VaR in Investment Management: Active Risk, Tracking Error & Beyond

Published: March 30, 2026

Article by Ryan O'Connell, CFA, FRM

Value at Risk (VaR) originated in bank trading departments, but the investment management industry has increasingly adopted VaR-based frameworks for setting risk limits and measuring active risk. For asset managers with benchmark-relative mandates, VaR provides a consistent method for allocating risk budgets, setting position limits, and meeting regulatory disclosure requirements. This guide covers how investment managers apply VaR differently from trading desks, the distinction between absolute and relative VaR, and the regulatory frameworks that govern VaR disclosures for funds.

How Investment Managers Use VaR Differently from Trading Desks

The investment management industry (the “buy side”) differs fundamentally from bank trading departments (the “sell side”) in how risk is measured and controlled. These differences shape how VaR is applied in each environment.

For a bank trader, risk is straightforward: the potential dollar loss on the mark-to-market position. For investment managers, risk is more nuanced — a manager with a mandate to beat the S&P 500 is primarily concerned with deviating from that benchmark, not with absolute losses.

Active Risk and Tracking Error: The Buy-Side Perspective

Investment managers with benchmark-relative mandates measure risk through tracking error — the volatility (standard deviation) of the portfolio’s returns minus the benchmark’s returns. This captures how much the portfolio’s performance deviates from its benchmark over time.

While tracking error and relative VaR are related, they are not identical concepts:

• Tracking error measures the dispersion (volatility) of active returns around the benchmark — it captures both outperformance and underperformance symmetrically.
• Relative VaR focuses specifically on the downside — the potential underperformance relative to the benchmark at a given confidence level (e.g., 95% or 99%).

For example, a portfolio with 3% annualized tracking error will have a higher relative VaR at the 95% confidence level — the exact relationship depends on the return distribution and confidence level chosen. For the calculation details, see our Tracking Error Calculator.

For the computational details of tracking error calculation, see our Tracking Error Calculator.

Using VaR to Set Position Limits

Traditional investment guidelines — notional limits, duration gaps, sector weight caps — don’t account for diversification or correlations. A 5% limit on emerging market equities and a 5% limit on technology stocks may appear conservative individually, but if both sectors are highly correlated, the combined risk could be much higher than intended.

VaR-based position limits address this problem by focusing on risk contribution rather than notional amounts:

• Total tracking error budget: The fund specifies a maximum tracking error (e.g., 3% annualized) for the entire portfolio.
• Position contribution: Each position’s limit is set based on its marginal contribution to portfolio risk — how much adding or increasing the position affects total tracking error.
• Consistent across asset classes: A VaR-based framework treats equities, bonds, currencies, and derivatives consistently, accounting for their different volatilities and correlations.

The advantage of VaR-based limits is that they adapt to changing market conditions. When correlations spike (as they often do in crises), the system automatically signals that the portfolio’s true risk has increased — something notional limits cannot capture.

VaR in Regulatory Disclosures

Among regulatory frameworks for investment funds, UCITS stands out for explicitly requiring VaR-based risk measurement. Other regimes (PRIIPs, AIFMD, Form ADV) mandate broader risk disclosures that may incorporate VaR but don’t require it specifically:

UCITS Global Exposure Framework (EU)

Under CESR/ESMA guidelines, UCITS funds using derivatives must calculate global exposure using one of three methods:

• Commitment approach: Converts derivative positions to equivalent underlying exposures. Suitable for simple strategies.
• Relative VaR: The fund’s VaR must not exceed twice the VaR of an unleveraged reference portfolio. The reference portfolio must reflect the fund’s investment policy.
• Absolute VaR: The fund’s VaR must not exceed 20% of NAV (using a 99% confidence level, 20-business-day horizon). Used when no appropriate reference portfolio exists.

This framework represents the most direct regulatory application of VaR in investment management. Complex UCITS strategies (absolute return, leveraged, derivatives-heavy) must demonstrate that their VaR stays within prescribed limits.

PRIIPs KID (EU Retail Products)

Since January 2023, the PRIIPs Key Information Document (KID) superseded the UCITS KIID for most EU retail investors. PRIIPs uses a Summary Risk Indicator (SRI) that incorporates both market risk (volatility-based) and credit risk. While not a pure VaR measure, it draws on similar historical volatility concepts.

AIFMD (EU Alternative Funds)

The Alternative Investment Fund Managers Directive requires disclosure of leverage (using both gross and commitment methods), liquidity arrangements, and risk profile to investors. While AIFMD does not mandate VaR specifically, many alternative fund managers use VaR internally and disclose it to sophisticated investors.

Form ADV Part 2A (US)

US investment advisers registered with the SEC must disclose material risks of their significant investment strategies, methods of analysis, and types of securities in Form ADV Part 2A (the “brochure”). This is a narrative disclosure requirement, not a VaR mandate. However, many advisers to institutional clients voluntarily include quantitative risk measures like VaR or tracking error as part of their risk disclosure.

Absolute VaR vs Relative VaR

The choice between absolute and relative VaR depends on the investment mandate:

A fund manager with an absolute return mandate (targeting positive returns regardless of market conditions) should focus on absolute VaR. A fund manager benchmarked to the Russell 2000 should focus on relative VaR and tracking error. Using the wrong measure can create a false sense of security — an active equity manager with low absolute VaR but high tracking error may be taking significant benchmark-relative risk.

Performance Measurement Under a Risk Budget

VaR-based risk budgeting extends beyond setting limits — it provides a framework for evaluating manager performance on a risk-adjusted basis.

The information ratio measures how efficiently a manager converts active risk (tracking error) into active return (excess return over benchmark). A manager who generates 2% excess return with 4% tracking error is performing similarly to one who generates 1% excess return with 2% tracking error — both have an information ratio of 0.5.

For calculation details, see our Information Ratio Calculator.

In Jorion’s textbook example, the Ontario Teachers’ Pension Plan allocated “risk units” to each manager based on their tracking error budget. The fund’s analysis found that top-quartile active managers historically generated excess returns of approximately 0.65 times their tracking error — an information ratio of 0.65. Managers who consistently fell short of this benchmark became candidates for replacement or reallocation.

This approach avoids micromanaging investment decisions. As long as managers stay within their risk budget, they have flexibility to execute their strategy without approval for each trade. For a deeper framework on allocating risk budgets across managers and asset classes, see our article on Risk Budgeting.

Common VaR Mistakes in Investment Management

Even sophisticated investors make systematic errors when applying VaR:

1. Using absolute VaR for benchmark-relative mandates

A fund manager benchmarked to the MSCI World Index should measure tracking error and relative VaR, not absolute VaR. Absolute VaR ignores the benchmark entirely and can provide false comfort — a portfolio closely tracking its benchmark will have low tracking error even if absolute VaR is high.

2. Ignoring benchmark-relative risk when selecting managers

Chasing managers with the highest returns without examining their tracking error can concentrate portfolio risk. If all selected managers are overweight technology stocks, the combined portfolio may have far more active risk than intended.

3. Treating tracking error as static

Tracking error changes with market conditions. A manager with 2% tracking error in calm markets may exhibit 5% or more during volatile periods. Regular monitoring of ex-ante tracking error is essential.

4. Confusing ex-ante and ex-post tracking error

Ex-ante (forecast) and ex-post (realized) tracking error often diverge. A manager may have low historical tracking error but significant forward-looking active risk due to recent portfolio changes.

5. Comparing managers across mismatched parameters

Comparing VaR or tracking error across managers using different time horizons (daily vs monthly), confidence levels (95% vs 99%), or benchmark definitions produces misleading results. Standardization is essential for valid comparisons.

Limitations of VaR for Investment Managers

Benchmark appropriateness: A small-cap value fund benchmarked against the S&P 500 will show high tracking error even if the manager is simply investing according to mandate. Benchmark selection is a governance issue that affects all relative risk measures.

Style drift detection: VaR captures the risk of current positions but doesn’t flag gradual changes in investment style. A manager drifting from value to growth investing may maintain consistent tracking error while fundamentally changing the portfolio’s character.

Non-mark-to-market assets: Real estate, private equity, and venture capital investments are difficult to include in VaR calculations because they lack daily market prices. Many pension funds exclude these asset classes from their VaR systems or use proxy indices.

Short history: Emerging markets, new asset classes, and alternative investments may lack sufficient return history for reliable VaR estimation. This is particularly problematic for Monte Carlo and historical simulation methods.

Model assumptions: VaR assumes stable correlations and (for parametric methods) normally distributed returns. Both assumptions often fail during market crises — precisely when accurate risk measurement matters most.

For these reasons, VaR should be complemented with stress testing, scenario analysis, liquidity limits, and concentration limits. No single risk measure captures all dimensions of investment risk.