Today I want to talk a bit about trading system complexity and trading system filters. I also want to show you a filter that can improve just about any day trading system.
Riveting stuff I tells ya!
Filters are the easiest method of cleaning up marginal edges. For example a common filter for stock systems is to only consider trading stocks above their 200 SMA (this improves just about every stock system in the known universe)
Just don’t overuse them.
Here’s the thing. The more complicated your system, the less freedom you have to add extra rules.
Think of trading systems like Jenga. Extra rules make them inherently unstable
What happens when you add too many rules and conditions to a trading system?
Your results start looking AMAZING at certain times, and then terrible at others. They start working on some markets and not others… they get a real Jekyll and Hyde feel to the results. And they tend to take far fewer trades, so you have less data to work with and therefore less confidence.
This is not a good thing. Systems like this are fragile, and carry what is known in the business as “tail risk”. Certain option writing systems are known for tail risk, making years of consistent profits and then wiping out in a week.
Tail risk = bad. Got that? Good. So what’s the solution?
The solution is to start with a lean, mean, pared down ruleset. In fact you should spend a lot of time trying to isolate EXACTLY the market behaviour you are trying to capture, and throwing everything else away.
My friend Max improved one of my own entry techniques by SIMPLIFYING my own system. Removing a rule or two, gave him the freedom to add other rules where he needed to.
And his system outperforms my original. After looking at it I have to concede it wasn’t just dumb luck, it is legitimately better.
Complexity is laziness, and simplicity is the ultimate sophistication.
Consider the following trend following futures system which has compounded at 12.5% CAGR for the last 40 years.
50 of the most liquid futures contracts, ES, CL, NG, GC, ZB, ZS, etc
Any instrument trading at a higher price than 1 year ago, BUY.
Any instrument trading at a lower price than 1 year ago, SHORT
Exit rules: When price closes below 4 times the Average True Range (100 period) exit the position.
Position Sizing: Volatility weighted position sizing
I shit you not – that is the whole system. And it does nearly as well as the actual institutional ones I show in the System Building Masterclass that hundreds of billions of dollars are traded with.
Anyway… back to this filter.
This is ticker ADD, aka the Advance Decline (in the bottom panel of the chart), and it is the number of stocks going up in price less the number of stocks going down in price.
Let me show you how to use it. Check out my Price Action Masterclass here
The idea is that if you are day trading you want to trade long on the days that trading long has an edge.
So we use a simple filter to not trade any day with a bearish divergence.
You can see the bearish divergence last Friday in the S&P500 which happened right at the start of two days of weakness. Not a good time to trade the long side.
The rule is simple: bearish divergence, and no long trades that day.
As long as your system isn’t too complicated already, this is worth testing on your systems.
Enjoy your day