them from turning into losses. For example, with the up move today I closed Wednesday /ES options I had sold when they reached around 0.40 of premium left ($20). I had made about 85% of the possible max profit and I didn’t think it was worth the risk of leaving them on all day today and tomorrow to pick up that last $20 and avoid the ~$3 charge to buy them back.
Thanks John for sharing. February drop was bad and unexpected.
1X credit to cut loss really is too little, which would necessitate a longer expiry to allow such a cut loss strategy.
Still feeling my way around and will be trying to regularly sell a monthly SPY put with 10.5% below strike with an annualised yield of about 15% on initial margin req. ( i actually will fund a 1.5 X initial margin marquirement for safety for a annualised yield of 10%). Not much but safety first.
If I’m interpreting you correctly, you would sell the monthly May 240 put tomorrow for example which would have ~30 days to expiration, for ~0.30. If you use 3x notional leverage eventually, that would only be about a 4.5% annual return selling such a far out of the money put, and that’s assuming you keep the entire credit all the time. That’s fine for learning, but I would suggest you are taking risk without a lot of potential return and may want to go a bit closer to the money in the future. Also make sure your commissions aren’t going to eat up too much of your potential profit.
3X leverage means, say, SPY is 270, strike is 248, initial margin is USD3231 and you have USD9693 (3X initial margin) as a back up? Do you mean this?
For me, im doing, say, SPY 270, strike is 248, initial margin is USD3231 and i am having USD4846 ( 1.5 X initial margin) as a back up instead.
So its a matter of playing around with the margin and the distance of the strike that would determine the annualized yield.
Further OTM, less margin required but less premium
Nearer OTM, more margin required but higher premium
I am going for the “Further OTM, less margin required but less premium “
For the same expiry
I look at leverage in relation to the amount of stock you could potentially be taking of delivery of if your put went in the money, not in relation to the margin. So say you sell the 248 SPY put. That means worst case you are buying (or have the same risk as buying) 100 shares at $248/share. So the ‘notional’ value is 248*100=$24800. So say you have a $100k account. If you sold 4x 248 puts, that would be ~1x notional leverage. 3x notional leverage would be 12x 248 puts. The range I have heard suggested and used myself is 2x-6x notional leverage for your entire account when you are selling options like this. 6x would be quite aggressive and I would only do that if implied volatility was very high and I thought it was a great opportunity or you have a very high risk tolerance. I was targeting more like 2x at the start of the year when volatility was at historic lows and currently a bit over 2x, similar to ERN.
Then to calculate yield, look at how many puts you can sell for some level of leverage, the premium you might collect from that in a year, and then divide by your account value. Or ERN posted a more direct formula in a previous comment. Then realize that in reality you might not keep all that premium, so depending on what you’re doing and how much the market accommodates you, you might keep between 25% and 75% of the premium you sell over time. Margin only enters in for capital management reasons. Ie, you don’t want to be using all your margin so that you’re in danger of getting a margin call and being forced to liquidate your positions if you suffer some losses. I would expect to use between 20-40% of your available margin at the low end of leverage and maybe up to 60% at the high end, but the relationship between leverage and margin usage will depend on what type of account you have.
Thank you John. Your reply is very clear. Based on what you mention, i am doing a ~ 6X leverage.
1 contract , SPY May18’18 241 put , premium $0.35, 6X leverage, ~10.2% pa, underlying 268.89
So with a larger leverage, the strike can be further out.
Im still 2 weeks old to options, so still wondering how a spike in volatility will hit my margin. Will wait for one before i add more funds .
Thanks again, difficult to find a good forum without people saying rude remarks to one another
I calculate leverage based on the notional. Example: ES future at 2650. Sell a put at 2575. Multiplier of 50. So one short put has potential maximum loss of 2575×50=$128,750. If I hold $60,000 in funds in the account then I have a little bit more than 2x leverage for one short put.
Commissions are something you should monitor pretty closely. I use Interactive Brokers and they are quite reasonable for the ES Futures options!
Hi Big Ern, great article! I just newly discovered FI a couple months ago, but have been actively manage my own money using options. I also just started dwelving more into the world of futures. Just wondering how did you manage your positions during the Feb 5th, 2018 drop? Looking forward to exchange some ideas.
I got hit pretty hard and lost on Feb 5, but percentage-wise less than the market. I made some money back since because implied vol was so high. But still, not a pleasant experience so far this year.
ERN, please help me understand your long-term confidence in the Sell Put strategy. Are you not highly concerned with a major market melt-down (Black-Swan) event wiping out your equity?
I am intrigued by your approach – but can’t ease my concern for the long-tail event. Any supporting comment will be appreciated.
The big left tail event would have to all happen in 1-2 days. Unlikely. Even during the GFC the decline of 57% was over 17 months.
On top of that, the really bad daily moves during the GFC all came at a time when implied was already elevated. I simulated this strategy during 2008/9 and it worked pretty well.
Would like to know how you fared in December, when I look at the PUTW ETF it got wrecked. Did you get wiped out?
You seem to refer to “bond funds” and “longer-duration bonds” sort of interchangeably here; am I correct to assume that you’re only holding bond funds (and if so, why those particular ones?) and not the bonds directly via IB?
I read ERN’s article as being pretty clear about a fund, but I am somewhat interested in this as well. I personally hold treasury bonds directly currently through TD with a large portion of the cash in my account. I have them in a ladder skewed towards the short end of the spectrum. I looked at the tax free muni’s, but their yield is really low on the shorter durations such that the after tax rate of the treasuries was about the same, but I would think the treasuries have less risk. I imagine sometimes it would be advantageous to hold the bonds directly and sometimes more so to hold the fund. For example, holding TLT with the same value as the bonds I hold would have lost me a significant amount of money over the last 3 days while the market also moved lower. However my bonds barely moved at all. But if bonds and the S&P had moved opposite each other as frequently happens holding TLT would have worked in my advantage. In all fairness I am holding shorter duration bonds than what TLT does. I just trade TLT frequently so it’s an easy comparison.
Yup, I invest in bond funds only. Individual bonds have high transaction costs for us ordinary retail investors. 🙂
Really? Do you have to pay to buy a bond at IB? I don’t pay anything directly as a fee to buy bonds through TD. As far as I can tell they just mark up the price slightly and show you that in the form of a lower yield to maturity. You get a volume discount as well, so the more you buy the higher the yield. If I wanted to sell the bonds back they would ding me on the way out as well. However this seems to work out to about 0.04-0.05% only buying them (I just bought some this morning and that is where they’re marked at relative to what I bought them for). On BCHYX, the expenses are 0.5%, NAD is 0.95%, NAC is 0.98%, PFF is 0.47%, so it seems to be about 10-20x higher for the funds. If you were buying longer duration bonds (more than 1 year), comparatively the fees would be even lower on the individual bonds as the bond fund is going to be taking that out every year where as I should only be paying once when I buy something. I am buying in $50k-$100k lots typically.
While I can’t speak to the costs of IB vs TD, another thing to consider in the context of this post is how holding bonds at TD differs from holding something with IB for margin-related purposes. I’m assuming this is a factor too, although it doesn’t specifically answer your questions if I’m reading this right.
I’m not sure what the difference would be there… At TD the margin on treasury bonds is 5% of the face value, at least in my PM account. However it reduces your cash balance by the full value of the bond. You would be charged margin interest if you took your cash value below 0. This works quite nicely with the type of futures trading strategies ERN has talked about in several posts. I can have a 2-3x leveraged position in the ES futures and still hold a large percentage of the net liquidating value of my account in interest-bearing bonds while only using ~15-25% of my available margin.
If you’re still doing this, John, can you give some sample numbers (including PM requirement) on a hypothetical $500K account so I can see how the numbers add up? Thanks!
Say you have $500k in your account. You use somewhere between 75-85% of your $500k to buy bonds, so say $425k of bonds. If you buy treasuries bonds, that’s about $21k in margin (or option buying power), so you have about $479k remaining to put towards put selling. At TD, if you sell futures they transfer the amount of margin required to hold the futures position out of your regular account into a futures account. It’s around $5500 per put (varies by broker), so say you want to run 2x leverage – you’d sell about 7 /ES puts for about $38.5k in margin. This would leave you with about $36k in cash in your account and ~$440k in margin / option buying power. I’ve been using SPX lately which requires about $29k in margin per put but doesn’t have the hit to my cash balance – it all happens in my normal brokerage acount (and I get 2% on my cash balance). So to do ~2x leverage with SPX you’d be selling 3-4 puts, let’s say 4. So that would be about $116k in margin used leaving you with about $363k in margin available and $75k in cash.
I’ve been doing about half muni bond funds and half treasuries lately, so that would use more margin – TD is holding about 1/3 the notional value in margin on the muni bond funds I’m in, so in this example, say $212k of muni bonds for about $71k in margin, $212k of treasuries for ~$11k in margin and $116k for the SPX puts. That would use $198k of margin leaving you with $302k free.
In general you should have plenty of margin / option buying power remaining when doing this. You need to know what your allocation targets are for both amount of bonds to own and number of puts to sell and not exceed those to avoid getting yourself into trouble. You would not want to use the extra buying power for other things, especially if those things were positively correlated to the S&P.
I’m getting confused when you talk about cash, John. At the end of the first paragraph, you say that leaves $75K in cash. $29K margin/put * 4 puts = $116K margin used for puts (as you said) + $21K margin used for treasuries (as you said) equals $137K margin used and $500K – $137K = $363K BP leftover (as you said). What is the $75K in cash?
Also, how do you get 2% on cash–is that the treasury interest rate?
What additional considerations do you make for a worst case scenario (correction) regarding PM requirement and beta test? Would you ever buy a unit or two to cut PM–even for a shorter duration?
Thanks for the concrete example. That is extremely helpful!
Too many levels of comments… Can’t reply directly anymore.
Even though you have margin or PM, when you buy something (like a bond) that uses up some of the cash you have in your account. If you continue buying marginable things, you can buy more than the cash value of your account but whatever you buy beyond the cash value of your account you will pay margin interest on. You don’t want to do that with this strategy. In TD, I look at the value of my ‘cash and sweep vehicle’ to keep tabs on how much cash I have remaining. TD lets me direct my cash sweep (the remaining amount of cash I have) into a mutual fund that pays 2%, so it’s not sitting entirely idle. However if I buy something the next day, I don’t have to sell some of the mutual fund to do that – it’s automatically done by TD. I think IB has a similar feature at the moment, but I think it’s a lower rate.
My considerations for a worst case scenario are encompassed in my leverage and cash management. Keep leverage low enough and keep enough cash on hand to deal with draw downs. I don’t like put spreads. If the market drops and you’re losing money, that’s an even worse time to be buying a put to define your risk as compared to buying on trade entry. I would only do that in a really crazy scenario, but you’re limiting how much you can make back of your loss by doing that so I might just close the trade early.
Good point! I believe funds, ETFs and individual bonds are all marginable.
Once I stop working (without the compliance hassle from work) I will definitely check out holding individual bonds directly!
How about bid-ask spreads? I am not interested in Treasuries, but exotic low-volume CA Muni bonds!
There is no published bid-ask spread. That is effectively the markup (or lower yield to maturity) that you’re quoted. I think you’re just trading with the broker which in this case is functioning as a bond dealer. So they make you a market in whatever you are interested in of their inventory and you can either buy at the price offered or not. Also, it seems like someone is looking at each trade. I’m not sure if it’s manually done or is computerized, but the trades don’t execute instantly. To look at the spread, what I’ve done is looked at the yield on a 1 year bond that I can buy at TD and compared to the published rate on that day and noticed a slight difference (around what I quoted before) or looked at where they marked the bond after I bought it and it was marked slightly lower (that same 0.04-0.05%). Then if you sold it back to them, you’d get charged about the same again. So I think they’re marking in the middle of the effective bid-ask spread and you have to buy at the ask and sell at the bid.
They do have corporate and muni bonds as well as the treasuries I’ve been trading. They only sell from AAA to A rated, so the yields might be lower than some of the stuff you’ve been looking at. BCHYX for example lists it’s average credit rating as BB so I’d expect a higher yield. They do have some CA muni bonds – just quickly looking I found, for example: SAN FRANCISCO CALIF CITY & CNTY ARPTS COMMN INTL ARPT REV REV REF BDS AIRPORTS REV ALL BONDS Ser 2011B, Non Callable, CUSIP: 79765A7U4. That has ~3 years to expiration, they have $450k of them and it yields 1.652% to maturity. They list the minimum buy at $50k, although sometimes what they list overnight is higher minimums than during the day – not sure why. Orders only execute during the normal trading day. If you go longer duration, yields go up to a max of 4.3% in the A-rated munis that they have at the moment.
I scanned through the Muni bonds they offer at Fidelity. No bids, just ask prices. And the yields aren’t that great. Not really anything that would beat my Nuveen closed-end funds with 5-6% tax-free yield.
Are you talking about NAC? If I’m calculating this wrong please let me know (I’m just getting into bonds), but looking at the last year it looks like you would have lost about 4.8% after including the distributions and price change. Going back 3 years, it looks like you would have made 0.6% per year on average. So even though it has paid distributions at a rate of 4.7% over the last year given the year ago price, it seems like the total return in the recent past is worse than the ~1 year treasuries I’m buying at the moment even after considering the federal taxes on the treasury bonds. And you could get more yield in the treasuries by going longer term which is probably more comparable to the duration of NAC (but that is a guess – I can’t find an average duration).
Going longer term in Treasury bonds exposes you to the same duration risk, i.e., you lose money when yields go up.
Also, keep in mind that I pay close to 50% in marginal taxes right now (federal+state+Obamacare). The 2.3% yield on the 1-year T-bill is then only 1.15% after tax for me.
If I get 5.3% yield on the Muni fund and I lose, say, 4.7% for a net gain of 0.6% I would still beat your T-bill. That’s because I regularly (quarterly) harvest losses from the Muni funds and the 4.7% loss generates a 2.35% offset to ordinary income (0.4 times my Section1256 gains plus $3000 of ordinary income each year with carry-over), so the 0.6% gain according to your calculation is actually a 0.%6+2.35%=2.95% after-tax return.
And at some point, the bond market might stabilize again and I get to keep a lot more of the 5.3%.
But your tax situation might be different. Either way, I wish you best of luck with your T-bills
Actually my previous comment was somewhat misleading as I was comparing past performance of NAC to current treasury yields while prior treasury yields were obviously lower than they are now. To compare to NAC, if you had also bought a 3 year treasury bond 3 years ago (when rates were much lower), it would have returned something like 0.5-0.6% per year after taxes depending on your tax bracket, so about the same as NAC. If you had bought a 10 year treasury 3 years ago, you would now be down money on it – I think about 0.3% per year on average over 3 years (not taking into account what you would have paid in taxes on interest received). If you had bought a 10 year 1 year ago, I think you would currently be down about 3.3% on it (also not including taxes on the interest). And if you had bought 1 year bonds 3 years in a row, you’d be up ~0.7% total or ~0.2% per year on average including the effect of taxes.
Again, do the same exercise with tax loss harvesting and a 50% marginal tax rate on ordinary income. You’d come out ahead with the Muni fund investment.
I’ll have to think about the tax loss harvesting. Haven’t done that before. However you don’t have a 50% marginal rate for treasuries – you don’t have to pay state taxes on them. I’m in the same marginal bracket as you.
Ah, that’s right! “Only” 40% marginal rate for those! 🙂
But check it out: rotating between the CEFs and harvesting losses makes a lot of sense and can offset some of the short-term gains from the 1256 contracts!
After doing the calculations for tax loss harvesting, I get your point that you have quite a large margin for taking losses on the muni fund before dropping down to the yield of the treasuries. For the NAC example going forward, it looks like you could lose about 7.5% of principal before dropping to the total portfolio return a 1 year treasury bond would currently provide (I’ve been doing short duration as that seems to be the more attractive area of the yield curve to me currently). A bond ladder of 1 year treasuries would have been better over the last year as rates started to rise, but only barely.
After reading some more about closed end funds and looking at the historical performance of NAC, the two things that still worry me are the bad performance in a rising rate environment (as we’ve been in for the recent past and quite possibly will be in for at least the near future) and how the fund performed in 2008. I see NAC lost about 40% of its value at one point in 2008 before bouncing back. It looks like about half of this was the discount to NAV increasing. This would seem to be a fairly undesirable characteristic in a position that is supposed to be offsetting an equity position which would have also been taking some substantial losses around that time. It is a bit tempting at the moment though given it’s towards the high end of its historical discount to NAV. It would be a big boost to overall portfolio returns if it bounced back in price.
You got that right! Rising rates are poison! And illiquid investments got hammered during the GFC, but I’m not sure this will repeat in every garden-variety recession.
But if you have the guts to have this fund around the time the rate hikes level off you can make a lot of money! And it’s all on top of the options trading! 🙂
The Nuveen funds you mention below are all leveraged 40%. Isn’t only the non-leveraged (60%) part of the muni return tax free?
It’s a nice “hack” in the US tax system. If the leverage is “packaged” into a fund like those offered by Nuveen, the entire interest income is indeed tax-free.
If you had leveraged Munis yourself, you’d run into the issue with “payments in lieu of dividends” and potentially partially taxable income.
What bond fund are you currently using?
Muni Mutual Fund: ABHYX
Closed-end funds: NVG, NZF, NMZ
Various Preferreds: Goldman Sachs, State Street, Wells Fargo, PBCTP, Morgan Stanley, Ally, Allstate, etc.
Right: I don’t hold any bonds directly. I use bond mutual funds and ETFs or Closed-End Funds. BCHYX, NAD, NAC are the most prominent ones. PFF and PFFD for the preferred shares.
Hi! I just started reading your blog and love the detailed analyses! Two questions regarding the put option writing:
1) From what I’ve researched, options on SPX (not the SPY options on the ETF) are settled in cash and Section 1256 applies to these as well. Is that correct? My broker does not allow me to write naked puts on futures, but I can write naked puts on SPX.
2) How are you computing the risk per annum of 6.4% on all your trades?
1: I’ve never traded the SPX options (not allowed right now due to compliance rules at work). But it sounds correct; they should be settled in cash. But double-check with your broker and the exchange.
2: This is just the annualized realized standard deviation, i.e., monthly times sqrt(12).
Forgive my ignorance but given the use of leverage, I presume you close out your losses prior to assignment of the futures contract? I work in finance but newer to put writing so right now doing only cash covered puts w/o leverage.
I guess the original thought I had was to write at a steeper 5%+ discount to weekly SPY/QQQ and be happy to get assigned those ETFs if the trade went the wrong way (since this is money that would have been allocated to equities anyway and historically 5%+ weekly down legs are more short lived outside of a recessionary turn). Had then thought to write covered calls on those long SPY/QQQ positions I was assigned.
Your concept intrigued me though so I was just wondering 1) about how/when you decide to cut bait and (presumably) buy to close your position on the e-minis and 2) if anyone has explored put writing then covered call writing upon assignment–would imagine that is more conservative but again I’m newer to this and could absolutely be missing something critical.
Any thoughts you guys have would be greatly appreciated!!
With the futures, the margin doesn’t change if you get assigned (at least at my broker) and your directional risk doesn’t change after assignment (you would have the same deltas). So the reason to close out your puts before assignment if they’re in the money or close the futures after assignment would be if you didn’t want to continue to have that level of directional risk. For example if you wanted to be long at that level of leverage at that price level, you could just leave the position on. However personally I would likely close the position and reestablish a new options position in the next expiration cycle with the same delta I always use in order to get back to a higher probability of success.
With ETF options, if you get assigned it can become a cash management issue. The cash to buy the stock will be deducted from your account’s cash balance. If you’ve already used a large amount of your cash balance to buy bonds, this could cause you to incur margin interest charges. Some brokers also charge high assignment fees. For example, TD charges $20 for assignment per option strike on their basic rate plan which is pretty terrible. By the way, commissions are negotiable if you have enough money and do enough business with your broker.
Covered calls are the synthetic equivalent of naked puts. You can do either with the same result ignoring any trading fees. However if you’re using equity options (like on SPY), the covered call will use a bunch of your cash. For that reason and for trading fee efficiency I’d prefer the naked put and would try to avoid assignment.
I’ve been trying this short term put selling strategy for the last month or so. It’s been working quite well which makes sense as the market has more or less gone sideways. I have made money on every put I’ve sold. I’ve also thrown in a few short calls. But as an example of closing losing positions early, yesterday (Thursday) I had on options that I’d previously sold for ~$4 that were expiring today (Friday). The market dropped enough that almost all my options were in the money. One of those options went as high as ~$19. Others went up even more. Then the market reversed back up and finished the day unchanged. I ended up making a nice amount of money for the day. Today I closed all those options for a profit as the market moved up. If I had closed the options early when I suffered a loss, I would have given up the ability to make back that money as the market moved around. However it’s also possible I would have then sold new options at lower strikes in the next cycle which could have helped recover some of that loss.
I work at a big bank and looks like futures might be covered securities, so would have to go through the hassle of pre clearing all of them which wouldn’t really work.
In lieu of the futures contracts though, can I synthesize that with cash settled index options?
I looked at CBOEs website and there is the XSP which is 1/100 of the SPX so the same exact size as SPY.
Since they’re cash settled, XSP would get the preferable 1256 Tax treatment of 60% long term and 40% short term cap gains.
They’re also European style so won’t have to worry about it early assignment.
If I’m thinking of this correctly, cash settled simply means netting the current index price vs the strike price at exercise date. So presumably wouldn’t I have the same leverage benefit as S&P futures and avoid the cash management issue that options on SPY present?
Trying to think of a good workaround to futures and this seems like it might be it. Am I missing something?
Sorry 1/10th the size. Fat fingered that. My bad.
You’re right that XSP is cash settled and thus has no assignment risk other than your position being converted to cash. There is no underlying stock to be assigned to you – you can only trade the options. I seem to remember the cash settlement has some oddities to it, but I’ve never traded it. However I personally would not trade it as the volumes in the options are really low and I would worry about the liquidity. I probably would have been in the 263 puts on Wednesday and the open interest is 60 contracts. Compare that to SPY which has about 20000 open contracts. SPX has 1550 open contracts and ES has 1300. The bid/ask spreads look equivalent to SPX and ES after adjusting for the notional value multipliers though (SPY looks a bit tighter). I suppose you could try trading it and see if it’s easy to get your orders filled near mid price.
Also note that the margin is higher than for ES. It’s basically the same as for SPY at least in my portfolio margin account, or 2x the margin on ES. In a regular margin account it would be even higher. It think it would be ~3.7x the ES margin but I’m not 100% on that. You could still achieve 3x leverage but it would use more of your available margin.
Thanks! For some strange reason, a naked short put requires more margin than a long ES future.
I think we have the same approach: simply close out the assigned ES future with a delta of 1 and replace with a new out of the short put with a delta much smaller than 1. THat’s what I do every week! 🙂
If I get assigned (2-3 times a year) I’m usually so frustrated and risk-averse that I prefer not to have the high delta of a long ES plus short call (at the money or – even worse – out of the money).
If I get assigned I close the ES contract, then sell another way out of the money put on the ES. Try to make the money back over the next weeks (sometimes months).
You mentioned that you’re writing on the MWF settled options now. Just wondering how our of the money are you targeting with those settling in 2-3 days vs the weekly settled?
Do you know where you could get historical data on the prices of put options contracts? I assume you could use the historical VIX, S&P 500 and T-Bills data to calculate the theoretical prices of the options using Black-Scholes, but I’m curious what was actually available in the marketplace? Thanks.
Certainly not for free. And it would be very messy data: All the different strikes and expiration dates for all the different trade dates.
That’s kind of what I gathered from google searches. How then do you backtest your strategies? You mentioned above having clean options pricing data going back quite a ways.
When evaluating the value proposition of a particular option, how important is the ratio of the market price to the theoretical price? Do you have a heuristic for comparing implied volatility to VIX?
It’s not an exact science! But this is how I would do it:
1) from experience I know that if the VIX is at X then the implied Vol (IV) for the options I normally sell is at Y. Sometimes X=Y, sometimes Y>X, for example after large vol spike. In January when the market rallied Y<X. I "estimated" the relationship between X and Y and apply that to the back-test.
2) simulate how far out of the money would have been the puts I'm selling at each previous option selling date.
3) hold options to expiration and realize the P&L then.
It's an approximation only, though!
I have played around with calculating theoretical options prices using Black-Scholes to simulate past strategy performance. It is very difficult – not to calculate prices, but to get input data that is trustworthy. The option prices you come up with are completely dependent on your input variables, and it is pretty tough to accurately model implied volatility lacking detailed historical data. I have used historical VIX data, but that does not account for the variation in volatility across different delta options (the volatility smile). It also doesn’t account for the variation of volatility across different option expirations. So while your estimate of large losses will probably be in the ballpark as those depend more on the difference between the underlying’s price and the strike price of the option sold, the question of how much you make in between drawdowns can only be a guess, and that is the real question. I’ve tried to approximate the inputs and I get P/L vs time curves that look similar in shape to the actual data ERN plotted from his own trading in this post, but I think it can only ever be an approximation.
I never look at theoretical value vs actual option value. I do look at the level of the VIX relative to its past levels as an indicator of when volatility is high or low, but after you’ve been trading for a while you can just glance at the option prices and see the same thing.
Not an easy task. But I proposed a plan, see above.
The challenge is that the IV of your options is not exactly equal to the VIX index, because a) the VIX is for longer-dated options than what I sell, b) there is a vol smile/smirk, so the IV of my options is most likely not equal to the VIX. For back-testing one has to make some assumptions… 🙂
I have a few questions coming from the perve of a layperson.
Why was 3X leverage chosen? Is this the optimal amount of leverage or is it possible to get higher returns with higher leverage ?
Would it be possible to use this strategy for one’s whole portfolio and attain a higher withdrawal rate?
I think the article explains the choice of 3x leverage pretty well. My additional take on leverage is that it is about balancing risk versus reward. If you are not losing money, more leverage is great – you will make more money. But if you have a loss, you will obviously lose more and you have no way of knowing when the market will drop so it’s not possible to time this. With selling options, you already have the possibility of encountering very large short term losses relative to your potential short term gains so balancing risk vs reward via leverage is fairly critical to the survival of your portfolio. If you use too much leverage, a sudden drop in the market can wipe out a massive chunk of your portfolio. For example, in my own backtesting of this strategy (which I only trust as a very rough approximation), if I used 3x leverage through 2008 I might have had a 35% draw down but still possibly made money that year. If I used 6x leverage, I might have had a 55% drawdown. Going up to ~15x, I probably would have lost all my money. In 2015, I might have lost all my money using 12x leverage. 20x leverage is around the limit just based on margin rates for ES futures although your broker might limit you to less contracts than that to prevent a fat finger trading error.
I have personally used 1x-6x leverage in my own trading in the past (with similar options selling strategies), and that is the broader range I have heard recommended by professional traders. I’ve only gone to 6x in very extreme situations. 2x-4x would be a more normal range. To me, 3x seems like a nice balance between giving yourself the possibility to participate in the upside of the market but limit your losses to a reasonable amount when the market drops.
Whether you use a strategy like this for your entire portfolio is up to you and your level of risk tolerance. While you have some diversification in so far as you are trading an index, you will not have strategy diversification so if anything goes wrong, it goes wrong for your entire portfolio.
Thank you for giving me such a detailed answer. My questions are out of curiosity to understand the mechanism behind the decisions. I have no doubt that Karsten didn’t just pick 3x levergate out of nowhere and it’s probably the most optimal.
I’m curious as to why further leverage is bad. Is further leverage diminishing in returns? E.g. if over the long-term underlying gains 7%, 3x gains 14%, 5x gains 16%
I’m just making numbers up but basically is higher leveraging getting you very little extra return for a relatively higher risk if ruin?
Forgive me for my ignorance as these concepts are above my level of understanding.
Further leverage is bad when you lose money. If you have too much leverage when a down move in the market comes along, you can lose all your money in a very short period of time. If you are early in your career and looking to take tons of risk with a small amount of money, maybe that’s not the end of the world. You hopefully learn from it and move on. If you have already retired, obviously that’s extremely bad. Now you need to go find a new job. However on the flip side, increasing your leverage will linearly increase your gains when you’re making money. For example, let’s say I used an average of around 1.5x leverage over the last two months and made around $30k. If I had used 3x leverage, I would have made $60k if I had made all the same trades. I just would have traded double the number of contracts for each trade. One day during that time I was down $11k in one day. If I’d used 3x leverage I would have been down $22k.
So higher leverage gets you more money on winning trades, but your losses will also be larger by the same ratio. Because the losses, while hopefully infrequent, can be much larger than all the little individual gains you collect, you need to choose a leverage to use that doesn’t cause you to go broke when you have a loss (and statistically you should expect to have losses – selling the 14 delta put should mean you lose money ~14% of the time given a large enough sample size and the probability that the underlying’s price touches the strike price of the put during the trade and thus causing you a temporary loss will be about double that or ~28%). Unfortunately you can’t know what will happen in the future so some safety margin needs to be built in based on evaluating past results.
To be clear, if you somehow knew your strategy would be overall profitable and that your largest loss would be X, you could pick a leverage such that you maximized your total return while avoiding going broke. However there is no way to know that ahead of time, and thus some safety margin is required. Additionally, the higher the leverage the higher the day to day swings in your portfolio so unless you have nerves of steal, there is a psychological advantage to a bit of moderation in how much leverage you use. You’re more likely to stick with it if your portfolio isn’t giving you an ulcer. I hope that helps a bit.
Thanks for the explanation John. I kinda already got the double-edged sword feature of leverage and avoiding leverage that would cause a Wipeout.
I was hoping that 3x, or as you mentioned 1x to 6x leverage is recommended because someone simulated or back tested and showed that these are the safest levels.
I guess there’s no way to definite know except maybe 3x appears to be safe enough. Although nothing is 100% safe , it gives the best return for low risk of ruin.
ERN has mentioned in the comments that he did some simulated back testing for this strategy. He also has implemented the strategy for a while as he talks about in the article, so some of the advice on leverage comes from his practical experience. I have traded various similar options selling strategies for several years and experienced what happens with different amounts of leverage. I have also done my own simulated back tests of this strategy. In my simulations running from 2005-2018, you would have had unacceptably large draw downs as you increase leverage above 3x, like losing more than half your money. The wider leverage range I mentioned comes from Tastytrade. They have a bunch of ex floor traders who have a lot of trading experience with short options strategies. They have also done a lot of back testing of various options strategies using real historical options price data (not simulated) although I don’t remember seeing them try to study this particular question.
So to summarize, the 3x leverage guideline comes from a lot of simulated and real world data points.
Oh okay, that makes sense.
In your simulations from 2005-2018, with higher leverage, the drawdowns were significant, did they end up recovering?
Do most scenarios have the end portfolio wealth less than just 3x leverage?
Higher leverages did simulate as recovering and earning more money until the leverage level where all the money in the account was lost in a draw down. I looked at up to 10x, and that was still predicted to make increasing amounts of money past 2008. However 10x had a ~92% loss in 2008, so given that my simulation probably needs to be interpreted with a very wide tolerance on the results and that you don’t know what will happen in the future, I’d say that would be very dangerous to actually implement.
Looking at 3x up to 10x, the higher leverages always produced a higher ending value. This is simply because in the model the same trades are being made on a given day, just with different quantities based on the leverage specified. So if 3x made money over the time period I looked at, 6x should make double. However losses would be larger in magnitude in the same way. So the trend of higher profits would continue until you reached a leverage where the biggest loss in the series took the value of the portfolio to zero.
Also I suppose it’s worth pointing out that I’m assuming starting with a $1M account in 2005, so there are ~3 years of gains before the big loss in 2008. If you had a much smaller account and started right before 2008, you might be forced to stop trading with a smaller percentage loss. That is, if your account dropped low enough in value that trading 1 futures contract was too much, that would be the same as losing all your money from the stand point of being able to continue to trade.
Thanks John for indulging me on this topic. Your replies are very enlightening.
Nice explanation! Let me know if you ever want to write a guest post on this! 🙂
Thanks – I take that as a big compliment coming from you, as your posts are very good. However I’m not sure I have much more to add to what you’ve already covered in your articles and in the comments. However I’ll keep commenting as I enjoy discussing this stuff.
I agree: great comments, John! You could definitely write a guest post or two (or maybe you already have since I’m new here).
Great post by you too, ERN. I find your stuff outstanding.
I just wanted to note two things here.
First, in a few studies, TT has found optimal allocation to be about 30% for straddles and/or NPs. That’s roughly 3x.
Second, you make a good point about the three years of gains leading up to 2008 boosting the account value. I think this suggests the argument that when looking at risk in terms of maximum drawdown percentage, you should take that as a percentage of initial account equity since that largest drawdown could theoretically happen from Day 1.
John, you do a great job explaining everything. I think ERN mentions below he typically writes puts with a 0.05-0.10 Delta, just curious what your chosen Delta is?
I’ve been testing a few levels and seem to find myself gravitating toward mid teens which would typically be a little over a 1 standard deviation event
I’m glad you find my comments useful. I’ve been selling the ~0.14 delta puts with this shorter time to expiration strategy. If volatility was higher, I might lean towards lower deltas as you’d still be able to collect about the same money as in low volatility at the 14 delta put, but at some point the trading fees become too large a portion of the money you’re collecting and you still have about the same risk in a big down move. I’ve been targeting selling something for 2.00 or more. On the other hand, going too close to the money doesn’t leave a lot of room for the market to move around which would make your account value fluctuate more. I’ve also been sticking to slightly longer days to expiration. I’ve been selling 2 cycles out instead of the next one, so if it’s Monday and I close puts, I might sell Friday puts instead of Wednesday. I’d look to close those when I’d made most of the money on them and move to selling in the next cycle instead of waiting for them to expire and make that last little bit of money. From my ‘back testing’ simulations that seems to reduce the volatility of returns while still making about the same overall return as holding to expiration. Only time will tell if the same thing happens in reality.
Also, since my last post I found a bug in my simulations that slightly over stated profitability. Thus, when I previously said that 10x leverage still eventually recovered and made money I would now amend that to something more like 6x leverage went bust in 2008 and 5.5x recovered although with a massive drawdown. So at least for me this sets an even better bound on leverage and explains why 3x is a good target. But again all this simulation stuff should be taken with a huge grain of salt and the input parameters can vary what predicted returns are quite a bit.
Interesting on the off cycle. Just curious, at what level do you close out the position? 80% of premium? 90%? Some dollar level like $0.20?
I’ve played with various levels both in real trading and in simulation. I’ve done anywhere from 50% to 95% of max profit depending on how things have been going for me. My simulation shows that closing at a fixed percent of max profit is not very good, but that closing at the end of the day on a day when your options are above some profit limit tends to be good. I’m interpreting that as showing that if the market has an up day, the short term puts tend to swing towards a higher percent of max profit on that day and trying to capture all of that is much more profitable over time instead of just closing at a predetermined percentage. But again, I haven’t verified that in the real world yet.
I sell very short-dated options now. With only 1-2 trading days. I simply keep them to expiration and then roll into the next batch that same day.
ERN, if you’re holding to expiration do you then sell your next batch after hours (after 3pm PST) or do you wait until the next morning? I’ve been avoiding trading after hours unless something interesting is happening as the bid/ask spreads are wider. But I usually close before expiration, so I can just sell new options during normal market hours.
I’m selling the next batch during the day if the old options look like they’ll expire worthless. Otherwise, I’ll wait until 5 min to expiration, sell the new contracts and (if necessary) sell the enderlying ES futures a few seconds after I get them. There is an additional 15 min of trading after the 4pm close with enough liquidity for the underlying. But liquidity is poor for the ES options after 4pm.
You got that completely right! Doing too much leverage in a strategy with negative skewness can create a loss big enough that you can never recover from it. See what happened to the XIV ETF strategy. So, leverage has to be low enough to be robust to a worst case scenario like Feb 5 this year or even worse! 🙂
Very nice explanation! Thanks for sharing! Glad we agree on the leverage level! 🙂
It’s a matter of risk vs return trade-off. Since I sell options quite a bit out of the money with a Delta between 0.05 to 0.10 I am comfortable doing a bit of leverage. But not too much because the 0.15-0.30 Delta can go to 3.00 if the market tanks.
Also keep in mind that this all before taxes. 3x leverage becomes 2x leverage after paying marginal taxes. 🙂
I’m sorry if I am just super ignorant for asking this again.
I understand conceptually what John explained about leverage but when you say “too much leverage”, is there a way to quantify what is ” too much”?
Is 4x considered too much? Potentially causing irrecoverable portfolio damage?
Johns back testing seems to propose that 5x is the upper bound, albeit that doesn’t mean it’s the most optimal.
I personally have a risk model that calculates the loss from a large equity drop. Large equity drop in terms of multiples of standard deviations: Both derived from the VIX and past realized equity vol. I try to stay under a 10%-12% loss for the really, really catastrophic equity losses, say, those that happen only once a decade or so.
Thank you once again for explaining your process. I think it’s still a bit over my head. I don’t even know where to begin so I’ll need to do a bit of self-studying. I am itching to try this out once I can spare the minimum capital required and I’ll probably just stick to 3x leverage.
I made a bit of a mistake. I was meant to say “I have no idea where to begin in regards to risk modelling”.
If you’ve never traded options before, I would suggest starting with the minimum possible – 1 contract. Get a feel for it for a while before you ramp up the leverage.
That’s is very prudent advice. I am hoping to study a bit more, maybe even getting the derivatives textbook by Whaley that ERN recommends. I’m not going to jump into derivatives before I learn a bit more but I am very excited about this strategy.
Karsten, I have a few questions regarding the practical implementation.
Since futures options and futures trade almost 24 hours but the bond funds we keep as a cushion trades during market trading hours, what happens during a margin call?
Would IB liquidate your options/futures since it’s marked to market after hours when we would rather be selling the bonds to maintain our position?
From what I’ve read, it sounds like you’ve managed your risk enough so that you’ve never had to face it but I was just wondering what the practical implementation be during a margin call especially during market crises.
Thanks again for writing this article. I am so excited to one day implement it.
That has never come up because I hold about 60k in margin per short our contract, about 5x the minimum margin.
It’s my understanding that the margin call would come whenever your cash falls short bur the forced liquidation would occur some time around the end of the trading day. So you should have time to sell other assets before then. But keep in mind that those other assets are marginable, usually up to 75% of their value.
Forgive me if I’m misunderstand the math. Isn’t it 45k of we’re applying 3x leverage at current index level?
I think I maybe misunderstanding the 60k portion. Sorry for the stupid question.
Looking at this another way, the market has overnight price limits that only allow it to move +/- 5%. This was hit after the 2016 US election. The futures markets dropped 5% and then sat there. So roughly, say you have $45k per contract and you need ~$5k in margin to hold the position, so you have $40k of money to lose before a margin call. You would need to see a drop of 40,000 / $50 per point = 800 points on the index which would be ~29% at the current value (800/2760). So it’s extremely unlikely that your overnight margin call scenario would ever happen. You would have to have multiple massive down days in a row first to get down into that -30% range. In that case your broker would possbibly start raising the margin requirements to hold the contract, but you would be dealing with this during the day, not at night. For example I think my broker raised their margin requirements in February when the market dropped and has recently relaxed them.
During the day, there are other price limits. They are at 7%, 13% and 20% and are set to coordinate with circuit breaker provisions of the NYSE. You can read about this at the CME’s website. Search for faq sp 500 price limits. They can restart trading during the day at the lower limits but at 20% they stop trading for the day (and overnight until the next morning), at least as I read the rules.
John, thanks again for your explanation. You’re contribution to this comment section has been immense. I really appreciate it.
That’s good go know about the various price limits. I’m doing my research and looking at understanding the worst case scenarios and just the downside in general to inform myself before trying this strategy out.
Karsten, I’ve been practicing using IB demo account and I think I have the basics of choosing the option with the desired delta and IV but I have a question regarding price.
What is your process on selecting the premium price? I’m assuming we’re placing a limit order?
Are you choosing the bid, ask, last price, or something else as the price you would like to sell?
Bob, I would always use a limit order. If you want to sell something right now you can sell at the bid price. Someone is offering to buy that option at that price so if you put in an order to sell at that price they will buy from you. If you put an order in to sell a put at the ask price, you won’t sell anything until the price of the index (or whatever you’re trading) moves down enough that someone is willing to buy at your price. With most of the /ES options I’ve been trading, there’s usually a spread between the bid and the ask of around 0.15. The tick size on the /ES options is 0.05, so I’ll usually put in an order to sell at 0.05 above the bid and usually get filled right away unless the index is rising. If I’m selling while the market is dropping, I might put it 0.05 below the ask knowing that the index might drop enough while I’m entering the order that someone will buy from me at that price. A quick example – I want to sell a put with a bid of 2.00 and an ask of 2.15. If I want to sell immediately, I put in an offer to sell at 2.00. If I put in an offer at 2.05, it’s likely I will also sell immediately but not as certain. If I put in an offer at 2.10, I probably won’t sell until the price of the index moves down a bit causing the put to be worth a bit more (at which point the bid probably would have moved up to 2.05).
If you put in a sell order for a put at a price higher than the ask, nothing will happen until the index drops enough that your offer price ends up between the bid and what everyone else is asking for that put. That can work, or you might never sell anything at that price. The ‘last’ price is the price the last time someone traded that option. It could be 5 seconds ago, it could be 5 days ago. I wouldn’t base my decisions on that.
Assuming you have a platform with realtime quotes, you can watch the bid and the ask move around as the index moves around to get a feel for how all this works. Sometimes it’s a moving target and you’ll have to adjust your order to get filled.
Also, given that you’re asking this question I’m guessing you haven’t done much trading. Nothing wrong with that – we all are new at some point. I’ll just point out that if you want to start out smaller to get a feel for this stuff before jumping in the deep end with futures, you could start out trading SPY options which would be 1/5 the size of /ES options. There are some other issues there with margin, taxes and assignment possibilities that are discussed further up in the comments but it’s something to consider.
John, I’m going to say it again, you’re such a valuable contributor to this thread.
You’re right, I’ve never traded any options. I’ve only ever traded individual stocks and index ETFs. With stocks, I’ve done my analysis and know the price I want to get in or out so I don’t really let the bid/ask affect my limit order.
While I was playing with the IB demo, I realised I had no idea how to “price” an option. I was assuming that there was a mechanical way of choosing the price because we think the market is efficient enough to set a fair price. I wasn’t sure if it was generally the bid, ask, last or something else.
You’ve really cleared it up for me. I’m sure I’ll have more questions once I actually start trading.
I don’t have enough spare capital to deploy yet.
In the meantime, I’ve ordered the Whaley textbook. Hopefully it’s not above my head.
Happy to help Bob.
I would agree that the market is mostly efficient, so an option is priced accurately at any given moment as long as enough people are trading it (which you can deduce by looking at the bid/ask spread). A tight bid/ask spread is desirable – usually around 0.10-0.15 for /ES options. After hours when no one is trading, the spreads will widen out to maybe 0.30 wide or so. I would not try trading then. Sometimes the bid/ask spread will also widen if the market is making crazy moves. You can try to trade then but it can be more difficult. Anyway, the ‘fair price’ would in theory be halfway between the bid and the ask. However most of the time you have to give the counter party (who you’re trading with, probably a market-making firm) a little something to get them to take your trade. Usually 1 tick or 0.05 below midprice in the /ES options is enough.
You can do the same kind of price analysis for the S&P that you would for any other stock if you want, but with these short term options that really just means you’ll end up waiting around a lot until the market happens to move the way you think it should. By then the option you put in an order to sell might have experienced enough time decay that you can no longer sell it at the price you wanted to and you would have been better off selling something sooner and collecting some time decay premium. So I would summarize the thinking for this strategy as: the market and options are always fairly priced, so always keep selling options and collecting premium. You’ll come out ahead compared to trying to time your entries. It’s kind of the same as what a lot of index investors say: time in the market beats timing the market.
I have a question regarding those crazy moves or huge drops.
Karsten wrote that some of the best times for shorting puts are after a big initial drop because really OTM puts are selling for rich premiums since everyone is scared.
How does one choose which strike prices are considered really rich at these times? It it that 0.15 or lower delta options have a really high premium attached?
Just to clarify the mechanics of choosing the right put to sell, am I correct that we’re looking for a put with delta less than 0.15, and IV higher than VIX? And making sure that annualised gross return is attractive e.g. circa 7%?
When the market drops, all the puts sell for more money. Just keep selling the 10-15 delta put.
I would not look for an IV higher than the vix unless you only occasionally want to sell puts. If you do that, you will very likely make less money. I would just sell a put with a delta in the 10-15 range.
Thanks again John, I have learned so much from you. I am beyond excited to add option selling to my arsenal.
Do you trade options on individual stocks whether it be puts or covered calls? I mainly want to follow Karsten’s short put on ES but now I’m also interested in squeezing extra juice out of my individual holdings?
Is 10-15 delta still the thing to look for when selling covered calls ?
IB displays the time value (in %, annualized) and I like the 5-7% time values. With a little bit of leverage.
Also, I’d put in a limit order at the mid-point between bid and ask if there’s a gap of more than one tick.
Karsten and John, I have a question regarding holding the bond portion of this strategy.
From what I remember, Karsten holds preferred stocks now because bond returns are expected to be low, especially as rate rises.
I thought the bonds perform two functions. 1) Juice up returns compared to cash, and 2) Being an uncorrelated asset to cushion during those large drawdowns.
I’m guessing prefered stocks have a higher stock correlation. Is it a better strategy to find extra return for less diversification, or would holding bonds be a godsend during large drawdowns?
How far can one take this, would a high yield ETF be OK as the “bond” portion?
I have some preferred ETFs (PFF and PFFD). But the bulk is still in Muni Bond funds and closed-end funds.
I find the HY bond funds too correlated with the equity drawdowns.
Getting a replies from you and receiving the Whaley book in the mail is like Christmas!
I’m not sure if I can upload images but that link has a screenshot of different options.
From that screenshot, would I be choosing the 2775 or 2780 strike?
Also does choosing the strike based on time value change just after a large market drawdown where you mention you can sell very OTM puts because everyone’s scared?
EIther one works. The 10-delta or 5-7% yield are only rough guidelines. You have the discretion to pick what you feel most comfortable.
Most definitely, the numbers would be quite different if the index had just fallen by 5-10% over the last few days. In that case, the 10-delta put would be way out of the money! 🙂
I see. So after a big drop, we’re still choosing a 10-delta put but the strike price for a 10-delta put will just be way out the money and I’m assuming higher yield? I think this is becoming a lot clearer. 🙂
Thanks ERN! I feel like I am pestering you but you’re the most knowledgeable person in these topics. I’ve honestly never come across anything like your analytical writing regarding finance topics.
I shudder to think what my FI understanding would be like if you never decided to start blogging. I would probably be running out of money with the 4% rule or something.
Exactly. Just one example:
June 24, 2016. The trading day after the Brexit vote. I sold puts with an expiration 6/30/2016 with a strike 1850. The ES future was at 2033 at that moment. Around 9% out of the money. The annualized yield was 7.25%, the delta only 0.05. Alternatively, I could have done a 10-delta for an even higher yield.
(I keep an Excel sheet with all my trades: close to 9,000 so far!)
In those situations, is it better to go with the safer strike price with a reasonable yield or would a 10 delta put still be okay?
Perhaps it’s hard to say because every “crisis” is different?
I was so happy to see a new SWR article today. Nominated your blog for every relevant category for the Plutus. You are by far the greatest blog writer I have ever seen.
My understanding is that shorter dated options offer richer premiums. If that is true, why has the weekly put index (wput) underperformed the monthly put index (put) for the past 10 years? This paper compares them all nicely but does not really address this issue other than increased transaction costs. Could that be the reason?
I doubt that t-costs are the reason. Personally, I have had better experiences with the shorter options, especially the Monday, Wednesday, Friday options.
One explanation for the results from CBOE is that most abrupt drawdowns are just long enough to cause a whipsaw in the Friday to Friday options. But the monthly options are saved by an eventual recovery. Might be a reason to go with the shorter than weekly expiration!
Maybe they need to create a new PUT index for intra-weeks. The “PUTern” LOL Thanks for your help! Any chance you would do mentoring to earn some side income in your retirement?
Haha nice word play!
Without a legal structure (RIA or the like) I’d not be comfortable giving for profit money advice. I’m thinking about setting up something once we settle down early next year! Stay tuned!
Super. I would hate for you to “work” in your retirement but if you are having fun doing it and on your own schedule I guess its not really work at that point.
Hoping you can post more on option selling as well.
What do you guys think about moving further back in the days til expiration in order to harvest a steeper time decay (theta) of the option contract?
ERN: seems like you do the M/W/F close and use very low Deltas to combat having to be more directionally correct and give you cushion. Transaction costs would be higher which is a negative but wouldn’t stepping back to open and closing earlier than expiry allow you to capture the steeper part of the time decay curve where you don’t have to as correct in your directionality?
John, I think your response to one of my earlier posts essentially showcases that you’re doing this to a small degree where you skip the next expiration cycle (so sell next Wednesday expiration’s this Friday) and then exit before expiry.
Just wondering if you guys have thought about moving back a bit further to 20-30 days to expiration when selling the puts and then closing out the position earlier? Could still only need maybe 14 days in the trade (certainly longer than 3!) so maximizing your “bets” but you also may not have to be as correct in your market call.
Good Theta Decay charts here:
Hi Jason, I am new to options trading but essentially had the same question as you (see a few posts above). All of my backtests show selling puts at longer expiration (30+ days) give you better returns than shorter expiration (7 days or less). My “guess” is that Ern prefers taking shorter dated positions because it reduces his overall exposure and lets him capture short term increases in vol which maybe outweighs the theta decay on a risk adjusted basis? In theory if someone had large enough buying power they could do what you are suggesting and in fact there was a podcast I heard last week where a guy was doing what you are talking about (https://optionalpha.com/selling-options-every-day-19957.html) Would love to chat more with you or anyone else interested in testing some of these.
Brad- came to the same conclusion after backtesting. If sufficient buying power to hold multiple positions at once, net returns should be higher over time. of course, exposure is higher too, but typical 90 DTE SPX/XSP 10 delta put corresponds to a strike 10-12% OTM, unlikely to get hit at expiry outside of a bear market or a 1987 event. I manage at 80%, which typically occurs at 30-45 DTE in a rising or flat market. Manage risk by using only 25-30% of total buying power at a time….Although a 1987 event would hurt for sure, I don’t think it would be a knock out scenario for me…
Jason and Brad, go look at options prices for SPY or /ES. Look at the ~14 delta put (to keep somewhat equal risk) at different days to expiration and what those puts are selling for. Then calculate how much premium you can sell over a year when you repeatedly sell those expirations. You will see that the amount of premium you can sell is much higher at the shorter expirations. It’s about 2.3x higher for 2DTE compared to 30DTE based on Friday’s prices. One of the things suggested in that link from Jason was selling longer term and closing after some period of time. If you look at 90DTE to 60DTE, you collect about half what you would just selling the 30 DTE put.
DTE / premium / 365 days of premium
2 0.14 25.55
4 0.25 22.81
6 0.34 20.68
30 0.92 11.19
62 1.5 8.83
90 2.03 8.23
The other relevant points are that the risk in short term vs long term options are reversed. With monthly options you have about 4x the volatility risk (vega) and about 1/4 the gamma risk as with weeklies. (Gamma is rate of change of delta – it’s basically the risk that the option is moving into and out of the money and rapidly changing how sensitive it is to movement of the underlying’s price.)
Given all that, the real question still boils down to what expiration is best for actually making the most money the most consistently. Just being able to sell more premium in the short term options doesn’t necessarily mean you’ll collect all of it due to movements of the market. But from everything presented in ERN’s article and in the previous comments, I conclude (and obviously ERN has as well) that the answer is that something at the shorter end of the spectrum of days to expiration will make you more money.
Totally agree with you. The shorter dated options not only earn have higher premiums on an annualized basis but they have less draw-down and risk. So the real question which keeps me up at night is why do all the backtests and the white papers show selling 30 DTE puts earn higher returns vs. weekly put selling? http://www.cboe.com/framed/pdfframed?content=/micro/buywrite/put-oleg.pdf§ion=SEC_OPTIONS_PRODUCTS&title=An%20Analysis%20of%20Index%20Option%20Writing%20with%20Monthly%20and%20Weekly%20Rollover
Frankly I have been shocked as some of the sloppy backtesting software available for options so perhaps its an artifact of bad backtesting software? If I can ever find some free time I will try to manually backtest both the weeklies and the monthlies using Think or Swim’s lookback feature to see how it plays out.
curious to hear what backtesting software/platforms others have used.
i’ve used TOS lookback selectively, but it is time consuming to do manually. also doesn’t have extended data (option deltas) from the last big bad bear….at least it didn’t the last time i looked…
The back testing software I wrote for myself shows the opposite – that shorter term options make more money. I only half trust it though – I’m calculating historical options prices off of the vix and the underlying’s movement so I have to make some adjustments to try to predict the volatility of a particular delta option relative to the vix. I’ve been thinking about buying historical option data to try to get a better data set to work with. You can pick up data for the SPX for around $400. However my actual results have been matching my predicted results closely enough for the last few months that I’m not too incentivized to spend the time.
The CBOE’s study could be showing a difference because they were using at the money puts instead of out of the money. Tastytrade did a back test several years ago that showed that weekly puts outperform monthly puts by about a factor of 2.5x which is what you’d expect just based on the premium, but then they misinterpreted the results and said it didn’t. They compared 1x leverage in weeklies to 4x leverage effectively because you can sell 4x the weeklies for each monthly which ‘wasn’t fair’ according to them. Of course that means they took 4x the notional risk with the monthlies, so of course they made more money there.
That’s similar to the backtest I used. Not based on actual option data, but derived prices from the VIX.
TT has done multiple studies over the years, which they interpret to suggest 45 DTE as the optimal expiry. I have not seen a single study where they conclude weeklys outperform. Their biggest argument against weeklys is the exploding gamma, which leads to high standard deviation of returns. They’ve pretty much got me convinced.
However, ERN has made one outstanding point with regard to the [much] more frequent resetting of IV basis. That could cut both ways, but I really feel it’s only relevant in the increasing direction as has been discussed.
The CBOE study uses at the money options. I use slightly out of the money puts.
Hard to model the actual difference in annual premium collected selling rolling 2 vs longer DTE puts without running the actual backtests. My view, as I mention above, is that the 90ish DTE SPX/XSP/SPY puts are rarely ITM at expiry and thus a bit easier to model than shorter duration puts.
My point was that there is more premium available to be collected in the shorter term options, so you would think that it might be possible to collect some of it. As it turns out (in my experience), you basically keep about the ratio of sale price in weeklies vs monthlies (or whatever expiration) in final profits. So if you’re only keeping 75% of the monthlies, you’re keeping about 75% of the weeklies but that’s still ~2.5x more.
There’s no difference in modeling the longer duration puts vs shorter in terms of whether they’re ITM or not at expiration. Not sure what you’re saying there… In terms of probability of them being ITM, in theory they have the same likelihood at 90DTE as 2DTE if you’re selling equivalent delta options.
although in theory a 10 delta put should have the same probability of closing ITM regardless of DTE, in reality the 2 DTE 10 delta closes ITM more often than a 90 DTE 10 delta put. i presume this is simply due to the higher vega in the longer dated option, but i may not understand correctly…
I don’t think vega would have anything to do with it. It seems to make intuitive sense to me, but I’m not sure why.
Black-Scholes does a poor job of modeling DOTM options (with regard to the greeks). I wouldn’t be surprised if it also does a poor job of modeling very short-term options.
Alternatively, maybe it’s just a matter of gamma. As a single example, going back to 3/21/19, an 11.2-delta put expiring the next day has a gamma of 0.77 (29 points OTM) whereas a 10.8-delta put expiring 92 days later has a gamma of 0.07 (304 points OTM). Keep in mind, too, that there’s a third-order greek (“speed”) that describes rate of change of gamma with respect to underlying price.
FWIW I tried to backtest selling naked puts at both the weekly and monthly expirations from 2016 to present (I believe weekly options did not come out until late 2015). The results:
-Weeklys generated $1,766 in total premium per contract on SPY (15 delta)
-Monthlys generated $2,764 in total premium per contract on SPY (15 delta)
So although the weekly options have larger annualized premiums, they also had more losing trades which reduced the total return during this time period. Incidentally, the monthly options at 15 delta did not have any losing trades during this time period (so a longer study is needed).
I guess my take away from all this and trying to follow BigErn’s strategy of using the weekly options is to sell 30-45 day puts but if your buying power allows for it try to sell one monthly option each week or every few days to capture short term premium increases. This may be the “best of both worlds”.
(disclaimer – these backtesting tools are a bit flaky so I dont have too much confidence in these results)
Thanks for doing this! Very intriguing! And the monthly puts didn’t lose money in February 2018?
At the 15 delta it was a win. At all other deltas it would have been a lose.
Shorted a 271 strike on January 19, 2018 and at expiration on February 16, 2018 the price was around 272 so it barely squeaked out a win.
Very well said! Thanks for weighing in! 🙂
I prefer selling very short-term puts. The chart in the link you provided is the reason why: time decay becomes more pronounced the closer you move to the expiration.
Stupid question but I’m testing IBs paper account. How do I track the short put position?
As soon as the trade goes through, the portfolio page will have a P&L column but it turned negative right away.
How do I track how I’m going? Do I have to just manually update it on a spreadsheet increased of IB’s mobile app portfolio page?
When I was trying things out with IB’s paper account, I didn’t get to the point where I felt like I had any level of mastery of their interface, so I’m curious how other people ended up doing this too.
All I ended up keeping track of, manually, was what trades I’d made and the resulting effect on the account balances.
Russ, I managed to figure it out. For some reason the P&L doesn’t seem to be correct when viewing your total portfolio but it is accurate when you view the actual put option itself.
See this screenshot. Shows the ‘cost basis’ as -88 which was the premium received and now the put is worth -25, so unrealise profit is 63.
I think another reason why the paper account is confusing is because they give us a paper account of 1 million, and also IBs interface is confusing as well haha.
I will know maybe tomorrow once my funds go through and start some live trading.
I started with an account of $10,000 and one short put. Never did the paper account.
I recently closed my IB account. Their platform and data access fees just bothered me too much. For paper trading and backtesting I would highly suggest ThinkorSwim through TD Ameritrade. I beleive you can get a free paper trading account from the and their platform lets you go back in time to simulate profit/loss for different options positions you enter. This is called their “Think Back” feature and you can watch videos on Youtube on how to use it. I know do all of my options trading on RobinHood which has free commissions and I use ThinkOrSwim for research and charting.
I also hate the data feed fees at IB! It used to be free when you traded a certain minimum.
But I have a high trading volume (20+ Put option contracts, 3x per week), so the low commission for ES put options ($1.42 total for IB plus exchange) helps.
I will see my short as negative value (negative positon, negative value). And the value will (hopefully) get less negative over time and expire worthless at expiration.
The P&L usually turns negative right away because you paid a commission ($1.42 per ES Put). But over time I usually make money. 🙂
Karsten, I see, it’s all because of the commission that makes it negative.
Also I realised for some reason the portfolio-wide view shows a different P&L value to the one shown if I click into the actual option’s info page. The latter seems to have an accurate P&L value.
Makes sense! Good luck!
So I’ve been trading for the past week.
The Good: Every put I have sold has expired worthless and thus I’ve kept all the premiums
The Bad: The premiums I’m collecting are very low. Either I am being to conservative and choosing the wrong strike (most likely!) or premiums are less rich at the moment?
Can Karsten, John, or anyone else more knowledgeable help me through this example.
https://imgur.com/a/MpjafnF (3 screenshots)
Underlying is 2906
Puts expiring in 3 days
If I want a buffer of ~2% and implied vol to be greater than VIX then I would choose the 2850 strike
Premium collected is likely $0.60
To get a good yield, I would have to sell puts with strikes at 2880 or 2885 but that only gives me a ~20 point cushion.
If I keep doing the same thing as I’ve been doing all week, I would sell the 2850.
What’s the “right” strike for this scenario?
Bob, I’m not sure there’s a ‘right’ strike to sell. You just get different risk/reward trade offs with each. I can tell you I’m currently short the 2870 and 2875 puts that expire on Monday. I sold the 2870’s around 11am EST today for an average of 2.10 and the 2875’s around 3pm EST for 1.60 to round out my position, but I didn’t like selling the 2875’s. IV dropped today and prices were low / you couldn’t go as far out of the money as the day went on. In your example, I’d probably sell the 2880’s or 2875’s. I try to sell between the 0.10 and 0.14 delta, and usually I try to sell something for 2.00 or more as I have been buying them back in the 0.30-0.50 range if I can do that far enough before expiration. If I can’t sell something within those criteria, I’ll start looking at the next expiration.
Hi John, thanks again for imparting your knowledge.
You have pretty much answered my question.
As I was typing that , I realised “right” wasn’t the correct word but I hit publish anyway.
What would be more accurate would be to ask what strike Karsten would sell according to his strategy because the risk/return profile of the strategy in this article is what I want to mimic.
Once my short puts expire today, I’ll try selling less conservative strikes for Monday.
You will notice that over the weekend (Friday to Monday) you have three calendar days but only one trading day. Yields will be much lower than over a three day window Tuesday->Friday. In this particular example I would have likely sold at ~40 points out of the money: 2865.
~40 points was about 3% time value. You mentioned before that you like 5-8% time value. How do you decide which is the correct one for that time?
Writing a put on Monday for this Wednesday seems very profitable so far. Sold 2855 for $2.75.
I think one of my problems at the moment is not knowing how to benchmark myself since I’ve only just started. I’m worried about yielding too little. What should I be aiming for in terms of income for each week?
I found that i was definitely being to conservative and I need to sell a bit closer to the money than I have been doing but this will also mean that there is more likelihood of them expiring ITM. Once again, for benchmarking purposes, how often would a put expire ITM with this strategy?
Anyway, hope you’re enjoying your holiday. 🙂
I’ve been seeing roughly 0.3% per week with 3x leverage. I expect to see a substantial loss at some point, but I’ve only seen 1 option expire ITM in the last 5 months, and that one was a pretty tiny loss. The market has generally been rising though, so not too surprising. The theoretical back testing I did showed puts finishing OTM in the 90%+ range for 10-14 delta over the last ~12 years, so in the range of what you’d expect. Ie, if you sell a 10 delta put, you’d expect about a 10% chance that it expires ITM.
Regarding the 0.3% per week, would that be gross premium received or is that net profit?
Going off what Karsten wrote in the article where he would expect to only keep half the premiums, should I be trying to get 0.6%?
Putting numbers to it. $48,000 would be the amount of capital for one option if utilising 3x leverage at roughly current underlying. 0.3% would be $144 per week which would work out to ~15% p.a. which is in line with the blog post. Since we’re budgeting for keeping only half the premiums, I should be aiming to receive ~$288 in premiums per week?
Sorry if I’ve made a mistake somewhere
Thanks! Same experience here! 🙂
Karsten, 0.3% seems to be about 15% annualised which is in line with the returns you’ve written about in this post. You also mentioned that you expect to keep only half the premiums, does that mean that you’re receiving 0.6% in premiums if averaged weekly or do you expect to make 0.15% weekly in the long run?
Whenever I refer to time value yield I always mean gross yield. Say, 5-7% gross yield without leverage. Back in the old days that meant 20+% with 3x leverage but not I’ve scaled it down significantly.
The haircut from the occasional losses will get you a lower net yield.
Friday to Monday is tricky: It’s 3 calendar days, but only 1 trading day. That 3% nnualized is close to my target if taking into account that there’s only one trading day!
Oof so Friday was my first loss and set me back 2 weeks. While I admit it does get to me emotionally, I rationalise it and know it’s all part of the strategy.
Sold another one for Monday at 2855.
The reason I’m posting here is I don’t know anywhere else to discuss this and this comment section has become a default “forum” for option writing to me.
How did everyone else go last Friday?
It takes a while to get used to it, especially when you’re used to seeing the strategy just make money. When the market drops and vol spikes, it can look a bit scary. These were some pretty small drops though. At the moment I’m only down about 0.2% from my peak last week while the market is down ~1.3% in the same time frame, so not too bad.
Nice one John, so the strategy is working as intended. Similar returns as SP5000 with less vol.
As you and Karsten mentioned, writing puts with Monday and Wednesday expiries since the loss has been very lucrative.
It’s interesting that we have the choice of either selling at a strike with a huge buffer for the “usual” premium we accept or take higher premiums at “usual” strikes we accept. I don’t know if this is wise or just dear but I chose the former.
Another two options that expire worthless should make back the losses from Friday.
Overall I’m breakeven since starting.
Nothing seems to beat experience for me. I can read about it all I want but actually doing is how I learn so this has been an interesting but good lesson.
I’m in a similar boat, only started trying this out a couple months ago and Friday was the first time I didn’t see things expire worthless. I wasn’t paying attention though so when I got around to looking Friday afternoon I was like “huh what? exercised?” and promptly closed that position.
(As an aside — how potentially terrible was this lapse in attention? Should I be making more of a point to close positions before they get exercised?)
Just kept on doing basically the same thing since then though, over the span of these few months things are still positive.
It’s a two-edged sword! I’ve seen my positions under-water throughout the day only to recover by market close. I was worried for nothing. I should have just ignored the market and simply checked in a few minutes before close.
I usually sell the ES future a few minutes before market close if it’s clear that we’ll end up in the money (e.g. Feb 5).
Seems like Karsten already answered and he would be the one to listen to.
I have only had one option exercised so what I did was sell it immediately after I received the futures contracts. I aim to sell within that 15min window before the trading halt.
Not sure if that’s the right thing to do.
Well, welcome to the club. There is ablessing in disguise in losing a small amount early on. I’ve seen folks with a winning streak for multiple months take on more and more leverage and then get wiped out when the big market move came. 🙂
Better luck next time!
I had picked my 10/5 strikes at 2885. Futures closed at a few points above that, but were below my strike temporarily during the day. But it all worked out!
I agree. I am taking this as valuable experience and also a test of my nerve. I am happy to report that I am not panicking and I’ll just keep doing what I’m supposed to.
If I’m honest, the money I was making at the beginning did make me daydream about what sort of return I would be getting but now back to reality.
Can this now be the default ERN option writing discussion forum?
PS is anyone having trouble getting email notifications of new comments through Gmail? It shows that I have a notification from ERN blog but then when I click into the “social” folder there’s nothing there.
Wow I think mine were at 2905, I wish I had a stronger rationale for how I arrived at this number but… ¯\_(ツ)_/¯
Haha! Better luck next time!
Oof that was brutal. It made last Friday look like nothing.
Sold another put with 100 point cushion. Not sure if that’s enough but you weren’t kidding when you said you can sell way OTM puts for such high premiums.
I saw the ES go into the low 2700s but then we recovered again to 2772 at the end of Friday. Hope this worked out for you!
It’s all working out so far. I’ve clawed back 10% of my loss in the past week because the premiums for way OTM puts have been very rich as you wrote about.
Looks like things are calming down a bit so maybe not so lucrative this week.
This strategy is not lacking in the excitement department! 🙂
I can tell you, when the underlying is trading around the strike price and you watch the last few minutes before expiration it’s more exciting than any football or basketball game in overtime… 🙂
ERN, do you (or will you consider) posting summary table or chart of your performance results (daily or weekly) from the Put Selling strategy?
Haven’t done that yet. I will probably keep that a secret for now! 🙂
Would you consider publishing dates where you have made losses? With details like underlying, strike, delta etc?
No need to go into specific dollar amounts and no need to go that far back.
I think it would really help to see if we’re following the strategy correctly. To see if we’re significantly losing more.
I lost a small amout on Oct 12. But even that was only because of bad luck on the timing. Because we were traveling on the 10th (Wednesday) I had to write options Monday->Friday. Had I done M->W and W->F I could have avoided even that loss. Apart from that the only loss days this year where in late March and early February.
Hope it stays calm for the rest of the year!
These are my three losses since starting this strategy on 15th Sept.
Since then I have been a bit more cautious (scared!) and have sold strikes with 80 to 100 points OTM.
I would really value your critique on how I can improve choosing a better strike. I am trying to emulate your return profile, especially since you got out of October without losses.
When you’re in Sydney, free dental is on offer if you want 🙂
I think your dates are off by one day (due to the time difference). Also, only on Oct 9 (Oct 8 NYC time) the underlying is below the strike. Typos?
Just to give one example of my own experience, on Oct 22 (NYC time) I sold puts with strikes at 2640 to 2650. The ES future was at between 2765 to 2771. So, these are more than 100 poin ts out of the money.
Sure enough, on Oct 24 we had almost a “point landing” and the futures closed at around 2658. Phew!
We’ll be in Sydney Nov 16-21. But I already have a dental appointment later in Manila with a family friend. Thanks!
But will you come to the ChooseFI Sydney meetup on Nov 17?
Thanks for the reply Karsten.
Oops you’re right, a few mistakes on my part.
I have the Australian dates on my spreadsheet and forgot to account for that.
The October 9th date should be October 11th here, and October 10th NYC date, and the strike was 2850.
Is your method usually to choose a strike that is 100 or more points OTM, or was this only possible due to the higher volatility during this October?
Is choosing the strike much more dynamic and nuanced than simply choosing a “safe” cushion?
When I started, I used to use the delta to help choose the strike but that didn’t help me during October so my emotional brain has made me want to sell strikes with a certain points cushion.
I wasn’t aware of the meet-up until now, and just joined the FB group.
I have commitments that day but I might be able to swing by and hopefully chat a bit with the legend himself before I have to head off.
Choosing the strike is more complicated than picking a hard number. A good guideline would be to target a certain put premium. I try to be around $1.00 for a Monday to Wednesday or Wednesday to Friday and $0.75 for Friday to Monday. I’d also like to see the option Delta and how many standard deviations (multiples of VIX) the strike is out of the money.
In October I sold plenty of puts with 100+ points out of the money. In other months it would be much less.
Maybe see you on Saturday!
Best of luck!
I thought you did post a return chart in one of the two articles. A revised one current to date would be excellent!
I’ll do a blog post on how we did in October. With some additional background info. Should be out on Dec 3.
Karsten, sorry I didn’t get to drop my to the Sydney meet-up. I really wanted to come say thank you and let you know how much your blog has helped shape my FIRE plan.
Unfortunately I couldn’t get away from the commitments I had today.
I hope you make it back to Sydney for another meet-up in the future. 🙂
No problem! We are heading out of Sydney today but we should be back! 🙂
A couple of comments regarding choosing very short expiration period. In theory, ATM put’s time decay accelerates as it approaches expiration, and OTM put’s time decay starts to slow down a couple of weeks before expiration. With the recent market volatility, your existing short put positions would experience high delta and gamma risks when the market goes ugly, although you can take advantage of the increase in implied volatility if you don’t have positions on. The point being, the risk reward of the gamma risk would not justify putting positions on with only one to two weeks to expiration. You can still manage the risks if your positions are to be hit due to the volatility but that would be the subject of next discussion.
I notice that you do puts selling. Short puts and covered calls have identical P&L curve if they are done at the same delta. So, the question is whether it makes sense to sell covered calls on dividend stocks to create synesthetic dividend yields that are many times higher than the actual dividend yields. If the sustainable retirement withdrawal rate is, say 4%, and one could produce effective yield of say anywhere between 4 to 10% annual yields (depending on the dividend paying stocks), then at least it seems on paper that one would be able to produce perpetual “passive” income stream. Say, with a portfolio of $3M, one could produce anywhere between $120K to $300K annually without using any leverage and reward to risk ratio is pretty good. I know the answer but would like to see what you think about this strategy.
I sell puts 3 times a week, not weekly and not 2 weeks ahead. I made it very safely through the volatility in the 4th quarter and even made money while the S&P went down 19%. Precisely because I never had much Gamma risk. The options normally expired worthless because the term was so ultra-short, that the Gamma effect was always swamped by the time decay effect. For details see this:
I have entertained the idea of doing the covered call selling on individual stocks. Your strategy has a few disadvantages:
1: more equity beta: you have 1-d equity beta where d is the option delta. Since we’re dealing with an OTM Call with a delta below 0.5 your overall beta will be >0.5. I sell 2 to 2.5x leverage puts with a delta of around 0.05 to 0.1 each, so my position has a lower equity beta during normal times. But due to Gamma hat can change, I know. But I did very well in Q4. You probably didn’t with your covered calls.
2: All else equal, the premium is richer on the OTM puts than on the OTM calls (vol smirk).
3: I prefer the index options that are Section 1256 contracts on my tax return, while individual stock option profits are taxed as short-term gains.
4: by selling SPX put options I get to play with a large pile of cash. I invest that in muni bond funds (ABHYX yield 3.5%+ tax-free), closed-end Muni Bond Funds (NZF yield 6.0% tax-free) and Preferred Shares with yields around 6% all the way up to 8% (though not tax-free, but treated as qualifying dividends).
But again: I find the individual stock covered call writing interesting. I may explore it in the future.
Agreed that real covered calls would require more capitals. If you use regular margin or portfolio margin, the buying power does need a lot less capital, and hence larger ROC. I understand puts are usually more expensive than calls with the same delta due to volatility skew and realized volatility is very often less than the implied volatility as I have traded short puts over tens of thousands of times. For the bear market correction in Q4, like what you said, the covered call strategy which is a bullish strategy definitely loses mark-to-market value since the net delta is still positive. For this type of bear market, only negative delta or close-to-zero delta positions generate profits. To synthetically emulate your short put with small but positive delta, one would have to sell deep ITM calls which has much lower breakeven points.
I am not proposing the covered call as an active trading strategy. For active trading, I would use all the tools at my disposal, such as long/short puts, calls, strangles, straddles, ratio spreads, butterflies, condor, jade lizards, etc. But that is besides the point as 99% of the retirees have neither the aptitude nor the attitude, mental agility or stomach to trade options.
I am looking into it as a passive or quasi-passive income strategy for retirees, that needs minimal management. In the scale of 0 to 10 with 10 being active management and 0 being no management. Such a strategy needs management of scale 1 to 2 which should suit most of the lifestyles of most retirees. One can continuously and mechanically sell, say, 16 delta or 30 delta calls every 30 to 60 days. Or if one is more sophisticated, one can sell calls with larger negative delta (i.e., ATM or ITM) to hedge against the long stocks if the market and its technical chart warrant such a move. The idea for retirees is, instead of liquidating their stocks/bonds for incomes, with very little physical work and mental thinking, one could produce perpetual incomes (dividends, call premiums, capital gains if the stocks are called away). So effectively, one could generate effective synthetic dividend yields at least double to triple he/she can get just from the dividends. Again, this is an income strategy (not trading strategy) for retirees, and for sure his/her portfolio will suffer from mark-to-market loss if the underlying assets have large draw-down.
Since short puts and covered calls have the identical P/L curve if the net deltas for each are identical, and dividends are also got priced into the option premiums per Black-Sch-oles formula, I believe that the covered calls probably are more suitable for pre-tax IRA accounts and short puts are more suitable in regular taxable trading account that use lower buying powers with margins, like you have stated in your point 4.
I do appreciate that the index trading has the 60/40 tax advantage. The only reason I trade SPY instead of SPX is due to its poorer liquidity with larger ask/bid spread. The much shorter expiry period has faster theta decay for OTM options but with increased gamma risk accordingly (no free lunch) and that could necessitate more frequent position monitoring and management. Hence for retirees, 30 to 60 days expiry probably are the most optimal time to short premiums with minimal management for folks busy with their retirement activities.
These are my 2 cents.
Thanks for your views! Very good points!
I should stress that I also do mechanical options trades, only more often. I trade 3x a week, so it sounds “active” but it’s the same philosophy as yours. I merely roll my expiring puts to new ones with a certain target delta, usually around 5 to 10.
And I should also stress that something along the method you described has been on my mind for a while. Tell you what, I’ll try to do this in one of our retirement accounts (to deal with the short-term cap gains issue). And I got you to thank for to finally get me going on this. 🙂
Also, you seem to be very passionate and knowledgeable and experienced with this strategy. If you ever want to write a guest post on this please reach out!
Thanks for the invitation to contribute. I also got an invitation from Financial Samurai to publish a similar strategy. Please do let me think about how to proceed. I do plan on writing a small booklet on this simple and effective strategy that most retirees can quickly adopt and use without learning the whole 9 yards about options. There are Greeks and maths about options that most retirees or FIRE guys can’t care any less, and I am planning on simplifying it so that an average Joe and Jane can use and benefit. That reminds me when I studied computer science at college, there was no PCs available. Computers were expensive, hard to use, were only available for those selected “smart” ones until Apple IIs and IBM PCs became available for the masses. I guess I am dating myself here. And the rest is history.
Given the fact that the majority of the people in the US may not have enough to retire, unlike the FIRE community of which the people are still in the minority, I am guessing there is a need for simple-to-use strategy to produce retirement incomes to augment the income stream retirees have or don’t have.
It’s up to you. Maybe do the broad overview on the FS blog. And here you can do as much math as you want.
I would love to see a long-term comprehensive backtest on covered calls and short puts. I realize the goal here would be to generate yield at the lowest possible volatility but the skeptic in me thinks the average investor would be better off with just focusing on asset allocation. At the end of the day nobody cares where they get their yield from (yield is yield) as long as the risk profiles are about the same and if yields are about the same between buy and hold and covered calls for the past 10-20 years it would be hard to make the case for writing options (especially after trading costs). Just look at the ETFs and CEFs that use option writing and its hard to find one that has out-performed buy and hold.
I think Tastytrade has a back-test result from 2005 to 2018 and it shows that short strangles, or covered calls do produce better P/L and lower standard deviation in the P/L over 13 years, after accounting for commissions. To be fair though, the back-test study did not take capital gain tax into account. If you think about it, it makes sense. When you are long stocks, your delta is 100. The covered calls, or strangles have smaller delta and hence have lower directional risks. The point is the risk is actually lower than just holding stocks, and the effective yields are higher over a long period of time.
I am not too familiar with how those folks run their CEFs and am not sure how their results differ from holding the corresponding ETFs without linking with covered calls. The point is if you can do it yourself, why pay somebody to do it for you.
A reliable and clean backtest would be here:
But also check the other links that others have pointed out here.
Also: this is not about outperforming. If you have the same average return as the index but with lower vol and lower drawdowns, that’s gold in the context of Sequence Risk.
But I should stress: It’s better to implement yourself and not buy the ETF. You save the expense ratio and you can do other creative stuff, i.e., muni bond funds, etc.
Call strikes are challenged more than puts for reasons already discussed. The slides on the TT link below shows some SPY data since ’05. Tax consequences of covered calls getting exercised if the strike is hit often left out of the discussion of covered call strategies.
There is no tax consequence if you do it in an IRA account.
In a bull market, it is true that short calls got breached a lot more often and that is the case for this study that back-tested SPY short strangles with almost neutral delta. However in 2018, positions with net long delta got hurt.
Thanks for the link!
Would you please summarize your put strike selection to me again! I am very interested in your strategy, but am only familiar with selling puts on ETFs and large caps, with 30-45 DTE, and IV percentiles> 35-40% ( I don’t use delta, but instead look to sell puts at a price below “support”, usually looking at the computer generated support lines on Finviz.com. The delta is usually around 0.15 to 0.25, but again, I don’t look at delta primarily). I believe you were saying you were looking for a premium of $1 for Monday thru Weds, Weds thru Friday , and $0.75 for Friday thru Monday. You also mentioned that you were looking for a delta of 0.05 to 0.10. Is that pretty much it? I see you don’t care much about negative gamma risk or volatility, which has limited my thinking in the past. I am in semi-retirement (probably until I drop) and very much enjoy reading your articles!
That is pretty much what I’m doing, yes. I do this pretty stoically and regularly.
I concede that my method, like every option selling strategy, has Gamma risk. I don’t quite understand why people would consider this a more worrisome Gamma risk than the 30-45 DTE selling. Sure, my strategy occasionally goes from 0.05-delta to 1.00-delta. But only for 2 trading days. Then I sell the next set of puts at a lower strike. I’d prefer that over having a 1.00 delta for 30 days. That could be very painful if the market keeps dropping (think August 2015, January 2016, October and December 2018).
Best of luck with your option-selling and (semi-) retirement! 🙂
Does the premium $1 and $0.75 fluctuate with price of the underlying? What is your range of typical leverage and the factors that contribute to the leverage fluctuations?
The premium changes with underlying market conditions. Most importantly it responds to
1: the price of the underlying (option delta)
2: the level of implied voltility (option vega)
3: time decay (option theta)
I used to sell futures options (50x multiplier) and keep around $50k as margin cash. That’s way more than the exchange mandated minimum margin (usually in the $10-12k range)
Currently, I sell SPX index options (100x multiplier) and keep around $110-120k per short contract. That’s again way more than the CBOE minimum margin requirement of ~$22k.
Hope this helps!
Is the reference to approx $1 premium on Mon and Wed and $0.75 on Fri, assuming SPX’s 100x or ES’s 50x?
This is the premium per 1x. You multiply that by 50 or 100 respectively for the ES and SPX.
Why did you switch from futures options to SPX index options?
This is a very good point.
There’s a total-exposure risk associated with the 45 DTE approach (the approach I use).
If for example I were to sell puts M/W/F @ 45 DTE and wait for theta decay to do its thing until profit targets are reached, those positions are all exposed to a potential VIX spike. I have to size the individual positions such that I’m not overexposed should the market stay flat or drift down slightly and I end up with 3 weeks of open puts.
The tradeoff is [potentially much] higher return on capital in the form of early management. If 16 delta 45 DTE puts reach 20% max profit on average 2 days into the trade (even faster for lower deltas), I collect 20% of the premium for ~4.4% (2 / 45) the duration risk. That’s a significant boost! Even if we underperform and only reach 10% after 2 days, that’s still over 2x improvement vs the near-linear (as measured in days) decay for the 1-2 DTE options.
I’m going to try the short-term SPX writing the rest of this month and will report back with the results and corresponding trade log. I’m fascinated by the reduction of concurrent exposure and the much faster results (wait a day or two to see if you’re profitable vs up to 21).
SpintTwig – could you share the results of your short-term SPX test for the month? And I’d be very curious to hear if you’ve executed the early sale strategy you outlined and what the % premium collected were over the 2-day hold period (or whatever period you used).
May I ask if I calculate IB margin requirements correctly based on formula below when current ES price is 2795, strike 2760 and premium is 1.20$
Index Options 1
Put Price + Maximum ((15% * Underlying Price – Out of the Money Amount), (10% * Strike Price))
1.2 x50 + MAX( (0.15 x 2795 x 50 – 35×50 ), (0.1 x 2760×50))
= 1.2 x50 + MAX (20960 – 1750, 13800) ± 17k
However Mr. Ern started trading with 10k, while also saw latest recommendation for 1 put option 30k. I assume I do some mistake in calculations.
Futures use SPAN margining. See the CME website for details. 1 ES future requires around $5500 of margin at the moment, and 1 put is similar. Individual brokers can require more margin if they choose. While the exchange will let you trade with that minimal amount of margin, you should have more cash than that in your account so your broker does not close your position the first time you have a small loss.
Quite intriguingly, the short put requires more (!) margin cushion than even a simple long-ES-future. I stopped trading ES puts (50x) and now do the CBOE index puts (100x). The margin requirement for the ES puts was, if I remember correctly, around $10-12k per contract (50x) and now it’s around $22k for the contract with twice the size.
Yes, futures margin requirement is simple number and is easy to understand: Cash – (Contract Price – Current Price)*50 > Min Margin.
I understand options margin is more dynamic and calculated by the formula above. However I was not sure if I understand the below correctly:
Underlying price – current ES contract price times 50?
Out of the Money Amount – current ES contract price minus Strike price and times 50
Strike Price – strike price times 50.
Is that correct?
Hi folks, it’s Thomas here again. I apologize if some of my scribbling does not make perfect sense. I’m not native English speaker nor do I have background in finances.
So, if there is someone interested to start selling ES Mini put’s on IB, be aware that indeed margin is about 12k per one put, so as Mr. Ern recommended, 30-35k might be needed to start.
So far I am playing with paper trading and discovered unpleasant thing, which I’m not sure is true in Live account: after 4pm EST when you expect your put option to expire worthless and you sell next batch, my margin jumped to 25k. I assume “expired” option is not liquidated immediately, thus the margin increase because of new sale. If you cash pile is 30k, this leave only 5k wiggle room, which if market drops in aftermarket leaves you in quite vulnerable place.
In the morning seems expired option was liquidated and margin is back to 12k. Thus not sure how long margin is locked for “expired” contract.
Maybe this means that with small portfolio it’s better to buy back expiring OTM put, before selling new one.
In practice that should not be an issue. I’ve seen that, too, when the new put was already sold and the old one expired, there was still a margin requirement. Even at 1:15 pm Pacific time, 15 minutes after the expiration. But when you get your statement for the day the expired option is no longer a margin drag.
But I should also stress that now I’m selling SPX options (100x instead of 50x) and I’m now using around $120k in margin cushion per option. The equivalent of $60k per ES option. I’ve grown a bit more risk averse when the portfolio got bigger and the paychecks stopped coming. 🙂
Ern, you mentioned that you started shorting 1 ES put with 10k. Later recommendation was to have margin about 30-35k. Today, I understand, you would use 60k of margin per 1 ES put. Is there any reason of this increase because of market conditions or it is just your own preference of having bigger cushion?
I believe I’ve read all the comments, however maybe I missed this part – what is the reason for switching to SPX?
Finally, do you count your annual revenue as % of underlying notional?
Don’t do 10k today. Back in 2011 the S&P was less than half of what it is today. And so were the margins. That plus my aim of taking less risk now: no more paychecks, much larger option trading portfolio.
Greetings Ern. Well, I have been feeling pretty good about my options strategy (like you, selling puts 3x/week), especially this past week when the markets lost but I managed a nice percentage annualized gain.
I am 57, not yet retired, and a part of me wonders, instead of having the vast majority of my investments in equites and a small portion in options, why not the reverse (especially since my portfolio is already sufficiently large that I think that I have already “won” and can readily attain a nice annualized gain with options, say 8% target by selling rather far out of the money. All my trading is within tax free and tax deferred accounts). I typically sell puts on iShares SPY fund (cannot trade futures or use margin in retirement accounts ), and the annualized gains are, to me, substantial.
Have you had similar thoughts; i.e., effectively swapping your equities for options in the majority of your portfolio (especially since, given the size of your portfolio, it seems that you have already “won” in the sense that steady gains – no need to aim for double digit returns as can happen in a strong bull market – can likely fund your retirement plans)? Given the likely superior risk-adjusted returns of derivatives, for me, only the inertia of my long history with equities is preventing me from taking on substantially greater options activity.
PS – I use Schwab, and they (and I think others) offer 3/years, 500 free trades every time a $100,000 asset transfer is made to a Schwab account. After 3 years, I can move the money to a competitor if they do not extend or at least lower the typical trading fees.
Of course you can create synthetic long or short stocks using options. There are many different ways to do this. It is a misconceived notion to sell way-out-of-the money (like way less than 10% delta) puts, because you don’t get pay enough to take the risk and it is very typical for inexperienced trader to overlook the notional risks. If you have not experienced large drawdown doing this regardless of the implied volatility, consider yourself lucky. It is a lengthy subject, and I am not going to talk about too much here. For IRA account, if the implied volatility is low and if you are bullish, you could sell a poor man covered call (i.e long diagonal call) which has much less capital requirement than short put. Likewise, you could do puts if you are bearish.
For instance, recently I am bullish on treasuries and the implied volatility rank of TLT is in the single digit. Instead of selling puts, I do a long call spread in which the short call extrinsic value is higher than the extrinsic value of the long ITM call. By doing that, I can offset the time decay of the long call position using the short OTM call. I just closed the position yesterday after one week with 30% ROC.
Thanks for the input, Multimega. Appreciate it.
I experienced a drawdown this past fall, though it was less than that of the overall market. I am, as of now, not as profitable as the S&P benchmark, but I have earned and continue to garner a solid return with much less volatility than most equities (my Delta is usually ~ 0.12, but, given the very recent increase in IV, it is already lower for this coming week).
I have seen the arguments that the payoff from far out-of-the-money puts is insufficient for the notional risks, but the mathematical explanations for that have never been pursuasive; i.e., my “benchmark standards” are the equity indices, and a long drawdown from selling puts would, for equity holders, likely be an even steeper drawdown over the same period. Furthermore, the more frequent selling one does, the greater the likelihood that the long-term expected payout (equal to expected winning percentage*actual payout) will approximate the expected odds for success.
It is for this reason that I am, as of now, uninterested in various spreads and other protective strategies; that is, I see the overall risk of selling puts as less than that of holding equities (though I am not, at this time, prepared convert a majority of my holdings from equities to options).
Well, I follow the ” misconceived notion to sell way-out-of-the money” and I compensate the low yield with a little bit of leverage. Seems to work just fine for me.
All very good points! I certainly increased my option trading percentage in the portfolio. Exactly for the same reason as yours: if you already won, why keep rolling the dice and keep hoping for double-digit equity returns. I like the slow and steady income from options.
Right now I do this in a taxable account at IB. But I’ve heard good things about Schwab, so best of luck to you. Sounds like a great plan to move once the “honeymoon” is over. I also heard that RobinHood now allows options at zero commission.
Regarding your comments: Well, I follow the ” misconceived notion to sell way-out-of-the money” and I compensate the low yield with a little bit of leverage. Seems to work just fine for me.
I have neither the time nor bandwidth nor inclination to go to great length to explain. Go at it whatever your think it works for you.
@multimega: I never asked for an explanation. What makes you think I did? I understood you perfectly well the first time.
There are many ways to skin the cat. I am not trying to convince you at your age to learn too many new strategies. Do whatever that suits you. 12% delta is less than 1 SD, and most of the time your puts (almost 90% of the time) would expire OTM. Also the realized volatility is often less than the implied volatility, so the probability is on your side. I am just pointing out one may need to look at the risk/reward ratio.
…”I am not trying to convince you at your age to learn too many new strategies.”…
LOL! Thanks for looking out for the welfare of this old man. Don’t wish to exhaust my brain learning new things!
…”Do whatever that suits you”…
As if I would or could do anything else.?
Bye for now. Off to have my prune juice and pureed vegetables for dinner.
Haha! Off the the 4:30pm dinner special and then it’s Bingo night!
Question for this forum: Why not also sell a 16-20 delta call in addition to the put (ie sell a strangle)? The margin is the same. Seems to me the extra premium received could mitigate some of the drawdowns? Thanks in advance for your input.
Hi Luc. I trade within a retirement account so I can only sell cash secured puts or covered calls (by law, no margin in retirement accounts). I sell a fixed # of put contracts each week. I also have various Schwab indexed funds in this account, though they are not optionable (insufficient open interest). But I am not interested in selling covered calls as my strategy is as follows:
– Sell fixed # of SPY put contracts each week at approx 0.1<=Delta<=0.13
– If benchmark rises at expiration, sell a few index funds shares inorder to attain the greater cash required to secure a higher strike price for the next put contract sale (and in doing so, taking a little profit from the gain).
– If benchmark falls at expiration (and I am not in the money), purchase a few index fund shares in order to take advantage of the price decrease (with the reduced benchmark price, I need less cash to secure the lowered strike price following the benchmark drop).
So in this account, I am partially invested in options and partially in equities. I like it that way, as to offer strangles would mean that I would not yield the benefit of a strong upside in equities for the covered calls, and that risk is already present in my current scheme on the put side. In short, I (so far) like the balance that I have struck.
But do you get enough income from this to make it worthwhile? Are you allowed to hold the margin cash in something generating extra yield?
Interesting question, Ern, re. whether I can hold the margin cash into something generating higher yield. I always assumed, no (cash-secured put, after all), but I plan to check with the broker to see.
Whether income gained is worthwhile worthwhile? I think so. I changed my strategy not too long ago from selling puts on QQQ with a higher delta than today (~0.22 before, vs ~ 0.1< = Delta <=0.13 on SPY today), and this typically nets me between 8 and 9+% annualized yield. Have not yet gone into the money, but will see how often that happens. Meantime, I recently created a spreadsheet to track every trade, so I will have a better data on this over the course of time.
Once retired (2 – 3 years), I want to continue this activity with margin, much as you described. Meantime, while you incur the cost of margin and trading fees, I am trading for free during this honeymoon period at Schwab (3 years left on it; I had been trading 1/week at standard fees before moving additional funds into the account) and, given that this is a Roth account, the gains are tax free. Will let you know over time how this continues to work out; meantime, it's also a good warmup for the aforementioned expansion of this strategy post-retirement.
Clarification re. above – The “annualized yield” is based on the premiums from trading 3x per week, annualized over 52weeks.
Nice! Well, if you face the risk of getting physical delivery of the underlying then it might be wise to keep lots of extra cash. But most people write options with 45 DTE and then roll them well before the expiration. Then, just hold enough cash for the occasional loss, and the rest in something higher-yielding.
I’ve done this before at various times, although mostly on longer expiration options. Obviously it works great when the market goes sideways and helps when the market drops. I got tired of it in 2017 when the market was going up all year. I was constantly getting tested on my short calls and didn’t make much money all year.
That would be exactly my concern! Thanks for confirming! 🙂
Good qustion: I personally don’t do this two-sided option selling (strangle) but others I know do and they like it.
I think it would violate my principle of never shorting the market. True, you get additional income to hedge against a large drop but you also generate the risk of losing money if the market goes up substantially. That and the lower IV for OOM Calls kept me from doing this so far.
I totally agree with “or even worse.”
There’s a saying “your worst drawdown is always in front of you.” Larry Williams used to talk about position sizing based on a drawdown 1.5x the worst drawdown ever seen in a historically backtested period.
Very good point. I’ve probably dodged a bullet when I was still running this with 3.5x leverage (or even more very early on). I like the relative safety of what I’m doing today! 🙂
Calling DOTM put selling a “misconceived notion” rubbed some people the wrong way. I agree with them from a data standpoint. I have seen different backtests (in addition to doing some of my own) that conflict. Some conclude it is a bad idea. Others conclude it is a good idea. This leaves your only grounds for calling it “misconceived” personal trading experience. Remember, though, that you are only a sample size of one.
Personally, I favor selling greater than 10 delta. However, capping leverage at 3x seems quite acceptable to me regardless of where you sell. I do think it noteworthy that from a portfolio margin standpoint, selling more contracts at a lesser delta and fewer contracts at a larger delta for the same initial leverage are different because that leverage will grow faster in the case of more contracts. I doubt this would run into issues at 3x, though.
I still don’t know the right answer or whether a right answer exists, and I certainly don’t think we have enough conclusive data at this point to reject any related approach outright.
My number one reason for DOTM puts: Everybody else else is way too close to ATM, so there’s more alpha for me there! 🙂
Please ignore. I’m not sure how to subscribe to the comments besides making a comment so that’s what I’m doing.
Any updates on this strategy from the past couple of weeks? These big down days and the vol spike must be causing some blow outs? I was doing this but stopped as I could not handle it in the volatile period at the start of Jan 2019. Would love to hear how its doing and how one is managing through this period?
Losses on Jan 24, Jan 31 and quite substantially on Feb 24. But still only down 1.5% YTD for the entire account (puts trading plus fixed income portion). Much better than the S&P500.
Also, I made all the option premiums this week on Wednesday and Friday. So, recovery started already! 🙂
I would have been down about 8% from Wednesday the 19th to Friday 2/28 following this method using 2x leverage and selling around the 10 delta put (but I have modified what I’m doing a bit).
I do 5 delta now, that should explain the difference.
When you say Delta 5 is it same as -0.05 on options scanner?
Yes. Puts have a negative Delta. So when I say 5-Delta it’s the -0.05 on the IB trading screen. 🙂
If you only had $20k of cash to use would you still implement this strategy on a lower level?
One thing you could do is to trade vertical spreads, i.e., short a put with strike X and buy a put at strike Y where Y