Thursday, March 4, 2010

Starting Quant Trading

Quantitative trading is supposed to be for big hedge funds. I'm trying to prove it wrong. Starting with two systems, coming from James Altucher.

System 2.5x2, modified 3x2 system, described here (caution: quant portfolios on Stockpickr.com are not being updated for a long time, all trades there are wrong). I changed system to my liking: buy if stock closed down two days in a row and is down in the morning of the third day. Sell stock if it closed up and is up next day near closing.

System 3.5x2, modified 4x2 system, described here. My take: buy if stock closed down three days in a row and is down on the fourth morning. Sell the same way as in 2.5x2 system.

Of course, both systems frontrun originals. But I'm adding several more conditions.

I'm only trading stocks, which are in my database and backtested against both systems. If stock has better return for 2.5x2 system, I buy it on the third day, if 3.5x2 system is better, then I buy it on the fourth day. There is a program which runs search every morning and sends me candidates. Of these, I choose several stocks to buy. I'm allocating 15% of my portfolio to this system.

I ran paper trading for couple of weeks to make sure it might work. Now is the time for real test.

Today buys
2.5x2: Darling International Inc (DAR), Intel Corp (INTC).

No buys for 3.5x2.

Seeking Alpha Certified

No comments: