Trading System Investigation: Series 1 - Part 4

A deep dive into a short term trading system found online. Is it worth trading?

So far in this series, we have spent a bunch of time coding, backtesting, and robustness testing the trading system under investigation. Which is great and provides a lot of learnings and interesting results; but what if this system doesn’t fit your trading personality or goals? Maybe the drawdown is too high, or the system trades too often, or you don’t like losing trades and the winning percentage is too low, or your broker has a more expensive commission structure and you require a higher expectancy per trade? Things like this should be considered when you are developing a trading system. The trading system we have been developing over the past few articles may be a perfect fit for some and a nightmare to trade for others. If the trading system doesn’t work for your situation then it’s going to be really difficult to trade.

Having a trading system with positive expectancy is very important, but if you cannot stick to the trading system consistently over time because it doesn’t match your personality, lifestyle, or situation then that system is no good for you. I don’t care if you have a system that makes 50% a year, if it requires you trade on 1 minute bars and you have a day job, then this system is worthless to you unless you can find a way to automate it or quit your job to trade it. While this is a more extreme example, the premise applies to all trading systems, including the system we have been working on in this series. If you can’t trade the system consistently, for whatever reason, then the system is worthless to you.

It’s is very important you tailor trading systems to your specific needs so you can stick to them for the long term. In this article, we are going to go over a couple scenarios/situations that you, the reader, may be experiencing. We are going to then look at a few methods of tailoring the system to your needs so you can trade this system, if you desire.

Also, it is very important, if you are going to modify a system it still needs to go through an In Sample and Out of Sample approach and robustness testing! If you “tweak” this system for whatever reason without going back to an In Sample and Out of Sample approach then it is very possible the system could become curve fit. For that reason, in this article I am just going to show the In Sample (2007 to 2018 on the S&P 500, consistent with the previous parts of this article series). Then, if you choose to work on this system further based on an idea in this article, the Out of Sample will still be available for you to use. As a reference to compare to, here is the baseline In Sample system performance: