The Sunk-Cost Fallacy: When 'I've Already Put So Much In' Drives the Next Decision

Most people have sat through a disappointing film because they had already paid for the ticket, or finished an unappetising meal because it was expensive. The money is gone either way; staying or eating does not bring it back. Yet the past cost still pulls on the decision. Economists call this the sunk-cost fallacy: letting costs that have already been incurred, and cannot be recovered, influence choices about the future.

In investing, sunk costs are everywhere, and they often come wrapped in reasonable-sounding language: 'I can't stop now', 'I've already averaged down twice', 'I've paid premiums for five years'. This article explains the research behind the idea and how to make decisions that look forward rather than backward.

What the research found

In a 1985 paper titled 'The psychology of sunk cost', published in Organizational Behavior and Human Decision Processes, Hal Arkes and Catherine Blumer reported a series of studies on the effect. In one well-known field experiment, people buying season tickets to a university theatre were randomly given different prices: some paid the full price and others received discounts. Over the following months, those who had paid full price attended more performances than those who had received discounts, at least early in the season. The amount already paid influenced later behaviour even though it could not be recovered.

Arkes and Blumer suggested that one reason for the effect is a desire not to appear wasteful: abandoning something you have invested in can feel like admitting that the original investment was a mistake. That connects the sunk-cost fallacy to loss aversion, described in Kahneman and Tversky's prospect theory: walking away crystallises a loss, while continuing keeps the hope of recovery alive.

How sunk costs show up for Indian investors

• Averaging down repeatedly on a falling stock mainly to lower the average cost and 'recover faster', rather than because the case for owning it has strengthened.
• Continuing to hold a losing F&O strategy, or rolling positions forward, because of the losses already taken.
• Keeping an underperforming mutual fund because of the years already invested in it.
• Continuing to pay for a stock-tips subscription or course because of what it has already cost, even when it has not helped.
• Continuing an insurance-cum-investment policy mainly because of the premiums already paid, without comparing the future costs and benefits of continuing against the alternatives.

The last example deserves care. Surrendering some policies early can involve surrender charges, loss of insurance cover, or tax consequences, so continuing is sometimes the better choice. The point is not that continuing is wrong; it is that the right comparison is between future costs and future benefits from today onwards, including any exit charges, and not the total already paid.

A worked illustration

Vikram bought 100 shares at ₹500. The price fell to ₹350, and he bought 100 more to bring his average cost down to ₹425. It fell to ₹250, and he is considering buying 200 more to reduce his average to about ₹337. His reasoning is that he will then 'only' need the price to reach ₹337 to break even.

Notice what that reasoning is focused on: the average cost, which summarises money already spent. It says nothing about whether the company is likely to be worth more than ₹250 in the future. If the business is sound and the fall reflects temporary pessimism, buying more could be a reasonable decision. If the business has deteriorated, buying more increases his exposure to a weakening company. The decision should rest on that question, not on the average cost.

When adding to a falling position is a deliberate choice

It is worth being fair to averaging down. Some long-term investors deliberately plan to buy more of a diversified fund or a high-conviction holding if prices fall, and a regular SIP automatically buys more units when prices are lower. The difference from the sunk-cost fallacy is that these decisions are made in advance, within set limits, and for reasons about the future. Averaging down becomes a trap when it is improvised in response to losses, when it keeps pushing a single position beyond the share of the portfolio you originally intended, or when its main purpose is to make the average cost look better rather than to buy something you believe is worth more than its price.

Why it is so hard to let go

Sunk costs are tied to identity and effort as well as money. Time spent researching a stock, arguments made in its favour, and commitments shared with friends all raise the emotional price of changing course. Abandoning the position can feel like abandoning a part of yourself. Recognising that the past investment of time and money is already spent, and will remain spent whatever you do next, can make it easier to evaluate the future honestly. Some investors find it helps to say this out loud: 'That money is already spent. The only question is what to do with what I have now.'

Practical self-checks

• Ask the forward-looking question: from today, what are the expected costs and benefits of continuing versus stopping? Leave out anything that has already been paid and cannot be recovered.
• Use the fresh-start test: if someone handed you this position today for free, would you keep it, add to it, or sell it?
• Separate averaging down from rescuing. Before adding to a losing position, write down what new information makes the investment more attractive now. If the only reason is lowering your average cost, pause.
• Set limits in advance. Decide the maximum you are willing to allocate to any single idea, and do not raise that limit simply because the position has fallen.
• Get an outside view. Explaining your reasoning to someone with no stake in the decision often reveals when past costs are doing the talking.

The takeaway

What you have already spent is gone, whatever you choose next. The sunk-cost fallacy is the tendency to forget that and let past commitments steer future choices. Good investment decisions ask only one question: given everything known today, what course of action is most likely to serve my goals from here?