Skip to content

Advanced Financial Management · Portfolio Management

Diversification and Efficient Frontier (Markowitz Theory) for CA Final AFM

Updated 5 October 2026 · Fact-checked

Diversification spreads money across assets so that unsystematic risk cancels out, leaving only systematic risk. Markowitz showed that the efficient frontier holds portfolios with the highest return for each level of risk. To solve questions, compute portfolio return, covariance, variance and standard deviation, then find the minimum variance weights.

Understand Diversification and Efficient Frontier

Every security carries two kinds of risk. Systematic risk (market risk) comes from factors that hit the whole market: interest rates, inflation, policy changes, recessions. You cannot remove it by holding more securities. Unsystematic risk (specific risk) comes from one company or industry: a strike, a product failure, a management change. It can be reduced by diversification.

Why does diversification work? Returns of different securities do not move in perfect step. When one falls, another may rise or fall less. The weaker the co-movement, the more the ups and downs offset each other. This co-movement is measured by covariance or correlation (ρ). Correlation lies between −1 and +1. Only when ρ = +1 is there no risk reduction. For any ρ below +1, portfolio standard deviation is less than the weighted average of the individual standard deviations.

Portfolio return is simply the weighted average of the individual expected returns. Portfolio risk is not. It depends on the weights, the individual variances and the covariances. That is the core of Markowitz theory (Modern Portfolio Theory). Markowitz assumes investors are rational and risk-averse, judge a portfolio only by expected return and variance over a single period, and prefer more return to less for the same risk.

If you plot every possible portfolio of risky assets with risk on the x-axis and return on the y-axis, you get the opportunity set. The leftmost point is the minimum variance portfolio. The upper part of the curve from that point onward is the efficient frontier: for each risk level it gives the highest return, and for each return level the lowest risk. Portfolios below the frontier are inefficient. Portfolios on the lower part of the curve are dominated by a portfolio with the same risk and higher return.

Which efficient portfolio you pick depends on your risk tolerance. Markowitz theory says the optimal portfolio is where your indifference curve touches the efficient frontier. A more risk-averse investor sits lower on the frontier. Diversification cannot remove systematic risk, so total risk never falls below that level in a well-diversified portfolio.

Key rules to remember

Total risk
Total risk = Systematic risk + Unsystematic risk
Diversification reduces only the unsystematic part.
Portfolio expected return (two assets)
Rp = w1 × R1 + w2 × R2, with w1 + w2 = 1
A simple weighted average. Correlation does not affect it.
Covariance and correlation
Cov12 = ρ12 × σ1 × σ2, so ρ12 = Cov12 ÷ (σ1 × σ2)
If the question gives covariance, use it directly. Do not multiply by ρ again.
Portfolio variance (two assets)
σp² = w1²σ1² + w2²σ2² + 2 × w1 × w2 × ρ12 × σ1 × σ2
Standard deviation σp = √σp². Square the weights in the first two terms.
Minimum variance weight (two assets)
w1 = (σ2² − Cov12) ÷ (σ1² + σ2² − 2 × Cov12); w2 = 1 − w1
Gives the lowest-variance mix. Works for any ρ. A negative weight means short selling.
Perfect negative correlation (ρ = −1)
w1 = σ2 ÷ (σ1 + σ2); w2 = σ1 ÷ (σ1 + σ2); σp = 0
Risk-free portfolio is possible only in this special case.
Bounds on portfolio risk
σp ≤ w1σ1 + w2σ2, with equality only when ρ12 = +1 (for non-negative weights)
Shows that diversification benefit exists whenever ρ < +1.

How to solve Diversification and Efficient Frontier questions

Use this sequence for any two-asset diversification or Markowitz numerical. Keep all inputs in the same unit (either % or decimals) throughout.

  1. 1List the inputs: expected returns, standard deviations, weights, and either ρ or covariance. Note which one the question gives.
  2. 2Convert to covariance: Cov12 = ρ × σ1 × σ2. If covariance is given, use it as is.
  3. 3Compute portfolio return: Rp = w1R1 + w2R2.
  4. 4Compute portfolio variance using the full three-term formula, then take the square root for standard deviation.
  5. 5If the question asks for the minimum variance portfolio, find w1 from the formula, then w2 = 1 − w1, and recompute return and risk with these weights.
  6. 6Check reasonableness: σp must not exceed the weighted average of σ1 and σ2 unless ρ = +1 is not the case you assumed. For ρ = −1 the minimum risk should come out as zero.
  7. 7Interpret: state the diversification benefit, whether the portfolio is on the efficient frontier, and what remains (systematic risk). Write the conclusion in one line.

Quickest way: Minimum variance in four lines

When to use it: When the question asks for the least risky mix of two assets and gives σ values and ρ.

  1. Write σ1², σ2² and Cov12 = ρσ1σ2 first.
  2. Compute w1 = (σ2² − Cov12) ÷ (σ1² + σ2² − 2Cov12) and w2 = 1 − w1.
  3. Find variance in one go: σp² = w1²σ1² + w2²σ2² + 2w1w2Cov12.
  4. Take the square root and compute Rp. If ρ = −1, skip the variance work: w1 = σ2 ÷ (σ1 + σ2) and risk is 0.

Common mistakes in Diversification and Efficient Frontier

  • Calculating portfolio risk as the weighted average of standard deviations.

    Because portfolio return is a weighted average, students assume risk is too.

    Fix: Always use the variance formula with the covariance term. The weighted average of standard deviations is only the upper limit, reached when ρ = +1.

  • Putting ρ in the covariance slot, or multiplying by ρ when covariance is already given.

    The formula has 2w1w2ρσ1σ2 and the covariance version looks similar.

    Fix: Check the data first. If covariance is given, the term is 2 × w1 × w2 × Cov12 with no further ρ or σ.

  • Forgetting to square the weights or the standard deviations.

    Rushing under time pressure.

    Fix: Write the formula as three separate terms and fill each one in turn. Then check that the variance is positive and below the larger σ².

  • Treating the minimum variance portfolio as the best portfolio.

    Students link lowest risk with optimal.

    Fix: It is only the starting point of the efficient frontier. The best portfolio depends on the investor's risk preference. Say so in theory answers.

  • Claiming diversification removes all risk.

    Mixing up the zero-risk case at ρ = −1 with the general case.

    Fix: State that diversification removes unsystematic risk only. Systematic risk remains. Zero risk is a special case needing ρ = −1 and the exact weights.

  • Mixing units, such as σ = 20 with weights in decimals and the result in the wrong scale.

    Data are given in % and some steps convert to decimals.

    Fix: Pick one unit at the start. If you use %, variance is in %², so take the square root to return to %.

Worked examples

Example 1

Case: An investor holds Asset A (expected return 12%, standard deviation 20%) and Asset B (expected return 18%, standard deviation 30%). The correlation between them is 0.5. She wants the minimum variance portfolio. Find the weights, expected return and standard deviation, and comment on the result.

Show the solution
  1. Variances: σA² = 20² = 400; σB² = 30² = 900 (in %²).
  2. Covariance = 0.5 × 20 × 30 = 300.
  3. Weight in A = (900 − 300) ÷ (400 + 900 − 2 × 300) = 600 ÷ 700 = 0.8571. Weight in B = 1 − 0.8571 = 0.1429.
  4. Expected return = 0.8571 × 12 + 0.1429 × 18 = 10.2857 + 2.5714 = 12.857%.
  5. Variance = (0.8571² × 400) + (0.1429² × 900) + (2 × 0.8571 × 0.1429 × 300) = 293.88 + 18.37 + 73.47 = 385.71 (%²).
  6. Standard deviation = √385.71 = 19.64%.
  7. Comment: the portfolio risk of 19.64% is lower than even the less risky asset A (20%). This is the benefit of diversification, since ρ is below +1. The weighted average of standard deviations would be 0.8571 × 20 + 0.1429 × 30 = 21.43%, which overstates the risk.

Answer: Minimum variance portfolio: about 85.7% in A and 14.3% in B, expected return about 12.86%, standard deviation about 19.64%.

Example 2

Case: Two stocks, X (expected return 8%, standard deviation 10%) and Y (expected return 14%, standard deviation 15%), have a correlation of −1. A fund manager wants a portfolio with no risk. Find the weights, the return and verify that the risk is zero.

Show the solution
  1. For ρ = −1, wX = σY ÷ (σX + σY) = 15 ÷ 25 = 0.6, and wY = 10 ÷ 25 = 0.4.
  2. Expected return = 0.6 × 8 + 0.4 × 14 = 4.8 + 5.6 = 10.4%.
  3. Covariance = −1 × 10 × 15 = −150.
  4. Variance = (0.6² × 100) + (0.4² × 225) + (2 × 0.6 × 0.4 × −150) = 36 + 36 − 72 = 0.
  5. Standard deviation = 0, so the portfolio is risk-free.
  6. Interpretation: this mix gives a risk-free return of 10.4%. To avoid arbitrage, the risk-free rate must equal 10.4%. If it is lower, one can borrow at the risk-free rate and invest in the mix for a riskless profit. This special case shows that perfect negative correlation can remove all risk, but real stocks rarely have ρ = −1.

Answer: Weights: 60% in X and 40% in Y. Expected return is 10.4% and standard deviation is 0%.

Exam tips

  • In case-based MCQs, read for the correlation value first. ρ = +1 means no diversification benefit, ρ = −1 allows zero risk, and anything between gives partial benefit.
  • In numericals, show the covariance, the three terms of the variance and the weights separately. Marks are given for each step even if the final figure slips.
  • For theory questions, keep a fixed structure: define systematic and unsystematic risk, explain how diversification works through covariance, then describe the efficient frontier and the Markowitz assumptions.
  • Always end a numerical with a one-line interpretation, such as risk lower than either asset or the portfolio lying on the efficient frontier.
  • Differentiate clearly between the minimum variance portfolio (lowest risk point) and the efficient frontier (the whole upper curve from that point).

Practice questions from Portfolio Management

Diversification and Efficient Frontier in other exams

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

Diversification and Efficient Frontier: frequently asked questions

What is the difference between systematic and unsystematic risk?

Systematic risk comes from market-wide factors such as interest rates, inflation and economic cycles, and affects all securities. It cannot be diversified away. Unsystematic risk is specific to a company or industry and can be reduced by holding a diversified portfolio.

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

Compute the covariance as ρσ1σ2. Then weight in asset 1 is (σ2² − Cov12) ÷ (σ1² + σ2² − 2Cov12), and weight in asset 2 is 1 minus that. Use these weights to find portfolio return and standard deviation.

What is the efficient frontier in Markowitz theory?

It is the set of risky-asset portfolios that give the highest expected return for each level of risk, or the lowest risk for each level of return. It starts at the minimum variance portfolio and slopes upward. Portfolios below it are inefficient.

Can diversification reduce risk to zero?

Not in general. It removes unsystematic risk, but systematic risk stays. Zero risk is possible only in a special case such as two assets with a correlation of −1 held in the right proportions.