Thanks for your response. I parsed it to be specific in answering your questions. Thanks again for your replies, etc..
"Why systems fell apart?" so you took 10 years after thinking that you are smart enough to answer above question to actually answer it. Wowwww!! I think it is one of the most overlooked issue in strategy development and trading it. Would you mind dropping some hints as to what are the underlying reasons you figured that caused "well developed systems (not curve-fitted, highly optimized systems)" to fail after performing as expected with live trading for some time. Maybe it will save few of us a decade of time to figure it out Thanks in advance if you have anything thoughtful to offer.
To the original post: First of all, the line between trader and programmer in HFT is blurred. Nonetheless, it would be very difficult to come up with an accurate number for average income of a high frequency trader, but I would hazard to say that most successful ones make north of $300k/year in taxable income. This, of course assumes that the trader is in a private firm, HF or IB. In my experience, anywhere from 10-30% of net profits is what the actual trader / trade developer gets paid. If you are self-funded, obviously you keep it all. Also, HFT can in fact be done from home - you simply remote into a colo-ed server. No different than being in an office. At my old firm we could run most strategies from a laptop during snowstorms and around holidays. These were ULLDMA strategies, so the user interface was minimalist.
I probably didn't make myself clear. It took me 10 years to completely solve the problem of systems deteriorating over time. I solved the problem that curve fitting and over optimized methods create. I think I'm average for those that put forth the effort. Most don't solve it because they don't want to apply the time and effort. I know others that have done it quicker and others that are still working on it even after 10 years. And no, I won't give you any hints sorry. My sacrifice has a price and most in here can't afford it.
Thanks for your reply. and no worries about not giving any hints, I am smart enough and i will discover all the pieces of the puzzle myself
i would like to share on hint well a couple. 1.) when you find a system that works make sure that it has at least a variety of profitable settings that will all work. example settings 1-20 all make money then instead of taking he most profitable setting use the middle like setting 10. my idea is that you give up short term profitability for longer term dependability. 2.) when a system exceeds it's worst draw-down period just hand on because it will usually deliver about half your loss back. then decide to stop trading it of change settings. but don't just bail on it at historical low equity, that's called being a parachute boy. m
I learned, on this site, not to give out any information. Your hints are tame so I guess they are reasonably safe. (The key word there is reasonably, watch.) 1.) I test my individual trade settings based on a win rate over a minimum period of 3 years, of intraday charts, where that win rate is greater that 55% and where the winners are approximately twice that of the loosing trades. I've found these settings to be just as accurate on longer term charts. 2.) I've found my worst drawn down has never exceeded the margin for one contract for whatever contract I am trading so stopping trading it has never been as issue. I guess it would be like needing a parachute running along the Jersey Shore. As long as gravity holds up I'm pretty safe.
here here we are of kindred blood that agree 55 and i mostly like the pf to be 1.25 or greater but when i see it to big i worry it's just mush. i am happier with less performance and on close orders. but i don't trade those types systems anymore. im doing something different now other than contemporary timing models. it's funny while some are still doubting it can be done at all the industries few greats are evolving beyond it and retiring the seemingly impossible to others. lol dino-SORE m
I just wish I would have figured this out 20 years ago. Now my kids will have all the fun. (if they don't screw it up). I relate to the dino-sore.