Nearly 40% of day traders quit within their first month. By year two, about 80% are gone. After five years, roughly 7% remain.

Here's the part that should bother you: most of them didn't quit because their approach was broken. They quit because it was boring, flat, or slightly down — which is exactly what a working system looks like most of the time.

The dropout curve has a shape

Nobody quits during a winning streak. Almost nobody quits during a dramatic crash either — a crash at least feels like a story. People quit in the middle. Month four of a sideways account. The seventh small loss in a row. The stretch where nothing happens and your friend's meme coin is up 300%.

Man Group looked at what happened to investors who abandoned trend-following strategies after the underperformance of 2016–2018. Walking away cost them roughly 20% of subsequent upside. Same pattern in 2023: investors pulled about $7 billion from AHL's managed futures funds near the trough — right before the 2024 rebound.

These aren't retail gamblers. These are allocators with research teams, who had seen the math, agreed with the math, and still couldn't sit through the boring part.

Why the middle feels unbearable

Loss aversion is the usual suspect — losses hurt about twice as much as equivalent gains feel good. But I think there's a second force that gets less attention: the middle of a system's life gives you zero feedback.

A discretionary trader gets a dopamine hit every trade. Win or lose, something happened, and they did it. A systematic trader in month five of a flat stretch gets nothing. No confirmation the system works. No proof it's dead. Just waiting.

Your brain reads that silence as evidence of failure. It isn't. It's just what most of the equity curve looks like up close.

My own BTC strategy wins 21.9% of its trades. That means long stretches — sometimes ten or twelve trades — where every single position closes red. The 6-year backtest returned +4,909%, but nearly all of that return came from a handful of trends. If you'd quit during any of the flat stretches, you'd have paid the full cost of the system and collected none of the payoff.

The mistake, named precisely

The common error isn't picking a bad system. It's judging a system on a sample size that can't possibly tell you anything. Three losing months is noise for almost any strategy with a real edge. Traders treat it as a verdict.

And here's the cruel mechanic: quitting doesn't just lock in the drawdown. It usually leads to switching — to whatever strategy looks best right now, which by definition just finished its good stretch. So you exit one system at its low and enter another at its high. Repeat this three times and you've underperformed everything you ever touched.

What to actually do

Decide your quitting criteria before you start, in writing, when you're calm. Something like: "I abandon this system only if the live drawdown exceeds 1.5x the worst backtest drawdown, or the logic stops matching the market structure it was built for." Boredom, impatience, and a flat quarter don't appear anywhere on that list — so they don't get a vote.

If the system hasn't violated its pre-written failure criteria, your only job is to keep running it. That's the whole discipline. Not the entry, not the analysis — the sitting.

This is honestly the strongest argument for automation I know. A bot doesn't experience month four. If you want to see what a system that survived its own boring middle looks like, the full 6-year backtest data and verification scripts are at v33systematic.com.

Want a system that doesn't flinch in the boring middle? See the full backtest data and methodology.

See the strategies