You're bad at cutting losses. And bad at walking away when you shouldn't.

The sunk-cost fallacy is the cheap one — continuing because you've spent money that didn't work out. The expensive version is the holding fallacy: continuing because you've spent time that didn't work out. Same mechanism underneath both. You don't want to waste what you've already put in, so you keep putting more in. The cheap version costs you dollars. The expensive one costs you years.

You can beg, borrow, or steal more dollars. You can't do that with time.

If you've kept a struggling hire for two years because you didn't want to waste the two years already spent, you know what this looks like. The math says the opposite. The next two years are paid out of everyone's remaining time, not the past. Letting the past decide the future is the fallacy at work.

The MBA curriculum probably taught you the sunk-cost version correctly. They missed the temporal cousin — what the literature calls escalation of commitment. The decision in front of you is whether to spend another $50K or another twenty-four months. By definition, neither the $200K already spent nor the eighteen months already logged are relevant to that decision. Evaluate the next $50K on its own terms. Evaluate the next twenty-four months on theirs.

The rational case for cutting losses in dollars is identical to the rational case for cutting losses in years. Most operators have internalized the first. Almost nobody has internalized the second.

It doesn't show up on a P&L. It shows up as the next planning cycle, the next quarter, the next performance review. By the time the cost is visible, you've already committed another year. The cascade is real — the underperformer absorbs leadership bandwidth that should be going to the rest of the team, the rest of the team notices, two good people start shopping their resumes, and the next-quarter hiring plan quietly doubles in cost. None of that was caused by the original bad hire. All of it was caused by the holding fallacy keeping them in place.

The same logic applies to a vendor relationship that isn't working, a product line that's underperforming, a market you've been trying to crack for two years and can't. The hire case is the clearest version — the true cost of a bad hire is usually measured in the months you spent keeping them, not the salary you paid.

Three moves that work.

Force a forward-looking frame at every review. Before any decision, write down what you'd do with zero prior investment. If you'd take the same action, hold. If you wouldn't, the past is deciding — it's a fallacy. Most owners discover, on paper, that they'd have walked away six months ago.

Build kill criteria before commitment. Every hire, vendor, or project needs defined exit conditions written down before it starts. "Revenue below $X by month Y" or "performance below expectations for two consecutive quarters" is enough. The point is to make the walk-away decision before the walk-away moment, when you still see clearly — the same discipline as phasing big moves as three small ones. Once you've lived with the position for a year, the holding fallacy is already baked in.

Name the opportunity cost out loud. The fallacy wins because nobody writes down what the next two years would buy if reallocated. A clear-eyed write-up of "if we cut this, here's what the next 24 months could fund" forces the comparison the fallacy is hiding. Often the opportunity — the second senior hire, the new market you shelved, the rebuilt system — is more valuable than the position you've been protecting. The position feels concrete; the opportunity is hypothetical. That's the asymmetry the fallacy exploits.

Sunk cost says: the dollar is gone, don't let it decide the next dollar. Holding fallacy says: the year is gone, don't let it decide the next year. Same math, different units, same discipline. Evaluate the next commitment on its own terms — not on what came before.