A bank earns 1.2% on a $50 billion loan book. The CFO holds capital against three different things at once: routine charge-offs already priced into the loan rate, the rare quarter where defaults spike to 4x normal, and the once-a-decade scenario where the model itself is wrong. Three numbers, three buckets, three different tools. Confuse them and you either over-hold capital and earn nothing, or under-hold and blow up.
Every financial firm exists to take risk. A bank that lent only to government-guaranteed borrowers would earn the risk-free rate and pay no shareholders. The point of risk management is not to eliminate risk. It is to make sure the firm takes the risks it understands and gets paid for, and hedges or refuses the rest.
Three jobs sit inside that mandate. Risk identification finds exposures the firm has, including ones nobody set out to take. Risk measurement quantifies size and probability, usually in dollars at a confidence level. Risk management decides what to do (accept, avoid, transfer, or mitigate) and monitors the residual.
Common mistakes
- Treating expected loss as a worst-case number. EL is the AVERAGE annual loss, priced into spreads. A loan book with 1% PD and 50% LGD has EL of 0.5%, but plausible bad-quarter losses run 2-3%.
- Confusing reserves with capital. Loan-loss reserves cover expected losses; equity capital covers unexpected losses. A bank with adequate reserves but undercapitalized is one bad quarter from insolvency. Trap: a question describes a bank with reserves equal to EL and asks if it is well-capitalized. The answer is no.
- Adding silo VaRs to get firm VaR. Standalone VaRs sum only when correlations are 1. With correlations below 1, aggregate VaR is less than the sum. With correlations above 1 (which can happen for tail co-movements in stress), aggregate is more. Most candidates default to addition and miss the diversification or stress-correlation question.
Bottom line
- Risk management is not risk avoidance: firms take risk to earn return; the job is choosing which risks, sizing them correctly, and hedging unwanted residuals (zero risk earns only the risk-free rate).
- Expected loss (PD × LGD × EAD) is priced into spreads and absorbed by reserves; unexpected loss (volatility around EL) drives economic capital; stress loss beyond UL drives the capital buffer.
- Reserves cover expected loss; equity capital covers unexpected loss, so adequate reserves alone never make a bank well-capitalized.
- Risk classes: market, credit, liquidity (funding + market), operational, business, reputational, legal, and model risk, each with different drivers and tools.
Exam shortcut
When a question gives you PD, LGD, EAD, the answer is almost always EL or capital. Multiply for EL; UL is the standard deviation of losses, NOT the product. The trap distractor reverses the EL multiplication or treats EL as the capital number. Memory aid: "EL feeds reserves, UL feeds equity, stress feeds the buffer." Three layers, three pools.
The full lesson (about 3,423 words, 23 min read) adds 2 worked examples, all 6 common mistakes, a self-check, free in the app.
Learning objectives
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10
- 11
Browse all free FRM Part I lessons or jump into free FRM Part I practice questions.