Reflections On Backtesting and Tuning Low Frequency Strategies
long, long, long thoughts on backtesting in low frequency space.
Welcome back friends.
Unfortunately, I'm in the business of trying to make 2bn with 2 Sharpe instead of 15m with 15 Sharpe. And most(?) of the time that implies working with lower frequency strategies (for us it's daily and little bit of hourly).
đź’ˇ
A big part of my job is:
> PM has as an idea
> I backtest the ideas and present the findings
> rinse and repeat
> PM has as an idea
> I backtest the ideas and present the findings
> rinse and repeat
My job title might as well be “Chief Jupyter Notebook” or “Backtest Warrior”. So yeah I would like to think of myself of someone that's slightly above the amateur level when it comes to backtesting. This article is purely just gonna be a rant and and thoughts on how I approach tuning/tweaking strategies, and how I view backtesting especially in the low frequency space. Let's get it.
This post is for paying subscribers only
Already have an account? Sign in.