Why I don't just trade my best system

A member asked this week's Q&A question. Here's the honest answer, with the math.

Each Thursday I’m going to take one question from the Pro member inbox and write the answer up here on the public newsletter. The questions tend to be good, the answers might be useful to people who haven’t asked yet, and I think it makes the Stats Edge thinking visible to readers who are still deciding whether systematic trading is for them.

This week’s question, from a Pro member:

“Why run three systems instead of just trading the single best one?”

The short answer: because “the single best one” is only the best one looking backward. Looking forward, you do not know which of your systems is going to be the best one over the next twelve months, and the cost of betting on the wrong one is your account.

Let me show you what I mean.

Markets go through regimes. Some years are trend years. The strong stuff keeps getting stronger, breakouts work, momentum compounds. In a trend year, a breakout system or a momentum system carries you. Mean reversion underperforms.

Some years are mean-reversion years. The market chops sideways, names that look weak bounce, names that look strong fade. In a mean reversion year, the Starter System I just published carries you. Momentum underperforms.

And then some years are the worst version of the puzzle: nothing works. The market is choppy, regime filters keep switching on and off, every system you run is in a drawdown at the same time. Those years are rare, and they end.

If you knew in advance which year was coming, you would run one system. But you do not, and neither do I, and neither does anyone else.

So the way I do this is run several systems whose edges come from different things. Some buy strength on breakouts. Some buy weakness in uptrends. Some are swing systems on weekly bars. Some are day trading systems flat by the close. The point isn’t to have ten systems for the sake of ten systems. The point is that the edges are different enough that they don’t all break at the same time.

When System A is in a drawdown, System B is hopefully not. The gains from B pay for the pain in A. Across the year, the combined equity curve gets smoother than any single system would have been on its own.

Here’s what that looks like in numbers. The Starter System I published last week has a 26 percent worst drawdown across 26 years. The combined Investing book, which is three uncorrelated systems running together (the Starter being one of them), has a 15 percent worst drawdown. Roughly the same returns. Roughly half the pain.

That smoothness is not free. It costs you the ceiling. If a single system has a banger year and triples, the combined book will not triple. You averaged it down by running the diversifying ones alongside. The trade you’re making is upside ceiling in exchange for survivability. And survivability is everything in this business, because the only edge that matters is the one you actually stay with through the drawdown.

This is why I do not just trade my best system. The best system from the last 26 years might not be the best one over the next 26. And even if it is, I cannot tolerate the drawdown of running it alone, which means I would cut it during the worst stretch and lock in the worst of it.

Running three or five or ten systems alongside each other is the discipline I run because it’s the discipline that lets me sleep through the bad months. And being able to sleep through the bad months is what produces the long-term return.

If you take nothing else from this email, take this:

The goal isn’t to find the one perfect system. The goal is to build a portfolio of systems whose drawdowns don’t line up.

For the people running the Starter System on its own, that is totally fine. One engine is a real way to be a systematic trader, and the Starter backtest holds up over 26 years on its own. But understand that you are getting one engine. The smoother ride with the same return profile is what Pro is built to deliver, and that’s the honest reason it’s the upgrade.

Next week’s Q&A I’ll take a different question. If you’ve got one, hit reply. I read every one.

Michael

Michael Nauss, CMT, CAIA, CDMS. Founder, Stats Edge Trading.