When Past Effort Becomes a Reason to Continue: Sunk Costs and Market Commitment

The research question: is the reason about the future or the past?

A market thesis can accumulate more than financial exposure. It can accumulate research time, public statements and a sense of personal ownership. When new evidence arrives, abandoning the thesis may then feel like invalidating all of that effort. The research question is whether past investment changes willingness to continue even when it does not improve the prospective case. This should be distinguished from rational persistence: an unfavorable recent outcome does not, by itself, prove that continuing is a mistake. Arkes and Blumer studied the sunk-cost effect across experimental and field contexts, examining the tendency to continue after resources had already been committed. Their work included evidence about concerns over appearing wasteful. [1] These findings offer a framework for asking how prior investment influences judgment. They do not establish that any particular losing position is irrational, or that immediate exit is always correct. The relevant comparison is between future alternatives given current evidence, including genuine switching costs and constraints.

Responsibility can change the meaning of negative feedback

Staw investigated escalating commitment in a simulated business-investment decision, manipulating personal responsibility and decision consequences. [2] The study illustrates how negative feedback can interact with ownership of an earlier choice rather than simply causing a reversal. Its participants and task were not live gold traders. The transferable question is whether defending a prior decision becomes entangled with evaluating the next one. That question requires measurement rather than an assumption that persistence always reflects ego. Garland examined how sunk costs influenced decisions to allocate further resources to an ongoing project. [3] Together, these studies show why a forward-looking choice can become sensitive to information that is psychologically salient but economically unrecoverable. The boundary matters: a past action may have produced useful knowledge, an asset or a contractual obligation that affects future options. Calling every historical consideration a sunk-cost error would be as careless as treating every prior expense as a reason to continue.

A trading thesis can acquire social and personal attachments

Imagine a hypothetical analyst who publicly argues that gold will rise, spends several days gathering supporting evidence and then encounters an important contradiction. The correct response is not predetermined. The analyst should examine the contradiction and the original thesis. But if the stated reason for maintaining exposure becomes 'I have already spent too much time on this view,' the justification has shifted away from prospective evidence. The time cannot be recovered by making the next decision less carefully. Informational cascades around a viral forecast can intensify this problem by adding an audience. A person may feel pressure to remain consistent with a group or with their own earlier public statements. That is a plausible social mechanism, not a claim that every public forecast produces escalation. A study would need to compare conditions with and without public commitment while controlling the information available. It should also distinguish reputational concerns from legitimate changes in the underlying opportunity.

Review continuation and exit with the same evidential standard

A useful review asks what would justify the current choice if the earlier commitment were absent. This counterfactual question can expose a historical attachment, but it is not a complete decision rule. Existing positions may carry transaction costs, tax consequences or constraints that a new position would not. The reviewer should explicitly separate those genuine forward consequences from unrecoverable effort. Otherwise a simplistic 'start fresh' instruction could discard relevant information while claiming to remove bias. The disposition effect provides a related lens on reluctance to realize losses, but the two concepts are not identical. Sunk costs concern the influence of prior investment on continuation; the disposition effect describes a pattern of realizing gains and losses. A study should define which behavior it tests. Labelling every delayed exit with both terms adds vocabulary without increasing explanatory precision. The decision's information set, alternatives and timing remain necessary for a meaningful interpretation.

Within NSI, a proposed interdisciplinary framework, this topic could be studied through timestamped decision explanations in a controlled task. Researchers could distinguish new evidence, changing constraints and references to prior effort, then compare those categories with subsequent choices. Independent coding and prespecified criteria would reduce the risk of labelling any unsuccessful persistence as bias. NSI should not be presented as a validated medical or financial prediction system, and no physiological signal alone would establish sunk-cost reasoning. The conclusion is that persistence deserves a prospective explanation. Past work can provide knowledge, but past expenditure is not automatically a reason to spend more. The primary studies make this distinction scientifically interesting without prescribing the correct outcome of every market decision. A sound educational review gives continuation and exit equal scrutiny, identifies genuine future costs and leaves room to revise a thesis without turning revision into a judgment about personal worth.