10-Delta Short Strangles

[tayte]

Plus they're selling the options that are simply 4% OTM, which doesn't take probabilities into account at all. I pulled up a 30DTE SPX option chain (it's Saturday, 7/17, so the 16Aug) and SPX last closed at 4327. 4% lower than that is 4154. CBOE's calculation methodology says they have to choose the strike that's at least 4% away, so we have to pick the 4150 strike. It's Delta is 22 and Prob.ITM 24.0%. Maybe we'll see about the same probabilities on the Call side, right?

4327 x 1.04 = 4500. The 4500 strike's Delta is 7! And it's Prob.ITM is 6.7%. I guess a trade like that would work, but doesn't that artificially skew the trade to the Put side? Maybe that works just as well as "probability-centering" the strangles like I do, I don't know. (I just made up that term, maybe it's silly.)


Thoughts on my analysis of the CBOE Short Strangle Index?
Take care,
Mike

Hi Mike, thanks for taking the time to explain your thoughts in detail. How did you derive the 6.7% from the 4,500 strike? If you estimated this from delta alone, then it is not accurate.

I cut my teeth making the market on JGB options for a while, more than a decade ago. Even back then, there wasn't much easy money to be made trading options, regardless of expiration lengths.

In the long run, things don't stay out of line long enough for you to easily beat buy & hold benchmarks. You haven't explained how you expect to have an edge in vol forecasting, and finding spots where implied is significantly more expensive wrt your own estimate, i.e. where you expect an overwhelming majority of participants to be wrong. That's really your only legitimate chance at beating buy & hold SPX. Maybe you do and just haven't disclosed it, if that's the case then all the power to you.

 

[caroy]

Interesting idea with the flies, is that mostly what you trade option-wise? How do you make them work in general? And is this a "100 pt wide lotto fly"?

View attachment 263659

AMZN earnings are on 7/29, is the idea that you would put a lotto fly on either side of current spot (whether with or without a strangle?), betting that the price move after earnings would land on/in one of them? Would you center them at +/- the Expected Move? I don't know what that's predicted to be, but would your position (without a strangle) look something like this?
View attachment 263660

You profit if AMZN makes the Expected Move, but not so much if it stays put or moves too far? Or are your centerpoints closer so that the triangles intersect and profit stays green in between?

I can see that working for an earnings play, but how would it work as protection for a strangle, and about where would you place the flies? I tried modeling it on CLF that I've been referring to in this thread, and this is the best-looking curve I could come up with:
View attachment 263663
That's an approximately 10Δ short strangle with a $1-wide iron fly $1 outside each strike. I thought I'd see the characteristic peaks of the iron butterfly at each end of the plateau, but that's the best I could get out of it. And if I moved the flies further out it just looked like this:
View attachment 263664
That lowers the BE and does provide a bit of extra downside protection, but not what I was hoping to see. Plus it's 10 contracts in and 10 out; at 65 cents per at TDA, that's $13, not insignificant against the $43 max profit of that trade. Have you done "strangles with flies" before?

I have a journal going called "weekly pin butterfly" that lays out the nuts and bolts of what I do I trade short term Thursday to Friday. I put on an array of butterflies usually ten positions in high dollar or high IV stocks and they are slightly directionally biased on the premise that all things being the same the underlying will gravitate towards the strike with the highest open interest. It's debatable whether "pinning" actually occurs. Those who advocate that it is real (myself included) argue it has to do with microstructure and market makers who are long options at the highest open interest have to gamma scalp their positions by buying stock as the underlying approaches the highest OI strike and selling as it rises above it creating the pinning effect. On expiration day you have the movement working for you with your long options and the theta bleed working for you at the short guts which increases the probability of the trade working for you. The greeks all become your friends if you have the movement correct. It's been working for me. Butterflies tend to spread their wings in the last couple of days with their biggest change in value. That being said I'm sitting all cash until Thursday afternoon usually around 2:30 PM CST I put them on and take them off usually by 10:00 AM CST of Friday. As I'm sitting with a decent chunk of capital in cash I'd like to add some weekly income trades to my style and that is where your strategy appeals to me.

The diagrams you have are about right. Selling the strangle inside the large wings low delta flies should produce the bat ears but then the diagonal down as you still have unlimited losses to the upside and are limited to 0 on the put side. You're correct in pointing out the cost of these. I'm eyeing higher priced SN's where the credit is large for a 5 or 10 delta strangle but this takes up a good deal of BP. But I could see myself doing this weekly in AMZN. I've been playing around last few weeks adding broken wing flies for income in SPX as one day plays. So far they've worked well. I like the income and they are defined risk. I can often roll up the broken wing pay a nickle lock in $200 on a 10 lot and leave a free fly on at 16 delta that could be a lotto ticket. And I like that SPX is cash settled so I don't have to stay around and babysit the screen all day. I sell a few iron flies on SPX as well looking for quick decay and that has worked well to for income.

 

[qlai]

So are you doing Covered Short Straddles/Strangles? But if the stock goes down you'll sell it but keep the straddle/strangle on? I had written this: {What kind of DTEs are you putting them on at, and if you choose to do a strangle what strikes do you choose?} Then I re-read and saw that you're legging in. Would you mind sharing an example trade?

Right, covered straddle/strangle because it was a bullish environment and I would rather get long than short if assigned. Depending on how wide the strangle and my view on the market, I may try to trade in and out of shares.

I haven't done these in a while, but the below one I remember because it was super juicy . It looks like I got into strangle first but then decided the risk was on the upside and bought shares.

 

[Theinkdon]

Hi Mike, thanks for taking the time to explain your thoughts in detail. How did you derive the 6.7% from the 4,500 strike? If you estimated this from delta alone, then it is not accurate.

You haven't explained how you expect to have an edge in vol forecasting, and finding spots where implied is significantly more expensive wrt your own estimate, i.e. where you expect an overwhelming majority of participants to be wrong. That's really your only legitimate chance at beating buy & hold SPX. Maybe you do and just haven't disclosed it, if that's the case then all the power to you.

Tayte, you give me WAY too much credit for knowing anything beyond the merest basics about options. I guess you don't use ToS, but you can add a column to the option chains they call "Prob.ITM," and that's where I got the 6.7% from. I think TastyWorks has a similar number. No Black-Scholes for me or calculating probabilities from delta or anything else. I'm simply an observe-and-act kind of trader. If TDA tells me a certain option has a 6% prob. of being ITM at expiration, I'll take that at face value. In fact, I've started mostly using that number in lieu of Delta as an estimate of probability, especially on the Put side, where almost all my breaches have been.

Also no, I don't claim to have any edge at all, and I'm not withholding anything. To be honest, I don't know enough to withhold anything! I don't try to forecast anything, I just mechanically implement the strategy. Maybe it's too simplistic, but I've observed after a fair bit of trial and error (in TDA paper-money accounts, mind you) that when selling very short duration strangles at 10∆ there's a very good chance they'll close automatically at 50% profit and I'll rinse and repeat, literally. If a trade behaved, I'll usually turn around and put another SS right back on it. For the few that don't behave (maybe 20%, I haven't tallied) I can usually roll the untested leg up or down to make the trade still work in that week, or roll out a week, always for a credit.

And I know I'm breaking all kinds of rules, but so far it's worked out, and I doubled a practice account with this strategy in 13 weeks. Dumb luck? Probably, but I'll keep doing it with PM and see what happens, and by September with real money. If it blows up it'll at least be a good education, but if it doesn't? Then maybe there's something to it, at least short of a market-wide event like March of last year or 2008 and 2000. But I like the rule-based nature of it, and the non-directionality of short strangles, which I suspect will do alright in any market conditions short of a system-wide shock like the COVID crash.
Cheers.

 

[caroy]

Sold the 5 Delta Strangle in Amazon on the open at 8.35. 3175-3770 so far so good. GTC order at 4.15 working. At 5.95 at the moment.

 

[Theinkdon]

Sold the 5 Delta Strangle in Amazon on the open at 8.35. 3175-3770 so far so good. GTC order at 4.15 working. At 5.95 at the moment.

Nice. You went all the way out to 5∆, huh? Looks like the trade is doing well for you so far. Did you happen to note the BP/collateral when you placed it?

 

[easymon1]

Are you guys looking for sideways action in the underlying?
What's your typical hold time?
What is your yield for a typical Winner per $100 ante?

"The short strangle option strategy is a limited profit, unlimited risk options trading strategy
that is taken when the options trader thinks that the underlying stock will experience little volatility in the near term.
Short strangles are credit spreads as a net credit is taken to enter the trade."

 

[easymon1]

<!-- wp:heading {"backgroundColor":"vivid-green-cyan","textColor":"black"} -->
<h2 class="has-black-color has-vivid-green-cyan-background-color has-text-color has-background"</h2>
<!-- /wp:heading -->

<!-- wparagraph -->
<p><strong> 0<strong>∆)</strong>. 20<strong>∆</strong> .</p>
<!-- /wparagraph -->

<!-- wp:heading {"style":{"color":{"back

 

[Theinkdon]

<!-- wp:heading {"backgroundColor":"vivid-green-cyan","textColor":"black"} -->

Sorry about that, didn't realize all the html code was going to come with it when I copied from my blog.
Deleted and re-posted.

 

[Theinkdon]

20-Delta Short Strangles Campaign starts today 7/19

In a TD Ameritrade Thinkorswim (ToS) paper-money account I was mostly selling Short Strangles (SS) with the strikes at about 20-delta (20∆). I was using high-volatility stocks whose option premiums implied returns of >1% per day (Premium ÷ Buying Power Reduction). The account started increasing very rapidly, much quicker than the concurrent strictly-10∆ SS strategy I was running on the same pool of stocks. So quickly that early in the 4th week I realized it was about to reach a doubling, and sure enough, Friday by 2PM it did. To put that in perspective, the 10∆ was "only" up 23% in the same 4 weeks. That's huge, of course, but the rate of increase of the 20∆ was mind-boggling. More of that backstory is in the first 3 posts in this thread if you haven't seen those.

So I'm going to run the 20∆ SS strategy again and see how long it takes to double, logging my trades here daily (hopefully).

Ground rules:
TD Ameritrade Thinkorswim paper-money margin account.
Start with $50,000, stop when/if NetLiq hits 100k.
Log end-of-day balances.
Measure drawdowns EOD to EOD.
Stocks will be high-IV, like almost GME & AMC high (but I don't trade those anymore).
Stocks will be >$5, >2 million shares avg. volume, and have weekly options.
Only selling short strangles (sell a Put and a Call in the same expiration, no protective wings).
Strikes will be the ones closest to 20% "Prob.ITM"* on ToS, whether higher or lower than 20.
Immediately place a Good till Cancelled Buy to Close order at about** 50% profit.
Trades will be 10DTE or less, and offer >1%/day return on collateral: (Premium ÷ BPR)/trading days.
Mon (& probably Tues) trades for this Friday. Wed or later, likely next Friday.
No stop loss orders, but I reserve the right to exit any trade at any time.
Position size 10%*** of NetLiq. Stay up to 75%*** invested.
First trade management technique will be to roll the untested leg toward the money.
That can include going all the way to a Short Straddle, but not inverted.
Next will be to roll out to the next week if possible for a credit. Otherwise go another week(s).
All rolls will be for a credit. Add up credits and change the BTC order to half of that.

That's all the rules/constraints I can think of. Let me know if there are any others I should implement. I'm trying to make this as rule-bound and repeatable as possible.

*Note: I call this a "20-delta" strategy because I started months ago with 10∆ where I was using delta as a proxy for probability, but I've come to trust more the Prob.ITM number in ToS. It might make a 1- or 2-strike difference sometimes.
**Note: I take out commissions, both entry and exit, when calculating "half-profit." TDA charges 0.65 per contract, so if I sold a strangle for $50 (0.50 Credit) I'd subtract (4 x 0.65) = 2.60 for a Net Credit of 47.40. Divide by 2 to get 23.70, then round down to 23, and the BTC goes in at 0.23. (It's a minor difference, 0.23 vs. 0.25, I just like to know that I got 50% of the net profit available.)
A related note: TDA has a Nickel Buyback program, where you pay no commissions on closing orders of 0.05 or less. Since it costs 1.30 to exit a strangle, it wouldn't make sense to exit at 0.06 because you'd effectively be exiting at 0.047. A 0.07 exit is effectively at 0.057, so if the BTC calculates to 0.06 or 0.07, I'll set it at 0.05.
***Note: 10% is far too aggressive but I chose it to minimize the number of active positions, and therefore the amount of time spent on this experiment.