OP, you do not understand software at all. No one buys a software program, you buy the developer, put them to work building the next version. I.e. you buy the talent not the product. That is so basic EVERYONE who writes code professionally knows that. Second if you do indeed have something novel and useful, the concept is easily reproduced. Hence the first point. Lastly, the backtest you present just smell of old toilet water. It is as if you cherry picked not only a time frame but also just wanted to show a pretty curve. All the people doing real work have trash cans full of stuff like that. We do it as to show what NOT to do and as comic relief to take a break. You don't show how you execute orders, manage risk, manage market disorder, slippage, commissions, drawdowns, volatility, dead markets, how it was developed, how it could be implemented live, and so much more. On the bright side if someone is dumb enough to work with you, then you do have a sucker on the hook, and you can try to reel them in, and sell your soul in the process. There are several people on ET with flat learning curves. They would make ideal partners. Together you can Seriously, just go back and keep on working. Do all the things that would make it useful, and maybe in a few years you will have something. Stop trying to promote concepts and dreams.
In short, I got a strategy that does great in backtests. You can take a glance below, goes from 10k to approximately 200M 90% CAGR What assets are you trading? On what Leverage? What platform or language do you use? Do you reinvest profits? How many trades?
Good Morning Aquarians, There is ONLY one answer for you: 1. Save up the capital you need to run this automated system live for 1-5 years, then and ONLY then try to sale the automated trading system to hedge fund or just simply make your millions with the system. There is no other answer for you.
Good Morning El OchoCinco, I agree. The software that creates the back test can be manipulated to show false back test reports. In real time it fails. Even a fairly decent back test, may fail in real time the next 1-3 years. Market conditions change. Automated trading systems work when there is +20 uncorrelated trading systems in a portfolio across different instruments. +20 uncorrelated trading systems in a portfolio = about +$500,000 initial investment. Cheaper and easier to just manual trade everyday.
Wanted 10 years. Catch a period of market recession. And leave some significant out of sample period.
How about having more than one strategy? If you only got one, gotta guard it with your life. If you have more, you can gamble. Like any gamble, be prepared to lose it, there's no other way to win otherwise. So sign the NDA and agree on an upfront payment + fee from trading it. Then disclose the strategy. If they are so enthusiastic about it, they will pay (should be a dip in the ocean for them anyway). If they back out, threaten to disclose it to OTHER funds in the beginning, general public in the end. Then, nobody makes money. So if they plan to steal it from you and silently use it themselves, that pig won't fly for long. Either they pay you and both keep the secret and make money or nobody makes money. Of course, you gotta have more than one and be prepared to gamble.
- What is the expectancy? - About 70% overall average. I double checked that and it's still there. Will triple check soon. - What are draw downs? - Didn't calculate. Depends how you use that 70% overall and how uniform it is distributed, will check eventually. - What instrument? - Not a single instrument but watching about 10,000 of them for the right conditions. - How long is forward test and how much does it deviate from backtest? - So far I tried it live for almost a month. It's ... up and down and then up. Can happen, even with an uniform 70% probability. Need more data, in the following months. - Was backtest automated ? - Yes, fully automated.