@guru, fascinating thread and a great read. The Holy Grail means different things to different people. Would you consider yourself to have found it, if it meets the following criteria? 1) You can make 50%+ profit a year, year after year, regardless of market conditions. 2) The system can withstand 2008/09, the Feb 2018, Dec 2018, and Mar 2020 crashes. 3) The strategy is scalable (say, up to a $5million account or more). 4) You do not have to be glued to the screen 7-8 hours a day. Happy trading.
Sure, I think those are great criteria for anyone wanting to determine whether they have the Holy Grail, so this is the stuff I look at as well. Regarding #2, there are still such issues as overfitting that allow to create strategies to work great during past market crashes while not necessarily surviving future ones. This is the type of stuff I’m still working on. Re #4, a trading system can become quite large over time and therefore requiring being maintained. Many top trading companies have large teams, so I’m not sure yet how far I can push mine on my own, even though I like to push things to the extreme. Additionally, sometimes the margin may be more of a problem than the options strategy itself (like on super-wide ratio spreads that I liked to use) while margin use is very difficult to model and backtest, so I’m trying to move away from those. Interestingly, and I may write more about this later, if you trade/know options, I’ve noticed that mixing up 1/-2 ratio spreads with -1/2 back ratio spreads (to cover risk holes) can result in 1/-3/2 butterflies that @dest used to utilize. And I believe he does have great alpha, so I also hope I’m on the right track, even though I’m not specifically aiming to trade butterflies, but noticed that they can be the result of other trades. Anyway, backtesting this stuff and staying within margin limits is what gives me headaches for now.
Interesting. I trade for a living and it's almost exclusively options. I don't claim to have quant level knowledge, nor am I aiming to learn that, but out of the four items I listed above, my strategies can cope with the first two. It's the scalability and the time/focus that is proving a challenge, and I am exploring different pathways for my own version of the Holy Grail. Anyway, I'm thoroughly enjoying this thread, and love the discussion and the detail that you are sharing, so thanks for that.
So here is how the Holy Grail may look like. Or at least one variant that my system recommended for today for SPY, with SPY at ~$336: (screenshots from TDA/ToS) If IV drops by -3%: If IV goes up by +3%: The P&L/risk graph shows that the chance of loss is quite small when compared to possible gain, with potential for highest gains when volatility goes up. Largest risk of loss is when volatility drops, but not necessarily because there are some skews at play here that may help on the upside beyond what the standard/above chart shows. While we've also seen IV going up recently even when the market goes up. My only problem is filling all the legs to enter such positions. There are 7 legs here across different expirys, with 1-4 qty each. This is still experimental work and I'm observing how all the pieces of the puzzle work together. This month is negative for me, but I'm also live-testing lots of setups, with some nice winners and some ugly losers. I haven't looked at my P&L recently (makes me lazy & complacent when too positive, while stressed when negative), just focusing on improving and defending each position, while considering any losers as the cost of R&D
This month is looking good, partially because the market is up which accelerates profits. I'd just make money slower in a slow but stable market. When reviewing my approach, the months like last September may be worst for me due to slowly downtrending market while VIX also stretched and not providing hedging ability, which luckily is rare. But I'm mostly happy and maybe surprised with one risk-free arbitrage trade I made last week that instantly produced $450. Not much, but I'd have bragging rights if I published the details, which I'm rather not going to. My system is now so sensitive to spotting options mispricing and figuring out how to combine them together, that it shows me such trades every day, just none of them will usually go through. Last week I've decided to try one that looked really good due to very wide bid/ask spreads and the combo not being too simple, and it did go through. But I had to buy/sell 200 contracts just to make the $450 ($520 - $70 commission), likely arbitraging between two different MMs, which is not scalable. And 2 seconds later MMs adjusted their prices, so I couldn't continue adding to those trades. I'm just thinking to myself, what are all those tens of thousands of puzzle-solving quants doing where they cannot spot basic arbitrage trades? They develop super-advanced algos trying to trick each other, yet they don't see low-hanging fruit. And they don't see it because they're looking at their big screens full of 1st and 2nd order Greeks
Doesn't this show that the market is efficient. For a capable programmer who builds such a system, he should make around 200K to 300K in the US being employed by mega corp.
I’m not sure how those things are related? Many programmers build advanced trading systems and make $200K+ working for mega corps, possibly/sometimes because the market is inefficient. The inefficiencies I’m finding are tiny and I’m not focusing on those, but many trading firms are in business due to various market inefficiencies. I’m focusing on larger inefficiencies in options, just picking up smaller ones along the way. Many algos supposedly fight for every cent in effort to make the market fully efficient, I just found some stuff they miss, but I wouldn’t be able to beat them in tasks they’re designed for. While I wouldn’t even pass 1st interview at any mega corp because I wouldn’t be able to solve their interview puzzles. Though I’ve had a quant and a math professor help me with some tasks, so the roles can be reversed.
Recently I'm spending most of the time doing option analysis and math, and putting on small number of trades to test different market conditions and see what "kicks in" and when. For now looks like my system performs best in bullish market with low volatility, which allows for buying cheaper hedges that work when volatility increases. I have different trades for high vol as well, but looks like not as reliable due to difficulty of predicting which way volatility will go. Math-wise, I did some "silly" work trying to show that certain types of options arbitrage may be possible. Still not sure, but the proof is in the pudding and I had 200 totally free SPX hedges that kicked in just when the market was dropping during last couple days. Couple weeks ago I got 200 SPX combos for -$0.01 credit each (after commissions), which now turned into $0.20+ value each. The unrealized profit was $5K yesterday and is still over $4K as the market started recovering: The margin use on these is fairly balanced as well, no risk there. One issue is that I don't know when to get out. The hedge/profit can grow exponentially if the market continues tanking, but I'd lose that profit when the market stabilizes. Another issue is that I cannot buy these too often, only during bullish market, when volatility and OTM prices are just right, and when my entry order(s) get fills, kind of like making market and waiting for other buyers/sellers. Otherwise I could just pay $0.05 for them. If exiting now, the price is marketable and liquidity sufficient to get out without problems.. For now it's just nice to see when free hedges work.
Grail quest was achieved by Sir Bors, Sir Percival, and Sir Galahad, and... https://www.arthurian-legend.com/sangreal-holy-grail/