Hi All -
Some of you know that I am a professional money manager for a small firm, focused on retirement accounts. We've been operating in a very qualitative, fundamental analysis fashion and I'm trying to take the firm into the 21st century by introducing some algorithmic trading to our client accounts.
I had been working on a backtester/optimizer similar to Quantopian for algorithmic trading for several months (prior to my employment, even) when I was turned on to Quantopian and realized that alot of my work had already been done for me. I coded one of the algorithms I was working on and found substantial excess return from SPY in Quantopian but had trouble reproducing the backtest in my in-house backtester.
The algorithm is not a typical high-frequency trade algorithm but rather a quarterly sector rotation. I used dividend adjusted data to calculate a sector ranking, which I uploaded into Quantopian using Fetcher. I buy and hold the two sectors with the highest relative strength and then rotate after 63 trading days.
Here are the results....