Seven Losses in a Row Is Normal. Doubling Isn't.

Nick Leeson's hidden error account was down £2 million at the end of 1992. Three years later it was £827 million, Barings Bank was gone, and a 233-year-old institution had been erased by one trader who kept increasing his size to get even.

He wasn't stupid. He was doing the thing almost every retail trader does after a bad week, just with someone else's balance sheet.

Here's the mechanic. You take a loss. It annoys you more than it should, because you had it right for a while and gave it back. So the next trade gets a little bigger.

Not double. Nothing dramatic. Just bigger, because if this one works you're back to flat, and flat feels like safety.

That's martingale thinking. You don't have to run a formal doubling system to be running one emotionally.

The math is unkind. Risk 1% of a $25,000 account, then double after each loss, and seven straight losers takes the whole account. Not most of it. All of it.

Scale it down and it's just as clear: after seven consecutive losses on a $10 base you're down $1,270, and you have to put $1,280 at risk to get back to even. You're risking more than you've already lost, on one trade, at the exact moment your judgment is at its worst.

Now the part most people miss. A run of six to ten consecutive losses isn't a tail event. It's a normal feature of most strategies. My BTC flip system wins about 22% of the time, and at that win rate a ten-trade losing streak isn't something I fear across a hundred trades — it's something I expect.

So the trader who sizes up to get even isn't betting against a rare disaster. He's betting against a Tuesday.

The formal version of this is the gambler's ruin result: with finite capital and no positive edge, the probability of eventual ruin goes to 1 as trades accumulate. Not "high." One. The only thing that rescues a martingale player is infinite money, and nobody has that. It's why casinos set table limits — they're not protecting you, they're protecting the arithmetic that makes them rich.

Why do smart people keep doing it? Because it works most of the time. That's the trap.

Nine sessions out of ten, sizing up after a loss does exactly what you hoped and you end the day green. You get paid for the bad habit over and over, right until the session where you don't. The feedback loop trains you into the behavior that eventually ends you.

There's a second reason, and it's worse. Sizing up after a loss isn't a strategy decision. It's an emotional one wearing a strategy costume.

You'll rationalize it — the setup was cleaner, the market was more oversold, you were "adding to a position." So ask yourself one question: would you have taken that size if the previous trade had won? If the answer is no, your size isn't coming from your edge. It's coming from your feelings about a number on a screen.

In my experience that's the clearest line between traders who last and traders who don't. Not skill. Not entries. Whether position size is a function of the account or a function of the mood.

Here's what to do about it. Write your size rule down before the week starts as a fixed percentage of equity, with a hard number attached. Then add one line underneath: size never increases after a loss.

Ever. Not a little. If you want to change your sizing, change it on a Sunday when nothing is bleeding, based on account equity — not on what happened Thursday afternoon. One sentence in a text file is the cheapest insurance in trading, and almost nobody buys it.

The deeper fix is removing the decision entirely. A system that sizes off account equity doesn't know it lost yesterday and doesn't care. It takes the next signal at the same risk it took the last one, which is exactly why it's still standing after the streak that would have ended you. We publish the full backtest data, losing streaks included, at v33systematic.com.