Economic Modelling · Mean-variance portfolio theory

Efficient Frontier and Diversification in Mean-Variance Theory

Updated 11 October 2026 · Fact-checked

The efficient frontier is the set of portfolios that give the highest expected return for each level of risk. Diversification lowers risk when assets are not perfectly correlated. To solve questions, compute portfolio mean and variance, find the minimum variance portfolio, keep the upper branch, then match it to the investor's indifference curve.

Understand Efficient Frontier and Diversification

Start with one idea. A portfolio's expected return is the weighted average of the returns of its assets. Its risk, measured by variance, is not a weighted average. It depends on how the assets move together, which is captured by covariance or correlation.

This is why diversification works. If two assets are not perfectly positively correlated (correlation below 1), the portfolio standard deviation is less than the weighted average of the two standard deviations. The lower the correlation, the bigger the gain. With correlation of -1, risk can be removed entirely with the right weights.

Total risk splits into two parts. Specific risk (also called unsystematic or diversifiable risk) belongs to individual companies and falls as you add more assets. Systematic risk (market risk) affects all assets and cannot be removed by diversifying. As the number of assets grows, portfolio variance approaches the average covariance, not zero.

Now plot every possible portfolio of risky assets with standard deviation on the horizontal axis and expected return on the vertical axis. The area of attainable points is the opportunity set (or feasible set). Its left edge is a curve. The leftmost point is the minimum variance portfolio. The efficient frontier is the part of the left edge above and including that point. Any portfolio below it has another portfolio with higher return for the same risk, so a rational investor ignores it.

Which efficient portfolio you pick depends on you. An indifference curve joins combinations of risk and return that give you the same utility. A risk-averse investor has curves that slope upwards and are convex. The optimal portfolio is where the highest attainable indifference curve touches the efficient frontier. A more risk-averse investor picks a point nearer the minimum variance portfolio.

The standard assumptions: investors are rational and risk averse, they care only about mean and variance of return over a single period, they agree on the inputs, and there are no taxes or transaction costs.

Key rules to remember

Portfolio expected return: E(Rp) = Σ wi E(Ri), with Σ wi = 1; Linear in weights. Short sales mean some weights are negative.
Two-asset portfolio variance: σp² = w²σA² + (1 − w)²σB² + 2w(1 − w)ρσAσB; w is the weight in A. Covariance σAB = ρσAσB.
General portfolio variance: σp² = ΣiΣj wi wj σij; Includes i = j terms, where σii = σi².
Minimum variance weight, two assets: w* = (σB² − σAB) ÷ (σA² + σB² − 2σAB); Weight in A. Set dσp²/dw = 0. This is the minimum for any σAB (except the degenerate case σA = σB with ρ = 1). Check 0 ≤ w* ≤ 1 if short sales are not allowed.
Perfect correlation bounds: ρ = 1: σp = wσA + (1 − w)σB; ρ = −1: σp = |wσA − (1 − w)σB|; With ρ = −1, σp = 0 when w = σB ÷ (σA + σB).
Equally weighted n-asset variance: σp² = (1/n) × average variance + ((n − 1)/n) × average covariance; As n grows, σp² tends to the average covariance. This is the systematic part.
Optimal portfolio rule: Optimal point: indifference curve tangent to the efficient frontier; Highest attainable utility. Slopes of the two curves are equal at the tangent point.

How to solve Efficient Frontier and Diversification questions

Use this order for any question on diversification, the frontier or optimal choice.

1Write down the inputs: expected returns, standard deviations and the correlation or covariance. Note whether short selling is allowed.
2Convert correlation to covariance if needed: σAB = ρσAσB. Keep variances, not standard deviations, in the formulas.
3Compute the portfolio expected return from the weights, then the portfolio variance using the full formula with the covariance term.
4If asked for the minimum variance portfolio, use w* = (σB² − σAB) ÷ (σA² + σB² − 2σAB), then compute its variance and return.
5Identify the efficient frontier: only the part of the curve with return at or above that of the minimum variance portfolio.
6For choice of portfolio, find where the highest indifference curve touches the frontier. Explain how risk aversion shifts the point.
7For diversification questions, separate specific risk (falls with more assets) from systematic risk (remains).
8State your assumptions and give units. Express risk as standard deviation if the question asks for it.

Quickest way: Fast route for the two-asset minimum variance question

When to use it: Use when you are given σA, σB and ρ (or σAB) and asked for the minimum variance portfolio, its risk or its return.

Compute σAB = ρσAσB.
Compute the numerator σB² − σAB and the denominator σA² + σB² − 2σAB.
Divide to get w*. Weight in B is 1 − w*.
Substitute w* into the variance formula once. Take the square root only at the end.
Quick check: the result should be no larger than min(σA², σB²) when 0 ≤ w* ≤ 1. If it is larger, recheck.

Common mistakes in Efficient Frontier and Diversification

Averaging standard deviations to get portfolio risk
Return is a weighted average, so students assume risk is too.
Fix: Always use the variance formula with the covariance term. The weighted average of standard deviations is only correct when ρ = 1.
Forgetting the factor 2 on the covariance term
The two cross terms are merged and the 2 is dropped under time pressure.
Fix: Write 2w(1 − w)σAB every time. Check with ρ = 1, where the result should equal (wσA + (1 − w)σB)².
Calling the whole feasible curve the efficient frontier
Students forget that the lower branch is dominated.
Fix: Mark the minimum variance point and keep only the part above it. Say why the lower part is inefficient.
Saying diversification removes all risk
Students overlook the difference between specific and systematic risk.
Fix: State that only specific risk is diversified away. Systematic risk remains, and variance tends to the average covariance.
Using σ instead of σ² in the minimum variance formula
The data is given as standard deviations.
Fix: Square the standard deviations first. Convert back to σ only at the end.
Placing the optimal portfolio where the indifference curve crosses the frontier
Tangency is not understood.
Fix: Choose the highest indifference curve that just touches the frontier. A crossing curve means a higher one is still attainable.

Worked examples

Example 1

Asset A has expected return 8% and standard deviation 10%. Asset B has expected return 14% and standard deviation 20%. The correlation is 0.25. Find the minimum variance portfolio (short sales allowed), and its expected return and standard deviation.

Show the solution

σA² = 0.01, σB² = 0.04. σAB = 0.25 × 0.10 × 0.20 = 0.005.
Numerator: σB² − σAB = 0.04 − 0.005 = 0.035.
Denominator: σA² + σB² − 2σAB = 0.01 + 0.04 − 0.01 = 0.04.
w* = 0.035 ÷ 0.04 = 0.875 in A, so 0.125 in B.
Expected return = 0.875 × 8% + 0.125 × 14% = 7% + 1.75% = 8.75%.
Variance = 0.875² × 0.01 + 0.125² × 0.04 + 2 × 0.875 × 0.125 × 0.005.
= 0.00765625 + 0.000625 + 0.00109375 = 0.0093750.
Standard deviation = √0.009375 = 0.09682, about 9.68%.

Answer: Hold 87.5% in A and 12.5% in B. Expected return is 8.75% and standard deviation is about 9.68%, which is lower than the 10% of A alone.

Example 2

Explain why a portfolio of 40 shares from different sectors is less risky than a single share, and why its risk cannot fall to zero. Then say which efficient portfolio a more risk-averse investor will choose.

Show the solution

Each share has specific risk linked to its own company, such as management or product events. These events are largely independent across companies.
In a portfolio, specific shocks partly offset each other. Their contribution to portfolio variance shrinks roughly in proportion to 1/n for an equally weighted portfolio.
All shares are affected by common factors such as interest rates, inflation and the economy. This is systematic risk. Covariances between shares are positive.
Portfolio variance for equal weights is (1/n) × average variance + ((n − 1)/n) × average covariance. As n increases, it tends to the average covariance, not zero.
So diversification removes specific risk but leaves systematic risk.
A more risk-averse investor has steeper indifference curves. Tangency with the efficient frontier occurs at a lower standard deviation, nearer the minimum variance portfolio.

Answer: Diversification removes specific risk, but the average covariance (systematic risk) remains, so risk does not reach zero. A more risk-averse investor chooses an efficient portfolio with lower risk and lower expected return, closer to the minimum variance portfolio.

Exam tips

Show the covariance calculation as a separate line. Method marks are often given for it even if the final number is wrong.
In written answers, name the assumptions: single period, mean-variance preferences, rational risk-averse investors, no frictions.
Be ready to sketch the opportunity set, mark the minimum variance point and shade the efficient frontier. Label both axes.
In multiple-choice questions, test the answer against the limiting cases ρ = 1 and ρ = −1 to catch errors quickly.
In computer-based questions, set up the weights and covariance matrix clearly and show the formula for portfolio variance before the result.

Practice questions from Mean-variance portfolio theory

Efficient Frontier and Diversification in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Efficient Frontier and Diversification: frequently asked questions

What is the efficient frontier in simple words?

It is the set of portfolios that give the most expected return for each level of risk. You cannot do better on return without taking more risk. Portfolios below it are inefficient.

How do I find the minimum variance portfolio of two assets?

Differentiate the portfolio variance with respect to the weight and set it to zero. This gives w* = (σB² − σAB) ÷ (σA² + σB² − 2σAB) for the weight in A. Check that the weights are within 0 and 1 if short selling is not allowed.

What is the difference between systematic and specific risk?

Specific risk is unique to one asset and can be diversified away by holding many assets. Systematic risk affects the whole market and remains after diversification. Investors are normally rewarded only for bearing systematic risk.

How do indifference curves decide the optimal portfolio?

Each curve joins risk-return points with equal utility. Curves further up and left are better. The optimal portfolio is where the highest attainable curve just touches the efficient frontier.