Sunk Cost Fallacy in Trading: Why Past Losses Shouldn’t Steer You

The sunk cost fallacy in trading shows up quietly. You bought a stock at ₹850, it slid to ₹620, and somewhere along the way the question changed from “is this still a good trade?” to “how do I get my money back?” Those two questions sound alike. They are not, and the gap between them costs traders more than most bad entries ever do.

This guide walks through what a sunk cost actually is, why our brains refuse to let go of it, and how it plays out in cash positions, averaging down and options. The rupee examples are hypothetical, built only to show the arithmetic.

What is the sunk cost fallacy in trading?

In economics, a sunk cost is a cost that has already been incurred and cannot be recovered. It is set against prospective costs, which are future costs you can still avoid by acting. Standard decision theory says only the prospective ones matter. Economists call this the bygones principle, or the marginal principle: the best choice depends on the alternatives in front of you now, not on what you paid to get here.

The fallacy is what happens when people ignore that rule. Research by Hal Arkes and Catherine Blumer, published in 1985 as “The psychology of sunk cost”, described it as a greater tendency to continue an endeavour once an investment in money, effort or time has been made. In plain words, throwing good money after bad.

For a trader, the sunk cost is the price you already paid, the brokerage already deducted, the premium already gone, even the hours spent on research. None of it changes what the stock does tomorrow. The market has no idea what your buy price was.

The market has no idea what your buy price was. Your decision today should not care about it either.

The textbook case, and the trading version

Wikipedia’s entry on sunk cost uses a neat factory example. A plant is projected to cost $100 million and yield $120 million in value. After $30 million has been spent, the value projection falls to $65 million. The rational move is to abandon it rather than spend another $70 million to finish. If the projection had fallen only to $75 million, carrying on would make sense. Notice that the $30 million already spent appears in neither decision. Only the remaining $70 million and the value it buys count.

Now the trading version, with made-up numbers.

Worked example 1: the stock you won’t sell

Say you bought 400 shares of a hypothetical XYZ Ltd at ₹850. That is ₹3,40,000 of capital. The stock now trades at ₹620, so the position is worth 400 × ₹620 = ₹2,48,000. You are sitting on a paper loss of ₹92,000.

The sunk cost voice says: “I can’t sell now, I’ll lose ₹92,000.” But here’s the thing. That ₹92,000 is gone whether you sell or not. What you actually own today is ₹2,48,000 worth of XYZ. So the honest question is: if you had ₹2,48,000 in cash right now, would you buy 400 shares of XYZ at ₹620?

If the answer is yes, because the business still looks sound and the price is simply cheaper, then holding is fine. If the answer is no, or “not really, I’m only holding to get back to ₹850”, then you are holding a stock you would never buy. Holding and buying are economically the same decision. Only the sunk cost makes them feel different.

We find this “would I buy it today?” test the single most useful habit against the fallacy. It strips out the purchase price, which is the one number that should not be in the room.

Worked example 2: averaging down to “fix” the loss

Averaging down is where sunk costs get expensive. Staying with XYZ, suppose you buy 200 more shares at ₹620 to bring your average down.

• New outlay: 200 × ₹620 = ₹1,24,000
• Total cost: ₹3,40,000 + ₹1,24,000 = ₹4,64,000
• Total shares: 600
• New average price: ₹4,64,000 ÷ 600 = about ₹773.33

On paper this looks like progress. Before, the stock had to climb from ₹620 to ₹850, roughly 37.1%, for you to break even. Now it only has to reach about ₹773.33, roughly 24.7% above ₹620.

But look at what happened to the risk. You now hold 600 shares instead of 400, a 50% bigger position in a stock that has already gone against you. If XYZ falls another 15% to about ₹527:

• Without averaging: 400 × (₹527 − ₹850) = a loss of ₹1,29,200
• With averaging: 600 × (₹527 − ₹773.33) = a loss of about ₹1,47,800

The lower average price made the loss feel smaller while making the actual rupee loss bigger. Averaging down is not always wrong. Long-term investors who have fresh, independent reasons to add can justify it. The test is whether you would make that ₹1,24,000 purchase as a brand-new trade. If the only reason is to repair the old average, that is the sunk cost talking. We go deeper into this in our piece on averaging down stocks.

Worked example 3: the option premium already paid

Options make the fallacy sharper because time works against the buyer. Suppose you paid ₹18,000 in premium for a call option position, and a few sessions later it is worth ₹4,500. The ₹13,500 drop is sunk.

Traders in this spot often say “it’s only ₹4,500 left, might as well hold to expiry and see”. That can be a reasonable call. It can also be the fallacy in disguise. The question is not whether ₹4,500 is small next to ₹18,000. It is whether you would pay ₹4,500 today for this exact contract, with this strike and these days to expiry. If you would not, the remaining premium is better off out.

Worked example 4: the research you can’t waste

Sunk costs are not only money. You spent three weekends reading an annual report, built a spreadsheet, and convinced a friend to buy the stock too. Then the quarterly numbers break the thesis. Selling now feels like admitting those weekends were wasted.

They were spent either way. The research has already done its job if it lets you see clearly that the thesis broke. Holding on to justify the effort just turns a lost weekend into a lost portfolio.

Why traders fall for it

The behavioural literature points to a few forces working together. The sunk cost entry lists four psychological factors that sit underneath the effect.

• Framing. People avoid risk when a choice is framed as a gain and seek risk when it is framed as a loss. A red P&L frames every decision as a loss, which nudges us towards gambling on a recovery.
• Over-optimism after investing. Once we have put money in, we tend to rate the odds of it paying off higher than we did before.
• Personal responsibility. The effect is strongest when we feel personally responsible for the original decision. A tip from a friend is easy to dump. Our own analysis is not.
• Not wanting to look wasteful. Stopping would mean admitting the earlier money was wasted.

The same entry mentions what is sometimes called the Concorde fallacy. The British and French governments kept funding the supersonic jet even after it became apparent there was no longer an economic case for it. Swap a government for a retail trader and a jet for a mid-cap that has halved, and the pattern is identical.

Selling a loser is not admitting the money was wasted. The money was lost the day the price fell. Selling just stops you from losing more of it.

Sunk cost and its close cousin, the disposition effect

If the sunk cost fallacy explains why we hold losers, the disposition effect describes the full pattern: investors tend to sell winning positions too early and hold losing positions too long. Hersh Shefrin and Meir Statman coined the term in a 1985 paper. They traced it to loss aversion, mental accounting against the original purchase price, regret avoidance and self-control problems.

Mental accounting is the bridge between the two ideas. The moment you anchor to your buy price, every decision gets measured against a number that should not matter. We covered the selling-winners side in our explainer on the disposition effect.

There is a related idea from aviation called plan continuation bias, the urge to stick with a plan even as conditions change. Wikipedia cites a 2004 NASA study that found aircrew showed this bias in 9 of the 19 accidents studied. A trader who refuses to change the plan because “I said I’d hold till results” is doing a milder version of the same thing.

How to keep sunk costs out of your trades

You can’t switch off the bias. You can build a process that gives it less room.

1. Decide the exit before the entry

Write down, before you buy, the price or condition that proves you wrong. A stop-loss set at entry is a decision made before any money is sunk, which is exactly when your judgement is cleanest. If you move it later, you should be able to give a reason that has nothing to do with your buy price.

2. Size positions so a loss is survivable

The bigger the loss, the louder the sunk cost voice. Here is a hypothetical example of sizing by risk rather than by conviction. With ₹10,00,000 of trading capital and a self-imposed limit of 1% risk per trade, the most you are willing to lose on one idea is ₹10,000. If you plan to buy at ₹500 with a stop at ₹475, your risk per share is ₹25, so the position is ₹10,000 ÷ ₹25 = 400 shares. A ₹10,000 hit is easy to accept. A ₹1,00,000 hit on an oversized bet is the kind that makes people hold on and hope. Our guide to risk management in trading covers sizing in more detail.

3. Run the “fresh eyes” review weekly

Once a week, go through each open position and hide the buy price. Look only at today’s price, the current chart and the latest news. Ask whether you would open this trade today at this size. Anything that fails gets trimmed or closed. It sounds simple. It is surprisingly hard the first few times, which tells you how much the purchase price was steering you.

4. Separate the review from the result

A bad outcome is not the same as a bad decision. Keep a trade journal that records why you entered, what would make you exit and what you actually did. When you look back, judge whether you followed the process, not just whether you made money. This takes the sting out of cutting a loss, because cutting it on plan counts as a good trade.

5. Watch your language

Certain phrases are red flags. “I’ll sell once it gets back to my price.” “I’ve already lost too much to sell now.” “Let me average and get out at cost.” Each one puts the sunk cost at the centre of the decision. When you hear yourself say one, run the would-I-buy-it-today test before doing anything else.

When holding a loser is the right call

None of this means you should dump every stock that is in the red. Prices fall for plenty of reasons that don’t touch the business: a broad market sell-off, a sector rotation, a block deal. If your original reasons still hold and you would happily buy at today’s price, holding is a forward-looking decision, not a sunk cost one.

The difference lies entirely in the reason. “The business is intact and the price is lower” looks forward. “I need to get back to ₹850” looks backward. Same position, completely different logic.

A quick checklist before you hold or add

• Would I buy this position today, at this price and this size, with fresh cash?
• Has anything changed in the thesis since I bought?
• Is my stop-loss still where I set it at entry, and if not, why did I move it?
• Am I adding to repair my average or because this is a fresh good trade?
• If I had never owned this stock, would it be on my buy list?

If the honest answers point to selling, the loss you lock in is not new. It already happened. Selling only frees the capital that is left to work somewhere better.

Disclaimer: This article is for educational purposes only and is not investment advice. All rupee figures and company names in the worked examples are hypothetical. Trading in stocks and derivatives involves risk of loss. Please do your own research or consult a SEBI-registered investment adviser before making investment decisions.