Loss Aversion And The Disposition Effect: Why Owners Hold Losers And Sell Winners Too Soon

Ask an owner why they still hold a stock down 40% and the answer is almost never "the fundamentals still support this price." It is closer to "I don't want to lock in the loss," a sentence that reveals, without the speaker intending it, that the decision is being driven by an accounting event rather than by forward-looking value. The stock does not know what was paid for it. The owner cannot forget.

The Theory: Prospect Theory In Brief

Kahneman and Tversky's 1979 paper, one of the most cited in economics, replaced the standard expected-utility model of decision-making under risk with one grounded in observed behaviour rather than theoretical consistency[1]. Three features of prospect theory matter for what follows. First, people evaluate outcomes as gains and losses relative to a reference point, not as final states of wealth, a hospital patient who learns their prognosis has improved from grim to poor experiences this as good news, even though their absolute state is still bad. Second, the psychological impact of a loss is substantially larger than the impact of an equivalent gain, a ratio generally estimated in subsequent research at somewhere between 1.5 and 2.5 to 1, commonly summarized as "roughly two to one." Third, people are risk-averse for gains and risk-seeking for losses: shown a certain gain versus a larger uncertain one, most choose the certain gain; shown a certain loss versus a larger uncertain one, most gamble to avoid the certain loss.

This last feature is the one most directly relevant to holding declining positions. An investor facing a locked-in loss if they sell now, versus an uncertain outcome if they hold, will systematically prefer the gamble, not because they have calculated that holding has positive expected value, but because prospect theory predicts exactly this preference reversal in the loss domain.

The Value Function, And Why Its Shape Matters

Kahneman and Tversky's value function is S-shaped: concave for gains, meaning each additional dollar of gain matters slightly less than the one before it, and convex for losses, meaning each additional dollar of loss also matters slightly less than the one before it, but the loss side of the curve is steeper throughout[1].

The convexity in losses produces a specific and counterintuitive prediction: because the pain of going from a $10,000 loss to a $15,000 loss is smaller than the pain of going from $0 to $5,000, a person already sitting on a loss has a weaker disincentive to risk making it larger than a person about to take a first loss. This is precisely backward from prudent risk management, which would suggest tightening risk as losses accumulate, not loosening it, and it is a specific, testable mechanism behind the familiar pattern of an investor or owner "doubling down" on a position that has already gone badly.

The Disposition Effect

Shefrin and Statman formalized the portfolio-level consequence in 1985, naming it the disposition effect: the tendency to sell assets that have risen in value while continuing to hold assets that have fallen[2]. The name is intentionally clinical, because the effect is a direct, almost mechanical, consequence of the value function described above: selling a winner locks in a gain, which feels good under the concave gain function; selling a loser locks in a loss, which feels bad under the convex loss function and, per loss aversion, feels roughly twice as bad as the winner felt good.

The rational benchmark this violates is straightforward. Absent tax or transaction cost considerations, the decision to hold or sell any asset should depend entirely on its expected future return relative to alternatives, and should be completely independent of whether the position happens to currently show a gain or a loss relative to its purchase price. The purchase price is a sunk historical fact with no bearing on future returns. The disposition effect is, in this sense, a close cousin of the sunk cost fallacy discussed elsewhere in this publication, expressed specifically through the lens of gains and losses rather than through spending already committed.

The Evidence

Odean's landmark 1998 study of over 10,000 individual brokerage accounts found direct evidence of the effect in real trading behaviour, not laboratory choices: investors realized gains at meaningfully higher rates than losses throughout the year, reversing only in December, when tax-loss selling motives temporarily overcame the psychological pull[3]. Critically, Odean also found that the stocks investors sold, the winners, subsequently outperformed the stocks they kept, the losers, meaning the disposition effect was not merely a behavioural curiosity but was actively costing these investors money relative to a strategy of holding or selling based on forward expectations.

The finding has been replicated across markets, asset classes and investor types in the decades since, including among professional traders and, with a smaller but still present effect, among some institutional investors. It is considered one of the most robust findings in behavioural finance precisely because it shows up in real transaction data with real money at stake, not only in survey responses or experimental choices.

Where The Reference Point Actually Comes From

The entire mechanism depends on a reference point, and it is worth being precise about how arbitrary that point usually is. For a purchased stock, the natural reference point is the purchase price, which is at least a real transaction. But the same mechanism operates around reference points that are considerably less grounded: the highest price a stock has ever traded at, a valuation an owner once heard mentioned for their business, the price a neighbour's house sold for, or simply the balance an account showed at its peak.

None of these carries any information about future value, and all of them can anchor the loss-aversion mechanism just as effectively as an actual purchase price. An owner who was once told their business might be worth $3 million, informally, years ago, can experience a genuine, current $2.2 million offer as a loss, complete with the full behavioural resistance that framing produces, even though no transaction ever occurred at the higher figure and the $2.2 million may be an entirely fair, well-supported number in current market conditions.

The Business-Owner Version

For an owner, the disposition effect operates on at least three levels beyond a securities portfolio.

Unprofitable divisions and product lines. A business line that was once profitable and has declined is evaluated relative to its historical peak, not relative to its current prospects, which produces the same reluctance to "sell at a loss" that Odean documented in brokerage accounts, expressed instead as reluctance to close, discontinue, or divest.

The business itself, at exit. This is the highest-stakes version. An owner who received informal interest at a higher valuation in a stronger market, or who watched a peer's business sell for a large multiple, anchors on that figure. A subsequent, genuine offer below it is experienced as a loss to be resisted rather than an outcome to be evaluated on its own terms, and owners in this position frequently hold past the point where waiting has positive expected value, hoping for a recovery to a reference point that may never return.

Real estate held for investment. Commercial or residential investment property purchased at the peak of a cycle is disproportionately likely to be held through a downturn specifically because selling would crystallize a loss, even where the capital would be better deployed elsewhere and even where the property's own prospects, evaluated fresh, do not justify continued holding.

Loss Aversion In Canadian Real Estate

Canadian real estate provides an unusually clean natural laboratory for this effect, because reference points are so public and so widely discussed. Every homeowner and investor in a given neighbourhood has ready access to what comparable properties sold for at the market's peak, which means the reference point is externally reinforced by every real estate conversation rather than existing only in the owner's own memory.

The behavioural pattern this produces in softening markets is well documented internationally and consistent with what Canadian data shows during periods of price moderation: sellers anchor list prices to recent peak comparables rather than to current market conditions, listings sit longer, transaction volumes fall more than prices do, because sellers unwilling to accept a "loss" relative to peak simply withdraw from the market rather than transact at the market-clearing price. For a business owner holding investment property as part of a broader portfolio, this means the asset most likely to be mentally anchored to an outdated reference point is also frequently the least liquid and the hardest to reprice through anything other than a direct, deliberate decision to do so.

The Genesove-Mayer Evidence On Housing

The claims above about real estate are not purely inferential. Genesove and Mayer's 2001 study of the Boston condominium market during a price downturn provides direct empirical evidence of loss aversion in a housing context, and its findings are unusually well suited to a Canadian audience because Canadian and American housing markets share enough structural similarity that the mechanism plausibly transfers[5].

The study found that sellers facing a nominal loss relative to their original purchase price set list prices 25% to 35% higher, relative to expected value, than sellers who were not facing a loss. These sellers also took longer to sell and, despite the higher list prices, achieved a higher realized sale price on average, but at a substantially reduced probability of selling at all within a given period. In effect, loss-averse sellers were gambling on a better outcome at the cost of liquidity, exactly the risk-seeking-in-losses behaviour prospect theory predicts, expressed through list price rather than through a portfolio trade.

The implication for a business owner holding investment real estate during a softer market is direct: the instinct to hold out for a price that avoids a "loss" relative to a purchase price or a remembered peak is not merely emotionally understandable, it is a documented, quantifiable market pattern, and it comes with a measurable liquidity cost. Whether accepting that cost is the right call depends entirely on the owner's actual need for liquidity, which is a question the loss-aversion instinct does not ask.

A Worked Case: The Two Divisions

A manufacturer operates two product lines. Line A was, three years ago, the company's most profitable segment, generating $180,000 in annual contribution margin. Competitive and input-cost pressure has eroded it to $20,000 annually and declining. Line B was launched eighteen months ago, is newer and less familiar, and currently generates $95,000 in contribution margin with a stable trajectory.

Asked where to focus limited management attention and capital, the owner's instinct, in our experience, gravitates toward defending Line A, "getting it back to where it was," rather than toward Line B, despite Line B currently generating almost five times the contribution margin. The reasoning is rarely stated in these terms, but the underlying logic is the disposition effect operating on a business unit rather than a security: Line A is evaluated against its own $180,000 peak and experienced as a loss to be recovered, while Line B, having no comparable peak in the owner's memory, is evaluated on its current, unremarkable-feeling trajectory rather than against the absence of a reference point.

The forward-looking question, ignoring both units' histories entirely, is simply which one produces better returns on the next dollar and hour of management attention. On the figures given, that is not close. The figures are illustrative, and any real allocation decision depends on growth trajectories, capital requirements and strategic fit that a single contribution-margin snapshot does not capture. The structural point is that the peak reference point for Line A was doing real work in the conversation before anyone examined whether it should.

The One Place Canadians Get This Right

It is worth noting, in fairness to investor behaviour generally, the one context where loss aversion is reliably overridden: December tax-loss selling. Odean's own data showed the disposition effect reversing specifically in December, as investors realized losses at higher rates to offset capital gains for tax purposes[3]. Canadian tax rules, allowing capital losses to offset capital gains in the current year, the prior three years, or indefinitely into the future, create exactly this incentive.

The interesting behavioural detail is not that Canadians harvest losses in December; it is that most only do it in December, rather than continuously whenever the analysis favours it. The tax deadline provides external permission to sell at a loss that the underlying economics alone did not provide, which is itself evidence for how much the disposition effect is really about emotional permission rather than analysis. Given permission via a clear external rule, the loss gets realized without much difficulty. Absent that permission, the same analytically identical loss sits unrealized for years.

Countermeasures That Actually Work

Set exit rules before the reference point exists. The most effective defence, structurally identical to the kill-criteria approach recommended elsewhere in this publication for capital projects, is deciding in advance, at purchase or at launch, what conditions would trigger a sale or closure, before any gain or loss has accumulated to anchor the decision.

Reframe every holding decision as a purchase decision. The question is never "should I sell this at a loss?" It is "if I did not currently own this, would I buy it today at its current price, given its current prospects?" If the answer is no, continuing to hold it is economically equivalent to buying it fresh, and the original purchase price is irrelevant to that choice.

Separate the emotional event from the decision. Realizing a loss and updating the strategy going forward are two different actions that do not need to happen in the same conversation. Some practitioners find it useful to formally "accept" a loss, journalling it, discussing it, treating the reference point as officially retired, before then making a fresh, unanchored decision about what to do next.

Use external deadlines deliberately. Since tax-loss selling shows that a clear external rule can override the effect where introspection cannot, the same principle can be applied deliberately elsewhere: a scheduled quarterly portfolio or division review, treated with the same "the deadline requires a decision" weight as the tax year-end, can substitute for the discipline that would otherwise depend on willpower alone.

The Advisor's Own Version

This bias is not confined to owners; it operates identically on the professionals advising them, and is worth naming because it is easy to miss from inside an advisory relationship. An accountant or wealth advisor who recommended a position, or who has watched a client hold one for years, develops their own reference point around it, and can find themselves reluctant to recommend crystallizing a loss for reasons that have more to do with their own discomfort delivering that advice than with the client's actual best interest.

The practical defence is the same reframing recommended throughout: an advisor reviewing a position should periodically ask, deliberately, whether they would recommend initiating it fresh today, independent of who currently holds it or how it got there. Advice that would not survive that test is itself running on the client's, or the advisor's own, loss aversion rather than on current analysis.

When Holding Is Correct

As with escalation of commitment, it would be a mistake to read this as an argument for reflexively selling anything currently at a loss. Genuine long-term investors with a sound original thesis and a long horizon are, correctly, expected to experience unrealized losses during normal volatility without that being evidence of anything having gone wrong. The test, again, is whether the case for continuing to hold rests on forward evidence, a still-intact thesis, a credible path to recovery grounded in something other than the price previously paid, or on the reference point alone.

A useful practical heuristic: if the same holding, at the same current price and prospects, arrived on your desk today with no purchase history attached, would you buy it? If yes, holding is not being driven by loss aversion, whatever the account statement shows. If no, the only thing keeping it is the number that used to be there.

The Limits Of This Analysis

Several caveats apply. The precise loss-aversion ratio, commonly cited as approximately 2:1, varies considerably across studies, populations and elicitation methods, with some research finding ratios closer to 1:1 under certain experimental conditions, and the "universal" 2:1 figure popularized in mainstream treatments overstates the consistency of the underlying evidence. The disposition effect itself, while very well replicated in securities trading, has been studied far less rigorously in the specific contexts of private business divestiture and real estate discussed here, and the extension is a reasonable inference from a well-established general mechanism rather than a directly measured finding in those specific domains.

It is also worth noting that some holding behaviour that looks like loss aversion may reflect entirely rational considerations this analysis cannot see from outside: genuine private information about recovery prospects, tax basis considerations that make realization currently unattractive, or contractual and relationship factors around a business or division that a pure numbers view omits. As throughout this series, the framework is a lens for examining decisions, not a verdict to be applied without knowing the specific facts.

Related Reading

Behavioural Finance Escalation Of Commitment: Why Businesses Keep Funding Projects They Should Have Killed The closely related mechanism: past commitment distorting a forward decision, examined in full. Fractional CFO & Exit Planning EBITDA Optimization Before An Exit Where an anchored valuation expectation most expensively collides with a real offer. Behavioural Finance Concentration You Can't See: Home Bias And The Canadian Business Owner's Undiversified Life Why the assets most likely to be held past a fair sale price are also the most concentrated ones.

Frequently Asked Questions

What is loss aversion, briefly?

Loss aversion is the finding, central to Kahneman and Tversky's prospect theory, that losses are experienced roughly twice as intensely as equivalent gains. It means the pain of losing $10,000 is felt more strongly than the pleasure of gaining $10,000, and this asymmetry shapes a wide range of financial decisions.

What is the disposition effect?

The disposition effect, named by Shefrin and Statman in 1985, is the tendency to sell assets that have gained value while continuing to hold assets that have lost value, driven by loss aversion. Terrance Odean's 1998 study of over 10,000 brokerage accounts found direct evidence in real trading data, and found that the winners investors sold went on to outperform the losers they kept.

Does this apply to selling a business, not just stocks?

The core mechanism, evaluating a current offer against a remembered peak reference point rather than against current market conditions, applies logically to any asset with a salient reference point, including a business at exit. This specific extension is inferred from the well-established general mechanism rather than directly measured in business-sale data, and should be treated as a reasonable hypothesis rather than a proven finding in that specific context.

Why do Canadians seem to realize losses more easily in December?

December tax-loss selling, permitted under Canadian rules that let capital losses offset capital gains in the current year or carry back three years or forward indefinitely, provides an external, rule-based justification for realizing a loss that overrides the usual psychological resistance. Odean's research found the disposition effect reversing specifically in this period.

Is a 2:1 loss-to-gain ratio scientifically precise?

It is a widely cited approximation rather than a precise constant. Estimates in the research literature vary depending on study design and population, with some finding ratios closer to 1:1 under certain conditions. The direction of the effect, losses weighing more heavily than equivalent gains, is far better established than any single numerical ratio.

How do I know if I'm holding something out of loss aversion or for good reason?

Ask whether you would buy the same holding today, at its current price and prospects, if you did not already own it. If yes, your reasoning is forward-looking regardless of what the position shows. If no, the purchase price or peak value is likely doing the work, not the current facts.

References

1. Kahneman, D., & Tversky, A. (1979). Prospect Theory: An Analysis of Decision under Risk. Econometrica, 47(2), 263-291.
2. Shefrin, H., & Statman, M. (1985). The Disposition to Sell Winners Too Early and Ride Losers Too Long: Theory and Evidence. The Journal of Finance, 40(3), 777-790.
3. Odean, T. (1998). Are Investors Reluctant to Realize Their Losses? The Journal of Finance, 53(5), 1775-1798.
4. Tversky, A., & Kahneman, D. (1991). Loss Aversion in Riskless Choice: A Reference-Dependent Model. The Quarterly Journal of Economics, 106(4), 1039-1061.
5. Genesove, D., & Mayer, C. (2001). Loss Aversion and Seller Behavior: Evidence from the Housing Market. The Quarterly Journal of Economics, 116(4), 1233-1260.
6. Barberis, N., & Xiong, W. (2009). What Drives the Disposition Effect? An Analysis of a Long-Standing Preference-Based Explanation. The Journal of Finance, 64(2), 751-784.

This article discusses general behavioural patterns documented in academic research and is provided for informational purposes only. The worked example uses illustrative figures. It is not investment, tax, or business advice; portfolio, divestiture, and exit decisions are fact-specific and should be confirmed with qualified advisors familiar with your circumstances.