Financial Planning
Sunk Cost Fallacy: Decide Using What Happens Next
Recognize the sunk cost fallacy and use an original stop-versus-continue worksheet that separates unrecoverable history from future consequences.

What you will learn
- Identify genuinely unrecoverable past costs.
- Compare future benefits and incremental costs consistently.
- Keep recoverable value, commitments and uncertainty in the decision.
The sunk cost fallacy is letting unrecoverable past spending overrule the future case for a decision. Avoiding it does not mean ignoring history; it means learning from the past without spending more merely to justify it. Compare what each feasible option can still change, including future costs, benefits and commitments.
What is an example of sunk cost fallacy?
A team might say, “We have already spent $40,000, so we have to finish,” without checking whether finishing is the best use of the remaining resources. The conclusion could still be right, but the historical spend alone does not establish it.
OpenStax’s cost discussion separates unrecoverable spending from costs a future choice can change. In their original research on the psychology of sunk cost, Hal Arkes and Catherine Blumer examined how prior investment can affect continuation decisions. This lesson uses a separate, fictional business exercise, not their experimental scenarios or results.
Compare an original software-project decision
Fictional business, one decision date, U.S. dollars. Past prototype spending of $40,000 is unrecoverable under the stated assumption. The team can use the same constrained capacity either to finish the prototype or pursue an alternative project. The figures below are uncertain planning inputs, not forecasts or customer outcomes. Assume the paths have comparable timing and no other costs for the first calculation.
Original forward-looking comparison
| Future item | Finish prototype | Pursue alternative |
|---|---|---|
| Expected future cash benefit | $18,000 | $9,000 |
| Future incremental cost | -$12,000 | -$4,000 |
| Expected future net benefit | $6,000 | $5,000 |
| Unrecoverable prior prototype spend | Excluded from the choice comparison | Excluded from the choice comparison |
Finishing has $1,000 more expected future net benefit on this narrow basis. Subtracting the same $40,000 history from both columns would not change the ranking, but putting it only into one option can make the comparison misleading. The purpose is to inspect what the choice can still change.
Test what happens when the future case weakens
Suppose new evidence lowers the prototype’s expected future benefit from $18,000 to $13,000 while the incremental cost stays $12,000. Its future net becomes $1,000, below the alternative’s $5,000. The historical $40,000 is unchanged. The new evidence, not a different view of the past, changes the comparison.
Do not use this example as an instruction to stop a real project. Contracts, safety, customer commitments, learning value and timing can change the relevant consequences. Make them explicit rather than placing them outside the table and then claiming the arithmetic settled everything.
Do not call every existing cost sunk
Cost classification worksheet
| Item | Question to ask | Treatment in the review |
|---|---|---|
| Unrecoverable past expense | Can today’s choice recover any of it? | Do not use it as a reason to add future spending |
| Refundable deposit | What amount is recoverable under each option? | Include the choice-dependent recovery |
| Asset with resale value | Can it actually be sold, when and at what cost? | Use supported future proceeds, not book value alone |
| Cancellation fee | Does stopping trigger a future obligation? | Include that incremental consequence |
| Existing commitment | Which payments remain unavoidable? | Keep them visible and distinguish avoidable amounts |
A fixed cost may be avoidable in a later period; an asset bought in the past may have recoverable value. The correct question is what differs between feasible paths, not which column in the accounts originally held the amount.
Make opportunity cost visible too
OpenStax’s opportunity-cost explanation concerns the next-best alternative forgone. In the worked example, using capacity to finish means giving up the alternative’s expected $5,000 net benefit. Do not add every possible alternative together or count the same cost twice.
Ask a fresh reviewer to read the forward-looking case without the historical amount first. Then restore the history for accountability and learning. If the preferred choice changes only because the old spending is shown, inspect the reasoning.
Use a stop-or-continue decision record
Record decision date, feasible paths, recoverable amounts, future costs and benefits, contractual questions, uncertainty, owner and next evidence point. Use the decision-journal lesson to preserve the rationale, and break-even analysis for a different operating-volume calculation.
Keep the forward-looking choice clear. Bring the alternatives, supporting records and open questions together in Finta, then ask Aurora to prepare a concise decision brief. The team can discuss what changes from today, make the choice and keep the follow-through connected to its reasoning.
Copyable AI prompt
Try this with Aurora AI in Finta, Codex, or Claude. Share only information you are allowed to use, and check the answer before acting.
Using [authorized project alternatives], draft a stop-versus-continue review. Separate unrecoverable past spending, recoverable value, future incremental costs, expected benefits, commitments and opportunity cost. Show assumptions and a downside case. Do not invent probabilities, cancel commitments or recommend personalized investments. Prepare an evidence-based draft for human owners.Finta Field Guide. By Finta Editorial Team. Sources checked November 15, 2026. Examples and worksheets are original fictional teaching exercises, not customer results or benchmarks. General education, not individualized accounting, financial, investment, legal or tax advice.