Disappointment aversion
Outcomes below what you expected sting extra, so you avoid bets that might let you down.
What it means
Disappointment aversion holds that people evaluate risky outcomes relative to a prior expectation — typically the certainty equivalent of the gamble itself — and weight outcomes that fall short of that expectation more heavily than elations above it. Unlike regret theory, the comparison is to one's own expected payoff rather than to a foregone alternative, so no information about other options is needed. The asymmetry generates first-order risk aversion even for small stakes and helps explain the certainty effect and preference for less dispersed payoffs. Models in this family endogenize the reference point as the expectation, linking them to reference-dependent preferences more broadly. It is a parsimonious way to capture why a gamble that could disappoint feels worse than its arithmetic suggests.
Where the reference point comes from
The device that makes the theory work is that the yardstick is not exogenous. In Gul's formulation the reference point is the certainty equivalent of the very gamble being faced, so the model has to be solved for a fixed point: the value that outcomes are compared against is itself the value of the prospect. Outcomes above that threshold count as elations, outcomes below it as disappointments, and a single parameter (Gul's beta) scales how much more the disappointments weigh. When beta is zero the model collapses to expected utility; positive beta produces a kink at the reference point. Because the kink sits at the expectation rather than at zero or the status quo, the same objective gamble can feel better or worse depending on what it led you to expect.
First-order risk aversion
The practical bite of the kink is first-order risk aversion. Under smooth expected utility the premium a person pays to avoid a small symmetric bet shrinks roughly with the square of the stake, so tiny gambles should be taken at nearly fair odds. Disappointment aversion instead makes the premium shrink only in proportion to the stake, so it stays economically visible even for small amounts. That is why the theory can rationalize turning down modestly favorable coin-flips, buying insurance at actuarially unfair prices, and the general distaste for dispersed payoffs, without invoking implausibly high curvature in the utility of wealth.
Why asset markets care
The largest payoff of the idea has been in finance, where standard models struggle to explain why stocks earn so much more than bonds. Ang, Bekaert, and Liu showed that a disappointment-averse investor, penalizing returns that fall short of the portfolio's own certainty equivalent, will hold far less equity and demand a larger premium, matching the historical gap even over long horizons. Routledge and Zin's generalized version lets the disappointment threshold sit at a fraction of the continuation value, which makes the pricing kernel more countercyclical and pushes both the mean and the volatility of the equity premium toward the data. Downside sensitivity, not raw risk aversion, does the work.
What the evidence shows
Direct measurement is thinner than the theory's popularity suggests. Choi, Fisman, Gale, and Kariv had subjects allocate money across states in graphical budget-line problems and fit a two-parameter form nesting Gul's model; they recovered substantial and highly heterogeneous disappointment-aversion coefficients across individuals, alongside conventional risk aversion. Abdellaoui and Bleichrodt elicited the model with the tradeoff method and likewise found aversion in the aggregate but wide dispersion, with a nontrivial share of people looking elation-seeking rather than disappointment-averse. The consistent finding is that the parameter is real but far from uniform, so a single population estimate hides more than it reveals.
Related but distinct
Keep three neighbors apart. Regret theory compares your outcome to what a foregone alternative would have paid, so it needs information about the option not chosen; disappointment aversion compares only to your own expectation and needs nothing external. Loss aversion also weights losses more than gains, but its reference point is usually the status quo or a purchase price, not the expected value of the gamble in hand. Prospect theory bends probabilities through a weighting function and can mimic some of the same choices; disappointment models get first-order risk aversion straight from the reference kink instead. The distinctions matter because each predicts a different way to soften the sting.
Examples
Promised a likely big win, you feel cheated by a merely average result — even though that result, judged on its own, is perfectly good.
A sales rep takes the flat salary over the commission plan that pays more on average, because most commission months would land below the number she is already counting on.
A patient picks the operation with a certain modest improvement over one that averages better but usually falls short of its headline promise. Anything under that promise would feel like failure.
A homeowner buys the extended warranty even though it costs more than the repairs it covers on average, because an appliance that fails far sooner than she expected would sting more than a premium she has already budgeted.
A studio greenlights a safe sequel with a loyal audience over an original script that projects higher on average, reasoning that a flop after high hopes hurts worse than a predictable, unremarkable return.
First described in Bell (1985); Loomes & Sugden (1986); Gul (1991).
Key references
- Routledge, B. R., & Zin, S. E. (2010). Generalized Disappointment Aversion and Asset Prices. The Journal of Finance, 65(4), 1303-1332. onlinelibrary.wiley.com/doi/abs/10.1111/j.1540-6261.2010.01571.x
- Choi, S., Fisman, R., Gale, D., & Kariv, S. (2007). Consistency and Heterogeneity of Individual Behavior under Uncertainty. American Economic Review, 97(5), 1921-1938. www.aeaweb.org/articles?id=10.1257/aer.97.5.1921
- Abdellaoui, M., & Bleichrodt, H. (2007). Eliciting Gul's theory of disappointment aversion by the tradeoff method. Journal of Economic Psychology, 28(6), 631-645. www.sciencedirect.com/science/article/abs/pii/S0167487007000670
- Ang, A., Bekaert, G., & Liu, J. (2005). Why stocks may disappoint. Journal of Financial Economics, 76(3), 471-508. www.sciencedirect.com/science/article/abs/pii/S0304405X05000164
- Gul, F. (1991). A Theory of Disappointment Aversion. Econometrica, 59(3), 667-686. www.econometricsociety.org/publications/econometrica/1991/05/01/theory-disappointment-aversion
- Bell, D. E. (1985). Disappointment in Decision Making under Uncertainty. Operations Research, 33(1), 1-27. pubsonline.informs.org/doi/abs/10.1287/opre.33.1.1