Understanding Expected Utility Theory

Expected utility bridges probability and personal preference. When you face a choice between outcomes you cannot guarantee, expected utility assigns a numerical score reflecting both the likelihood of success and how much satisfaction (or pain) each result delivers to you specifically.

This framework rests on a simple insight: $1 million means less to a billionaire than to someone with $10,000. Similarly, losing $1,000 stings more than gaining $1,000 feels good—a phenomenon called risk aversion. Expected utility captures this asymmetry by converting monetary amounts into utility scores before multiplying by probabilities.

The theory assumes rational decision-making. When choosing between two uncertain outcomes, the option with higher expected utility aligns with your preferences and circumstances, even if it has lower expected monetary value.

Expected Utility Formula

To find the expected utility of two competing scenarios, multiply each outcome's probability by the square root of its monetary value, then sum the results. The square root term models diminishing marginal utility—each additional dollar matters less as your wealth grows.

Expected Utility = (P₁ × √V₁) + (P₂ × √V₂)

  • P₁ — Probability of outcome 1, expressed as a decimal (e.g. 0.4 for 40%)
  • V₁ — Monetary value of outcome 1 in dollars
  • P₂ — Probability of outcome 2, expressed as a decimal
  • V₂ — Monetary value of outcome 2 in dollars

Applying Expected Utility to Real Decisions

Portfolio managers rely on expected utility to balance risk and return. A conservative investor and an aggressive trader facing identical investment opportunities will reach different conclusions because their utility functions differ. The conservative investor experiences greater utility loss from downside risk, shifting the expected utility calculation in favour of safer alternatives.

Business leaders use expected utility when evaluating capital projects. A manufacturing expansion might offer a 60% chance of $2 million gain and a 40% chance of $500,000 loss. Computing expected utility—not just expected monetary value—reveals whether the venture aligns with the company's risk appetite and strategic position.

Insurance and gambling illustrate the reverse: insurers sell policies where expected monetary value favours the customer, yet the policyholder's expected utility improves because they eliminate catastrophic downside. The maths work when both parties understand their own utility curves.

Common Pitfalls in Expected Utility Analysis

Avoid these mistakes when computing or interpreting expected utility:

  1. Confusing utility with monetary value — A $10,000 gain does not equal a 10,000-unit increase in happiness for everyone. Your utility function is personal. Two people with identical financial outcomes can have vastly different expected utilities depending on their wealth, goals, and risk tolerance. Use the same utility model across all scenarios you compare.
  2. Neglecting probability accuracy — Expected utility is only as reliable as your probability estimates. If you guess a 50% chance of success when historical data suggests 30%, your calculation misleads. Base probabilities on empirical evidence—past project outcomes, market data, or expert consensus—not intuition or optimism bias.
  3. Treating utility functions as universal — The square-root formula used here assumes a standard concave utility function representing diminishing marginal utility. Your actual utility curve may differ. Extreme wealth or losses can flatten the curve. Always sense-check whether the mathematical model fits your real preferences before acting on the result.
  4. Ignoring non-financial factors — Expected utility in finance focuses on monetary outcomes, but real decisions involve time, effort, stress, and opportunity cost. A low-utility investment might still be worth pursuing if it frees you to pursue higher-impact work, or vice versa. Use the calculator as one input in a broader decision framework.

When Expected Utility Diverges from Expected Value

A bet with a 50% chance of winning $100 has an expected monetary value of $50. Its expected utility (using the square-root model) is approximately 0.5 × 10 = 5 utils. Meanwhile, a guarantee of $25 has expected value $25 but utility √25 = 5—identical utility despite lower expected value.

This explains why people buy insurance. A house fire is unlikely (low probability) but catastrophic (high negative utility). Insurance's expected value is negative for the policyholder; its expected utility is sharply positive because it eliminates tail risk. Conversely, lottery tickets have terrible expected value and utility, yet appeal through the fantasy of a rare jackpot.

Expected utility theory predicts behaviour more accurately than expected value alone in real-world decisions. It explains why billionaires decline high-variance bets that poor people might accept, and why loss aversion is so powerful: losses loom larger in our utility function than equivalent gains.

Frequently Asked Questions

How is expected utility different from expected value?

Expected value multiplies each outcome by its probability and sums the results—a straightforward average. Expected utility does the same but first converts monetary amounts into utility scores, typically using a concave function (like a square root) that reflects diminishing marginal satisfaction. Expected value = 0.5 × $100 + 0.5 × $0 = $50. Expected utility might equal 0.5 × √100 + 0.5 × √0 ≈ 5, capturing the reality that gaining $100 when you have nothing matters more than losing $100 when you have substantial wealth.

Can I have negative expected utility?

Yes. If both outcomes carry undesirable consequences, or if you assign negative utility values to losses or unfavourable scenarios, expected utility can be negative. For example, a business facing a 70% chance of losing $50,000 and a 30% chance of losing $100,000 has negative expected utility. Negative expected utility signals that all available choices carry net harm relative to your baseline. In such situations, the rational move is often to minimize expected utility loss rather than maximize gain.

Why does the formula use a square root?

The square root captures diminishing marginal utility—the principle that each additional dollar adds less happiness as your wealth grows. A square-root utility function reflects realistic human preferences: gaining $1,000 when you have $5,000 is far more impactful than gaining $1,000 when you have $1 million. This mathematical form is one of several possible utility functions; others (logarithmic, exponential) may better suit specific contexts. The square root is a reasonable default for financial scenarios where outcomes span moderate ranges.

How does risk aversion change my expected utility calculation?

Risk aversion shapes your personal utility function. A highly risk-averse person experiences sharp utility drops from potential losses and mild utility gains from equivalent wins. This translates into a steep, concave utility curve. In contrast, a risk-seeking individual exhibits a flatter, convex curve—they get more excitement from gambling despite worse expected outcomes. The square-root formula assumes moderate risk aversion. If you are extremely risk-averse, your true utility curve bends more sharply, meaning you should weight downside scenarios even more heavily in your decisions.

Should I always choose the option with the highest expected utility?

Expected utility is a powerful decision-making aid, but not a law of nature. It formalizes your preferences and risk tolerance, revealing which choice aligns best with your values and circumstances. However, it assumes your probability estimates are accurate and that non-financial factors (time, relationships, health, learning) are secondary. Use expected utility to structure your thinking, challenge your assumptions, and avoid emotional biases—but ultimately, integrate its insights with qualitative judgment about your broader life and business goals.

What if I only have one uncertain outcome, not two?

Extend the two-outcome formula to include all scenarios. If you face three possible outcomes with probabilities P₁, P₂, P₃ and values V₁, V₂, V₃, compute expected utility as (P₁ × √V₁) + (P₂ × √V₂) + (P₃ × √V₃). The principle remains: weight each possibility by its likelihood and utility, then aggregate. For a single outcome with certainty, probability is 1.0 and expected utility simply equals the utility (square root of the value). More outcomes create richer decision models but also demand higher-quality probability estimates.

More finance calculators (see all)