827 Million in Six Weeks: The Bias That Turns a Small Loss Into a Catastrophe

Published 09 Sep 2026 · 6 min · PSYCHOLOGY

Six weeks to break a 233-year-old bank

On 17 January 1995 an earthquake devastated Kobe and the Nikkei collapsed. Nick Leeson, Barings' trader in Singapore, was holding positions built on the assumption that the market would stay calm. Instead of closing them, he bought more futures. The market was going to bounce, and when it bounced, everything would be covered.

It did not bounce. The losses hidden in account 88888 went from 200 million pounds to 827 million in six weeks — more than the bank's entire capital base. Barings, founded in 1762, ceased to exist. Leeson was arrested in Frankfurt on 2 March.

Here is the uncomfortable part for you: the mistake that broke Barings was not the first losing trade. It was the second decision.

The bias has a name: escalation of commitment

In 1976 Barry Staw published a study with a title borrowed from a Vietnam-era song: "Knee-deep in the Big Muddy". He had 240 business students allocate budget to a division that had already failed. The finding that changed decision psychology was this: the people who had personally made the original decision poured more money into the failure than the people who inherited someone else's decision.

And it gets worse. Those who had already sunk money into the project rated its probability of success higher than people looking at the exact same project with nothing at stake. Lost money does not just tie you to the position — it distorts how you read the chart.

That is sunk cost. In trading it has one very concrete, very ordinary face.

What it looks like in your account: averaging down

You buy at 100. It drops to 95. Your thesis "is still intact", so you buy more at 95 to lower your average. It drops to 90. You buy again. Now you are carrying three times your original size in the worst trade of your week, and the original stop no longer makes sense because closing there would hurt too much. So you move it.

Notice the elegance of the trap: every individual step looks rational. "Better price", "same thesis", "managing the position". But the sequence is a systematic increase in risk at exactly the moment the market is telling you that you were wrong. Averaging down is not management. It is paying not to admit an error.

Prop firm data confirms it from another angle. Between 90 and 94 percent of traders fail challenge phases, and the dominant cause is not a bad strategy — it is maximum drawdown breaches. Accounts that would have died at -1% die at -8% because somebody chose to defend a position instead of closing it.

Why the second decision hurts more than the first

The first loss is money. The second decision is identity.

When you add to a loser you are not buying the asset — you are buying the possibility of not having been wrong. That is why the bias spikes precisely when the decision was yours, exactly as Staw measured. The ego cannot close in red, because closing in red signs a confession.

This is the mechanism PSYCHO / The Trader Within exists to dismantle. A loss is not a verdict on you. It is the operating cost of a probabilistic business. The moment a trade's outcome stops being a note about your intelligence, closing it stops requiring courage.

The framework: the outsider rule

Before adding to any losing position, answer one question in writing: if I did not hold this position and I saw this chart right now, would I open at this price with this size?

If the answer is no, you are not averaging. You are rescuing. And a rescue is not a setup.

Three rules that make that question executable:

1. Full size defined before entry. If your plan allows scaling in, maximum risk is calculated on the complete position, not on the first entry. A plan that grows after entry was never a plan.

2. The stop never moves backwards. It can follow price in your favour. Never against you. The day you move a stop to avoid a loss, you already lost — you only negotiated the price.

3. One loss, one mental session close. Write the outcome and the outsider question in your journal. If you answered "no" and added anyway, flag that trade as a process error even if it ended green. A win earned by escalation is the most expensive lesson there is, because it teaches you to do it again.

That honest record — process versus outcome — is the core of The Trader Within, and it is the only thing separating a trader who loses 200 from one who loses 827.

Leeson had six weeks. You usually have six minutes. Spend the first one asking whether you would buy this from scratch.