Quote from bespoke:
Sigh....
I've seen a lot of your posts and let me tell you what your problem is. You have no real trading experience. You're trying to fit a strategy onto data. You will probably never get real forward results this way.
I started out just like you. I tested every indicator known to man on years of data for hundreds of stocks. Moving averages, CCI, stochastics, renko, kenko, kirilenko charts and the hundreds of other BS crap they use to try to sell you books and backtesting software. So many times I thought I had something but it actually got me nowhere. I tested shit from the moment I woke til I closed my eyes for months.
I did so much testing that my harddrive eventually died from the constant reading and writing of files (I had no backup I was poor). It ended up being a blessing in disguise because I was only able to recover some of the data and it was so disheartening to lose all that work that I said fuck it.
I sat in front of the screen pissed off and just watched the market and traded. You eventually see patterns that reoccur every day and you say to yourself "after this happens, this usually happens". You take that idea and you test it. Do this over and over until you find something that puts money in your pocket. And as you gain experience this process becomes easier and easier from the initial spotting phase to the testing phase.
I haven't been trading long, just 3 years. Spent the first few months doing what you did. Then my eyes opened and I haven't had a losing month since Feb or March 2008. I've probably had about a dozen edges over that time, of which only 5 or 6 still work (I'm sure the others could still work given the right market conditions). NONE of which use any indicators, well, besides an ATR based exit/stop system for one strategy. All of these ideas could be traded by hand without the use of any math or calculations that couldn't be done in your head, but the point of automation is to improve execution, discipline and monitor thousands of stocks for more opportunities. I'm sad to say that I'm actually negative for all the trades I've entered in manually over the entire period since I've been profitable, ha.
I guess the whole point of what I'm saying is you need screen time. Ya. More watching the market, less backtesting.
Another thing. You speak of HFT but I really don't think you know what that is. 50% of my orders probably last less than 5 seconds, 10% less than 1 second, and 5% are at 150 ms (that seems to be the fastest I can get a signal, send an order, receive confirmation, send out another order to exit, get filled for profit). And even though that may seem fast for you, I don't consider that HFT at all. You have a thread about moving averages... that's not HFT...
I don't know why I wrote so much. It will never happen again.
BTW I'm not saying that strategies that use complex math don't work. It's just that I probably can't compete with them so I'll just stick to my niche. All my strategies could each be explained in one sentence.
