If 80% is computer driven then there are a lot of computer driven models losing money, even if the 20% retail is losing 100% of the time it isn't big enough to make all the computer models winners.
Yup. Luck plays a role. Here's an interesting related talk: http://www.ted.com/talks/alain_de_botton_a_kinder_gentler_philosophy_of_success.html It's a good talk for losers to read, anyway. This guy says that doing stuff and telling people you're doing stuff is what increases your 'Luck Surface Area' - a term I like though I don't agree with the exact mechanics of how he describes it: http://www.codusoperandi.com/posts/increasing-your-luck-surface-area
I'd like to bump the thread with a couple of nitty-gritty questions: - What technology/software do you use to scan for stocks? - If different, what technology/software do you use to submit/manage orders? - Is there a backtesting platform that works well to test a strategy that scans the market and trades a whole basket of stocks? "Take GOOG off the platform" means to me that many of these successful strategies are not about staring at an individual stock (or even a few stocks) and trying to find a trade - rather it's about casting a wide net, finding 20-30 stocks that are doing something interesting and trading them with similar rules.
- Trade Ideas - who is your broker? use whatever they give you... - Amibroker (best retail one I know, you need IB)
Hi EricP , you are one of the most decent traders here and I was inspired by you in the old days of 2006, 2007 I want to discuss with you your assumption that any trader making more than 2 cents per share would be a rock star In my humble opinion it greatly depends on the stocks he is trading and the volume he is trading, for example let's say he is trading a stock around $20 he can easily (relatively speaking) make 3 cents per share if his daily volume is 200 shares but if his daily volume is 2 million shares (in the same stock) then no way he can make that and actually if he can make 1 cent per share he is really a rock star Please share with us your thoughts about that
Certainly, the price of the stock makes a big difference. It is much easier to average 3 cents per share while trading a $600 stock, rather than a $10 stock. For what it's worth, historically I think the average stock price of the stocks that I have traded have been around $40 (in case that puts my thoughts into better perspective). Also, for what it's worth, I'm referring to 3 cents <b>per share traded</b>, so 6 cents per round turn trade. For example, you buy 200 shares at $20, and sell those 200 shares at $20.06 => You have made a gross profit of $12.00 and have traded a total of 400 shares, or 3 cents per share traded. Obviously, commissions and various fees (SEC, ECN, etc) will lower this a bit further. The time duration of a trade certainly comes into play, as well. The longer the trade duration, the larger the average profit of a good trading method could be expected to be. For the numbers I quoted, I was assuming the person was a daytrader. Now, I will still contend that anyone averaging 3 cents per share or more (6 cents between buy and sell price) in their daytrading is a rock star in my book. I don't care if that trader is doing 200 shares per day, or 2 million shares per day, that is a very good per share average profit. Certainly, the guy doing 200 shares per day will have days when he/she has profits of 50 cents per share or more, but will also have days when his average trade is a loss of 50 cents or more per share. What matters is the average over a decent enough number of trades to be meaningful. If a guy can do those sort of numbers manually, then he ought to find a way to automate his trading to allow more trades to help diversify his trading and multiply his profits. Obviously, this would need to be within the context of good risk control. Any way you look at it, 1 to 3 cents per share traded are good number in my book.
EricP - Regarding automation: do you have any suggestions for back testing or scanning technologies that are appropriate?