Value at Risk (VaR): Measuring Market Risk Explained for CAIIB

Value at Risk is the single most important number a bank's treasury and risk teams watch every day. And it is a cornerstone topic in the CAIIB Risk Management elective. In one figure, Value at Risk answers a deceptively simple question: how much could this portfolio lose, with a given confidence, over a given period? This guide explains what VaR means. The three main methods to compute it, how confidence levels and holding periods work, and why back-testing and limits matter for market risk measurement under the Reserve Bank of India's framework.

Market risk — the risk of loss from movements in interest rates. Equity prices, exchange rates, and commodity prices — is unavoidable for any bank with a trading book. VaR turns this uncertainty into a managed, monitored number.

What Value at Risk actually means

Value at Risk is a statistical estimate of the maximum expected loss on a portfolio over a defined holding period at a specified confidence level. Under normal market conditions. A statement such as "the one-day 99% VaR is Rs 5 crore" means there is only a 1% chance the portfolio loses more than Rs 5 crore in a single day. Three parameters define every VaR figure:

• Confidence level — commonly 95% or 99%; a higher level gives a larger, more conservative VaR.
• Holding period — the horizon over which loss is measured (1 day for trading desks, 10 days under regulatory norms).
• Base currency and portfolio — the positions and currency in which loss is expressed.

Crucially, VaR tells you the threshold of loss, not the worst case. It says nothing about how bad losses can get beyond the threshold — that gap is filled by Expected Shortfall (ES), also called Conditional VaR. Examiners love this distinction, so remember: VaR is a quantile, ES is the average loss in the tail.

The three methods to calculate VaR

CAIIB expects you to compare three standard approaches to estimating Value at Risk:

The variance-covariance method is fast and elegant but relies on the normality assumption, which fails during crises when "fat tails" appear. Historical simulation needs no distribution assumption and naturally captures observed correlations, but it is only as good as the chosen look-back window. Monte Carlo is the most flexible — it handles non-linear instruments like options well — but it demands computing power and careful model design. A good answer in the exam names each method, its mechanism, and one strength and one weakness.

Confidence levels, holding periods and scaling

Regulators and banks choose VaR parameters deliberately. The Basel framework historically used a 99% confidence level and a 10-day holding period for market risk capital. Since daily data is easier to gather. Banks often compute a 1-day VaR and scale it to 10 days using the square-root-of-time rule: a 10-day VaR is approximately the 1-day VaR multiplied by the square root of 10 (about 3.16). This shortcut assumes returns are independent and identically distributed — a simplification you should be able to flag.

The choice of confidence level reflects risk appetite: a 99% VaR is breached on roughly 2-3 trading days a year, while a 95% VaR is breached far more often. Banks set VaR limits per desk and aggregate them at the firm level, escalating breaches to senior management. Under the Basel III / FRTB evolution, the regulatory measure has been shifting from VaR toward Expected Shortfall to better capture tail risk, a development worth noting for current CAIIB syllabi. You can review the prudential context in RBI's master directions on market risk on the Reserve Bank of India website.

Back-testing, stress testing and limitations

An estimate is only trustworthy if it is validated. Back-testing compares the daily VaR against actual profit and loss: if losses exceed VaR more often than the confidence level predicts. The model is too optimistic. Basel's "traffic-light" approach classifies models into green. Amber, and red zones based on the number of exceptions over 250 trading days, with capital multipliers rising as exceptions increase.

Because Value at Risk assumes normal markets, banks supplement it with stress testing and scenario analysis to capture extreme but plausible events that VaR misses. Other well-known limitations include VaR's lack of sub-additivity in some cases (a portfolio's VaR can exceed the sum of parts). Its silence about the size of tail losses, and its sensitivity to the look-back window. Knowing these limitations — and the remedies of ES, stress testing, and limit frameworks — separates a strong CAIIB answer from a weak one.

It is also useful to place VaR within the broader market-risk governance chain. A bank's middle office independently computes VaR daily, compares it against board-approved limits, and reports breaches to the risk committee. Stress scenarios — such as a sharp rupee depreciation.

A bond yield spike, or an equity crash — are run alongside VaR so that management sees both the everyday risk and the extreme risk. Capital is then held against the larger of the modelled measures. Understanding this governance flow, not just the arithmetic, is what examiners increasingly reward in the Risk Management elective.

Studying VaR for the CAIIB Risk Management elective

To master this topic, work numerical examples by hand, memorise the three methods with their trade-offs, and be ready to interpret a VaR statement in plain English. Practising application questions under time pressure is the fastest way to build confidence. Enrol in the structured CAIIB course on iibf.store, attempt risk-focused practice tests, keep an eye on policy updates via IIBF news, and read further explainers on the iibf.store blog.

📖 Also read: three lines of defence in banks.

Conclusion: Value at Risk distils complex market risk into a single, governable number — but only when you understand its methods, parameters, and limitations. Cement your understanding by solving problems and sitting timed mocks. Start now with a CAIIB Risk Management mock test on iibf.store, or join the complete CAIIB course to clear the elective with confidence.