Study guide

15.3 Operational Risk

FRM Part I Financial Risk Manager Study Guide Exam

Track your progress as you read

Finished sections get a check mark.

1 Introduction & Exam Overview

1.1 Current Exam Facts 1.2 Study Strategy & Domain Weights

2 Foundations: Risk Building Blocks & Governance

2.1 Building Blocks of Risk Management 2.2 How Firms Manage Financial Risk 2.3 Governance of Risk Management 2.4 Credit Risk Transfer Mechanisms

3 Foundations: Portfolio Models, Data & ERM

3.1 MPT and the CAPM 3.2 APT and Multifactor Models 3.3 Risk Data Aggregation & Reporting 3.4 Enterprise Risk Management

4 Foundations: Disasters, Crisis & Ethics

4.1 Learning from Financial Disasters 4.2 Anatomy of the Great Financial Crisis 4.3 GARP Code of Conduct

5 Quantitative Analysis: Probability Foundations

5.1 Fundamentals of Probability 5.2 Random Variables 5.3 Common Univariate Distributions 5.4 Multivariate Random Variables

6 Quantitative Analysis: Inference & Regression

6.1 Sample Moments 6.2 Hypothesis Testing 6.3 Linear Regression 6.4 Multiple Regression

7 Quantitative Analysis: Diagnostics & Time Series

7.1 Regression Diagnostics 7.2 Stationary Time Series 7.3 Non-Stationary Time Series 7.4 Returns, Volatility & Correlation

8 Quantitative Analysis: Simulation & Machine Learning

8.1 Simulation and Bootstrapping 8.2 Machine-Learning Methods 8.3 Machine Learning and Prediction

9 Financial Markets: Institutions & Derivatives Intro

9.1 Banks 9.2 Insurance Companies and Pension Plans 9.3 Fund Management 9.4 Introduction to Derivatives

10 Financial Markets: Clearing, Futures & Hedging

10.1 Exchanges and OTC Markets 10.2 Central Clearing 10.3 Futures Markets 10.4 Using Futures for Hedging

11 Financial Markets: FX, Forwards, Commodities & Options

11.1 Foreign Exchange Markets 11.2 Pricing Financial Forwards and Futures 11.3 Commodity Forwards and Futures 11.4 Options Markets

12 Financial Markets: Option Properties, Strategies & Rates

12.1 Properties of Options 12.2 Trading Strategies 12.3 Exotic Options 12.4 Properties of Interest Rates

13 Financial Markets: Bonds, MBS, Rate Futures & Swaps

13.1 Corporate Bonds 13.2 Mortgages and Mortgage-Backed Securities 13.3 Interest Rate Futures 13.4 Swaps

14 Valuation & Risk Models: VaR, Volatility & Ratings

14.1 Measures of Financial Risk 14.2 Calculating and Applying VaR 14.3 Measuring and Monitoring Volatility 14.4 External and Internal Credit Ratings

15 Valuation & Risk Models: Credit, OpRisk & Stress

15.1 Country Risk 15.2 Measuring Credit Risk 15.3 Operational Risk 15.4 Stress Testing

16 Valuation & Risk Models: Fixed-Income Valuation

16.1 Pricing, Discounting & Arbitrage 16.2 Interest Rates 16.3 Bond Yields and Return Calculations 16.4 Duration, Convexity & DV01

17 Valuation & Risk Models: Term Structure, Options & Greeks

17.1 Non-Parallel Term Structure Shifts 17.2 Binomial Trees 17.3 The Black-Scholes-Merton Model 17.4 Option Sensitivity Measures: The Greeks

15.3 Operational Risk

Key Takeaways Explain This Topic AI

Operational risk is loss from inadequate or failed internal processes, people, and systems or from external events, including legal risk but typically excluding strategic and reputational risk as primary definitions

Basel evolved from BIA and Standardized Approach through AMA toward the Standardized Measurement Approach (SMA), which blends a business indicator with historical loss experience

Loss distribution approaches combine frequency and severity—often via Monte Carlo—to produce OpRisk capital; data scarcity and truncation make modeling fragile

RCSA, KRIs, scenario analysis, and training complement capital models; insurance can transfer risk but introduces moral hazard and adverse selection

Heavy tails and power-law severity behavior imply that rare operational events dominate capital—granular average losses understate true OpRisk

Last updated: August 2026

Operational Risk

Operational risk (OpRisk) is the risk of loss resulting from inadequate or failed internal processes, people, and systems or from external events. Basel’s definition includes legal risk and generally excludes strategic and reputational risk as primary categories (though reputation damage often follows OpRisk events). VRM–7 tests taxonomies, capital approaches, loss-distribution mechanics, non-model controls, and insurance pitfalls.

OpRisk Categories (Basel Event Types)

Supervisors and banks map losses into seven Level-1 event types:

  • Event type Illustrative examples
  • Internal fraud Unauthorized trading, embezzlement, insider theft
  • External fraud Cyber theft, payment fraud, forgery by outsiders
  • Employment practices & workplace safety Discrimination claims, employee injury
  • Clients, products & business practices Mis-selling, unsuitable advice, antitrust, disclosure failures
  • Damage to physical assets Fire, natural disaster damage to premises
  • Business disruption & system failures Outages, utility failures, critical IT downtime
  • Execution, delivery & process management Failed trade processing, data entry errors, vendor failures

Cross-border cyber events, model-implementation errors, and third-party cloud outages often sit in external fraud, systems, or process categories depending on root cause—exam stems reward precise classification.

BIA, Standardized Approach, and AMA

Under Basel II-era frameworks:

Basic Indicator Approach (BIA) — capital = α × average positive annual gross income over a window (classically α = 15%). Simple, blunt, income-linked.

Standardized Approach (SA) — split activities into business lines, each with a beta factor applied to gross income; sum across lines. Still income-based, slightly more granular.

Advanced Measurement Approach (AMA) — banks use internal models (often LDA) subject to qualitative standards (use test, independent validation, scenario analysis, BEICF—business environment and internal control factors). AMA offered risk sensitivity but created comparability problems across banks.

Worked BIA sketch

If average relevant gross income = USD 2.0 billion and α = 15%, BIA capital = 0.15 × 2.0bn = USD 300 million. This number ignores whether the bank’s actual loss history is light or heavy—hence the push toward loss-sensitive measures.

SMA Rationale

The Standardized Measurement Approach (SMA) (Basel III finalization / Basel III endgame OpRisk reforms) replaces AMA for regulatory OpRisk capital in the standardized framework. Rationale:

Restore comparability and reduce model-driven capital variance across firms.

Link capital to a Business Indicator (BI) reflecting size/volume of business, adjusted by a Loss Component based on historical operational losses (with thresholds and averaging rules).

Reduce reliance on idiosyncratic internal models while still rewarding (to a degree) better loss experience versus pure income proxies.

SMA is not “zero modeling”—banks still need robust loss data collection for the loss component and for internal risk management beyond the regulatory number.

Frequency × Severity and Monte Carlo LDA

A loss distribution approach (LDA) models:

Frequency : how many OpRisk events occur in a period (often Poisson or negative binomial).

Severity : loss size given an event (lognormal, Pareto, or mixed distributions).

Annual loss = sum of severities over the random number of events. Because the convolution rarely has a closed form, banks use Monte Carlo :

Draw N ~ frequency distribution.

Draw N severities from the severity distribution.

Sum to get one annual loss path.

Repeat many times; read VaR/ES of the simulated annual loss for capital.

Worked LDA toy calculation

Suppose frequency ~ Poisson(λ = 4) per year and severity is fixed at USD 2 million for a stylized drill (unrealistic but exam-clear). Then expected annual loss = 4 × 2m = USD 8 million. With random Poisson frequency, variance of count is 4, so SD of annual loss = 2m × 2 = USD 4 million. In realistic models severity is highly skewed; a 99.9% OpRisk VaR can be tens of times EL because of rare large events.

If instead severity is lognormal with mean USD 2m and heavy right tail, Monte Carlo paths occasionally produce single events of USD 50–100m+, dominating the capital quantile. That is the economic message of OpRisk modeling: tails, not averages, set capital .

Data Issues

OpRisk data are ugly:

Sparse large losses — the events that matter for capital are rare.

Truncation and collection thresholds — small losses may be unreported; databases start above a threshold, biasing severity fits.

Reporting lag and classification error — root-cause coding drifts over time.

Heterogeneity — mixing retail process errors with mega legal settlements distorts a single severity fit.

External data and scaled industry data — needed for rare events but hard to scale to the firm’s size and controls.

Good practice: separate body and tail modeling, use scenario analysis for extreme cells, and document scaling assumptions.

Scenario Analysis

Scenario analysis elicits expert estimates of plausible severe losses (frequency and severity) for risk cells with thin data—e.g., “major cyber ransom shutting payments for a week,” “class-action mis-selling,” “building destruction at primary site.” Scenarios feed AMA-style models and remain essential under SMA for internal stress and contingency planning. Weaknesses: optimism bias, anchoring, and inconsistent facilitation. Strong programs use structured workshops, historical anchors, and independent challenge.

RCSA, KRIs, and Education

Capital is not the whole control framework:

Risk and Control Self-Assessment (RCSA) — business units identify risks, rate inherent/residual risk, and assess control effectiveness.

Key Risk Indicators (KRIs) — metrics such as system uptime, failed trade rates, overdue reconciliations, staff turnover in control functions, phishing click rates.

Training / education — reduces people-risk frequency (fraud awareness, conduct, cyber hygiene).

These tools improve the business environment and internal control factors that should, over time, show up in better loss experience.

Capital Allocation

Firm-wide OpRisk capital must be allocated to business units for accountability and pricing. Methods include proportion of BI or income, stand-alone LDA VaR with diversification haircuts, and Euler-style allocation on a firm model. Political sensitivity is high because OpRisk events are lumpy and often perceived as “someone else’s” failure. Transparent drivers (volume, complexity, control scores, loss history) reduce gaming.

Power Laws and Heavy Tails

Empirical OpRisk severities often exhibit power-law (Pareto-like) tails: P(Severity > x) ≈ C / x^α for large x. When the tail index α is low, mean severity may be finite while high quantiles explode, and sample means are unstable. Implication: ignoring the tail or fitting only a thin-tailed lognormal without a Pareto mixture understates capital. Exam cue: a few extreme legal or rogue-trading losses can dominate decades of small process losses.

Insurance: Moral Hazard and Adverse Selection

Banks buy insurance (BBB, cyber, property) to transfer OpRisk. Two classic incentive problems appear:

Moral hazard — after buying cover, the insured may underinvest in controls or take more risk because losses are shifted to the insurer (subject to deductibles, exclusions, and claims friction).

Adverse selection — firms with worse unobservable risk profiles may demand more cover; insurers respond with underwriting, exclusions, and pricing that can leave residual gaps precisely where risk is highest.

Insurance recoveries also have basis risk (policy wording versus actual loss), timeliness risk (slow pay), and counterparty risk (insurer credit). Regulators therefore limit how much insurance benefit can reduce OpRisk capital and require haircuts and documentation.

OpRisk on the FRM is equal parts taxonomy, capital evolution (BIA/SA/AMA → SMA), LDA mechanics, and governance—controls and incentives matter as much as the Monte Carlo engine.

Loading diagram... Operational Risk LDA Monte Carlo Loop