Economic Modelling · Rational choice theory and utility
Risk Aversion and Insurance Demand: Utility, Certainty Equivalent and Risk Premium
Updated 11 October 2026 · Fact-checked
A risk-averse person has concave utility, so E[U(W)] < U(E[W]). The certainty equivalent is the sure amount giving the same utility as the gamble. The risk premium is E[W] minus that certainty equivalent. Risk-averse people pay for insurance because a sure wealth gives more utility than a risky one with equal mean.
Understand Risk Aversion and Insurance Demand
Utility measures the satisfaction a person gets from wealth. In expected utility theory, you compare risky choices by their expected utility, E[U(W)], not by expected wealth.
The shape of U shows the attitude to risk. If U is concave (U''(w) < 0), the person is risk-averse: each extra rupee adds less utility than the one before. If U is linear (U'' = 0), the person is risk-neutral and only cares about the expected value. If U is convex (U'' > 0), the person is risk-seeking and likes gambles.
For a concave U, Jensen's inequality gives E[U(W)] < U(E[W]) for any non-degenerate random W. So a sure amount equal to the mean is preferred to the gamble. This is the root of risk aversion.
The certainty equivalent (CE) is the sure wealth that gives the same utility as the gamble: U(CE) = E[U(W)]. The risk premium is the amount of expected wealth the person gives up to avoid the risk: π = E[W] − CE. For a risk-averse person, π > 0.
This explains insurance demand. A person with wealth W facing a random loss will pay a premium up to the point where utility with insurance equals expected utility without it. If the insurer charges the expected loss (a fair premium), a risk-averse person strictly prefers full cover. The insurer can charge more than the fair premium, up to the risk premium, and the person still buys. The Arrow-Pratt measures quantify how strongly someone is averse to risk.
Key rules to remember
- Expected utility
- E[U(W)] = Σ p_i × U(w_i)
- Use the probabilities of each wealth outcome. Wealth is final wealth, not just the gain or loss.
- Risk attitude by curvature
- U''(w) < 0: risk-averse; U''(w) = 0: risk-neutral; U''(w) > 0: risk-seeking
- Assume U'(w) > 0 (more wealth is preferred).
- Jensen's inequality for risk aversion
- E[U(W)] < U(E[W]) for concave U and non-constant W
- Reverses for convex U. Equality holds for linear U.
- Certainty equivalent
- U(CE) = E[U(W)]
- Solve by inverting U. For U = ln w, CE = exp(E[ln W]).
- Risk premium
- π = E[W] − CE
- Positive for risk-averse, zero for risk-neutral, negative for risk-seeking.
- Maximum insurance premium
- U(w − P) = E[U(w − X)], so P_max = w − CE = E[X] + π
- X is the random loss and w is initial wealth. P_max exceeds E[X] for a risk-averse person.
- Absolute risk aversion (Arrow-Pratt)
- A(w) = −U''(w) ÷ U'(w)
- Larger A means more risk-averse at wealth w.
- Relative risk aversion
- R(w) = w × A(w) = −w U''(w) ÷ U'(w)
- Measures aversion to proportional risk.
- Approximate risk premium
- π ≈ ½ × A(w) × Var(W)
- An approximation for small risks around mean wealth w. Not exact.
How to solve Risk Aversion and Insurance Demand questions
Use this method for any question on risk attitude, certainty equivalent, risk premium or insurance demand.
- 1Write down the utility function U(w) and check U'(w) > 0.
- 2Find U''(w). Its sign gives the risk attitude. Compute A(w) = −U''/U' if asked.
- 3List the possible final wealth outcomes and their probabilities. Include initial wealth in each outcome.
- 4Compute E[W] and E[U(W)] = Σ p × U(w).
- 5Find the certainty equivalent by solving U(CE) = E[U(W)].
- 6Compute the risk premium π = E[W] − CE.
- 7For insurance, set U(w − P) = E[U(W without insurance)] and solve for the maximum premium P. Compare it with the actual premium and the expected loss.
- 8State the conclusion in words: buy or not, and why.
Quickest way: Shortcut with inverse utility
When to use it: Use when U is simple (log, square root, exponential) and the gamble has two or three outcomes.
- Write the inverse function of U first, for example U = √w gives w = U².
- Compute E[U(W)] once, using final wealth values.
- Apply the inverse to get CE directly.
- Subtract CE from E[W] for the risk premium.
- Check the sign: for a concave U, CE must be below E[W]. If not, you made an error.
Common mistakes in Risk Aversion and Insurance Demand
Taking U(E[W]) as the expected utility.
The notation looks similar and averaging wealth first feels natural.
Fix: Compute U for each outcome first, then weight by probability. Remember E[U(W)] ≠ U(E[W]) unless U is linear.
Confusing certainty equivalent with risk premium.
Both come from the same calculation and both are amounts of money.
Fix: CE is a level of sure wealth. The risk premium is E[W] − CE, the amount given up. Always label which one you have found.
Using the gain or loss instead of final wealth in U.
The question lists a gamble of ±₹ amounts.
Fix: Add the gain or loss to initial wealth before applying U. For log or square root utility this changes the answer.
Deciding risk attitude from U' instead of U''.
Students see an increasing function and call it risk-seeking.
Fix: Every sensible U is increasing. Only the sign of U'' tells you the attitude.
Forgetting the sign convention in A(w).
The minus sign is dropped, giving negative A for risk-averse people.
Fix: A(w) = −U''/U'. For a concave U it is positive.
Saying a risk-averse person will pay any premium for full cover.
The idea of preferring certainty is overextended.
Fix: The person buys only if the premium is at most P_max = E[X] + π. Above that, they prefer to stay uninsured.
Worked examples
Example 1
A person has utility U(w) = √w and wealth of ₹1,00,000. With probability 0.5 she loses ₹36,000 and otherwise loses nothing. Find the certainty equivalent, the risk premium and the maximum premium she will pay for full insurance.
Show the solution
- Final wealth is ₹64,000 with probability 0.5 and ₹1,00,000 with probability 0.5.
- E[W] = 0.5 × 64,000 + 0.5 × 1,00,000 = ₹82,000.
- √64,000 = 252.98 (approx) and √1,00,000 = 316.23 (approx).
- E[U] = 0.5 × 252.98 + 0.5 × 316.23 = 284.60 (approx).
- CE = 284.60² = ₹81,000 (approx; exactly, (√64,000 + √1,00,000)² ÷ 4 = (64,000 + 1,00,000 + 2√(6.4 × 10⁹)) ÷ 4 ≈ ₹81,000).
- Risk premium π = 82,000 − 81,000 = ₹1,000 (approx).
- Maximum premium P_max = w − CE = 1,00,000 − 81,000 = ₹19,000. Check: expected loss is ₹18,000 and 18,000 + 1,000 = ₹19,000.
Answer: CE ≈ ₹81,000, risk premium ≈ ₹1,000, maximum premium ≈ ₹19,000 (expected loss ₹18,000 plus the risk premium).
Example 2
A person has utility U(w) = ln w. Find the Arrow-Pratt absolute and relative risk aversion, and state the risk attitude. Then find the risk premium for a gamble that gives wealth ₹40,000 or ₹90,000 with equal probability.
Show the solution
- U'(w) = 1/w > 0 and U''(w) = −1/w² < 0, so the person is risk-averse.
- A(w) = −U''/U' = (1/w²) ÷ (1/w) = 1/w.
- R(w) = w × A(w) = 1. Relative risk aversion is constant.
- E[W] = 0.5 × 40,000 + 0.5 × 90,000 = ₹65,000.
- E[U] = 0.5 ln 40,000 + 0.5 ln 90,000 = 0.5 ln(40,000 × 90,000) = 0.5 ln(3.6 × 10⁹).
- CE = exp(E[U]) = √(3.6 × 10⁹) = ₹60,000.
- Risk premium π = 65,000 − 60,000 = ₹5,000.
Answer: Risk-averse; A(w) = 1/w; R(w) = 1; certainty equivalent ₹60,000; risk premium ₹5,000.
Exam tips
- Show U'' and its sign first. Examiners award marks for the correct risk attitude with reasoning.
- Always write the formula for CE and π before substituting. Label each answer clearly.
- For log utility, CE is the geometric mean of outcomes. This saves time for equal probabilities.
- Written questions often ask you to explain why insurance is bought. Mention concave utility, E[U] < U(E[W]), and a premium up to E[X] + π.
- In MCQs, check the sign of the risk premium. A negative value means risk-seeking, and a CE above the mean is a warning sign for an error.
Practice questions from Rational choice theory and utility
- Prospect theory includes a probability weighting function. Which description of its typical shape is correct?
- In expected utility theory, an individual with a utility function U(w) is described as risk averse. Which property of U must hold for all le…
- An individual with a strictly concave, increasing utility function U(w) has wealth that is uncertain. Which statement about her behaviour un…
- Which of the following is an axiom underlying the expected utility theorem of von Neumann and Morgenstern?
- An individual has utility U(w) = ln(w) and faces a 50:50 gamble giving final wealth of Rs 4,00,000 or Rs 1,00,000. What is the certainty equ…
Risk Aversion and Insurance Demand: frequently asked questions
What is the difference between certainty equivalent and risk premium?
The certainty equivalent is the sure wealth that gives the same utility as the gamble. The risk premium is the expected wealth minus this certainty equivalent. For a risk-averse person, CE is below the mean and the premium is positive.
Why do risk-averse people buy insurance?
Because their utility is concave, a sure wealth gives more utility than a risky wealth with the same mean. They will pay more than the expected loss to remove the risk, up to the expected loss plus the risk premium.
What does the Arrow-Pratt measure tell you?
A(w) = −U''(w) ÷ U'(w) measures how risk-averse a person is at wealth w. A higher value means a larger risk premium for the same gamble. Relative risk aversion multiplies A(w) by w.
Does a risk-neutral person buy insurance?
Only if the premium does not exceed the expected loss. Their risk premium is zero, so they gain nothing from removing risk. They will not pay a loaded premium.