Fact-checked by Grok 2 weeks ago

Expected shortfall

Expected shortfall (ES), also known as conditional value at risk (CVaR) or tail conditional expectation (TCE), is a used in to quantify the expected loss of a portfolio over a specified time horizon, conditional on the loss exceeding the threshold at a high confidence level, typically 97.5% or 99%. Formally, for a loss random variable X (positive values indicate losses) and confidence level \alpha \in (0,1), ES is defined as \mathrm{ES}_\alpha(X) = \frac{1}{1-\alpha} \int_\alpha^1 \mathrm{VaR}_p(X) \, dp, where \mathrm{VaR}_p(X) is the p-quantile of the loss distribution, providing an average of the losses exceeding \mathrm{VaR}_\alpha(X) (i.e., the worst $1-\alpha tail). ES addresses key limitations of by incorporating the magnitude of extreme losses beyond the quantile threshold, rather than merely identifying a cutoff point. It satisfies the four axioms of monotonicity (greater losses imply greater or equal risk), (risk of combined positions does not exceed the sum of individual risks), positive homogeneity (risk scales linearly with position size), and translation invariance (adding cash reduces risk by the same amount)—making it suitable for diversification and capital allocation in portfolios. These properties ensure ES promotes prudent risk-taking without incentivizing excessive concentration, unlike non-subadditive measures such as . The concept of coherent risk measures, including ES as TCE, was formalized by Artzner et al. in 1999 to evaluate risks in . ES gained prominence in regulatory practice through the framework, where it replaced in the internal models approach for capital requirements, calibrated at a 97.5% level over a 10-day horizon (adjusted for ), to better capture risks during stressed conditions. This adoption reflects its robustness in estimating capital needs for banks' trading books, with calculations incorporating both modellable and non-modellable risk factors.

Fundamentals

Definition

Expected shortfall (ES), also known as conditional value at risk or , is a that quantifies the expected of a in the worst (1-α) portion of cases, where α is the confidence level (typically 0.95 or 0.99). For a random variable X (where positive values indicate losses), the formal definition is given by the average of the (VaR) over the tail: \text{ES}_\alpha(X) = \frac{1}{1-\alpha} \int_\alpha^1 \text{VaR}_u(X) \, du, where is the at level u, defined as the u-quantile of the loss distribution, i.e., the smallest value such that the probability of exceeding it is 1-u. This form captures the severity of extreme losses beyond the VaR threshold by averaging the quantiles in the upper . For continuous loss distributions, expected shortfall admits an alternative expression as the tail conditional expectation: \text{ES}_\alpha(X) = \mathbb{E}[X \mid X > \text{VaR}_\alpha(X)]. This represents the expected loss given that the loss exceeds the α- threshold. In financial contexts, the convention treats X as the loss (positive for adverse outcomes), distinguishing it from profit-and-loss frameworks where negative values denote losses; adjustments for sign conventions ensure consistency in . Expected shortfall was introduced as part of the framework for coherent risk measures by Artzner et al. (1999), who proposed axioms such as and positive homogeneity to evaluate risk metrics, with ES emerging as a prominent example satisfying these properties. This development addressed limitations of VaR by providing a more comprehensive perspective.

Relation to Value at Risk

Value at Risk (VaR) at confidence level α is defined as the quantile threshold VaR_α(X) = inf{x | P(X > x) ≤ 1-α}, where X represents losses, providing a measure of the maximum expected loss exceeded with probability no greater than 1-α. However, VaR has significant limitations as a risk measure: it is not subadditive, meaning the risk of a combined may exceed the sum of individual risks, potentially discouraging diversification. Additionally, VaR ignores the severity of losses beyond the quantile threshold, offering no information on the magnitude of extreme tail events. Expected Shortfall (ES) addresses these shortcomings by calculating the average loss conditional on exceeding the VaR threshold, thereby capturing the expected severity of tail losses rather than just the boundary. Unlike VaR, ES is a , satisfying , monotonicity, positive homogeneity, and translation invariance, which ensures it promotes diversification and provides consistent risk assessments across portfolios. This makes ES particularly suitable for managing extreme events, as it integrates the full tail for a more robust evaluation of potential downside risks. The development of ES gained prominence following critiques of 's incoherence, as highlighted in foundational work on coherent risk measures, leading to its proposal as a superior alternative in . In response to the limitations exposed during the , Basel III incorporated ES through the Fundamental Review of the Trading Book, replacing for capital requirements to better account for tail risks. For instance, in fat-tailed distributions common in financial returns, ES yields more conservative risk estimates than by incorporating the heavier probabilities and magnitudes of extreme losses beyond the .

Properties

Mathematical Properties

Expected shortfall (ES) satisfies the axioms of coherent risk measures, including monotonicity, subadditivity, positive homogeneity, and invariance. Monotonicity implies that if one 's outcomes are always less than or equal to another's, then the ES of the first is less than or equal to that of the second. ensures that the risk of a combined does not exceed the sum of individual risks, promoting diversification. Positive homogeneity means that scaling a by a positive factor scales its ES by the same factor, while invariance states that adding a constant to all outcomes reduces the ES by that constant. ES is continuous with respect to the confidence level α, providing stability under small perturbations in α. More broadly, as a monetary risk measure, ES is Lipschitz continuous with respect to the L¹ norm on the space of random variables, ensuring robustness to changes in the underlying distribution. Among coherent risk measures, ES is the smallest one that dominates value at risk (VaR) at the same confidence level α and depends only on the distribution of the loss variable. This property positions ES as a tight upper bound on VaR while maintaining coherence. ES exhibits greater sensitivity to heavy-tailed distributions compared to VaR, as it averages losses beyond the VaR threshold, thereby penalizing extreme tail events more severely. For instance, under distributions with fatter tails, ES increases more than VaR, reflecting the higher expected severity of losses in the tail. A specific limiting case occurs when α = 0, where ES reduces to the unconditional expectation of the loss variable -X, ES₀(X) = E[-X].

Axiomatic Foundations

Expected shortfall (ES) is classified as a within the axiomatic framework introduced by Artzner et al., which defines desirable properties for assessing financial risks. A coherent risk measure \rho must satisfy four axioms: monotonicity, , positive homogeneity, and translation invariance. Monotonicity requires that if X (representing returns) is less than or equal to Y , then \rho(X) \geq \rho(Y), ensuring worse outcomes incur higher risk assessments. states \rho([X + Y](/page/X+Y)) \leq \rho(X) + \rho(Y), promoting diversification by capping combined risk at the sum of individual risks. Positive homogeneity demands \rho(\lambda X) = \lambda \rho(X) for \lambda > 0, reflecting of risks with position size. Translation invariance specifies \rho(X + c) = \rho(X) - c for constant c, adjusting risk linearly for additions. Expected shortfall satisfies these axioms, as demonstrated through its integral representation: for a confidence level \alpha \in (0,1), \text{ES}_\alpha(X) = \frac{1}{1-\alpha} \int_\alpha^1 \text{VaR}_u(X) \, du, where \text{VaR}_u(X) = \inf \{ z \in \mathbb{R} : P(X \leq -z) \leq u \} is the at level u. Monotonicity follows directly: if X \leq Y , then \text{VaR}_u(X) \geq \text{VaR}_u(Y) for all u, implying \text{ES}_\alpha(X) \geq \text{ES}_\alpha(Y). For translation invariance, adding c > 0 shifts \text{VaR}_u(X + c) = \text{VaR}_u(X) - c, so \text{ES}_\alpha(X + c) = \text{ES}_\alpha(X) - c. Positive homogeneity holds because \text{VaR}_u(\lambda X) = \lambda \text{VaR}_u(X) for \lambda > 0, preserving the integral scaling. Subadditivity is the most involved: it arises from the of coherent measures as suprema over probability measures, where ES corresponds to the set of measures with density bounded by $1/(1-\alpha), ensuring \text{ES}_\alpha(X + Y) \leq \text{ES}_\alpha(X) + \text{ES}_\alpha(Y) via on the tail expectations. In contrast, (VaR) fails , rendering it non-coherent. A classic counterexample involves two identical bonds, each with a 4% default probability leading to a 100-unit loss and otherwise 0, under a 95% level where \text{VaR}_{0.95} = 0 for each. Merging them yields a default risk with 8% probability of at least one default (loss of 100), so \text{VaR}_{0.95} = 100 > 0 + 0, violating and discouraging diversification. Expected shortfall extends naturally to the broader class of risk measures, which relax to convexity (\rho(\lambda X + (1-\lambda) Y) \leq \lambda \rho(X) + (1-\lambda) \rho(Y) for \lambda \in [0,1]) while retaining the other coherent axioms; coherent measures are precisely the positively homogeneous ones. ES is because its tail form inherits convexity from the operator, positioning it as a special case that fully satisfies coherence for \alpha < 1. Under law invariance and comonotonic additivity conditions, expected shortfall is the unique coherent risk measure for elliptical distributions, such as multivariate normal or Student's t, where risks exhibit symmetric dependence structures.

Examples

Numerical Illustrations

To illustrate the calculation of expected shortfall (ES), consider a simple discrete uniform distribution for losses L taking values \{-1, 0, 1, 10\}, each with probability $0.25. This example demonstrates how ES captures the magnitude of extreme losses in the tail. First, compute the Value at Risk (VaR) at the 95% confidence level, denoted \mathrm{VaR}_{0.95}, which is the 0.95 quantile of the loss distribution: the smallest value q such that P(L \leq q) \geq 0.95. Sorting the outcomes in ascending order gives -1, 0, 1, 10. The cumulative probabilities are P(L \leq -1) = 0.25, P(L \leq 0) = 0.5, P(L \leq 1) = 0.75, and P(L \leq 10) = 1. Since P(L \leq 1) = 0.75 < 0.95 and P(L \leq 10) = 1 \geq 0.95, \mathrm{VaR}_{0.95} = 10. Next, compute \mathrm{ES}_{0.95} using the formula \mathrm{ES}_p = \mathrm{VaR}_p + \frac{1}{1-p} E[(L - \mathrm{VaR}_p)^+], where (x)^+ = \max(x, 0). Here, p = 0.95, so $1-p = 0.05. The excess term E[(L - 10)^+] = \sum (l_i - 10)^+ P(L = l_i). Only l = 10 contributes: (10 - 10)^+ \cdot 0.25 = 0, and all others are negative, so E[(L - 10)^+] = 0. Thus, \mathrm{ES}_{0.95} = 10 + \frac{1}{0.05} \cdot 0 = 10. Alternatively, since the tail consists solely of the outcome 10 with probability 0.25 (exceeding the 0.05 tail probability), the conditional expectation E[L \mid L \geq 10] = 10, confirming the result. In this case, ES highlights the impact of the rare large loss of 10, pulling the risk measure to that extreme value despite its 25% probability, far beyond the distribution's mean of 2.5.
Outcome l_iProbabilitySorted Order
-10.251st
00.252nd
10.253rd
100.254th
Another illustrative case involves a binomial distribution for the number of losses L \sim \mathrm{Bin}(10, 0.1), where each trial represents a potential loss event with success probability 0.1, and L counts the occurrences (mean 1). For \alpha = 0.95, compute \mathrm{VaR}_{0.95} as the smallest integer k such that the cumulative distribution function F(k) = P(L \leq k) \geq 0.95. Using standard binomial probabilities: P(L=0) \approx 0.3487, P(L=1) \approx 0.3874, P(L=2) \approx 0.1937, so F(2) \approx 0.9298 < 0.95; P(L=3) \approx 0.0574, so F(3) \approx 0.9872 \geq 0.95. Thus, \mathrm{VaR}_{0.95} = 3. The tail probability P(L > 3) \approx 0.0128 < 0.05. For \mathrm{ES}_{0.95}, compute using the excess formula \mathrm{ES}_p = \mathrm{VaR}_p + \frac{1}{1-p} E[(L - \mathrm{VaR}_p)^+]. Here, E[(L - 3)^+] \approx 0.0145, so \mathrm{ES}_{0.95} \approx 3 + \frac{1}{0.05} \cdot 0.0145 \approx 3.29. This accounts for the discrete nature by focusing on losses strictly exceeding VaR, scaled to the tail probability. The unadjusted conditional E[L \mid L \geq 3] \approx 3.21 is close but includes the full mass at VaR=3. This step-by-step approach identifies the tail threshold via VaR and averages the losses beyond it, weighted appropriately. In both examples, ES quantifies the "expected" extreme loss in the tail, providing a more informative measure than VaR's mere threshold; for instance, while VaR flags the point where 5% tail risk begins (3 in the binomial case), ES averages the severity within that tail (\approx 3.29), revealing how much worse outcomes could be on average.

Financial Applications

Expected shortfall (ES) has been particularly valuable in analyzing equity portfolio losses during the 2008 financial crisis, where it consistently indicated greater tail risks than value at risk (VaR). For the S&P 500 index, extreme value theory-based estimates during the crisis period showed that ES at a 99% confidence level captured average losses exceeding VaR thresholds by significant margins, reflecting the severe drawdowns in equity markets triggered by the subprime meltdown. This discrepancy highlighted ES's ability to quantify the magnitude of extreme events, such as the over 50% decline in the S&P 500 from October 2007 to March 2009, whereas VaR often underestimated the potential severity beyond its quantile. In credit risk management, ES provides a robust measure for assessing loan default portfolios under stress scenarios, incorporating the expected losses conditional on exceeding severe thresholds. For corporate bond portfolios, stress tests involving heightened default rates and correlation shocks—simulating conditions like economic downturns—demonstrated that ES could double or triple as a percentage of portfolio value at 99% confidence, far surpassing baseline expected losses. This approach reveals the tail dependencies in default events, such as clustered failures in high-yield loans during recessions, enabling institutions to allocate capital more accurately for potential systemic credit deteriorations. ES also informs hedging strategies with derivatives to mitigate tail risks, guiding positions that minimize average losses in adverse scenarios rather than just quantile breaches. In portfolio hedging contexts, optimizing for ES reduction—using instruments like out-of-the-money options—yields more effective protection against extreme market moves compared to VaR-based hedges, as it accounts for the full extent of potential shortfalls and promotes subadditive diversification benefits. Following the 2008 crisis, central banks incorporated ES into regulatory frameworks to better capture systemic risks, replacing or supplementing VaR in assessments of market and credit exposures. The Basel Committee's Fundamental Review of the Trading Book (FRTB), finalized in 2019, mandates ES at 97.5% confidence under stressed historical periods to compute capital requirements, addressing VaR's shortcomings in the subprime exposures where it failed to signal the impending tail events. During the COVID-19 market crash in March 2020, ES measures for global equity portfolios highlighted tail risks more effectively than , with estimates showing ES levels 1.5 times or higher than VaR at 99% confidence for major indices like the amid unprecedented volatility. This underscored ES's role in stress calibration for regulatory capital amid pandemics and liquidity shocks. Overall, ES's superiority in crisis analyses stems from its sensitivity to fat-tailed loss distributions; retrospective studies during the revealed that ES projected losses higher than VaR at equivalent confidence levels for equity markets.

Computation Methods

General Formulas

The expected shortfall (ES) at confidence level $1 - \alpha, denoted \mathrm{ES}_{\alpha}(X), for a loss random variable X with cumulative distribution function (CDF) F, is defined as the conditional expectation of X given that it exceeds the value at risk (VaR) at level \alpha, adjusted for the general case where the distribution may have atoms. This measure captures the average severity of losses in the tail beyond the VaR threshold. For continuous distributions, the formula for \mathrm{ES}_{\alpha}(X) is given by \mathrm{ES}_{\alpha}(X) = \frac{1}{1 - \alpha} \int_{\mathrm{VaR}_{\alpha}(X)}^{\infty} x \, dF(x), where \mathrm{VaR}_{\alpha}(X) = \inf \{ x : F(x) \geq \alpha \} is the \alpha-quantile of X. An equivalent representation, derived via integration by parts or Fubini's theorem, expresses ES as the average of VaRs over the tail: \mathrm{ES}_{\alpha}(X) = \frac{1}{1 - \alpha} \int_{\alpha}^{1} \mathrm{VaR}_{u}(X) \, du. For general distributions with possible atoms at the VaR level, ES is \mathrm{ES}_{\alpha}(X) = \frac{1}{1 - \alpha} \left[ \int_{\mathrm{VaR}_{\alpha}(X)}^{\infty} x \, dF(x) + \mathrm{VaR}_{\alpha}(X) \left( P(X \geq \mathrm{VaR}_{\alpha}(X)) - (1 - \alpha) \right) \right]. For discrete distributions, where X takes values x_1 < x_2 < \cdots < x_n with probabilities p_i, the ES simplifies to the average of the losses exceeding the (1 - \alpha)-quantile, specifically \mathrm{ES}_{\alpha}(X) = \frac{1}{1 - \alpha} \sum_{x_i > \mathrm{VaR}_{\alpha}(X)} p_i x_i + \frac{\mathrm{VaR}_{\alpha}(X) (P(X \geq \mathrm{VaR}_{\alpha}(X}) - (1 - \alpha))}{1 - \alpha}. This empirical average of the tail outcomes directly follows from the integral formula when F is a step function. Monte Carlo provides a distribution-agnostic to approximate ES by generating independent samples from the distribution of X. To compute \mathrm{ES}_{\alpha}(X), draw N samples x_1, \dots, x_N, estimate \mathrm{VaR}_{\alpha}(X) as the empirical \alpha- of the samples, and then average the samples exceeding this threshold (with interpolation if necessary for the exact tail probability). This approach is particularly useful for complex dependencies or high-dimensional portfolios where analytical is infeasible. Historical simulation estimates ES using observed past loss data as a non-parametric proxy for the empirical distribution. Sort a sample of N historical losses in ascending order x_{(1)} \leq \cdots \leq x_{(N)}, identify the empirical VaR as x_{(\lceil \alpha N \rceil)}, and compute ES as the average of the largest \lfloor (1 - \alpha) N \rfloor losses, adjusted for any fractional tail. This method assumes that historical patterns repeat and requires a large for accuracy. Estimators of ES derived from these methods, such as the empirical tail average in historical or settings, exhibit asymptotic under mild conditions on the underlying distribution, meaning they converge in probability to the true ES as the sample size N \to \infty. Error bounds can be established using central limit theorems, with the variance of the scaling as O(1/N) for the tail conditional expectation.

Parametric Distribution Formulas

Expected shortfall (ES) for parametric distributions is derived from the general definition ES_α = \frac{1}{1-\alpha} \int_{\VaR_\alpha}^\infty x f(x) , dx, where \VaR_\alpha is the α-quantile and f(x) is the , assuming a known distributional form with parameters. This integral often yields closed-form expressions for common distributions used in , facilitating analytical computation of . These formulas assume the left tail for losses, with α typically close to 1 (e.g., 0.95 or 0.99), and distributions scaled appropriately. For the normal distribution N(μ, σ²), the expected shortfall is given by \ES_\alpha = \mu + \sigma \frac{\phi(\Phi^{-1}(\alpha))}{1 - \alpha}, where ϕ and Φ are the standard normal pdf and cdf, respectively. This follows from evaluating the conditional expectation beyond the VaR_α = μ + σ Φ^{-1}(α), using the properties of the normal density to integrate the tail. For the Student's t-distribution with ν > 2 degrees of freedom, location μ, and scale σ, the formula is \ES_\alpha = \mu + \sigma \frac{f_\nu^*(t_\alpha^*)}{1 - \alpha} \left( \frac{\nu + (t_\alpha^*)^2}{\nu - 1} \right), where f_ν^* is the pdf of the standardized t-distribution, and t_α^* is its α-quantile (noting the tail convention adjustment). The derivation integrates the t-density tail, leveraging the relation to the F-distribution or beta functions for the conditional moments. For low ν, this captures heavier tails than the normal case. The exponential distribution (λ), with 1/λ, has _α = -\frac{1}{\lambda} \ln(1 - \alpha) and \ES_\alpha = \VaR_\alpha + \frac{1}{\lambda}. This simple form arises because the memoryless property implies the excess over _α follows the same exponential distribution, so the conditional excess is 1/λ. For the LN(μ, σ²), the expected shortfall is \ES_\alpha = e^{\mu + \sigma^2/2} \frac{\Phi\left( \Phi^{-1}(\alpha) + \sigma \right)}{1 - \alpha}, with VaR_α = e^{μ + σ Φ^{-1}(\alpha)}. The derivation uses the moment-generating function of the underlying normal and the Mills ratio for the tail probability, common in option pricing contexts for modeling asset returns. Approximate forms exist for numerical stability when σ is large. The Laplace (double exponential) distribution with location μ and scale b has, for the upper tail relevant to loss distributions, \ES_\alpha = \mu + b \left(1 - \ln(2(1 - \alpha))\right). This is obtained by direct integration of the piecewise exponential density over the tail, highlighting symmetric heavy tails relative to . For the (Type II, generalized) with scale σ, shape ξ < 1, and threshold v, ES_α = VaR_α + \frac{σ + ξ (VaR_α - v)}{1 - ξ}, where VaR_α incorporates the survival function. If ξ ≥ 1, ES_α diverges to infinity, underscoring the inability to quantify tail risk for distributions with infinite mean, a critical issue in heavy-tailed financial modeling. The formula derives from the power-law tail integral. If v = 0, this simplifies to \ES_\alpha = \frac{\VaR_\alpha}{1 - \xi}. The generalized extreme value (GEV) distribution with location μ, scale σ > 0, and shape ξ ≠ 0 provides \ES_\alpha = \mu + \frac{\sigma}{\xi (1 - \alpha)} \left[ (1 - \alpha)^{-\xi} \gamma(1 - \xi, -\ln(1 - \alpha)) - \gamma(1 - \xi, y_\alpha) \right], where γ is the lower incomplete gamma function and y_α relates to the quantile; for ξ = 0, it limits to the Gumbel case involving exponentials. This form, used in extreme value theory for peak losses, integrates the GEV density, which unifies Fréchet, Weibull, and Gumbel types for tail extrapolation.

Non-Parametric Estimation

Non-parametric estimation of () relies on historical or simulated loss data without assuming a specific underlying , making it suitable for capturing empirical tail behaviors in financial returns. The most straightforward approach is the empirical estimator, which uses the sample average of losses exceeding the empirical () threshold. For a sample of n losses X_1, \dots, X_n, sorted in ascending order as X_{(1)} \leq \dots \leq X_{(n)}, the empirical VaR at level \alpha is X_{(k+1)} where k = \lfloor n \alpha \rfloor, and the ES is approximated as the mean of the tail losses: \hat{ES}_\alpha = \frac{1}{n - k} \sum_{i=k+1}^n X_{(i)} \approx \frac{1}{n(1-\alpha)} \sum_{i=k+1}^n X_{(i)}. This provides an unbiased under independent and identically distributed (i.i.d.) assumptions but can exhibit high variance in small samples or heavy-tailed data. To address the discreteness and variability of the empirical , methods smooth the empirical in the tail region for a continuous of ES. Scaillet (2004) proposed a -based that integrates the using a function K and h, given by \hat{ES}_p(X) = \frac{1}{np} \sum_{i=1}^n X_i A_h \left( \hat{q}(p) - X_i \right), where \hat{q}(p) is the kernel quantile estimator, and A_h is a smoothed indicator function. This approach reduces bias in tail estimation for dependent data under \beta-mixing conditions and improves finite-sample performance compared to the raw empirical method, particularly for moderate sample sizes. Assessing uncertainty in non-parametric ES estimates often involves bootstrap techniques, which resample the data to construct intervals focused on the tail. The moving block bootstrap (MBB), suitable for with weak dependence, generates B bootstrap samples of block length l, computes \hat{ES}_\alpha^{(b)} for each, and uses the empirical of these to approximate the variance \hat{\sigma}^2 = \frac{1}{B-1} \sum_{b=1}^B (\hat{ES}_\alpha^{(b)} - \bar{ES})^2 or intervals. Under conditions like continuous density and mixing rates, MBB consistently estimates the of the normalized ES, enabling reliable for financial applications. Historical simulation, a core non-parametric method, faces pitfalls in non-stationary markets where past data fails to capture structural shifts, leading to underestimation of tail risks. During the 2020 market crash, historical simulation and ES models, relying on the prior 250 trading days, systematically underestimated extreme losses as the crisis introduced unprecedented not present in recent history, resulting in widespread exceptions. This highlighted the need for adaptive weighting or filtering in empirical tails to mitigate regime-shift biases. Post-2020 developments have integrated to enhance non-parametric tail estimation, particularly using neural networks to model complex dependencies in extreme losses. Recurrent neural networks (RNNs), such as stateful variants, forecast ES by learning sequential patterns in historical returns, outperforming traditional empirical methods in out-of-sample accuracy for volatile assets like equities and cryptocurrencies. For instance, neural network-based captures conditional tail moments, providing robust ES estimates under non-stationarity, as demonstrated in applications to high-frequency data where conventional bootstraps falter.

Optimization and Uses

Portfolio Optimization

In portfolio optimization, expected shortfall (ES) serves as a coherent risk measure that captures tail risks more comprehensively than value-at-risk (VaR), enabling the formulation of mean-ES optimization problems where the objective is to minimize ES subject to constraints on expected returns. This approach contrasts with mean-VaR optimization, which often leads to non-convex problems prone to multiple local optima, whereas mean-ES optimization is inherently convex due to ES's properties as a , guaranteeing global optima and facilitating efficient solving via standard techniques. To make ES optimization tractable, linear programming approximations discretize the tail distribution by representing ES as the average of losses exceeding the VaR threshold across a finite set of scenarios, transforming the problem into a linear program that can be solved efficiently even for moderately sized portfolios. Scenario-based methods further enhance this by generating discrete stress scenarios from historical data or simulations, allowing optimization of ES as the expected loss over these scenarios weighted by their probabilities, which promotes diversification benefits inherent to ES's coherence. Despite these advances, computing ES for large portfolios remains computationally intensive due to the need to evaluate tail expectations over high-dimensional return distributions, often exacerbated by the curse of dimensionality in generation and optimization. Developments in the , including fast convex solvers and robust reduction techniques, have addressed these challenges by approximating ES with lower while preserving accuracy in portfolio weight allocations.

Regulatory and Risk Management Applications

Expected Shortfall (ES) plays a central role in the framework, where it serves as the key risk measure for determining capital requirements under the Fundamental Review of the Trading Book (FRTB). Specifically, ES is calculated at a 97.5% confidence level, replacing the previous reliance on (VaR) to better capture tail risks and reduce model risk. This shift was finalized in January 2016, but implementation has been delayed multiple times, with the EU postponing application to January 1, 2027, as announced in June 2025, aiming to enhance the robustness of capital charges during stressed market conditions. In the insurance sector, under the directive, ES has been proposed in analytical studies for calibrating solvency capital requirements (SCR), particularly for portfolios, offering a more conservative assessment than the directive's 99.5% standard formula. For instance, applying ES at a 99% confidence level has been shown in studies to increase the SCR for products by accounting for the severity of extreme losses beyond the VaR threshold. This approach ensures that capital holds are sufficient to absorb tail events in risks, though the core standard formula remains anchored to a 99.5% VaR, with ES informing sensitivity analyses and internal models. Beyond regulatory mandates, ES is integrated into (ERM) frameworks to monitor tail risks across organizations, enabling firms to quantify and mitigate potential losses from rare but severe events. In practice, ERM systems employ ES alongside other metrics to track portfolio exposures in volatile environments, such as or markets, fostering proactive and . Criticisms of ES in regulatory applications center on its potential procyclicality, where reliance on recent data can amplify capital volatility during economic downturns. In response, 2025 EU revisions to , including a November 2025 consultation on FRTB implementation, propose adjustments such as a multiplier to mitigate capital impacts and phase-in periods for certain components, with implementation delayed to January 1, 2027, and effects extending to 2029, aiming to balance risk sensitivity with systemic resilience.

Advanced Extensions

Dynamic Expected Shortfall

Dynamic expected shortfall extends the static concept to time-dependent settings, incorporating information from the filtration up to time t to assess future tail risks over a horizon \tau. Formally, it is defined as \mathrm{ES}_t(\tau) = \mathbb{E}[X_{t+\tau} \mid X_{t+\tau} > \mathrm{VaR}_t(\tau), \mathcal{F}_t], where X_{t+\tau} represents the loss of the portfolio at time t + \tau, \mathrm{VaR}_t(\tau) is the value-at-risk conditional on the information set \mathcal{F}_t, and the expectation captures the average loss beyond the VaR threshold given past observations. This formulation allows for forecasting tail risks in stochastic environments, addressing limitations of static measures by accounting for evolving market conditions. In time series applications, dynamic expected shortfall is often estimated using models like ARMA-GARCH, which capture mean reversion and conditional heteroskedasticity to model the dynamics of returns and volatility. Under such frameworks, the conditional distribution of returns is typically assumed to follow a skewed or fat-tailed form, enabling computation of time-varying ES as the integral of conditional quantiles or via . For instance, ARMA components handle serial correlation in returns, while GARCH processes model , yielding more accurate forecasts of extreme losses compared to unconditional estimates. These models have been particularly useful in volatility forecasting following the 2008 Global Financial Crisis, where heightened tail risks necessitated robust conditional measures for and capital allocation. For diffusion processes, such as those underlying in , the dynamic expected shortfall can be derived analytically under assumptions. Consider a dS_t = \mu S_t dt + \sigma S_t dW_t; the ES over horizon \tau conditional on \mathcal{F}_t is given by \mathrm{ES}_t(\tau) = S_t e^{\mu \tau} \left( \Phi\left( \frac{\Phi^{-1}(\alpha) + \sigma \sqrt{\tau}}{\sqrt{\tau}} \right) - \frac{\phi(\Phi^{-1}(\alpha))}{\alpha} \right) / \Phi\left( \Phi^{-1}(\alpha)/\sqrt{\tau} \right), where \Phi and \phi are the standard CDF and PDF, respectively, and \alpha is the confidence level. This formula adjusts for the drift and parameters, providing a for tail expectations in continuous-time settings. Multi-period extensions of expected shortfall incorporate path dependency for horizons \tau > 1, evaluating risks along simulated trajectories rather than point forecasts. In GARCH-based approaches, path-dependent ES is estimated by averaging losses exceeding the multi-period across simulated paths that reflect persistence and fat tails, often using skewed t-distributions to match conditional moments. This method avoids heavy computational burdens of full by leveraging analytical approximations, enhancing applicability in for longer horizons.

Backtesting Techniques

Backtesting expected shortfall (ES) models involves evaluating their predictive accuracy against historical loss data, typically by assessing the frequency, clustering, and magnitude of tail events. Unlike , which benefits from established unconditional coverage tests, ES backtesting faces inherent challenges due to its non-elicitability as a standalone , meaning no direct statistical test can consistently identify the true ES without additional information. This property, established by Gneiting in 2011, initially raised concerns about reliable ES validation, but subsequent research has developed robust methods by leveraging joint elicitability with VaR or focusing on tail-specific statistics. Kupiec-like tests for ES adapt the proportion-of-failures (POF) framework from backtesting by examining the likelihood ratio for the coverage of tail events, often through the number of exceedances beyond the VaR threshold. These unconditional tests verify whether the observed frequency of losses exceeding the forecasted aligns with the expected tail probability α, using a binomial likelihood ratio statistic analogous to Kupiec's 1995 POF test for . The Acerbi-Szekely statistic provides an adaptation for the magnitude, computing Z_2 = \frac{1}{T} \sum_t \frac{X_t \cdot 1\{X_t + \mathrm{VaR}_t < 0\}}{\alpha \cdot \mathrm{ES}_t}, where X_t are P&L values, T is the number of observations, and under the of correct specification, Z_2 \approx 1 (with Z_2 - 1 following approximately a mean-zero for large T), enabling rejection if deviations are significant. This test effectively captures under- or overestimation of tail magnitude without requiring parametric assumptions, making it suitable for non-parametric ES estimates derived from historical simulations. Christoffersen independence tests, originally designed for clustering, have been extended to ES by testing the conditional coverage of tail violations, ensuring exceedances are independent and not serially correlated. The test uses a approach to model the probability of consecutive ES breaches, computing a likelihood statistic that compares observed transition probabilities (e.g., from no-violation to violation states) against assumptions. If clustering is detected—indicating model underestimation during stress periods—the test rejects the null, prompting recalibration; empirical applications show it maintains adequate size in samples over 500 observations. Severity tests specifically evaluate the average magnitude of tail losses against ES predictions, addressing the measure's focus on conditional expectations beyond VaR. These tests compare the realized average of exceedances—defined as \frac{1}{N} \sum (X_t - \mathrm{VaR}_t)^- for the N observed tail events—with the forecasted ES - VaR (adjusted for the excess), often using t-statistics or orthogonal polynomial expansions to assess deviations. A recent duration-severity framework integrates this by modeling inter-exceedance durations (via Meixner polynomials) and severity sequences (via ), deriving moment conditions that test both frequency and average tail depth simultaneously; simulations confirm high power against misspecifications, with rejection rates approaching 100% for severe underreporting in large samples. The non-elicitability of ES is mitigated in hybrid VaR-ES tests, which jointly validate both measures using consistent scoring functions, as shown in Fissler and Ziegel's work. These tests employ Diebold-Mariano statistics on joint loss functions, such as S(v, e, x) = (1\{x \leq v\} - \alpha)(v - x) + \frac{1}{\alpha} (e - v) 1\{x \leq v\} (v - x) + (e - v), to compare model forecasts against benchmarks; passing requires non-rejection of superiority hypotheses, enhancing reliability for regulatory applications post-2015. Recent advancements include simulation-based p-values for ES backtesting in machine learning contexts, exemplified by e-backtesting methods that generate anytime-valid e-values through sequential betting processes on tail outcomes. These model-free approaches, detailed in Wang, Wang, and Ziegel (2025), use e-statistics like e_{ES}(x, r, z) = (x - z)^+ + (1 - p)(r - z) to compute p-values via , detecting underestimation with over 99% accuracy in simulations of non-i.i.d. data; they are particularly advantageous for ML-driven ES models, offering early flagging without distributional assumptions.

References

  1. [1]
    [PDF] Expected Shortfall: a natural coherent alternative to Value at Risk
    May 9, 2001 · We discuss the coherence properties of Expected Shortfall (ES) as a financial risk mea- sure. This statistic arises in a natural way from ...
  2. [2]
    [PDF] Minimum capital requirements for Market Risk
    • Frequency of observable prices. Global Expected Shortfall (ES): Equal weighted average of diversified ES and non-diversified partial ES capital charges for ...
  3. [3]
    [PDF] Coherent Measures of Risk - ETH Zürich
    Jul 22, 1998 · In this paper we study both market risks and non-market risks, without complete markets assumption, and discuss methods of measurement of these ...
  4. [4]
    (PDF) On the Coherence of Expected Shortfall - ResearchGate
    Aug 5, 2025 · PDF | Expected shortfall (ES) in several variants has been proposed as remedy for the deficiencies of value-at-risk (VaR) which in general ...
  5. [5]
    Coherent Measures of Risk - Artzner - 1999 - Mathematical Finance
    Dec 25, 2001 · In this paper we study both market risks and nonmarket risks, without complete markets assumption, and discuss methods of measurement of these ...
  6. [6]
    Minimum capital requirements for market risk
    Jan 14, 2019 · As in the January 2016 framework, the core features of the standard include: a clearly defined boundary between the trading book and the banking ...
  7. [7]
    [PDF] Comparative analyses of expected shortfall and value-at-risk under ...
    3 Thus, by definition, expected shortfall considers loss beyond the VaR level. Yamai and Yoshiba. (2002c) show that expected shortfall has no tail risk under ...Missing: III | Show results with:III
  8. [8]
    [PDF] On the coherence of Expected Shortfall - arXiv
    Hence WCEα is a majorant to TCEα ≥ VaRα. It is in fact the smallest coherent risk measure dominating VaRα and only depending on X through its distribution if ...
  9. [9]
    [PDF] Part 2 Axiomatic Theory of Monetary Risk Measures
    Any monetary risk measure is Lipschitz continuous with respect to. || · || ... p (X) is the p-Expected Shortfall of X under Q. This ρ is still a coherent ...<|control11|><|separator|>
  10. [10]
    [PDF] Adjusted Expected Shortfall 1 Introduction
    Aug 18, 2021 · However, the risk measure ESg is able to detect the heavier tail of X2 and penalize it with a higher capital requirement. This is because. ESg ...
  11. [11]
    [PDF] Seven Proofs for the Subadditivity of Expected Shortfall - ETH Zurich
    Oct 7, 2015 · Following Artzner et al. (1999), the key property that distinguishes any coherent risk mea- sure, such as ES, from VaR is subadditivity. More ...
  12. [12]
    [PDF] Accurate Evaluation of Expected Shortfall for Linear Portfolios with ...
    Important advantages of expected shortfall (ES) over value at risk (VaR) as a coherent risk measure ... An important special case of an elliptical distribution of ...
  13. [13]
    [PDF] Expected Shortfall
    This chapter presents the construction of Tail Value at Risk (TVaR) and the Expected Shortfall (ES), which, unlike Value at Risk, are coherent risk.
  14. [14]
    Binomial Probability Table n = 9 to 12 - Pindling.org
    (Download Excel Program) Cumulative Binomial Distribution Tables ; n, x, 0.01, 0.05, 0.1 ; 10, 0, 0.904382, 0.598737, 0.3486784 ...
  15. [15]
    Extreme Value at Risk and Expected Shortfall During Financial Crisis
    Jan 21, 2011 · This paper presents Peak Over Threshold's method and the generalized Pareto distribution of extreme value theory to measure Value at Risk and Expected Shortfall
  16. [16]
    [PDF] Credit Risk: Default, Migration and Correlation Shocks - MSCI
    First, we stress ratings migrations and defaults. Next, we stress correlation. From the loss distribution, expected shortfall at varying confidence intervals is ...
  17. [17]
    [PDF] Credit Risk Concentrations under Stress
    Oct 17, 2005 · It is therefore straightforward to calculate risk measures like Expected Loss, Value. atRisk or Expected Shortfall for the loss distribution ...
  18. [18]
  19. [19]
    [PDF] Explanatory note on the minimum capital requirements for market risk
    Jan 1, 2019 · Conceptually, the method is similar to a stress test – the capital requirement is based on the loss a bank estimates it would suffer under a ...
  20. [20]
    Conditional value-at-risk for general loss distributions - ScienceDirect
    An alternative measure that does quantify the losses that might be encountered in the tail is conditional value-at-risk, or CVaR. ... Rockafellar and Uryasev ( ...
  21. [21]
    [PDF] Conditional Value-at-Risk for General Loss Distributions
    Rockafellar, S. Uryasev, Optimization of conditional value-at-risk, Journal of. Risk 2 (2000), 21–41. [40] R. T. Rockafellar and R. J-B Wets, Variational ...
  22. [22]
    Historical Simulation, Value-at-Risk, and Expected Shortfall
    Download Citation | Historical Simulation, Value-at-Risk, and Expected Shortfall | This chapter introduces the most commonly used method for computing ...
  23. [23]
  24. [24]
    [PDF] Comparative Analyses of Expected Shortfall and Value-at-Risk
    We compare expected shortfall with value-at-risk (VaR) in three aspects: estimation errors, decomposition into risk factors, and optimization.
  25. [25]
    (PDF) Expected Shortfall – An Alternative Risk Measure to Value-at ...
    Aug 7, 2025 · ... historical simulation method when implementing Value-. at-Risk and Expected Shortfall risk models. There are two main reasons why historical.
  26. [26]
    [PDF] Conditional Value-at-Risk and Average Value-at-Risk: Estimation ...
    The average value-at-risk measure is also called the conditional value-at-risk or expected shortfall measure. (Because we discuss here “conditional” variants of ...
  27. [27]
    [PDF] IEOR E4602: Quantitative Risk Management - Columbia University
    Definition: The unconditional loss distribution is the distribution of Lt+1 given the time t composition of the portfolio and assuming the CDF of Xt+1 is given.Missing: _α^ | Show results with:_α^
  28. [28]
    [PDF] Financial Time Series Lecture 9: Value at Risk, Expected Shortfall ...
    A simple counterexample; see also Example 3.13 of Klugman, Panjer and Willmot (2008). Example 1. Suppose the CDF F`(x) of a continuous loss random variable X ...
  29. [29]
    [PDF] A Simple and Robust Approach for Expected Shortfall Estimation
    Apr 14, 2020 · Assuming the sample size is 250, the number of one year's observations, only the largest two losses are covered when estimating ES at the 99%.
  30. [30]
  31. [31]
    [PDF] Robust Portfolio Optimization with Expected Shortfall
    May 30, 2016 · Expected Shortfall is a coherent risk measure, it is also convex and hence ... in Expected Shortfall for standard elliptical distributions ...<|control11|><|separator|>
  32. [32]
    [PDF] Efficient calculation of expected shortfall contributions in large credit ...
    Nov 26, 2007 · The efficient computation of expected shortfall (15) and expected shortfall contribu- tions (17) is a challenging task for realistic portfolios ...
  33. [33]
    Forecasting Value-at-Risk and Expected Shortfall in Large Portfolios
    Feb 2, 2021 · Unfortunately, due to the curse of dimensionality, their accurate estimation and forecast in large portfolios is quite a challenge. To ...Missing: computing | Show results with:computing
  34. [34]
    [PDF] Basel IV: Revised internal models approach for market risk - PwC
    • Banks must calibrate the internal model to an expected shortfall measure and periods of significant market stress. • This metric will help to capture the ...Missing: central | Show results with:central
  35. [35]
    Solvency II solvency capital requirement for life insurance ...
    Oct 14, 2017 · In this paper, we examine what the effects are of using such a standard model when risks are calibrated using the expected shortfall instead.
  36. [36]
    Solvency II Solvency Capital Requirement for Life Insurance ...
    Jan 13, 2016 · This paper examines the consequences for a life annuity insurance company if the Solvency II Solvency Capital Requirements (SCR) are calibrated ...
  37. [37]
    The Case for Expected Shortfall in Tail Risk Management - Man Group
    Jul 23, 2025 · Its value is most obvious in cases where the originating distribution is asymmetric with long, fat tails (like the signals in Figure 6) or non- ...
  38. [38]
    Understanding Tail Risk: The Giant of Enterprise Risk Management
    Tail risk, characterized by events with low probability but high impact, is often neglected in traditional enterprise risk management (ERM).
  39. [39]
    [PDF] EBA report results from the 2023 market risk benchmarking exercise ...
    also includes the VaR results for MC simulation banks and the expected shortfall. 135. In terms of risks across different assets classes, the IQDs for VaR ...
  40. [40]
  41. [41]
    [PDF] EU Commission consultation on FRTB 2025 - Finance Watch
    Apr 22, 2025 · In the case of a prorata solution in applying the RFET, this means that a proxy will be used to compute the expected shortfall, which might lead ...
  42. [42]
    Commission proposes to postpone by one additional year the ...
    Jun 12, 2025 · On 12 June 2025, the European Commission (Commission) adopted a Delegated Regulation that postpones by one additional year – until 1 January ...
  43. [43]
    [PDF] Dynamic Semiparametric Models for Expected Shortfall (and Value ...
    Expected Shortfall (ES) is the average return on a risky asset conditional on the return being below some quantile of its distribution, ...Missing: numerical | Show results with:numerical
  44. [44]
    Dynamic semiparametric models for expected shortfall (and Value-at ...
    Having such a loss function is a stepping stone to building dynamic models for these quantities. We use recent results from Fissler and Ziegel (2016), who show ...
  45. [45]
    Extreme conditional tail risk inference in ARMA–GARCH models
    In this study, we investigate the estimation of extreme conditional Value-at-Risk (CVaR) and conditional Expected Shortfall (CES) within the framework of ...
  46. [46]
    [PDF] Forecasting Value at Risk and Expected Shortfall Using a ...
    A measure addressing this, and which can be viewed as a complement to VaR, is Expected Shortfall (ES), which is the conditional expectation of exceedances ...
  47. [47]
    Estimates of Dynamic VaR and Mean Loss Associated to Diffusion ...
    Finally, in Section 5, we consider the expected shortfall as well as a second order. VaR conditioned on the past, and obtain the corresponding estimations. 2.
  48. [48]
    Estimation of Multiple Period Expected Shortfall and Median ...
    Aug 7, 2025 · The first objective of this paper is to propose statistical methods for estimating multiple-period expected shortfall under GARCH models. In ...Missing: path- | Show results with:path-
  49. [49]
    [1507.00244] Expected Shortfall is jointly elicitable with Value at Risk
    Jul 1, 2015 · Expected Shortfall is jointly elicitable with Value at Risk - Implications for backtesting. Authors:Tobias Fissler, Johanna F. Ziegel, Tilmann ...
  50. [50]
  51. [51]
    [PDF] Backtesting Expected Shortfall - Zeliade Systems
    Dec 16, 2020 · Expected Shortfall vs Value-at-Risk. Value-at-Risk (VaR) has become a standard risk measure for financial risk management due to its conceptual.
  52. [52]
    [PDF] Backtesting expected shortfall: A quantitative evaluation - DiVA portal
    dSt = µ(St,t)dt + σ(St,t)dWt. 2.4.1 Geometric Brownian motion. An important example is the geometric Brownian motion(GBM). dSt = µStdt ...
  53. [53]
    [PDF] Backtesting the Expected Shortfall - Junior Management Science
    The traffic light test is a straightforward backtest, which detects financial institutions that underestimate their market risk fig- ures and thus exhibit too ...
  54. [54]
    [PDF] Backtesting Expected Shortfall: Accounting for both duration ... - arXiv
    May 12, 2024 · We propose an original two-part, duration-severity approach for backtesting Expected Shortfall. (ES). While Probability Integral Transform ...
  55. [55]
    [PDF] E-backtesting - Statistics and Actuarial Science
    Feb 26, 2025 · We conduct extensive simulation studies and data analysis to illustrate the advantages of the model-free backtesting method, and compare it with ...