SPY Short Put 45 DTE Cash-Secured Options Backtest
In this post we’ll take a look at the backtest results of opening one SPY short put 45 DTE cash-secured position each trading day from Jan 3 2007 through Jul 31 2019 and see if there are any discernible trends. We’ll also explore the profitable strategies to see if any outperform buy-and-hold SPY.
There are 54 backtests in this study evaluating over 170,500 SPY short put 45 DTE cash-secured trades.
Let’s dive in!
Contents
Summary
Systematically selling puts on SPY was profitable across all delta targets.
No option strategy outperformed buy-and-hold SPY with regard to total return.
Methodology
Strategy Details
- Symbol: SPY
- Strategy: Short Put
- Days Till Expiration: 45 DTE +/- 17, closest to 45
- Start Date: 2007-01-03
- End Date: 2019-07-31
- Positions opened per trade: 1
- Entry Days: daily
- Entry Signal: N/A
- Timing: 3:46pm ET
- Strike Selection
- 2.5 delta +/- 2 delta, closest to 2.5
- 5 delta +/- 2 delta, closest to 5
- 10 delta +/- 2.5 delta, closest to 10
- 16 delta +/- 3 delta, closest to 16
- 30 delta +/- 3.5 delta, closest to 30
- 50 delta +/- 4 delta, closest to 50
- Trade Entry
- 2.5 delta short put
- 5 delta short put
- 10 delta short put
- 16 delta short put
- 30 delta short put
- 50 delta short put
- Trade Exit
- 25% max profit or when DTE = 21, whichever occurs first
- 50% max profit or when DTE = 21, whichever occurs first
- 75% max profit or expiration, whichever occurs first
- 100% max profit or expiration, whichever occurs first
- 1x Stop Loss
- 2x Stop Loss
- 3x Stop Loss
- 4x Stop Loss
- 5x Stop Loss
- Max Margin Utilization Target (short option strats only): 20% | 1x leverage
- Max Drawdown Target: 99% | account value shall not go negative
Assumptions
- Margin requirements are always satisfied
- Margin calls never occur
- Margin requirement for short CALL and PUT positions is 20% of notional
- Margin requirement for short STRADDLE and STRANGLE positions is 20% of the larger strike
- Margin requirement for short VERTICAL SPREAD positions is the difference between the strikes
- Early assignment never occurs
Mechanics
- Prices are in USD
- Prices are nominal (not adjusted for inflation)
- All statistics are pre-tax, where applicable
- Margin collateral is held as cash and earns no interest
- Assignment P/L is calculated by closing the ITM position at 3:46pm ET the day of expiration / position exit
- Commission to open, close early, or expire ITM is 1.00 USD per contract
- Commission to expire worthless is 0.00 USD per contract
- Commission to open or close non-option positions, if applicable, is 0.00 USD
- Slippage is calculated according to the slippage table
- For comprehensive details, visit the methodology page
Results
Win Rate
Closing less-risky positions at 75% of max profit or expiration yielded the highest win rate. As trades were opened closer to the money closing positions at 25% max profit or 21 DTE began to improve the win rate.
Annual Volatility
An active subscription is needed to view this content. Already a subscriber? Log in. Not a subscriber? Sign up for just 99 USD a year. Learn more.
Worst Monthly Return
An active subscription is needed to view this content. Already a subscriber? Log in. Not a subscriber? Sign up for just 99 USD a year. Learn more.
Average P/L Per Day
An active subscription is needed to view this content. Already a subscriber? Log in. Not a subscriber? Sign up for just 99 USD a year. Learn more.
Average Trade Duration
Lower delta positions reach profit targets or are stopped out sooner than higher delta positions.
Compound Annual Growth Rate
An active subscription is needed to view this content. Already a subscriber? Log in. Not a subscriber? Sign up for just 99 USD a year. Learn more.
Sharpe Ratio
An active subscription is needed to view this content. Already a subscriber? Log in. Not a subscriber? Sign up for just 99 USD a year. Learn more.
Profit Spent on Commission
20.77% – the blended average percent of profits spent on commission across all short put strategies.
Total P/L
The higher the delta the greater the total return.
Overall
None of the 54 strategies outperformed a buy and hold SPY with regard to total return.
Discussion
Through this study we can compare two risk management strategies, namely early management and stop losses.
It appears that managing positions early, in general, outperforms using stop losses. It also appears that using stop losses, in general, outperforms holding till expiration.
Additional Resources
Private, Custom Backtests
Discover your edge with private, custom backtests for as little as 99 USD. Learn more or contact us for a quote.
Trade Logs
Visit the trade log store and download the data used in this and other backtests.
s1 signal
Trade S&P 500 options more efficiently using the s1 signal. Learn more.
s2 signal
Trade GLD options more efficiently using the s2 signal. Learn more.
s4 signal
Trade Russell 2000 options more efficiently using the s4 signal. Learn more.
mREIT Preferred Share Dashboard
High-yield, low-beta alternatives to cash or treasury bills. Learn more.
Consultations
Schedule a consultation to review your specific scenario and get direct answers to your questions.
-
SPY short put 45 DTE Bundle$269.99
June 10, 2019 @ 3:56 am
I have an IB account using Portfolio Margin (PM).
Moving from short put to a vertical spread reduces the margin requirement enormously.
Could this help?
June 10, 2019 @ 5:10 pm
Thanks for visiting Stephen.
It might.
When I was emulating BigERN’s strategy in parallel to mine I received a communication from IB that essentially read: “even though you’re technically compliant with margin requirements, we don’t like the risk you’re taking. You have 1 week to decrease your risk otherwise we’ll charge you an ‘exposure fee’ of 2% p.a. on notional exposure above ‘x'”.
If IB’s risk mgmt uses a 20% drop as the “worst case” risk on a broad-based index and the purchased put in the vertical is 25% OTM, it won’t help. IIRC FINRA sets a floor of 20% but institutions can be more conservative and use 25%, 30% or more in their risk calculations.
As for strategy performance, it’ll be significantly muted due to the cost of the put. The few times the protective put will be profitable won’t offset the majority of time it expires worthless.
June 11, 2019 @ 3:05 pm
Thanks for the reply.
Here are some figures taken from IB just now, to compare a short put with a vertical spread.
I’ve used SPX with 45 DTE and aimed for Delta = 5.
I used 5 put contracts and 20 spread contracts to get similar credit.
Short Long No. Price Credit Margin % ret. At Risk % ret.
2530 5 5.2 2,593 76,520 3.4% 1,265,000 0.20%
2530 2480 20 1.35 2,646 36,152 7.3% 100,000 2.65%
I hope this formats correctly!
Question: why would you NOT want to use the spread and cut your margin in half?
June 11, 2019 @ 3:11 pm
OK, disaster on the formatting.
Short put: 5 at 2530 credit = $2593, margin = $76,520 return on margin = 3.4%, at risk = $1,265,000
Spread: 20 at 2530/2480 credit = $2646, margin = $36,152 return on margin = 7.3% at risk = $100,000
June 11, 2019 @ 8:49 pm
It formatted correctly when I received the email notification of your comment. In the web page it’s another story 🙂
That’s a great question! I see the same on my end and it looks very plausible.
Let me look into using 5 vertical spreads on SPY with $5 (or on SPX with 50 pt) wings to replicate the credit received for a single naked put. Perhaps that’ll be my next post.
June 14, 2019 @ 3:53 pm
Spintwig, I’ve been catching up on your blog and BigERN’s as well because I’ve had a lot of trouble finding anyone else who invests like I do. I think the best way to describe what I do is that for VIX<16 I trade like you do, but with Put Credit Spreads on the SPX. I've modeled this going back to 1993 and it significantly beats the S&P buy and hold. The difference from what I've read of your strategy is that as the VIX startes to move up, I start to enter trades for shorter durations and at 17.5+ I trade only 2 DTE. Credit spreads with 2-3 strike width (10-15 SPX points). This allows to cycle your capital faster, while getting as much exposure to the higher premiums as possible.
You should really look into how well Credit Spreads improve your Return on Capital. I love your posts, BTW, I really appreciate how much effort you put into it. Thanks!
June 15, 2019 @ 12:03 pm
Thanks for stopping by, Jeff! Appreciate the feedback; glad you’re enjoying the content.
The 2 DTE credit spreads when VIX is over 17.5 – are those held till expiration?
I shift from SPY to SPX 45 DTE puts when VIX is over 17.5 and take profits early. It worked exceptionally well last month. Would love to formally backtest it through the GFC but ORATS doesn’t have the ability and OptionStack doesn’t have the historic data.
Definitely a fan of cycling capital faster.
The next post will be about the performance of vertical put spreads and will go live early next week. One of the sections will indeed cover capital efficiency. Stay tuned 🙂
June 15, 2019 @ 4:31 pm
I do hold the 2 DTE until Expiration. How high the VIX is determines how far I move out of the money. Right now it seems to range 5-7 delta. Going to expiration eliminates Theta even if the market is still moving down since you’ll expire worthless, and it also saves on the commission of closing out the position…especially since you’ll usually be able to open more positions in Credit Spreads than when you do a naked put.
I manually backtested about 1200 transactions and then used those pricing averages to apply to the days I did not manually backtest. The data says:
VIX < 16 use longer term contracts and close early, either like you do, or like I do after 20 trading days (28 calendar days)
VIX <17.5 I trade the weekly that is closest to 8 trading days and hold to expiration.
VIX 30 open a Put credit spread very close to ATM for 90 days and close at 50% profit when the volatility comes out of the market.
Doing this over the last two and a half years would have returned 136% or 37%annualized of the account full balance.
During VIX < 16 I only have max 60% invested and then that ramps up as the cycles get shorter
Looking forward to seeing your analysis! Last month was pretty awesome with the spike of volatility. I ended up with just under 8% total account gain for the month because of that!
June 15, 2019 @ 9:26 pm
Interesting!
After a quick glance at IB, the credit received on 2 DTE 10-15pt SPX credit spreads is ~$25 to $30 minus $3.26 in commission for the pair of contracts vs $50 for a naked put minus $1.63. Bump up the credit received a little to account for the +17.5 VIX (right now just under 16). About 10% of credit received is spent on commissions (ignoring slippage costs of SPX). Seems high, no?
The 90-day put credit spread is exactly the same duration range I concluded in the spintwig vs ERN post, though the idea was to use a 90-day call to capture the ride up from a correction (when VIX is likely to be at 30+) as opposed to using a near-ATM spread to capture the VIX crush. I like your approach much better and will likely do that going forward. Thanks for the idea!
Nice returns! I’m using a significantly more conservative approach as my numbers weren’t anywhere near that. I made 0.5% in profits which softened the overall down month of my underlying SCHB portfolio. Given my target is 0.8-1.0% p.a. I consider it a significant win / month.
Do you keep a trade log?
June 16, 2019 @ 10:36 am
Just to be clear when I use 2DTE trades, I’m looking at 2 Trading DTE. Your example appears to come from the June 17th Expiration at a strike of 2855 or 1% out from ATM (which I would say is 1 Trading DTE).
Just to evaluate the Return on Capital for a Naked Put vs a Credit Spread at 2855/2845, TD Ameritrade tells me that the Margin required for a Naked put is about $54,300. With a net Credit of $48.37 ($50 minus $1.63 commission) this is a Return of 0.089%.
The Credit Spread selling a Put at 2855 and buying one back at 2845 Gives a net Credit of $21.74 ($25 minus $3.26 commission), but with only $1000 of Margin required. This is a Return of 2.174% or about 24X of the Naked Put.
The Downsides of the Credit Spread is greed. I could look at this and say “wow! I can open 54 contracts for the same amount of Buying Power and get $1,173.96 vs $48.37, I’d be crazy not to!!!”. The Naked Put is far easier to defend and takes smaller loses with a big market move to the negative. Take this example: Friday June 14th, 2019 the SPX closed at 2886.98, and we went 1% out at 2855…suppose the market closes down 1.25% at end of Day Monday (which happened 5.55% when the VIX was between 14.5 and 16.5). The market would be at 2850.98 and the Naked put would experience a loss of $4,107.25, but the 54 Credit Spreads would take a total loss of $22,179.15.
Credit spreads allow you to trade farther OTM which reduces risk drastically as well as lets small accounts participate in this “revolver” method that I use to revolve your Capital faster as VIX goes up. I have started this very small with only about $17K invested between two accounts (one Roth IRA, one regular brokerage).
As far as the 90 day Call vs a Put, What I saw looking at the S%P Data is that with a Put or a Put Credit Spread that is within 5% of ATM, you get two gains: 1) the pricing drop as Volatility leaves the market (Vega), which also gives 2) the gains as the market recovers and moves away from your strike price (Delta). When I looked at selling calls, I kept running into the market recovering and rapidly challenging whatever strike price I chose and while I gained on the Drop in Volatility, I gave most or all of those gains back in Delta as the market moved closer to my Strike.
I do use a trading log for each transaction as well as to set the schedule for the next trade. I’m trying to boil all of this down to a mechanical investing process that takes emotion out of the equation of what to do.
June 16, 2019 @ 4:34 am
Jeff,
Why does using 2 DTE credit spreads to recycle your capital faster not also apply with lower values of VIX? Why the change to longer duration spreads?
June 16, 2019 @ 11:06 am
Hi Stephen, The Revolving of your capital does apply at lower VIX ranges, but what I found when modeling the last 26 years of market moves over 2 trading day periods is that your losses are greater at low VIX due to the pricing and how close to ATM you have to go. At VIX<12 for instance you get an average of $26.74 for a Spread of 10 strikes or 50 points wide at 2% away from ATM. These very short duration trades have a huge premium drop-off the farther you go OTM. During the last 5 years we have seen a lot more “Flash Crashes” that cause the market to move more than 2% in 2 days. Over the last 26 years when VIX is 15 or less, the market has moved down more than 2% in 2 days 36 times or 1.5% of the 2,398 days the market was in that range. Contrast that with VIX between 20 and 22 and you get an average Premium of $155.90 for a 10 Strike wide trade, but 4% OTM. Looking at the 26 years of data, the market has move more than 4% in 2 days when the VIX is between 22 and 24 only 2 times out of the 469 times that it has been in that VIX rage or only 0.43%. So at low VIX your pricing is lower (even more so with commissions taken out) and your risks of loss are more than 3X higher. So when the VIX is less than 16 I open a credit spread that is 40-72 DTE on the Monthlies with the intent to close after 28 days. This gives time for the trade to come back to positive.
A Great example of this was on May 1st, 2019 I opened a 51 DTE Put Credit Spread (when SPX was at 2849 and VIX was 12.92) at 2830/2815 for a credit of $180 with the intent to close it out on May 29th (28 days later). Well the market dropped and VIX went into a 18-21 range pretty quickly and I held the trade open. I started running several 2 trading DTE and 8 trading DTE trades during the elevated time. That first Credit spread on the worst day was showing a $952 loss on paper but since I had time I waited. It now shows a $43 gain were I to close it today, but I’m planning to let it expire on the 21st and keep all $180 of premium (92% chance).
An example of risk is that opening a 2 trading DTE on Thursday May 3rd would have seen a 2.07% drop to the end of the day Monday. Not a huge loss, to be sure, but a loss none the less. My average gain per 10 strike Credit Spread was under $10 for VIX17.5 the average gain is between $68 and $99. Again that’s due to moving farther OTM as VIX rises, so at VIX 25, I’d open a trade that is 4.5% away from the ATM price. It is very difficult to find times where the market has moved more than 4.5% in only 2 trading days.
Sorry for all the numbers and the wall of text I’ve written on the last couple of posts.
June 16, 2019 @ 11:13 am
Weird, something happened when I pasted my comments in. That last paragraph should read
“An example of risk is that opening a 2 trading DTE on Thursday May 3rd would have seen a 2.07% drop to the end of the day Monday. Not a huge loss, to be sure, but a loss none the less. My average gain per 10 strike Credit Spread was under $10 for VIX17.5 the average gain is between $68 and $99. Again that’s due to moving farther OTM as VIX rises, so at VIX 25, I’d open a trade that is 4.5% away from the ATM price. It is very difficult to find times where the market has moved more than 4.5% in only 2 trading days.”
June 16, 2019 @ 11:16 am
ARRGH! under $10 is for VIX17.5 is $68-$99.
June 16, 2019 @ 11:22 am
So the text I’m losing repeatedly is for when VIX is less than 14 (average trade is less than $10). VIX greater than 17.5 is 68-99.
It’s too bad I cannot edit these posts. Sorry for spamming so many comments Spintwig! 🙂
February 27, 2020 @ 5:58 am
Hey Jeff, wondering how your revolver method is holding up these past couple of days!
February 27, 2020 @ 2:08 pm
Is has been a really wild ride actually. Had total wipeouts of 3 positions on Monday that wiped out about 8 months of gains. I’m surprised by the extent of the move, but this is the reason to never have more than 33% expiring on a single day. I had my position on Wednesday get penetrated and I rolled it out to Friday. That is a potential wipeout depending on what happens by close tomorrow.
My historical backtesting shows several times where there were 40%+ losses that, because of sustained high volatility afterward, allowed the losses to be made up within 4-8 months.
I’ll admit to having a pretty serious gut check this week and frankly the markets performed out at the extremes from a historical standpoint, making much of my strategy un-useable while the VIX is over 30, so that’s where I am currently…waiting for the markets to settle a bit and then I’ll be diving back in and working my capital back up.
Before Monday I had returned an average of 6.10% per month since I started Revolver, so I have faith that I can be successful in the long term by staying with the program.
February 27, 2020 @ 4:55 pm
Jeff,
Hard times and getting harder as the markets close today!
Can’t be many option sellers enjoying this ride…
As I look back over the past couple of weeks when the markets were making new highs, whilst the virus was running wild in China. I ask myself why I didn’t step aside?
Chaps, this too will pass!
February 27, 2020 @ 5:13 pm
On the bright side, there are now more data points for when VIX is over 30 🙂
After the first 3.25% down day and VIX jumped to 22.5, I opened some short puts on SPY with a strike of 286 (10D; over 12% OTM). Not ITM yet but it’s heading there in a hurry! Let’s see how this plays out.
Took profits twice in Jan decreasing my stock AA by 10%, so there’s some relief there…
June 10, 2019 @ 10:03 am
The more I read various websites and backtests, the more I come to the same conclusion, in the long run, it is just not worth it. The time, the opening and closing of trades, tax implications, missed dividends. Buy and hold is also not for me due to big drawdowns, but I think what would be ok is buy and hold SPY + hedging with options and using slight leverage because of much reduced drawdowns. Sure this will trail the returns in big bull markets and in years when SPY is up 20 % you will get “only” 16 %. Totally fine with me. What I find most important is that in years when SPY is down 30 %, 40 %, 50 %, you will be down 4 % using protection. Makes you sleep a lot easier, makes being 100 % or 150 % invested much easier.. One can also try to hedge for free by buying 1 year out ATM put option and financing it by selling monthly or weekly calls against it. Or maybe financing 50 % of the protection cost which would put short weekly / monthly call much further OTM. All in all those simple strategy are mostly just too much hassle. After all if things would be so easy, a lot of big hedge funds would do it much before retail would. They have the money, the brains, the machines… 🙂 But ok if you find it worth it for 1% additional income it is ok..:)
June 10, 2019 @ 5:15 pm
Welcome back Tomaz! What are your thoughts on holding 80/20 or 60/40 equities/bonds (or another combination thereof)? That might be the simplest and least time consuming way to obtain the portfolio drawdown protection you’re seeking.
June 11, 2019 @ 3:25 am
Thanks! I love your articles, specially back-testing you do!
If I would to include bonds, I was thinking more in line of leveraging them up to same volatility level. Something in the line of risk parity portfolio. Because it has been proven time and time again that 80/20 is actualy 95/5 when it comes to risk. Stocks are much riskier and much more volatile and in 80/20 almost all risk comes from bonds. 60/40 would be ok I guess, but then again you can probably get the same results just being less invested in stocks and no bonds. But 150% funds could be an alternative. To hold like 90 % stocks and bonds levered up to 60 %. Then again all backtests for bonds are almost useless I think because we had like 30 years bonds bull market where yields came from 15 % down to almost 0 %.. So we for sure cannot see as good bonds returns in the future. Not even close.. I read good academic paper that bonds are ok only if you are in withdrawal state where you can just take money from bonds if stocks do poorly for some years. But other than that it is better to just invest in stocks and if volatility is too much to stomach either decrease alocation, so like 80 % stocks and 20 % cash or hedging. 80/20 with cash I think is also ok because you have always some dry powder if stocks tank and you can buy them cheaply. So after big sell-off you are 100 % stocks and then after recovery you go back to 80/20 etc.. As you can see I am still not 100 % decided what I will do with my cash but because I am risk averse I lean towards 100 to 150 % investment in stocks only but being 100 % hedged all the time. I will try to recover some heding costs by selling far otm weekly calls and just accept the fact that I will slightly lag returns during big bull markets chaning this for the piece of mind when markets sell-off big. I know I would not be able to sleep without protection, seeing 50 % of my capital evaporate in the next bear market.. Yikes 🙂
June 11, 2019 @ 9:18 pm
The risk-parity portfolio you’re describing sounds like Hedgefundie’s 40/60 UPRO/TMF approach over at Bogleheads.
Based on the screenshots it looks like he’s implementing it in M1 Finance due to the free “trades” and ability to auto-rebalance at the click of a button (quarterly rebalancing was deemed optimal).
Hedgefundie has allocated 15% of his portfolio to the strategy. Perhaps throwing a few dollars at it may be an idea. I’ve considered it but haven’t had enough conviction to implement.
June 12, 2019 @ 4:14 am
Thanks for that! A very lengthy post on BH forum, but looks promising. This is exactly what I had in mind.
June 12, 2019 @ 12:51 pm
I’ve seen similar studies (https://www.projectoption.com/short-put-management-study/), but nobody seems to want to compare their results with just buying and holding SPY, so thanks.
Have you considered backtesting strangles with the same methodology as you have here?
June 12, 2019 @ 11:52 pm
Thanks for stopping by, JEI
You’re welcome —
It appears ProjectOption has an affiliate program with TastyTrade and, of course, they’re selling options trading educational materials. The truth will hose their business model and their affiliate relationship lol. Me on the other hand, I’m not selling anything so nothing to loose and no agenda. Just nerding out 🙂
I haven’t. In Part 1 I reference some TT research on the topic. It’s concluded returns are muted in return for lower volatility. I’ve got some red in my trade logs from strangles, too (see part 4 for the log).
Is there a particular strangle setup you’re looking to explore? I’m open to running a few backtests and sharing the results.
June 13, 2019 @ 11:21 am
The strangle/iron condor setup I’d like to see compared to the S&P is the one that TastyTrade, OptionAlpha, etc are promoting, basically sell 15 delta options on either end, take profits early, and manage failures.
(e.g. replicate this one: https://www.projectoption.com/iron-condor-management-study/)
I suppose one major difference you won’t be able to replicate is that the underlying isn’t always SPY, it’s whichever ticker has the highest IV rank & liquidity on the day of the trade.
I’m very confused right now about the truth. On the one hand is all this indirect anecdotal evidence: my gf’s dad who trades options (and is a financial planner who generally recommends indexes), ERN putting 35% of his portfolio into his options strategy, and the abundance of blogs out there claiming to make money.
There’s also this study, which claims that both BXMD (selling 30 delta covered calls) and PUT (selling ATM puts) indexes outperform SPY:
http://www.cboe.com/micro/buywrite/wilshire-2016-benchmarks-30-years.pdf
I think the actual answer might be what you concluded above, “selling premium is best used as a tactic to generate returns in addition to a SPY portfolio.”
Another possibility is that beating the market is possible, but only by limiting risk and more efficiently allocating capital (e.g. spreads), using leverage through margin, and using portfolio margining so that your money is working while it’s working.
June 14, 2019 @ 2:36 am
What I dont like about the idea of selling premium as a tactic to generate returns in addition to a SPY portfolio is that both are actualy the same type of trade. They both make money when SPY goes up and on the other hand if we get another 2008 or even worse, you can go broke. You can loose 60 % of capital on SPY and additionaly 30 % by selling premium and you are screwed.. I think if one is long SPY one should seek strategy which goes the other way or makes money unrelated to SPY.. Something in the line of trend following, bonds, covered calls etc. Strategy that is uncorelated or even better negatively correlated. Maybe even barbell portfolio from Taleb. I read an interesting study not a long ago by Spitzernagel who suggest buying like 0.5 delta (not a typo) put options 1-2 months our for 0.5 % of portfolio and rolling them, but only when SPY is high, overvalued, has high tobbins Q ratio but other measures could be used too. Sure you will underperform most of the years, but when innevitable big sell-off happens those puts explode in value. 30 % OTM put actually increased 60 times and offered a complete protection in 20 % sell-off.. Now imagine 30-50 % selloffs.. You can actualy make more money on those long far OTM puts than you loose in main portfolio. Insurance is not cheap, but if you use it the right time, when SPY is making new highs months after month, it can pay off in the end. The hard thing with strategy like that is that you will underperform slightly month after month, but if you have patience you will be rewarded at the end.. And the more important for me is strategy like that can make you sleep soundly.. I could never sleep being 100 % invested without protection and a year like 2008 would put me into depression if I was to loose 50 % of my capital in a few months… Yikes.. 🙂
June 15, 2019 @ 10:40 am
As a mostly buy & hold investor, I don’t think of a selloff as losing money, especially in an index. I’m just going to wait for its eventual return, and it’s not an opportunity cost because I was never going to do anything else with the money anyway. Maybe I’ll think differently when I have to live off of it.
So, if there’s a massive selloff, and I can’t roll those puts, I just accept assignment and wait it out, right?
June 15, 2019 @ 12:04 pm
That is what I thought too… It is only when you look at 2008 and when options increase in value like 60x that you see why this is such risky strategy. Yes, you could wait it out if it would not be for a margin call and IB selling your portfolio to raise money. You will be executed at the most un-apropriate time. Gamma and vega is what kills you at that time. Stock owners can wait it out, option sellers cannot, because at the same time many things happen. Volatility goes through the roof, short options value increase in value 50x and at the same time brokers raise margin requirements.. I did some excel backtests because I am too stupid and too cheap to do real ones, but even in XLS you can see that you have like 50 % drawdown in a case like that. At the same time you have 50 % drawdown from stocks. Now calculate.. Your 100k portfolio is down to 50k just from stocks and your short options which you sold for 10k are now worth 50k. And you are short them…
Believe me, I am searching for a holy grails too and for the long time I thought selling premium might be it. But the more research I do the more I am convinced by what they say, it works well for 10 years than a catastrophic even wipes you out.. This is what happened to optionsellers.com … They were actually right on the trade at the end, the problem is they got margin call in between due to nat gas and crude huge divergence and they could just not ride it out.. And those were hedge funds guys that wrote some bestseller option selling books.
Why do you think premium exists in options? It is there for unforseen events too. It there would only be occasional 10 % falls, then there would not be such a hefty premium for options seller to exploit. It is there for unforseen events too, for black swans like Taleb calls them. There is no edge in selling options all the time. There is an edge if you sell VIX = 90, because it will eventually go back down. But selling premium all the time, even at VIX = 10 is just stupid.. It just take a lot of years to learn this, because such evens happen like once in a decade.. And selling premium is such a rush for us because we all like to be right and selling premium is the only trade where you can be right 99 % of the time, making money day after day, week after week, month afte month.. Until that one month that takes you out and shows you why there was a premium in option in the first place..
To each his own, I still plan to trade option, but now I will focus on reverse.. Buying them, for convexity, protection and huge payoff when they do.
And one more note. If you insist on doing it, do it the Spintwig way where you only sell only a few not leveraged way like most people do.. In 100k portfolio sell 3 puts and accept a very low return, do not try to bump up return by selling 6 or 10 because this is what burries you at the end.
October 9, 2019 @ 8:37 am
Agree on your notion of black swan events wiping out put option sellers. Catastrophic events like a terror attack on a French nuclear power plant, the big earthquake in San Francisco, war in the Hormuz straight… can not be predicted, but in those events VIX will likely hit 40 within a day. So writing put options when the VIX is below 17.5 or 20 seems an overly risky strategy…but not sure if your way of leveraging a portfolio and buying put options is the way to go…
June 15, 2019 @ 12:42 pm
Using dynamic underlying may work (will have to see if that’s something that can be built in OptionStack) but there’s significantly more risk if it’s not a broad-based index. The non-systematic risk that’s introduced by using sector indices or individual equities could in theory outperform. Then there’s identifying and attributing the performance between strategy and luck… what are the specifics on the strategy?
My next post will speak to the IC study.
As for the PUT index, yes, the index outperforms S&P but the fund has an expense ratio of 0.44% (waiver brings it to 0.38%). I’d like to see the the actual returns of the PUTW fund replicated back to 1986 (to match the index) with the 0.44% expense ratio factored in. Let’s use 0.03% for the S&P Index expense ratio (i.e: SCHX). We’ll ignore the fact indexing / ETFs didn’t exist as we know them back then.
There’s a ceiling to the practical effects of spread capital efficiency and/or managing early. I’ll elaborate in the next post 🙂
July 28, 2019 @ 2:29 pm
Spingwig, thanks for this article and site! Found it while trying to find backtesting results of selling weekly ATM calls on a stock like F. Am totally impressed with your methodology and this particular article which is the first have read of yours. Superb test, you gave the strategy every chance to succeed. Am surprised at the results, that a straight long beat them all, but that is valuable info.
Have you done same test on a few components of SPY? That would also be valuable. Would it be better or worse?
I am going to poke around your site more. CNBC Fast Money is their best show and Options part in particular, Carter often makes really good directional calls. BUT their strategies they play it with are so primitive and the same one over and over, a debit spread with 1×1 ratio. They should have you on seriously to show your results of various backtests of strategies.
My background is have been successful investing this century, but with large drawdowns. Mostly directional stock trading. Overall about a 20% CAGR, or about 3 times better than Buffet, but like him two 40%ish drawdowns.
Am looking for more income oriented lower risk methods going forward. Am guessing weekly options provide an opportunity.
In contrast to Options Action’s simply 1×1 debit spreads I do things like a 4×3 Calendars with 2 weeks vs 1 week, or 4 x 3 wks. As you make good money maybe even 40% if it sits still, you make money if it goes up and you are 100 shares long at a tiny cost/risk if it takes off. The only way you lose is if it goes down a fair amount and you lose a small amount of money, but up to 100% of risk if it crashes. The win % is about 66%. So if you avoid using the leverage aspect and retain some cash it is quite safe. I have no idea why CNBC never suggest such a trade or all the other ones that are so much better than their primitive debit spreads that lose money if it sits still and have less upside if it takes off.
Have you ever back tested such a strategy? or something similar to buying 500 F and selling each week 4 atm Calls? Would be most interested in results and also 4×3 calendars over long period.
Will now poke around you site.
Thanks
July 29, 2019 @ 2:58 pm
Thanks for visiting, Darp! Glad you’re finding the content useful. Appreciate the feedback 🙂
I haven’t done backtests on the SPY constituents… yet. I can certainly add some to the list.
If you know how I can get in front of the right people, I’m open to having the research exposed to a wider audience.
Haven’t done any asymmetric strategies such as 4 x 3; not opposed to looking into them.
October 9, 2019 @ 8:41 am
Spintwig!
Thanks a lot for all this nerdy value you provide to us readers. Have been looking for a website that goes a bit deeper than the usual option fanboys. Lucky to find you on Reddit and read your articles. Keep up the great work!
October 10, 2019 @ 12:36 am
You’re welcome! Glad to hear you’re enjoying the content; thanks for stopping by.
January 17, 2020 @ 12:09 pm
Hi Spintwig
Thanks for all the effort you put into this free information. I would be curious to see the correlation of starting IV with returns. TT is always pushing to go bigger when IV Rank is high, but I don’t know if that is really a good idea. Also, do you factor in any interest received from your collateral? Thanks!
OM
January 27, 2020 @ 6:52 pm
Hey OM – sorry for the delay in replying. Your comment was erroneously marked as spam.
Good callout. I intend to expand on this study by exploring performance during different IV regimes. The tricky part is quantifying performance. Suppose we open positions only when IV Rank is > 50. Two questions arise:
1) Is the resulting performance a consequence of higher IV Rank / subsequent vol crush or is it a consequence of timing luck?
2) If a position isn’t opened daily, we now have <100% time in the market. What's the best way to measure strategy performance against a benchmark that normally experiences market exposure 100% of the time.
TT's research is big on opening a single position/lot then rolling it over the duration of the backtest, which lends itself heavily to timing luck. Depending on which day the trade is opened can have a substantial impact on big-picture strategy performance. This study opens a position every trading day so there is much less timing luck variance.
As for interest on collateral, all the cash-secured studies are modeled as if each backtest is an ETF that implements the given option strategy. There is no collateral for interest to accrue. The leveraged backtests have a different methodology and do factor in the interest accrued on collateral. If you'er interested, feel free to take a look at the Options Backtest Builder. It's the tool I use to build the leveraged backtests and allows you to see under the hood at how everything's calculated.
Hope this helps!
February 27, 2020 @ 4:17 pm
Don’t chicken out now!!
I may have mentioned I lost my stomach for selling. I found a service for high probability option buys, and I had been doing really well… but I’ve also lost about ~20-25% from my highs over the last month.
_shrug_
Maybe now’s a good time to do a few credit spreads though.
February 27, 2020 @ 5:54 pm
Oh, not chickening out at all, my program has ALWAYS said that 30+ VIX is too unpredictable to trade consistently. My intent is to let a bit of the panic subside, but still trade for premium.
May 2, 2020 @ 11:36 am
Hi Spintwig
Thank you for the backtests.
on this page, the worst monthly return of SPY is 37%. However, on pagehttps://spintwig.com/spy-short-strangle-options-backtest-results/, it is 14%. Can you please check why? Thank you.
May 3, 2020 @ 10:20 am
Great catch songlake68!
The reason for the discrepancy is that I didn’t have cell referencing configured on that particular attribute when I refreshed the charts and tables’ benchmark data per https://spintwig.com/option-research-and-blog-updates/.
The correct value is -16.52%. I have updated all the SPY cash-secured and spread studies effectively immediately. It may take up to four hours for the CDN to update and propagate the changes across the globe.
My workflow has improved since those times and all newer studies bypass the issue entirely by way of improved data administration methodology.
May 20, 2021 @ 11:03 am
So buying and hold has a better CAGR but selling 16D puts has less volatility. Would selling naked puts at 16D offer a good combination of lowered volatility and comparable return of buy / hold?
May 20, 2021 @ 11:11 am
Depends what you mean by “good combination” and “comparable return.” Total return of the strategies vs buy/hold is in the second-to-last chart at https://spintwig.com/spy-short-put-strategy-performance/#Total_PL.
It’s essentially a question of “What are you trying to accomplish?” The various strategies can be reviewed and implemented to generate a risk and return profile that aligns with your financial goals and objectives.
January 31, 2022 @ 6:24 pm
The performance here hinges off the margin utilisation / leverage. Your tests run 20% only. You could easily outperform S&P by a large margin by simply increasing this backtest parameter. This test isn’t conclusive at all.
January 31, 2022 @ 6:28 pm
Here’s the latest leveraged version of this study: https://spintwig.com/spx-short-put-45-dte-s1-signal-options-backtest/
An older version with profit takers is at: https://spintwig.com/spy-short-put-45-dte-leveraged-options-backtest/