That's a lot of testing. I sometimes wonder if we enjoy the hunt more than catching the prey. Congratulations on your findings. Your sharing of your journey will encourage others.
Oops, thanks. I previously pasted images from a private messages exchanged with someone else, and didn't realized those images wouldn't paste properly with public access. I can't modify my previous post, so here is a copy of the relevant part of previous post, this time with images: ===== ... The performance for this account is up 50% during the last 2 months since I started trading these options. This is my options-only account, meaning I have other accounts for trading stocks and penny stocks, so this one doesn't show much activity until the last 2 months. Previously I only traded it a bit in March/April where I made some money from (nearly free) hedges, but later lost it betting on crude oil (selling $1-strike CL puts) that went negative, so I went back to working on my options trading program, which I then started utilizing less than couple months ago in July. Now I finally don't have to guess anything, just rely on my program. Here is IB's summary of my performance vs SPY since July 1st: ... =====
this is one of the best posts I have seen on elitetrader in recent years. congratulations on your success. why do you choose to focus on option instead of stock directly?
Thanks. I didn't find serious inefficiencies in stocks. Even when I combined 200 top stock strategies together, the backtested P&L looked very nice and going up, but once launched live I've only got flat-line results, barely breaking even. I also didn't find a way to limit losses, as having tight stop-losses resulted in being frequently stopped, while having wide stop-losses resulted in larger losses. In fact, my top stock strategies ended up with not having any stops, and the risk was managed by position size. For example I'd bet 0.5% of account on any one stock and didn't care if I've lost 100% of that position. This worked OK, but that's what produced flat-lined results, with no alpha. One thing I didn't manage to do with stocks, is to backtest small intraday intervals, less than 1-minute bars , like say 1-5 second bars with 2-level quotes, basically simulate tape-reading, where possibly some edges may exist. I needed to test history on 15000+ stocks over 14 years , and I barely could squeeze that data at 1-5 minute intervals in 500 GB RAM. Minute-based backtests could also take hours or days to complete, so I couldn't run too many of them or optimize them. Testing smaller bars/intervals would require a supercomputer. I do believe there may be ways to make money using modern portfolio theory, basically holding a basket of long/short stocks and rebalancing every so often. But this approach seems common and doesn't look like trading but more like investing, and I didn't think it would be worth spending more time on (I could be wrong). While in terms of options, I believe there are a few but serious flaws in how they're priced, or possibly flaws in their behavior, as well as how volatility behaves, as well as in volatility smiles and skews. Options still work better in combination with trading the underlyings, so that you can detail-adjust desired hedging, usually to stay delta-neutral but not necessarily. I do take risk and don't often trade delta-neutral because I don't use Delta (yet), but my system amazingly comes up with trades that when I look at the Delta then indeed it is near 0/neutral on some trades. The delta can change, so my system rather targets those future changes, not always the current delta. Finally and possibly most importantly, with options you can define risk without using stops. This means that you can hit theoretical stops many times but still stay in the trade, knowing that your risk is limited anyway and you won't lose more than the risk limit you've set. This simply isn't possible with stocks.
This is a very interesting point, so thanks for pointing this out. I can instantly see how this would be beneficial. On the flip side though, with options, having to get the timing right is critical. Suppose you figured in March that the market had bottomed and you wanted to start buying calls for the ride up. If you thought the recovery would be as fast as the drop, and hence bought 3400 ES calls or something similar for a month out, you would be right on direction, but not accurate enough with the timing. Hence, you wouldn't get the payout, whereas buying the lows outright in either ES or SPY would see you eventually hit that 3400 payday. Would you say that timing is another variable that complicates options, which makes it that much more fundamentally difficult than just having to get the direction right?
Definitely. I’ve only listed benefits of options, but could think of downsides as well. For the reasons you described, some of my option trades require that the stock moves in either direction within some time. Others require that it doesn’t move much. And yet others bet on the stock moving in specific direction. All this is combined with how volatility may behave during the time of being in each of those trades. Partial solution to the “time” problem is to trade long-term options like several months and up to 2 years out, as long as it can be cashed out sooner if profitable. Most basic example may be buying SPXU, SPXS, SDS, TECS, SOXS puts 1-2 years out. They can wait long time to become profitable, but in a bullish market you could collect profit quite quickly.
Mostly volatility bets, but many will make more money with specific direction and/or volatility targets. There is nothing without risk, so I may lose when both the direction and volatility don't go my way. I may also have periods of losses vs periods of profitability, just like if you'd buy a SPY put, it may suddenly gain value when the stock market drops, but if you wait too long then it may start losing value again.
have you thought about doing market making with delta hedging in option market since the option spread is huge?
Definitely thought of that. Though I think this market is crowded because Virtu earnings aren’t that great unless volatility picks up. In my case market making would mainly allow me to save money on my own trades, scale up, and hide my option combos. But I wouldn’t have time to work on this by myself either, so it’s something that may happen in the future if things go my way...