Skip to content

Economic Modelling · Measures of investment risk

Downside Risk Measures and Shortfall Probability Explained

Updated 11 October 2026 · Fact-checked

Downside risk measures look only at outcomes below a target or threshold. Shortfall probability is P(X < L). Expected shortfall is the average shortfall below the target, E[(L − X)+]. Below-target semi-variance is E[((L − X)+)²]. To solve, set the target, find the distribution of the return, then compute the required quantity.

Understand Downside Risk Measures and Shortfall Probability

Variance treats gains and losses alike. An investor, or an insurer with a liability to meet, does not mind returns above the target. They worry about returns below it. Downside risk measures capture this by looking only at the bad side of the distribution.

You first choose a target or threshold L. It could be a guaranteed return, a liability to be met, or zero. A shortfall occurs when the outcome X falls below L. The size of the shortfall is (L − X) when X < L, and 0 otherwise. We write this as (L − X)+.

Three measures follow. Shortfall probability answers: how likely is a shortfall? Expected shortfall (in CM2 also called the expected shortfall below target, or mean shortfall) answers: how big is the shortfall on average, counting zero when there is none? Below-target semi-variance squares the shortfall, so large shortfalls are penalised more heavily.

Shortfall probability ignores how bad the miss is. A 0.1% miss and a 50% miss count the same. Expected shortfall and semi-variance fix this. They are useful when returns are skewed or asymmetric, where variance can mislead.

Do not confuse expected shortfall as defined here with the tail-based measure related to Value at Risk (TVaR), which is the average loss given that the loss is beyond the VaR level. Read the question to see which definition it uses. If it gives a target L, it means the target-based form.

Key rules to remember

Shortfall probability
P(X < L)
L is the target. For a continuous X, P(X < L) = P(X ≤ L). For a normal X, use Φ((L − μ) ÷ σ).
Shortfall (one outcome)
(L − X)+ = max(L − X, 0)
Zero when the target is met.
Expected shortfall below target
E[(L − X)+] = Σ over x < L of (L − x) × P(X = x)
For discrete X. For continuous X, integrate (L − x) f(x) from −∞ to L. Includes zeros when the target is met, so it is not conditional on a shortfall.
Conditional expected shortfall
E[L − X | X < L] = E[(L − X)+] ÷ P(X < L)
Average size of the shortfall given that one occurs.
Below-target semi-variance
E[((L − X)+)²] = Σ over x < L of (L − x)² × P(X = x)
Measured about the target L, not about the mean. The square root is the target semi-deviation.
Normal shortfall probability
P(X < L) = Φ((L − μ) ÷ σ)
Valid only if X is normally distributed with mean μ and standard deviation σ.

How to solve Downside Risk Measures and Shortfall Probability questions

Use this method for any question on shortfall probability, expected shortfall or semi-variance.

  1. 1Identify the random variable X (return or final value) and the target L. Check units: rate of return or rupee amount.
  2. 2Write down the distribution of X. It may be a discrete table, a normal, or a lognormal.
  3. 3List the outcomes below L. Outcomes equal to L give no shortfall.
  4. 4For shortfall probability, add the probabilities of those outcomes, or use Φ for a normal.
  5. 5For expected shortfall, multiply each shortfall (L − x) by its probability and add up.
  6. 6For semi-variance, square each shortfall before multiplying by its probability and adding.
  7. 7Check the definition asked: unconditional or conditional on a shortfall. Divide by the shortfall probability only if conditional.
  8. 8State the answer with units and a short comment on what it means for the investor.

Quickest way: Table method for discrete outcomes

When to use it: Use when X has a small discrete distribution and the question gives a target.

  1. Draw a table with columns x, probability, shortfall (L − x)+, probability × shortfall, probability × shortfall².
  2. Fill the shortfall column with zero for every x ≥ L.
  3. Sum the probabilities of rows with a positive shortfall to get the shortfall probability.
  4. Sum the last two columns to get the expected shortfall and semi-variance.
  5. Sanity check: expected shortfall should not exceed the largest shortfall, and semi-variance ≥ (expected shortfall)² only holds in the sense E[Y²] ≥ (E[Y])².

Common mistakes in Downside Risk Measures and Shortfall Probability

  • Measuring the semi-variance about the mean instead of the target.

    Ordinary variance is about the mean, so students carry the habit over.

    Fix: Read the definition. Below-target semi-variance uses L, and the shortfall is (L − x).

  • Including outcomes equal to or above the target in the sums.

    Students sum over all outcomes without checking each against L.

    Fix: Mark each row as shortfall or no shortfall first. Zero-shortfall rows add nothing.

  • Dividing expected shortfall by the shortfall probability when it is not asked.

    Students confuse unconditional and conditional expected shortfall.

    Fix: E[(L − X)+] includes zeros. Divide by P(X < L) only if the question asks for the average shortfall given that one occurs.

  • Using Φ((μ − L) ÷ σ) for the shortfall probability.

    The sign of the standardised value is reversed.

    Fix: Shortfall means X < L, so use Φ((L − μ) ÷ σ). A target above the mean must give a probability above 0.5.

  • Thinking shortfall probability measures how bad losses are.

    It sounds like a complete risk measure.

    Fix: It only gives the chance of missing the target. Use expected shortfall or semi-variance to capture severity.

  • Applying the normal formula to skewed or lognormal returns without conversion.

    Students apply Φ automatically.

    Fix: If final value is lognormal, take logs: P(S < L) = Φ((ln L − μ) ÷ σ) using the parameters of ln S.

Worked examples

Example 1

An investment return X (in %) has this distribution: −10% with probability 0.1, 0% with probability 0.2, 5% with probability 0.3, 10% with probability 0.25, 20% with probability 0.15. The target return is L = 5%. Calculate (a) the shortfall probability, (b) the expected shortfall below target, (c) the below-target semi-variance.

Show the solution
  1. Outcomes below 5% are −10% and 0%. The return of 5% meets the target and has no shortfall.
  2. Shortfalls: for −10%, 5 − (−10) = 15. For 0%, 5 − 0 = 5.
  3. (a) Shortfall probability = 0.1 + 0.2 = 0.3.
  4. (b) Expected shortfall = 15 × 0.1 + 5 × 0.2 = 1.5 + 1.0 = 2.5 (percentage points).
  5. (c) Semi-variance = 15² × 0.1 + 5² × 0.2 = 225 × 0.1 + 25 × 0.2 = 22.5 + 5 = 27.5 (%²).
  6. As a check, the conditional expected shortfall is 2.5 ÷ 0.3 = 8.33 percentage points, which lies between 5 and 15.

Answer: (a) 0.3 (b) 2.5 percentage points (c) 27.5 (%²)

Example 2

The annual return on a fund is normally distributed with mean 8% and standard deviation 12%. An investor needs a return of at least 0%. Find the probability of a shortfall. Take Φ(0.6667) = 0.7475.

Show the solution
  1. X ~ N(8, 12²) and the target is L = 0.
  2. Shortfall means X < 0.
  3. Standardise: z = (0 − 8) ÷ 12 = −0.6667.
  4. P(X < 0) = Φ(−0.6667) = 1 − Φ(0.6667).
  5. 1 − 0.7475 = 0.2525.
  6. Interpretation: about a 25% chance of a negative return in a year.

Answer: Shortfall probability ≈ 0.2525, or about 25%.

Exam tips

  • Write the target L and the word shortfall explicitly at the start. It earns method marks and avoids sign errors.
  • Draw the table for discrete questions. Examiners award marks for the working even if arithmetic slips.
  • In written answers, comment on the weakness of shortfall probability: it ignores the size of the loss. Link it to expected shortfall and semi-variance.
  • Compare measures when asked: variance penalises upside, semi-variance does not. For symmetric distributions the rankings are often similar, but for skewed returns they can differ.
  • In the computer-based paper, compute shortfall measures from simulated data with a logical condition, for example the mean of pmax(L − x, 0) in R.

Practice questions from Measures of investment risk

Downside Risk Measures and Shortfall Probability: frequently asked questions

What is the difference between semi-variance and shortfall risk?

Shortfall probability only counts how often you fall below the target. Semi-variance squares the size of each shortfall and averages it, so bigger misses count much more. Semi-variance therefore reflects severity, while shortfall probability reflects frequency.

Is expected shortfall the same as Tail Value at Risk?

Not in general. Target-based expected shortfall averages the shortfall below a chosen target L. Tail Value at Risk is the average loss given that the loss exceeds a VaR level. Check which definition the question uses.

Why use downside measures instead of variance?

Variance counts gains above the mean as risk, which investors do not mind. Downside measures focus on outcomes below a target. This suits skewed returns and liability-driven investors.

How do I calculate shortfall probability for a normal return?

Standardise the target: z = (L − μ) ÷ σ. The shortfall probability is Φ(z). If the target is above the mean, z is positive and the probability exceeds 0.5.