A sunk cost is money already spent and non-recoverable. Because it cannot be changed by any future decision, it should not influence go-forward choices — though the 'sunk-cost fallacy' causes many people to let it.
Sunk costs are economic history. Whether you've spent $50,000 on a marketing campaign that produced no leads, or $500,000 renovating a location that underperforms, those dollars are gone regardless of what you decide next. The rational decision is to evaluate only future costs and benefits — forward-looking cash flows — when deciding whether to continue, pivot, or exit. The sunk-cost fallacy is the human tendency to continue investing in a losing position because of what's already been spent: 'We've put $200K into this — we can't stop now.' This logic is economically irrational. The relevant question is: does the next dollar invested generate more than a dollar of expected return, given current information? Past spending doesn't change that calculation. For business owners, sunk-cost thinking most commonly appears in: (1) holding onto underperforming equipment rather than selling it, because of what was paid; (2) continuing a money-losing product line because of development investment; (3) staying in an expensive lease because of fit-out costs; (4) avoiding refinancing decisions because of origination fees already paid on the existing loan. In loan restructuring and workout situations, lenders explicitly apply sunk-cost thinking. The question is not 'how much have we already lent?' but 'given the current situation, is extending more credit, restructuring, or exiting our best forward-looking option?' Forbearance and workout decisions are made on prospective expected recovery, not historical loan balance.
The sunk-cost fallacy is making decisions based on past unrecoverable expenses rather than forward-looking expected value. It leads to irrational continuation of losing strategies, underperforming investments, or failed products. The rational corrective: ask only 'what are my expected future returns and costs from this point forward, independent of what I've already spent?'
Origination fees paid on an existing loan are sunk costs — they should not prevent refinancing if the forward-looking NPV of refinancing is positive. The analysis should compare future interest savings (over the remaining loan life) against new origination fees and any prepayment penalties. Fees already paid are irrelevant to that calculation.
Rarely, but there are edge cases. Sunk costs may be relevant for tax purposes — if the cost creates a tax loss, the tax benefit is a future cash flow that should be considered. They may also be relevant for accounting (impairment decisions require recognizing sunk costs as losses). But for operational go/no-go decisions, sunk costs should be excluded from the analysis.