When I first started backtesting, I set out to improve upon a strategy I was already trading manually.
(Spoiler alert – I was able to improve it immediately.)
That worked well, but I quickly realized how inefficient that process was.
I was essentially backtesting trying to find a particular tree in a large forest.
I eventually found the tree I was looking for, but it dawned on me that I was spending a lot of effort ignoring equally good and even better trees throughout the forest.
I realized that I needed to change my process to create profitable strategies from a backtesting first approach.
After a lot of trial and error, I ended up with an efficient process I call: Backtest Wide but Trade Narrow.
This is the opposite of what most traders do when they backtest.
When I have an idea for a strategy, I create a backtest that finds every instance of the signal with very little filtering.
I end up with a backtest with sometimes tens of thousands of trades in it.
When you try to apply rules to your strategy at this point in the backtest, it’s cumbersome. You make uneducated guesses. Stabs in the dark.
The smart way to apply rules is to take your trades list and use Excel or some statistical software to determine the rules.
So my backtest becomes very easy to generate.
The valuable part of my approach is determining what subset of the tens of thousands of trades I’ll end up trading.
If you’re familiar with SQL, imagine querying a large database table without an index applied.
Your query will eventually return results, but it could take hours or even days to return. Heck if the table is large enough it might NEVER return – your query might be effectively impossible to run.
But, of course, if you apply the right index to the database table your query will run INSTANTLY.
It’s the same with my Backtest Wide but Trade Narrow approach to creating trading strategies.
To learn more, sign up for my newsletter below where I teach the process of creating your own profitable trading strategies. 👇
