Loss aversion

Definition

Loss aversion is the empirical regularity that losses hurt roughly twice as much as equivalent gains satisfy — the central asymmetry of prospect theory (Kahneman-Tversky 1979), measured across decades of experiments at a loss/gain sensitivity ratio near 2:1. In markets it drives the disposition effect, panic liquidation at drawdown lows, under-allocation to volatile assets checked too often (Benartzi-Thaler’s myopic loss aversion), and the systematic purchase of overpriced downside protection.

How it works / structure

• The prospect-theory machinery: value is felt over CHANGES from a reference point, not levels; the value function is concave for gains, convex for losses, and ~2x steeper on the loss side — small frequent losses feel worse than their aggregate warrants, and a position “down 10%” is a different psychological object than the same portfolio arrived at differently.
• Myopic loss aversion (the horizon interaction): Benartzi-Thaler (1995) — the more often an investor checks a volatile asset, the more loss-events they experience, and the less of it they hold; equity’s daily loss frequency (~47% of days) vs its annual (~25%) makes checking frequency an allocation input — their model fit the equity premium with annual evaluation.
• Platform counters (engine-executable): evaluation- window design (positions graded at thesis horizons, not tick-by-tick — the UI’s reporting cadence is a behavioral parameter), drawdown budgets set in COLD state (risk-max-drawdown-budget — the entry’s premise is that live pain overrides plans unless the plan is mechanical), and sizing at levels survivable without override (risk-fixed-fractional — the behavioral case for small f is exactly this bias).
• The pricing footprint: loss aversion is a candidate explanation for the volatility risk premium and expensive OTM puts (indicator-realized-vs-implied-vol — insurance overpricing as preference, not error).

When it applies

Sizing and drawdown-budget design (the bias defines the override risk those entries manage); UI/reporting cadence decisions; interpreting user overrides at drawdown lows (the documented capitulation pattern); understanding why systematically selling insurance has carried premium (strategy-iron-condor economics have a behavioral leg).

Risk profile & failure modes

• Plan override at the low: the bias’s signature market damage — liquidating at maximum drawdown, converting a planned-for excursion into a realized worst case; the counter is mechanical execution, not resolve.
• Checking-frequency spiral: drawdowns increase monitoring, which increases experienced loss events, which increases the urge to act — myopia compounding itself.
• Protection overpayment: persistent long-put programs purchased for comfort carry documented negative expectancy (the insurance premium) — comfort is the product, and it is priced.
• Misreading the ratio as universal: the 2:1 figure is a population average from experimental contexts; individual and contextual variance is wide — diagnostics over assumptions.

Evidence & limits

Prospect theory is among the most replicated results in behavioral science (1979 paper; Nobel 2002); myopic loss aversion has experimental and field support. Its asset- pricing implications (equity premium, insurance pricing) are credible candidate explanations, not settled attributions — the platform cites them as interpretations, labeled.

Falsifiable-thesis examples

Illustrations only, not signals:

• “Reducing this account’s mark-to-market display frequency from daily to weekly will reduce its mid-drawdown override rate” — falsified by the before/after override log.
• “This account’s realized behavior implies a loss/gain sensitivity above 1.5 (fitted from exit asymmetries)” — falsified by the fitted parameter.

Cross-references

• The market expression: bias-disposition-effect
• The structural defenses: risk-max-drawdown-budget, risk-fixed-fractional, port-allocation-frameworks (evaluation windows)
• The pricing footprint: indicator-realized-vs-implied-vol (insurance premium)
• The horizon interaction: strategy-buy-and-hold (whose evidence assumes holding through the loss days)