You have for now understood that your simulated results were not attainable in real trading, or better said – were the result of a very calm market (we are all genius in bull markets) and….. overleveraging, but - seems to me - find a bit difficult to face reality. In time, you'll accept it. So, if you like the concept, you can still do it, but with more realistic target….. Think about it. If the difference between implied and realized is – for example – 5% per annum, only way to make 100% is apply 20x leverage. There’s no other way. And no broker will allow you to decide the leverage you like (for very good reasons...) So, accept that your proposed short strangle have probably a positive expected value (over time), because DOTM option are overvalued. This overvaluation is NOT A FREE LUNCH, it is the prize you receive for exposing yourself to the risk of blowing up (others call it …. because of the skewness of the expected P&L). And quite frankly, I strongly disagree with Magic’s last message, ……….. no one could improve the edge or find “alpha” with strike selection (otherwise anyone would do it, end the edge would disappear), not with a different hedging scheme (look at Euan Sinclair books, if you fancy to know better) and not even forecasting deltas (if one is able to forecast, she/he should trade stocks or futures, not options…). Also, understanding counterpart’s reasoning is useful in certain cases (es if you trade OTC), but it’s useless (and impossible) if you trade AMZ or AAPL or almost any other option in a regulated exchange…. (knowing that a market maker will probably build a a conversion/ reverse conversion with the DOTM options is maybe interesting, but not very useful in practice).
Thanks for doing all that work, and for including "normal" stocks. Those actually seem to make money over time using this strategy, aside from the March 2020 drawdown and the constraints of fills at Midpoint and no commissions. Does the backtesting software allow you sell at the Bid? Or does it allow you to include commissions? Commissions are usually a small bite of the premium collected, maybe 5 to 10% on 10-delta trades, much less percentage-wise on 20-delta trades because premium is so much higher. And Bid/Ask spreads aren't usually all that wide on stocks I've been trading, even way out at 10-delta; for 30Jul, the BB spreads are 1 cent in the 10-20 delta range on both the Put and Call sides. But BBBY spreads are wider at 2-5 cents. And RIDE spreads are 3-4. Etc, so that's probably a bigger impact than commissions. But as I've said before (and I think here on ET), I'm not convinced by backtesting. As a gross indicator, sure, but what you modeled isn't quite what I'm doing, and I'm not sure that the way any given person trades, even with strict rules, can be accurately reflected in backtesting. For instance, did you use a 50% take-profit rule? That's easy enough to model, I've read. But did you allow for any trade adjustments? That's where I think it gets harder. How about any kind of stop-loss? And what about when a trade comes off (as either a winner or loser), turning around and opening another trade, either on that symbol or another one? Because not only am I diversifying across about 20 different symbols, I'm also diversifying in time (at least a little bit) by continually rolling into new trades, and NOT just putting on a batch of trades Friday or Monday and then watching them succeed or fail. No offense meant, and I really do appreciate you taking the time to backtest these particular symbols, but I just suspect there are too many variables, even in what I've laid out as a pretty mechanical system, to make accurate back-testing possible without a LOT of work put into setting up the model. I'm excited though that your backtesting showed a net profit on the "normal" stocks. If you could survive a March 2020-type event then it looks pretty promising on those kinds of symbols Which brings me to my biggest concern with backtesting: none of us here likely trades in a vacuum, completely oblivious to market conditions. (Strict buy-and-holders excepted.) It's not too hard to see Feb/March 2020-type events as they're happening and move to the sidelines (I try to use {at least in theory} a 10% S&P decline as a signal to move to cash, and a 10% increase as confirmation of a reversal.) The S&P's highest close was on Feb 19 at 3386. Feb 27 it closed >10% below that. That's 8 calendar days, 6 trading days. Every single day since 2/19 the market was down, so we'd have been hearing & reading that in the financial media. Not too hard to look at the S&P's chart on 2/27 and think, "I should maybe get out for a while or at least scale back my trading until this thing sorts itself out." But even if you didn't do that, starting about 2/21 you'd probably start seeing red in this account as the downward market pressure pushed your stocks down into and past their Put strikes. That alone would've made you look at the broader market and reevaluate if this was a good time to be trading. (Or maybe skew your trades toward the Put side, an idea I've been thinking about for those kinds of market conditions but hadn't verbalized yet.) So anyway, whether we "believe" in backtesting or not, there's probably no better measure of a strategy's likelihood of success than trading it through all kinds of market conditions, which is why I keep doing this. When/if I'm approved for Tier 3 options at TDA in mid-to-late August I'll do it with real money, real fills, real commissions, and real BPEs, and we'll see what happens. Thanks for chiming in on my Journal! Mike
So you're saying no alpha can exist otherwise someone would instantly harvest it and then it would be gone? Sounds like the story about the EMH proponent who claims there's no way a $20 bill could be sitting on the ground because someone would've picked it up already. There are a myriad of factors that can influence the direction of delta1 in the market. Momentum and mean-reversion are both rather simple and they have statistically significant results in certain timeframes. So utilizing those can add a tailwind to any existing vol strategies or hedging systems you are running. Not to mention more esoteric methods or rare disruptions where deltas are dislocated from fair value. Also naive to say you should always be trading the underlying if you have a delta thesis. Look at Dustin's journal. The guy is paying vol edge but making higher return on capital than trading spot because of the leverage. There are scenarios close to binary outcomes where a stock is likely going to be worth around $20 or $80 and it's trading at $60. R:R is going to be way more efficient in certain cases structuring in vol. --- One last thing since I am winding down on this thread. When you guys find vol worth trading you should be in local strikes. More often than not I am the one buying junk wings "theta traders" are selling and spreading them against stuff that actually has vega/gamma. Here's daily PnL on a CMG position I've been working on this week. Imagine trying to collect $8K off 5-10 delta strangles. Even if you got them to drop $2 overnight on a stock like this, that's 40 short strangles. Let's even be more generous and say your strikes are 5% otm and you still managed to net that same $2. Your jump risk on a freak 15% news/dilution event is going to be ~$1800 * (.15 jump - .05 strike distance) * 40 lots = $720K. There are better ways to go about trying to make money.
If S&P moved up to 10% down, BBBY will down -20%, and this is x15 REG-T loss. If you use 50% or 100% Reg-T margin for these straddles you wiped out after S&P moved down by 5%, one day, no time for any adjustments, margin call and finish You can survive if used only 10-15% from REG-T margin, but profit on capital invested will be 5-10% per year. Better sell straddles before earnings day, and buy back next day, this is more promising strategy.
Thanks again for the challenges. Before I respond to your individual points I'd like to make a statement to all that might help clarify where I'm coming from in general: Part of my psyche is that I find it difficult to just blindly accept "conventional wisdom." For good or bad, I sometimes have to prove things to myself. (Having said that though, I'm a firm believer in the maxim, "Learn from the mistakes of others so you don't have to repeat them yourself." But so far no one's told me they've tried 10-delta, or any-delta, short strangles to this extent.) And I think especially in the financial arena we're often told as gospel that certain things won't work for certain reasons, and I'm just not convinced that those things are in fact holy scriptures. The most common one is that you can't beat simply buying and holding the S&P index. But you can, and lots of people are; you just have to be willing to put in the work to do it. I'm not one to shy away from work if there's a tangible benefit, and I've been putting a TON of effort into this. I'll do the same when it's with real money and either prove (at least to myself) that it works, or I'll stop because I've seen that it doesn't work. I'm in no way saying or implying that I'm smarter than any other traders, but I've observed something that seems to work and I'll keep at it until I'm convinced or not. "Break down this trade into short a call or short the put. Are you expected to make money over long term shorting either? If so why?" I'm guessing "the answer" is short puts because stocks tend to always rise? But if short calls couldn't make money over the long term, would the Call Credit Spread (Bear Call Spread) be a strategy that anyone would ever do? And just so I'm clear on this: I've just about given up on my ability to predict a stock's direction. I've seen it over and over where I do a bunch of research and analysis, pick a direction, place a trade, and the stock immediately goes the other way. So I LOVE the non-directionality of short strangles. And the fact that the BP for one side of the trade pretty much covers the other side just means you get to collect twice the premium on the same margin/collateral. "Why 10Delta? Can you do better with other strikes? Are 10 delta strikes really overpriced? If so why aren't there more sellers in the market driving the prices of these down till the expected returns is 0." That's part of what I'm still trying to figure out by using 20∆ for this challenge on ET. 10∆ just "felt right" to me when I sort of discovered it, that 9:1 odds felt to me like a good place to be given the returns offered. But after trading it to a doubling and then doing the same at 20∆, I'm starting to think that 20∆ might be "better." (Granted, both were done in a TDA PM account that may have issues around fills and collateral, but at least it was an apples-to-apples comparison.) And I can't prove it, but I don't think there's much chance of "the edge being traded away" because 10∆ is just a snapshot in time. A day later those strikes are at 5 or 15∆ or whatever. And option premiums need to sort of maintain their Black-Scholes implied values, so it would be hard for one strike to get crazy out of whack with its neighbors. If it did you'd see that in the option chain and go either up or down. I suppose that if everybody on WSB decided to sell 10∆ strangles on GME at 3PM on Fridays you might see some price depression at those strikes, but they couldn't go below their further-OTM neighbors because of the arbitrage opportunities that would create. In short, I don't think 10∆ or 20∆ or 33.33∆ gives me an edge, it's just a comfortable place (for me) to trade from. And if it offers even a third of the returns I've been seeing, then it's a lucrative strategy also. "Why is the margin requirements so high? Or is it to be expected because black swans occur more frequently?" Sorry, not following you here. According to feedback, the margin requirements I've been seeing in TDA PM are much LOWER than they should be, maybe on the order of 3x. Are you saying that they're higher at 10∆ than, say, ATM or somewhere else? I've read through how TDA calculates margin requirements for a given trade, and while I don't remember a lot of it, isn't it very formulaic? Based on stock price, IV, strike price, and premium collected? If you did a Put Credit Spread with the short put at 10∆ and the long/protective put a strike below it, does the long put somehow affect the margin requirement for the short put? I don't think it does, except of course that the overall trade has a much lower Max Loss. "Try to answer all these questions to understand your strategy better." I think I've been doing this, and will continue to, but let me bounce a silly analogy off you that just popped into my head: Say I have a favorite fishing hole, and I start to notice that I tend to catch more fish there the day after a rain, but less if it didn't rain the day before. Do I need to understand WHY the fish bite at my hook more one day over the other, or can I simply apply that observation by only fishing the day after a rain, and therefore make my probability of catching fish on each outing higher? That's where I'm at with fixed-delta short strangles. I'm not trying to understand WHY they work (or worse, why they SHOULDN'T or CAN'T work), I'm simply taking advantage of my observations that they DO work. If I were the fisherman, maybe next I'd try fishing only before noon on the day after a rain, or only after 4PM, and see if one or the other led to higher probabilities. That's where I am with 10∆ and 20∆, and so far 20∆ is winning. To close this reply, let me state for the record: I'm less interested in the theory/gospel/conventional wisdom of WHY 'x'-delta short strangles can or can't work, and more interested in testing my OBSERVATIONS that they do work, and then eventually tweaking the strategy to maybe work even better.
I'm going to stop trading and posting about this strategy for now until I can do it with real money, but I WILL be back! Might as well go ahead and close out this first week though. The 20∆ short strangles account (TDA paper-money) on high IV stocks returned 12% in 1 full week, with no drawdowns to speak of and no blown-up positions: Starting at the bottom of the Filled Orders screen, I was able to buy back the BBBY short STRADDLE early in the day for 0.50 against its total credit of 0.95, so after commissions I was able to retain about 45% of its max theoretical profit. I immediately placed another short strangle on that. BLNK continued to drop in price and the trade had a "P/L Open" of about -$1130 when I decided to roll it out to next week as a 30P/32C for another 0.28 in credit. I had taken in a total of 1.19 from the initial strangle plus the roll to a short straddle, which I think means I could've exited the trade then as a scratch (isn't that how that works? Took in $1190, showing a loss of 1130?) The PLUG trade stayed positive all day, as high as $184 I saw at one point, so I could've closed it then for a decent profit. But that didn't meet my capricious metric of 1%/day [with ~5k collateral on this trade I needed $250 out of it]. So I rolled out to a next-week 26.5P/28.5C. Total credit now 1.41. Thanks to everyone who chimed in on this thread, it's given me a lot to think about!
i believe i will. Need to check earnings though and the date on that. I'd like to be out of it before earnings.
For sure. Take advantage of the juiced premiums due to earnings coming up, but don't hold it THROUGH earnings. Let me/us know how it goes if you do.