FRM Exam Part II · Range of Practices and Issues in Economic Capital Frameworks
Economic Capital Risk Measurement by Risk Type
Updated 11 October 2026 · Fact-checked
Economic capital is the loss buffer a bank needs at a chosen confidence level over a set horizon. Banks model each risk type separately: credit with portfolio loss models, market with VaR or ES, operational with frequency and severity, and IRRBB with earnings or value sensitivity. Then they aggregate the results.
Understand Risk Measurement by Risk Type
Economic capital is the capital a bank thinks it needs to absorb unexpected losses at a chosen confidence level and horizon, usually one year. It is an internal measure. Regulatory capital is a rule-based measure. The two can differ a lot.
Banks do not use one model for everything. Each risk type has different data and a different loss shape. So they build a separate measure for each and then combine them. The key point for the exam is that the methods are not directly comparable.
Credit risk is usually the largest block. Banks model a portfolio loss distribution from PD, LGD, EAD and default correlation. Common approaches are structural-type models, CreditMetrics-style models and actuarial models like CreditRisk+. Economic capital is typically the loss at a high percentile minus expected loss. Data limits are real: defaults are rare, and correlations and LGDs are hard to estimate, especially for large corporate and sovereign exposures.
Market risk is the easiest to quantify because prices are observed daily. Banks use VaR or expected shortfall by historical simulation, parametric or Monte Carlo methods. The horizon is usually short, such as 10 days, so it must be scaled or adjusted to the capital horizon. Illiquid positions need a longer liquidity horizon. Models can understate tail risk.
Operational risk has the weakest data. Common approaches are the loss distribution approach (separate frequency and severity distributions combined by simulation), scenario analysis, and scorecards. Internal loss data is scarce for rare, severe events, so banks add external data and expert scenarios. Severity is heavy-tailed, so results are very sensitive to the tail assumptions.
Interest rate risk in the banking book (IRRBB) is often measured by sensitivity of economic value or net interest income to rate shocks and simulated rate paths. Hard parts are modelling non-maturity deposits, prepayment options and behavioural assumptions. Some banks include IRRBB in economic capital; practice varies. Other risks, such as business, strategic, reputational and model risk, are handled with simple methods or judgement, or left out.
Key formulas to remember
- Economic capital (credit, typical form)
- EC = Loss at confidence level α − Expected Loss
- Capital covers unexpected loss only. Expected loss is covered by provisions and pricing.
- Expected loss
- EL = PD × LGD × EAD
- Per exposure, over the horizon.
- Horizon scaling of VaR (square-root-of-time)
- VaR(T days) ≈ VaR(1 day) × √T
- Valid only under i.i.d. returns and constant positions. It is a rough approximation.
- Operational loss distribution approach
- Annual loss = sum of N individual losses, N ~ frequency distribution, each loss ~ severity distribution
- Combined by Monte Carlo. Capital is read at a high percentile of the aggregate loss.
- Aggregation with correlation (two risks)
- EC_total = √(EC₁² + EC₂² + 2ρ × EC₁ × EC₂)
- A variance-covariance shortcut. Total is below the simple sum if ρ < 1.
How to solve Risk Measurement by Risk Type questions
Use this method for any question on measuring economic capital by risk type.
- 1Identify the risk type in the question: credit, market, operational, IRRBB or other.
- 2Recall the standard method for that type and the key inputs it needs.
- 3Check the confidence level, horizon and whether capital is unexpected loss only or total loss.
- 4Look for the data limit the question is hinting at: rare defaults, short market horizon, scarce loss data, deposit behaviour.
- 5If a calculation is needed, do expected loss, percentile loss or scaling step by step and keep units consistent.
- 6If aggregation is involved, check the correlation assumption and whether the sum or a diversified figure is asked.
- 7Pick the option that states the method and its limitation correctly. Reject options with absolute words like always or never.
Quickest way: Match risk type to method and weakness
When to use it: Use for conceptual multiple-choice questions where you must pick the correct statement.
- Credit: portfolio loss distribution; weakness is default correlation and sparse data.
- Market: VaR or ES; weakness is horizon scaling and tail risk.
- Operational: frequency and severity or scenarios; weakness is scarce tail data.
- IRRBB: economic value or earnings sensitivity; weakness is behavioural assumptions.
- Eliminate any option that says one method fits all risk types or that results are directly comparable.
Common mistakes in Risk Measurement by Risk Type
Treating economic capital as total expected plus unexpected loss.
Students mix up the loss percentile with the capital figure.
Fix: Remember capital covers unexpected loss: percentile loss minus expected loss, unless the question defines it otherwise.
Assuming the same horizon and confidence level apply to all risk types.
Market VaR is quoted at 10 days, credit at one year.
Fix: Check each measure's horizon and confidence level, and adjust to a common basis before aggregating.
Saying market risk models have no data problems.
Market data is plentiful, so it looks easy.
Fix: Recall tail risk, illiquid positions, liquidity horizons and the weakness of square-root-of-time scaling.
Adding risk-type capital figures and calling it diversified capital.
Simple summing is quick.
Fix: A plain sum assumes perfect correlation. Diversified capital is lower when correlation is below 1.
Believing operational risk models are as reliable as market models.
Both produce a percentile number.
Fix: Operational data is scarce and heavy-tailed, so results are highly uncertain. Expect scenario and external data to be used.
Ignoring IRRBB because it is not a trading risk.
Students focus on trading book VaR.
Fix: Remember IRRBB arises from the banking book, with deposit and prepayment behaviour as the main modelling difficulty.
Worked examples
Example 1
A bank's credit portfolio has a 99.9th percentile one-year loss of USD 480 million and an expected loss of USD 120 million. What is the credit economic capital, and what does it cover?
Show the solution
- Economic capital = percentile loss − expected loss.
- EC = 480 − 120 = USD 360 million.
- Expected loss of USD 120 million is covered by provisions and pricing, not capital.
Answer: USD 360 million, covering unexpected loss at 99.9% over one year.
Example 2
A bank has stand-alone economic capital of USD 300 million for credit risk and USD 400 million for market risk. Assuming correlation of 0.5, what is the aggregated capital using the variance-covariance formula?
Show the solution
- EC_total = √(300² + 400² + 2 × 0.5 × 300 × 400).
- 300² = 90,000 and 400² = 160,000.
- 2 × 0.5 × 300 × 400 = 120,000.
- Sum = 90,000 + 160,000 + 120,000 = 370,000.
- √370,000 ≈ 608.3.
Answer: About USD 608 million, below the simple sum of USD 700 million because correlation is less than 1.
Exam tips
- Expect questions asking which method suits which risk type, or which limitation is correct. Learn the pairings above.
- Read whether the question asks for capital or for total loss. Subtract expected loss for credit capital.
- For operational risk, think frequency, severity and scarce tail data. Questions often test the data limit.
- Watch for absolute words in options. Practices vary across banks, so always or never is usually wrong.
- For aggregation, check correlation. Total below the sum signals diversification.
Practice questions from Range of Practices and Issues in Economic Capital Frameworks
- A bank's EC framework is built by a central risk team. Business-line heads complain that the numbers are a 'black box' and ignore them when …
- A bank computes economic capital for credit, market and operational risk separately and then adds the three figures to obtain total capital,…
- A bank's one-year economic capital at 99.9% is 800 million using a model with estimated standard error of 10% on the tail loss estimate due …
- Which practice best addresses data quality risk in an economic capital framework?
- A bank's internal audit review finds that the economic capital model is used for pricing and limit setting, but the model's assumptions, inc…
Risk Measurement by Risk Type in other exams
The same ground in other exams, if you are preparing for more than one or want another angle on it.
Risk Measurement by Risk Type: frequently asked questions
How do banks measure operational risk economic capital?
Most commonly with a loss distribution approach. The bank models loss frequency and loss severity separately, simulates annual losses, and reads capital at a high percentile. Scenario analysis and external data fill gaps in internal loss data.
Why is credit risk harder to model than market risk?
Defaults are rare and the horizon is long, so there is little data. Default correlation and LGD are also hard to estimate. Market risk has frequent price observations.
Is interest rate risk in the banking book always in economic capital?
No. Practice varies across banks. Where it is included, it is typically measured through economic value or earnings sensitivity, and behavioural assumptions on deposits and prepayments are the main difficulty.
Does economic capital equal regulatory capital?
No. Economic capital is a bank's internal estimate of the capital needed for its own risks. Regulatory capital is set by rules such as Basel, and the two can differ.