Sunk Cost Fallacy Trading: Cut the Dead Trade

The sunk cost fallacy in trading has a tell. You stop asking whether a position is still good and start asking how much you have already put into it.

Those are different questions. Only one of them has anything to do with future returns.

A trade that is down 12% does not owe you anything. The market does not know your entry price. Yet the account balance you have already committed quietly becomes the reason you stay.

What Is the Sunk Cost Fallacy in Trading?

The sunk cost fallacy is the tendency to continue an action because of resources already spent, rather than because of expected future value.

Economists call the spent money “sunk” for a reason. It is gone. It cannot be recovered by any decision you make now. A rational choice looks only at what happens next.

Humans do not work that way. Arkes and Blumer demonstrated this in their 1985 study, where people who had paid more for a theatre subscription attended more performances in bad weather. The tickets were already paid for either way.

In trading, the sunk cost is not only money. It is the hours of research, the conviction you posted publicly, and the opportunity cost of everything you did not buy instead.

The Concorde Problem, in a Trading Account

Britain and France kept funding the Concorde long after it was clear the aircraft would never turn a profit. The money already spent was the argument for spending more. Behavioural economists still call this the Concorde fallacy.

Your account runs the same programme on a smaller scale. A position is down 15%. Cutting it means admitting the research was wrong. Holding means the loss stays theoretical.

So you hold. Then you average down, which converts a bad trade into a bigger bad trade. Each addition raises the cost of admitting the mistake.

This is escalation of commitment. The deeper you are, the harder the exit feels, and the deeper you get.

Where It Shows Up in a Trading Week

The fallacy rarely announces itself. It arrives dressed as patience.

Widening a stop. The price approaches your stop, so you move it. The stop was your maximum acceptable loss. Moving it means the loss already taken is now negotiating on the trade’s behalf.

Averaging down without a plan. Adding to a loser can be legitimate if the plan specified it in advance. Adding because the loss feels too large to accept is something else.

Refusing to switch strategies. You spent three months building a system. It has underperformed for six. The build time is sunk, but it keeps you from turning the strategy off.

Holding a bag for the narrative. The thesis broke months ago. You still hold, because selling makes the research worthless.

This differs from the disposition effect, which is about the pain of realising a loss. Sunk cost is a reasoning error about resources already committed. They compound each other, but they are not the same bias.

The Question That Breaks the Loop

There is one test that cuts through all of it.

Would you open this position today, at this price, knowing nothing about your entry?

If the answer is no, you are not holding a trade. You are holding a receipt.

The question works because it deletes the anchor. It forces a fresh evaluation using only current information. Most traders find the answer uncomfortable, which is precisely why it is useful.

Try it on your worst open position right now. Then ask a harder version: if you had this cash instead, is this the trade you would pick?

How Rules Remove the Decision Entirely

The question above is a manual fix, and manual fixes fail exactly when you need them. Under loss, judgement is the least reliable tool you own.

The structural fix is to decide the exit before the position exists.

Write the invalidation first. Not a price you hope holds, but a condition that proves the idea wrong. A close below the structure that justified the entry. A trend filter flipping. A volatility band breaking.

Then commit to it mechanically. An exit that executes without your input cannot be talked out of firing.

Three rules cover most of the damage:

A hard invalidation level, set before entry. Defined by market structure, never by how much you are willing to lose emotionally.
A time stop. If the thesis has not played out in N candles, the setup has expired. Close it regardless of P&L.
A scheduled strategy review. Judge systems on a fixed calendar and a full sample, not on how much effort they cost to build.

How to Apply Sunk Cost Fallacy Fixes in Arrow Algo

Arrow Algo’s visual block builder turns each of those rules into blocks you drag onto a canvas. No code is involved.

For the hard invalidation, drop a comparison block that checks price against your structural level. Wire it straight to a close-position block. The exit fires on the condition, not on your mood.

For volatility-aware stops, use an ATR block feeding a stop distance. This scales your exit with current conditions rather than a fixed percentage that becomes meaningless when ranges expand.

For the time stop, use a Timer block counting candles since entry. When it reaches your limit, the position closes. A trade that has not worked in the window you allowed is a trade whose thesis expired.

For strategy-level discipline, run your backtests over the full sample and compare them against a benchmark. Arrow Algo pulls historical data directly from Binance, Coinbase and HyperLiquid, so you evaluate on the exchange’s own candles. The equity curve does not care how long the build took.

The deeper point is simple. Automation does not make you immune to the bias. It moves the decision to a moment when you are calm, then removes your ability to override it when you are not.

What Are the Key Takeaways?

Sunk cost fallacy trading means holding positions for what you have already spent, not for what you expect to earn.
Money, research hours and public conviction all function as sunk costs.
Averaging down without a pre-set plan escalates commitment and makes the exit harder.
Ask whether you would open the position fresh today at this price. If not, close it.
Define invalidation before entry, using market structure rather than emotional tolerance.
Time stops close trades whose thesis has expired, regardless of profit or loss.
Automated exits cannot be renegotiated mid-drawdown, which is the entire benefit.

Disclaimer: This content is for educational purposes only and does not constitute financial advice. Trading involves significant risk and you should only trade with capital you can afford to lose. Past performance is not indicative of future results. Always conduct your own research before making any trading decisions.

Ready to build your own automated trading strategies without writing a single line of code? Start for free at Arrow Algo and join thousands of traders who've made the switch to systematic trading.