Why does money you have already lost change what you do next?

Why this matters. Every framework you have ever read says ignore sunk costs. Nobody does. It is worth understanding why the advice fails rather than repeating it louder.

The money is gone. It cannot be recovered by this decision or any other. And yet the size of what you have already lost is one of the strongest predictors of what you do next.

The finding, and its limit

Replicated: Arkes and Blumer (1985) established the sunk-cost effect across repeated experiments, and Staw (1976) showed the escalation pattern — people commit more resources to a course of action precisely after it has gone badly, and more so when they chose it themselves. That last clause is the part that matters here. Escalation is strongest when the decision was yours. An open position you selected is the maximum-exposure case.

Why the standard advice does not work

"Ignore sunk costs" asks you to treat the loss as irrelevant. But the loss is not psychologically irrelevant — it is the thing your reference point is anchored to. Telling someone to ignore their anchor does not move the anchor.

What moves the decision is changing what the decision is about. The useful question is not "should I hold this?" — which drags the entry price in with it — but: would I open this position today, at this price, at this size, knowing what I now know? Same economics, different reference point, and it is answerable.

The artefact

A fresh-eyes note, written as though flat. One paragraph: the case for the position as it stands today, with no mention of entry price and no mention of what it has cost so far. If the paragraph is hard to write without those two facts, the position is being held by its history rather than its case.

Keep the note. Later it becomes evidence about you: how often did the fresh-eyes case turn out to exist?

Try it now

Write the fresh-eyes paragraph for something you hold. Count how many times you reach for the entry price while writing it, and delete each one.