Disposition Effect¶

Explain why investors sell winners too early and cling to losers: the purchase price is a reference point, so gains sit in the concave risk-averse domain of the prospect-theory value function and losses in the convex risk-seeking one, measured as the PGR minus PLR gap.

Core Idea¶

The disposition effect is the empirical regularity in individual investor trading by which investors sell assets that have risen above their purchase price (winners) at systematically higher rates than they sell assets that have fallen below their purchase price (losers), even when tax incentives and return-momentum considerations both favour the opposite pattern. Named and documented by Shefrin and Statman (1985) and given canonical empirical confirmation by Odean (1998) using discount-brokerage account records, it is measured operationally as the ratio of gains realised to gains available (PGR) exceeding the ratio of losses realised to losses available (PLR) across a large cross-section of trading accounts.

The mechanism is a direct application of prospect theory to the asset-holding decision. The purchase price functions as a reference point; an asset currently above that price places the investor in the gain domain of the prospect-theory value function, where the function is concave and risk-averse — producing a preference for locking in the certain gain by selling. An asset below the purchase price places the investor in the loss domain, where the function is convex and risk-seeking — producing a preference for holding in hope of recovery rather than crystallising the certain loss. Mental accounting compounds the effect: an unrealised loss is an open mental account whose closure, triggered by a sale, forces acknowledgement of failure and the attendant regret; an unrealised gain can be enjoyed as a latent asset without closure's psychological cost. The asymmetry in realisation rates thus reflects not a view on future prices but a structural asymmetry in how the value function and the mental-accounting ledger treat gains and losses relative to the acquisition anchor.

Structural Signature¶

Sig role-phrases:

the purchase-price reference point — the acquisition price serving as the cognitive anchor against which gain and loss are reckoned
the current-price location — the sign of current-price-minus-purchase-price, placing the asset in the gain domain or the loss domain
the gain-domain risk-averse stance — concave value function above the anchor, favouring locking in the sure gain by selling
the loss-domain risk-seeking stance — convex value function below the anchor, favouring the gamble of holding for recovery over crystallising the certain loss
the open-versus-closed mental account — the unrealised loss as an open account whose closure by sale forces acknowledgement of failure and incurs regret
the perversity condition — both tax treatment (realise losses, defer gains) and short-horizon momentum push the opposite way, so the pattern cannot be rationalised as tax- or return-smart
the realisation-asymmetry observable — PGR > PLR (proportion of gains realised exceeds proportion of losses realised), the measurable signature in any trading record
the cognitive-substrate constraint — requires an owner with prospect-theoretic reference-point cognition and a discretionary hold-or-sell decision; absent that valuer there is no effect to find

What It Is Not¶

Not a tax-smart or return-smart strategy misread as a bias. The pattern runs against both the tax code (which rewards realising losses and deferring gains) and short-horizon momentum (which favours holding winners and shedding losers), so selling winners while clinging to losers cannot be rationalised as either. That double perversity is precisely what isolates the reference-point asymmetry as the residual driver rather than a defensible plan.
Not the endowment effect. The endowment effect is an elevated valuation of a good simply because one owns it, with no reference to its price history; the disposition effect is an asymmetry in realisation rates keyed to whether the current price sits above or below the purchase price. Ownership alone is not enough — the effect needs an acquisition anchor and a sign relative to it.
Not the sunk-cost fallacy. Sunk-cost reasoning throws good money after bad — continuing to fund a failing venture because of unrecoverable past investment. The disposition effect is about whether to close a position, driven by the gain/loss domain of the value function and the regret of crystallising a loss, not by escalating commitment to recover prior outlays.
Not a forecast that losers will recover. Holding a loser is not a reasoned bet that the asset will rebound; it is a reluctance to close the open mental account and acknowledge the loss. An investor who held because of a genuine, evidence-backed recovery thesis is making a forward-looking call, not exhibiting the effect — the diagnostic is exactly that the holding survives the absence of any such thesis.
Not the prospect-theory cause itself. PGR > PLR is the measurable symptom — a realisation asymmetry about an anchor, computable from any trading record. The concave-then-convex value function, mental accounting, and regret are the proposed explanation; the statistic documents the gap, it does not by itself prove the cognitive mechanism, which requires the human valuer behind the numbers.

Scope of Application¶

The disposition effect lives across the household- and investor-behaviour subfields of behavioural finance; its reach is bounded to settings with a human owner, a salient purchase-price anchor, and a discretionary hold-or-sell decision, and the looser cross-domain "holds its losers, sheds its winners" analogues belong to the parent primes (loss_aversion, reference_dependence), not here.

Equity-trading research — the home turf: documented in discount-brokerage records (Odean's PGR > PLR), replicated on Finnish (Grinblatt & Keloharju), Israeli, Chinese, and German account data across retail populations.
Real-estate markets — homeowners refuse to sell below what they paid, so the effect underwrites the "lock-in" collapse of housing sales volume as prices fall, coupling transaction volume to the price level.
Mutual-fund and 401(k) behaviour — account-level analyses find the same realisation asymmetry in fund-switching and in the timing of contributions and withdrawals.
Professional and institutional trading — the effect attenuates but does not vanish among seasoned traders (Coval & Shumway), so it serves as a measure of how far experience dampens, rather than erases, the reference-point machinery.
Prospect-theory-based behavioural finance — a workhorse empirical anchor for value-function models of investor behaviour, sitting alongside the equity-premium puzzle and the house-money / break-even effects.

Clarity¶

Naming the disposition effect splits a single observed action — the decision to sell or hold — into two components that behavioural finance must keep apart: the forward-looking sell decision, which a wealth-maximiser should govern by expected return, tax position, and portfolio balance, and the backward-looking contamination introduced by where the current price sits relative to the purchase anchor. Without the label, a portfolio full of clung-to losers and prematurely-dumped winners reads as a sequence of individual judgement calls, each defensible on its own story. With it, the analyst can ask the diagnostic question that the raw trading record otherwise hides: is this investor's reluctance to sell a defensible view about the asset's prospects, or merely an unwillingness to crystallise a loss and close the mental account? The effect also makes the perversity precise — because both the tax code (realise losses, defer gains) and short-horizon momentum push the opposite way, an observed PGR > PLR cannot be rationalised as either tax-smart or return-smart, isolating the reference-point asymmetry as the residual driver.

This in turn sharpens the practitioner's lever. If the distortion is anchored to the salience of the purchase price, then the remedies that bite are precisely those that suppress or remove that anchor — automatic rebalancing, target-date funds, advisor mediation, account displays that hide the acquisition cost — rather than exhortations to "be more rational." Localising the failure to the reference point tells the designer where to intervene, and the PGR/PLR gap supplies a measurable handle on whether the intervention worked.

Manages Complexity¶

Individual investor trading presents behavioural finance with a sprawl of seemingly unrelated puzzles: the tax-inefficiency of clung-to losers, the underperformance of dumped winners, the housing market's "lock-in" where sales volume collapses as prices fall below what owners paid, asymmetric fund-switching in 401(k) accounts, the persistence of a residual effect even among seasoned professionals. The disposition effect compresses this scatter to a single regularity — realisation rates are asymmetric about the purchase-price anchor — and hands the analyst one measurable quantity that captures it: the gap between the proportion of gains realised and the proportion of losses realised (PGR − PLR), computable from any trading record. Rather than re-deriving a separate story for each market, asset class, or investor population, the analyst tracks just this gap and reads off the qualitative outcome — winners shed early, losers held — and reasons about its magnitude from a small set of parameters: how concave-then-convex the value function is about the reference point, how psychologically costly closing the mental account is, and how strongly tax and momentum considerations (which both push the opposite way) fail to override it. The same PGR/PLR instrument transfers across equities, mutual funds, and real estate, so cross-market comparison becomes the reading of one slope rather than the collation of unlike anomalies. The branch structure is correspondingly simple: above the anchor, the concave gain-domain stance favours selling; below it, the convex loss-domain stance favours holding — two cases keyed to the single sign of current-price-minus-purchase-price, collapsing a high-dimensional "why did this investor sell or hold" question into a one-parameter location relative to the acquisition reference.

Abstract Reasoning¶

The disposition effect licenses a set of behavioural-finance inferences, all keyed to the sign of current-price-minus-purchase-price and to one measurable handle, the PGR − PLR gap (proportion of gains realised minus proportion of losses realised).

Diagnostic (decompose a sell/hold into forward-looking and anchor-driven parts). The signature move is to split an observed trading record into the forward-looking decision a wealth-maximiser should make and the backward-looking contamination from the purchase anchor. The analyst reasons FROM "this investor clings to losers and sheds winners" TO "is this a defensible view of the assets' prospects, or an unwillingness to crystallise a loss and close the mental account?" The discriminating evidence is the perversity test: because both the tax code (realise losses, defer gains) and short-horizon momentum push the opposite way, an observed PGR > PLR cannot be rationalised as tax-smart or return-smart, so the analyst reasons FROM "the pattern survives both overrides" TO "the reference-point asymmetry is the residual driver" — isolating the anchor as the cause rather than future-price beliefs.

Predictive (sign relative to the anchor → sell or hold stance). From the location of current price relative to the purchase reference, the framework predicts the realisation stance via the prospect-theory value function. The analyst reasons FROM "the asset is above its purchase price (gain domain, concave, risk-averse)" TO "a preference to lock in the certain gain by selling"; FROM "below the purchase price (loss domain, convex, risk-seeking)" TO "a preference to hold for recovery rather than crystallise the certain loss." The prediction is keyed to a single sign, and it extends to aggregate market behavior: reasoning runs FROM "house prices fall below what owners paid" TO "sales volume collapses as owners refuse to sell at a loss" (the lock-in phenomenon).

Interventionist (suppress the anchor, predict the gap shrinks). Because the distortion is anchored to the salience of the purchase price, the framework predicts the sign of a remedy by whether it removes that anchor. The analyst reasons FROM "automatic rebalancing, target-date funds, advisor mediation, or account displays that hide acquisition cost" TO "the reference point is suppressed, so PGR − PLR narrows"; FROM "mere exhortations to be more rational, leaving the purchase price salient" TO "no effect." The PGR/PLR gap doubles as the measurement of whether the intervention worked — localising the failure to the reference point tells the designer where to act and supplies the metric.

Boundary-drawing (the cognitive substrate, and the measurement transfer). The inferences require an owner with reference-point cognition exhibiting the prospect-theoretic value function, an asset whose purchase price serves as an anchor, a discretionary hold-or-sell decision under uncertainty, and the absence of a fully rational tax/momentum override; the analyst reasons FROM "a fully passive portfolio with no discretionary trading, or a market aggregated without individual-level reference-point cognition" TO "no disposition effect to find." Within that substrate the PGR/PLR instrument is computable from any trading record and transfers across equities, mutual funds, and real estate, so cross-market comparison becomes the reading of one slope. The same boundary marks the concept's composition: stripped of behavioural-finance vocabulary it is the joint action of loss aversion, a reference point, mental accounting, and regret in a financial-decision setting, so those component primes travel to non-financial substrates while the disposition effect as such — the named PGR > PLR regularity and its portfolio remedies — stays within personal finance.

Knowledge Transfer¶

Within economics and finance the disposition effect transfers as mechanism, and the carrier that makes the transfer literal is the PGR/PLR instrument. Wherever there is an owner with a salient purchase-price anchor and a discretionary hold-or-sell decision, the same diagnostic (compute the gap between the proportion of gains realised and the proportion of losses realised), the same prediction (winners shed early, losers clung to), and the same remedies (suppress the acquisition anchor) carry intact — the currency changes but the structure does not. So the template moves without translation across the home domain's subfields: from the founding equity-trading records (Odean's discount-brokerage accounts, Grinblatt and Keloharju's Finnish data) to real-estate markets, where the same reluctance to sell below the purchase price produces the "lock-in" collapse of sales volume as house prices fall; to mutual-fund and 401(k) behaviour, where fund-switching and contribution timing show the same asymmetry; to professional and institutional trading, where the effect attenuates but does not vanish (Coval and Shumway), confirming that experience dampens rather than removes the reference-point machinery. The asset class, the market, and the trader population all vary; the PGR − PLR slope reads the same regularity in each, so cross-market comparison is the reading of one number rather than the collation of unlike anomalies.

Beyond the home domain the transfer splits cleanly into three kinds, and honesty requires keeping them apart. As an instrument, the PGR/PLR statistic transfers literally wherever its precondition holds — any setting with a recorded acquisition price, a population of held positions, and observable realisations admits the gain-realised-versus-loss-realised ratio as a well-defined measurement; nothing about it is metaphorical, and the only boundary to mark is that the statistic measures a realisation asymmetry about an anchor and cannot be over-read as evidence of the prospect-theoretic cause without the cognitive substrate behind it. As a causal mechanism, however, the named effect is substrate-bound: its engine is a reference-point-anchored value function (concave-then-convex), an open-versus-closed mental account, and the regret of crystallising a loss — machinery that presupposes a human (or human-like) valuer holding a discrete asset. Strip that valuer away and the "disposition effect" does not travel; what travels is the more general pattern it instantiates. When the cross-domain lesson is genuinely needed, it should be carried by the parent primes — loss_aversion, reference_dependence, mental_accounting, regret, sunk_cost_and_irreversible_commitment — which really do recur across non-financial substrates as co-instances (a manager refusing to abandon a failing project, a gambler chasing losses, a person staying in a sunk relationship). Those are not "disposition effects"; they are sibling instances of the same parents, and the honest move is to attribute the cross-domain recurrence to the parent mechanism, not to the named financial effect. Finally, as analogy (the loosest and most common cross-domain use), invoking "a disposition effect" for any system that holds onto its failures and lets its successes go renames the components (asset → project/policy, purchase price → original commitment) and borrows the shape of the story while dropping the PGR/PLR measurement and the prospect-theory cause that give the original its predictive force — illuminating by resemblance, but pattern-matching, not the mechanism travelling, and worth marking as such (see Structural Core vs. Domain Accent).

Examples¶

Canonical¶

Terrance Odean's (1998) study "Are Investors Reluctant to Realize Their Losses?" is the defining empirical demonstration. Odean took the trading records of 10,000 accounts at a large U.S. discount brokerage over 1987–1993 and, for each day a sale occurred, counted the winners and losers in the account. He formed two ratios: the Proportion of Gains Realized, PGR = realized gains ÷ (realized gains + paper gains), and the Proportion of Losses Realized, PLR = realized losses ÷ (realized losses + paper losses). Investors realized about 14.8% of their available gains but only about 9.8% of their available losses — PGR ≈ 0.148 versus PLR ≈ 0.098, a gap far too large to be chance. Crucially the asymmetry reversed only in December, when tax-loss selling briefly dominated, confirming the pattern runs against, not with, the tax incentive.

Mapped back: Each stock's cost basis is the purchase-price reference point, and whether the day's price sits above or below it is the current-price location. Selling winners at 14.8% reflects the gain-domain risk-averse stance; holding losers, realizing only 9.8%, reflects the loss-domain risk-seeking stance. That the gap holds outside December — despite the tax code rewarding the opposite — is the perversity condition, and PGR ≈ 0.148 > PLR ≈ 0.098 is the realisation-asymmetry observable itself.

Applied / In Practice¶

David Genesove and Christopher Mayer (2001), studying the Boston downtown condominium market through the 1990s price bust, showed the effect governing real-estate behavior. Owners who had bought near the late-1980s peak and now faced a nominal loss behaved sharply differently from owners sitting on gains: the loss-facing sellers set higher asking prices, were less willing to cut them, and consequently sat on the market far longer and transacted less often. Genesove and Mayer estimated that sellers expecting a nominal loss set list prices that incorporated a substantial fraction of that shortfall — refusing to price at market and thereby suppressing their own sale probability. This links individual reference-point reluctance to the aggregate "lock-in" collapse of housing transaction volume in falling markets.

Mapped back: The original condo purchase price is the purchase-price reference point, and the post-bust market value below it fixes the current-price location in the loss domain. The owners' refusal to sell at a loss — holding for recovery rather than crystallising it — is the loss-domain risk-seeking stance driven by the open-versus-closed mental account. The observed gap in sale hazards and pricing between loss-facing and gain-facing owners is the realisation-asymmetry observable expressed as volume, and it persists against the return-smart move of selling, satisfying the perversity condition.

Structural Tensions¶

T1: Measured symptom versus prospect-theory cause (what the PGR/PLR gap does and does not prove). The disposition effect's evidential force rests on a statistic — PGR > PLR — that is computable from any trading record without asking the investor anything, and that literal transferability is its great virtue: no interviews, no self-report, just realised-versus-available ratios read off account data across equities, funds, and housing. But the same detachment that makes the gap portable also strips it of its explanation. The number documents a realisation asymmetry about an anchor; it does not by itself show that a concave-then-convex value function, an open mental account, or regret is producing it — any process that shed winners and clung to losers would leave the same fingerprint. The instrument travels precisely because it is agnostic about the cognition, which is exactly why it cannot, alone, confirm the cognition. Diagnostic: Is the PGR/PLR gap being read as a measurement of a realisation asymmetry, or over-read as direct evidence of the prospect-theoretic mechanism behind it?

T2: The perversity condition as clean isolator versus its own regime-dependence (December breaks the test). What makes the diagnosis unusually clean is that both the tax code and short-horizon momentum push the opposite way, so an observed PGR > PLR cannot be rationalised as tax-smart or return-smart — the reference-point asymmetry is left as the residual driver. Yet the very condition that isolates the cause is not permanent: Odean's own data show the asymmetry reversing in December, when tax-loss selling briefly dominates and the perversity flips. The isolating test therefore holds only in the regime where the rational overrides stay weak; where they strengthen, the same investor's record stops satisfying the condition and the effect goes measurably quiet — not because the machinery vanished but because a stronger incentive temporarily out-pulled it. Diagnostic: In the window being examined, are the tax and momentum incentives actually pushing against the observed pattern, or has a seasonal override (like year-end tax selling) suspended the perversity test?

T3: A single fixed anchor versus a migrating reference point (the one-parameter compression's load-bearing assumption). The framework's power is that it collapses a high-dimensional sell/hold question into one sign — current price minus purchase price — and reads the stance off that alone. That compression buys its simplicity by fixing the reference point at the acquisition cost. But an investor's psychological anchor need not stay there: a position that once peaked well above cost can re-anchor to its high-water mark, so a "winner" now below its peak is felt as a loss though still above purchase, and the gain/loss sign the model keys on silently changes. The concave-then-convex prediction is only as good as the claim that the purchase price is the operative anchor; where the reference migrates, the same PGR/PLR reading can misclassify the very domain it is meant to diagnose. Diagnostic: Is the investor reckoning gain and loss against the purchase price the model assumes, or against a different, drifted anchor (peak value, break-even target, expected return)?

T4: Reluctance to close versus a genuine recovery thesis (the diagnostic hinges on something the record cannot show). The effect is defined as holding a loser out of unwillingness to close the mental account — not a reasoned bet that the asset will rebound; an investor with a real, evidence-backed recovery thesis is making a forward-looking call and is exempt. But the trading record shows only the holding, not the reason for it, and both the biased clinger and the thesis-driven holder leave the identical footprint of an unrealised loss carried forward. The concept's boundary thus runs through exactly the variable the data omit, and the PGR/PLR statistic — powerful because it needs no self-report — cannot on its own distinguish the bias it names from the rational patience it excludes. The cleaner the instrument, the blinder it is to this particular seam. Diagnostic: Is there an articulable, evidence-backed reason this loser is still held, or only the absence of one — a reluctance to crystallise the loss dressed as patience?

T5: Anchor-suppressing remedies versus the information the anchor legitimately carries (removing the cost basis cuts both ways). Because the distortion is anchored to the salience of the purchase price, the remedies that bite are those that suppress it — automatic rebalancing, target-date funds, displays that hide acquisition cost. Suppressing the anchor does narrow PGR − PLR, but the purchase price is not pure noise: it is precisely the quantity a tax-loss-harvesting or capital-gains calculation legitimately needs, and it is where the perversity test itself is read. A display that hides cost basis to defeat the reference-point pull also hides the input that makes a genuinely tax-smart realisation possible, and erases the very anchor an analyst uses to compute the gap. The remedy trades a behavioural distortion for a loss of decision-relevant information, and the more thoroughly it hides the anchor, the more it disables the rational uses of the same number. Diagnostic: Does suppressing the purchase-price anchor remove only its distorting pull, or also the information a tax-aware or return-aware decision would rightly use?

T6: Experience dampens versus experience erases (is the residual effect a correctable error or a structural feature). Among seasoned and institutional traders the effect attenuates but does not vanish (Coval and Shumway), and that persistence carries two opposed readings. If experience merely dampens, the reference-point machinery is a structural feature of a human valuer that sophistication can suppress but not remove, and the residual PGR − PLR among professionals is the real, irreducible signal. If experience is on a path toward erasing it, the residual is transitional — noise from incomplete learning — and the "true" disposition effect belongs to naïve retail investors. The measurement cannot settle which, because a small persistent gap is consistent with both a floor that learning asymptotes to and a slope still descending. How one reads the professional residual decides whether the effect is a bias to be trained away or a substrate property to be designed around. Diagnostic: Is the residual gap among experienced traders a stable floor the machinery imposes, or a still-shrinking remnant that further learning would close?

T7: Autonomy versus reduction (its own named regularity or the finance instance of its parent primes). "Disposition effect" is a named, canonically measured behavioural-finance regularity with its own instrument (PGR > PLR) and its own founding records (Shefrin–Statman, Odean, Genesove–Mayer). Yet its cross-domain cargo is not proprietary: strip the vocabulary and it is the joint action of loss_aversion, reference_dependence, mental_accounting, regret, and sunk_cost_and_irreversible_commitment in a financial-holding setting. A manager refusing to abandon a failing project, a gambler chasing losses, a person staying in a sunk relationship are not "disposition effects" but sibling instances of the same parents, and it is those parents — not the named financial effect — that genuinely recur off the trading floor. Within personal finance the named regularity and its PGR/PLR handle earn their own study; beyond it, invoking "a disposition effect" borrows the finance label for what is really the parents travelling. Diagnostic: Resolve toward the parent primes (loss aversion, reference dependence, mental accounting, regret) when the lesson must travel outside asset markets; toward the named disposition effect when diagnosing a specific investor's realisation record in situ.

Structural–Framed Character¶

The disposition effect sits at mixed on the structural–framed spectrum — a genuine cognitive regularity (prospect-theoretic reference dependence) at its core, but pinned to financial-holding decisions and carrying a bias/perversity charge. On evaluative_weight it leans mildly framed: the effect is a described empirical regularity, but it is treated as a distortion — the "perversity condition," the remedies that "bite," the implicit judgment that selling winners and clinging to losers is a mistake against tax and momentum — so the concept carries a corrective, bias-diagnosing valence heavier than an evaluatively silent mechanism-name. On human_practice_bound it is framed: the entry's own cognitive-substrate constraint says the effect "requires an owner with prospect-theoretic reference-point cognition and a discretionary hold-or-sell decision; absent that valuer there is no effect to find" — it runs only in a human (or human-like) mind making financial decisions, not observer-free in nature. Institutional_origin points the same way: the purchase-price anchor, realisation rates, and the tax/momentum overrides are features of financial-market institutions, and the named regularity itself is behavioural-finance furniture (Shefrin–Statman, Odean). Vocab_travels fails at the effect-specific layer — the PGR/PLR statistic and the purchase-price anchor are pinned to asset markets, and "a disposition effect" applied to a failing project is analogy that renames the components. Import_vs_recognize is bimodal: within finance the mechanism is recognized intact across equities, funds, and real estate (the PGR/PLR instrument reads one slope everywhere), while beyond asset markets the recurrence is carried by the parent primes as sibling instances, not by the named effect.

The portable structural skeleton is not proprietary and is not a single prime: stripped of behavioural-finance vocabulary the disposition effect is the joint action of a reference-anchored, concave-then-convex value function, an open-versus-closed mental account, and the regret of crystallising a loss — exactly what it instantiates from its umbrella primes loss_aversion, reference_dependence, mental_accounting, regret, and sunk_cost_and_irreversible_commitment. Those parents carry the "holds its losers, sheds its winners" pattern cross-domain as genuine co-instances (a manager clinging to a failing project, a gambler chasing losses); the PGR/PLR instrument, the purchase-price anchor, and the portfolio remedies are the domain accent that stays home and keeps the entry domain-specific. The cross-domain reach belongs to the loss-aversion/reference-dependence parents, not to "the disposition effect." Its character: a genuinely structural prospect-theoretic cognitive regularity dressed in financial-market vocabulary and framed as a correctable bias, structural in the reference-dependence-plus-loss-aversion skeleton it instantiates and held at mixed by the human valuer and asset-market institutions that constitute it.

Structural Core vs. Domain Accent¶

This section decides why the disposition effect is a domain-specific abstraction and not a prime, and it carries the case for its domain-specificity — there is no separate section for that.

What is skeletal (could lift toward a cross-domain prime). Strip the asset markets and a thin relational structure survives: outcomes are reckoned against a reference point rather than absolutely, the value function is concave (risk-averse) in the gain domain and convex (risk-seeking) in the loss domain, an open commitment resists the closure that would force acknowledgement of failure, and the regret of crystallising a loss deters that closure. The portable pieces are abstract — a reference point, an asymmetric loss-averse valuation about it, an open-versus-closed account, and regret at closing. That structure is genuinely substrate-portable, which is exactly why the disposition effect is best read not as a single parent but as the joint action of catalog primes it instantiates: loss_aversion and reference_dependence (the concave-then-convex value function about an anchor), mental_accounting (the open-versus-closed position), regret (the cost of crystallising a loss), and sunk_cost_and_irreversible_commitment (the pull to not abandon). This reference-dependence-plus-loss-aversion complex is the core the disposition effect shares — and it is what genuinely recurs, as sibling instances not metaphors, in a manager clinging to a failing project, a gambler chasing losses, a person staying in a sunk relationship — but it is not what makes the disposition effect distinctive.

What is domain-bound. Almost all the distinctive content is behavioural-finance furniture and none of it survives extraction intact: the purchase-price reference point as the acquisition anchor; the PGR > PLR realisation-asymmetry observable (proportion of gains realised exceeding proportion of losses realised) computable from a trading record; the perversity condition that both tax treatment and short-horizon momentum push the opposite way, isolating the reference-point asymmetry; the sell/hold decomposition into forward-looking and anchor-driven parts; and the portfolio remedies (automatic rebalancing, target-date funds, hiding acquisition cost). These are the worked instrument, the diagnostic, and the canonical cases (Odean's PGR ≈ 0.148 versus PLR ≈ 0.098 across 10,000 brokerage accounts; Genesove–Mayer's Boston condo lock-in) the discipline actually studies — all specific to an owner with a salient purchase-price anchor and a discretionary hold-or-sell decision. The decisive test: remove the human valuer and the asset-holding decision — a fully passive portfolio, or a market aggregated without individual reference-point cognition — and there is no effect to find; the PGR/PLR statistic and the purchase-price anchor have no referent outside asset markets, and "a disposition effect" applied to a failing project renames the components while dropping the measurement and the prospect-theory cause, which is the tell that what crosses is the parent complex, not this named regularity.

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. The disposition effect's transfer is bimodal. Within economics and finance it travels intact as mechanism, carried by the PGR/PLR instrument: the same diagnostic (compute the realisation gap), prediction (winners shed early, losers clung to), and remedies (suppress the anchor) hold across equity trading, real-estate lock-in, mutual-fund and 401(k) behaviour, and professional trading, the PGR − PLR slope reading one regularity in each. Beyond asset markets the transfer splits: the PGR/PLR statistic transfers literally wherever a recorded acquisition price and observable realisations exist (but measures only a realisation asymmetry, never the cause), while the named causal effect is substrate-bound to a human valuer holding a discrete asset — invoking "a disposition effect" for a failing project or a sunk relationship is analogy that borrows the shape. And when that bare structural lesson is needed cross-domain — a system holds onto its losses and lets its gains go because outcomes are judged against a reference point under loss aversion — it is already carried, in more general form, by the primes the disposition effect instantiates: the anchored value function is loss_aversion / reference_dependence, the open position is mental_accounting, the deterrent to closing is regret, and the pull to not abandon is sunk_cost_and_irreversible_commitment. The cross-domain reach belongs to those parents, which recur off the trading floor as genuine sibling instances; "the disposition effect," as named, carries the PGR/PLR instrument, the purchase-price anchor, and the portfolio remedies that stay home in asset markets and should.

Relationships to Other Abstractions¶

Current abstraction Disposition Effect Domain-specific

Parents (2) — more general patterns this builds on

Disposition Effect is part of Frame of Reference Prime

The Disposition Effect contains a frame of reference centered on the purchase price that classifies an unrealized position as a gain or a loss.
Disposition Effect is part of, typical Loss Aversion Prime

The canonical behavioral-finance account usually contains loss aversion around the purchase-price anchor, while the observed realization gap alone does not uniquely establish that mechanism.

Hierarchy paths (4) — routes to 4 parentless roots

Disposition Effect → Frame of Reference → Viewpoint

Show alternative paths (3)

Not to Be Confused With¶

Endowment effect. An elevated valuation of a good simply because one owns it, with no reference to price history. The disposition effect is an asymmetry in realisation rates keyed to whether the current price sits above or below the purchase price — ownership alone is not enough; it needs an acquisition anchor and a sign relative to it. Tell: does the bias arise from mere ownership (endowment effect), or from the gap between current price and what was paid (disposition effect)?
Sunk-cost fallacy. Throwing good money after bad — continuing to fund a failing venture because of unrecoverable past investment. The disposition effect concerns whether to close a position, driven by the gain/loss domain of the value function and the regret of crystallising a loss, not by escalating new commitment to recover prior outlays. Tell: is the decision to invest more to justify past spending (sunk cost), or to hold rather than sell to avoid booking a loss (disposition effect)?
House-money and break-even effects. Sibling prospect-theory trading regularities: the house-money effect is increased risk-taking after gains (playing with "found" money); the break-even effect is increased risk-taking when down, to get back to even. The disposition effect is specifically the realisation asymmetry (sell winners, hold losers) about the purchase anchor. They share the reference-point machinery but describe different behaviors. Tell: is the pattern about risk appetite shifting after wins/losses (house-money/break-even), or about the rate of selling winners versus losers (disposition effect)?
Anchoring bias. The general cognitive tendency to rely on an initial reference value when estimating a quantity (a first price offer anchoring a negotiation). The purchase price is an anchor, but the disposition effect is the specific loss-averse realisation asymmetry it produces in hold-or-sell decisions, not the broad estimation bias. Tell: is the referent the general pull of a reference value on a judgment (anchoring), or the specific sell-winners/hold-losers pattern about the purchase price (disposition effect)?
The parent primes it instances (loss_aversion, reference_dependence, mental_accounting, regret, sunk_cost). The substrate-neutral machinery — a reference-anchored, concave-then-convex value function, an open-versus-closed account, and regret at crystallising a loss. These recur off the trading floor as genuine sibling instances (a manager clinging to a failing project, a gambler chasing losses); the disposition effect is the asset-market instance with a PGR/PLR handle. Tell: strip the human valuer and the asset-holding decision and what remains — holding losses and shedding gains under loss aversion — is these parents, not the disposition effect. (Treated more fully in earlier sections.)

Neighborhood in Abstraction Space¶

Disposition Effect sits in a crowded region of the domain-specific corpus (7^th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Financial Markets & Valuation Models (11 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12