Very impressive. I admire your perseverance. I know you have been working at it for a long time. Can you forward testing it to verify? I think a pension fund or college endowment fund would love it. Good luck.
I was a quant for a while, and was involved with 100s of thousands of backtests. Speaking as someone who really knows backtesting, I can tell you their predictive value is almost zero. Now there is also post discovery. How has the strategy done after it was discovered? That is a little more predictive of the strategy working, but even then, something that worked post discovery can just stop working. The backtester I worked with was very sophisticated. It would not just buy something on the first of the year, and hold for a certain amount of time. It would simulate buying everyday of the year, and take an average of all the days.
Thanks @ironchef , I'm also impressed that you remember (me working at it a long time). I've quit trading more times than Mark Twain has quit smoking (to quote him: "Giving up smoking is easy... I've done it hundreds of times"), staying away from it afterwards is hard. Eventually from all the failures, I gathered techniques. This last strategy uses several techniques from the past which individually or in different context didn't work but in the right combination do the trick. I'll test it extensively both on historical market data and live, at the moment I'm just a bit exhausted and taking a break.
Yeah well @shMark, I dunno about hundreds of thousands of backtests. I ran a lot too, I worked as a quant as well and ran a bazillion for others as well. A backtest is only as good as the strategy behind it, if it's hopelessly retarded (like technical analysis mystical voodoo star alignment) then you can run a trillion backtests and still won't get anything. On my backtests, the vast majority of them failed, proving that the market model and strategy behind them is not working. Like the good ol' Black-Scholes "guess the volatility" and delta-hedging afterwards. Works wonderful in theory if you can add obscene amounts of spread to those options you're selling (or subtract if you're buying). But that ship has left the port a long time ago in real life, only way to make money now is be a market maker and add that spread by means of selling it many times over with opposing positions which cancel each other until you're left with one that needs to be hedged and approximately make a bit of money or lose a bit of money, doesn't matter because the spread accumulated so far covers it. What I wanna say is that the models and strategies I use are Black-Scholes like (or quantitative finance type) so if shit is going to work, it's got a solid mathematical reason behind it. It's still looking for the needle in the haystack but at least it's not requiring mystical voodoo star alignment when it works.
Thought about it, took me 15 years to get to a trading system that significantly, consistently and reliably beats the market. I don't wanna spend the next 15 years marketing it so let's speed it up. Reckoned it doesn't need that much money to be traded individually, about $100k would suffice. Which makes the system usable by regular net-worth individuals. I intentionally left high because $100k is not high net worth. It's just net worth that most people worth the label of net worth have already or will have. Therefore having reduced the target from multi billion funds to regular individuals, how much would I charge for the system. Which, I repeat, includes the software to trade the system automatically and a PDF/plaintext description of the algorithm so you have tens of thousands to spend better/custom software, be my guest. Previously I've listed $500 as an anchor point just to get me motivated but I just raised that 10x. So $5000 for a system with no limits on capital, sales and whatever you wanna do with it and also very importantly, a system that works and beats the market by a 2x-4x margin. Includes $500 worth of support in the first year, afterwards if you still have problems or questions with your software, it costs you $500 to talk to a knowledgeable human rather than the much uglier alternatives.
Another thing I recokend is that I need a partener. I can handle development as I did for 15 years and can handle marketing and sales but I'm fed up to a level that can't be described in words of doing all this myself. Several questions arise. 1) What's in for the partner? Depending on how much we make, 10% if selling the whole business for $500k, 25% if selling for $5M, equal partnership for $50M or more. 2) What do I expect from him initially? Apart from general help with the work and just slapping me when 4 weeks pass with no progress whatsoever, a ca$sh investment inversely proportional with the reward. That is, if you're getting 10% and I did 90% of the work then you multiply my pay by 1/10%, which is 10x. I'm putting $500 / month into this so you do the calculus. Whatever percent you get, you pay inversely proportional. Since I'm on this site since like 10+ years, I'm using it preferentially in my IPO. As I stated in some post which I don't care to find now, I don't care about naysayers as the like of those are in the billions. I care about guys who are interested, drop me a private message if you wanna take this further. If this doesn't work, in a few days I'll press the atomic button and post on Hacker News, backed by the $85B venture capital fund Sequoia Capital and put my ass on the fire: https://news.ycombinator.com/news I'm done fucking around, it's time for serious business.
But, as usual, it's another one-trick pony: options. And I reckon, it's a delta neutral strategy. So it wouldn't even matter if the market is going up or down. Your cash cow will be cranking out money regardless. To that I say, go figure.