Skip to content

FRM Exam Part II · Basel III: Finalising Post-crisis Reforms

Basel III Operational Risk Loss Data and Governance Requirements

Updated 11 October 2026 · Fact-checked

Under the Basel III Standardised Approach (SMA), a bank's own loss history sets the Internal Loss Multiplier. The loss component is 15 times average annual operational losses over ten years, using net losses above a threshold. ILM = ln(e − 1 + (LC ÷ BIC)^0.8). Capital = BIC × ILM. Data must be complete, reviewed and disclosed.

Understand Operational Risk Loss Data and Governance Requirements

Basel III replaced the old internal-model approaches for operational risk with one Standardised Approach, often called the SMA. It has two parts. The Business Indicator Component (BIC) measures how big the bank's business is. The Internal Loss Multiplier (ILM) adjusts that size-based figure up or down using the bank's own loss history. Capital = BIC × ILM.

The loss history enters through the Loss Component (LC). LC = 15 × the average annual operational risk loss over the previous ten years. The factor of 15 is fixed by the framework. A bank with heavy past losses gets a higher LC, a higher ILM and more capital. A bank with light losses can get an ILM below 1.

Because the capital number depends on the loss history, the data itself is regulated. Losses are counted net of recoveries. Banks collect them above a minimum threshold: the standard is €20,000, and supervisors may allow €100,000 for banks in the middle size bucket (BI above €1 billion and up to €30 billion). Banks need ten years of good-quality data. On first implementation, supervisors may accept five years. A bank without even that must use the BIC alone, which means ILM = 1.

Smaller banks (BI up to €1 billion) have ILM = 1 by design, so loss data does not change their capital. They still must collect it for risk management and disclosure. Supervisors can also set ILM = 1 for larger banks at national discretion.

Governance ties it together. Loss data must be comprehensive, identify the dates of occurrence, discovery and accounting, and be independently reviewed. Banks disclose their ten-year loss history and the ILM publicly. This lets markets see the data behind the capital number.

Key formulas to remember

Loss Component (LC)
LC = 15 × average annual operational risk loss over the previous 10 years
Losses are net of recoveries and above the loss data threshold. The factor 15 is fixed.
Internal Loss Multiplier (ILM)
ILM = ln( e − 1 + (LC ÷ BIC)^0.8 )
If LC = BIC, ILM = 1. If LC > BIC, ILM > 1. If LC < BIC, ILM < 1. Floor at LC = 0 is about 0.54.
Operational risk capital
Capital = BIC × ILM
RWA = capital × 12.5.
BIC marginal coefficients
12% on BI up to €1bn; 15% on BI above €1bn up to €30bn; 18% above €30bn
Coefficients apply to each slice, not to the whole BI.
Loss data threshold
€20,000 standard; €100,000 may be allowed for bucket 2 banks at national discretion
Applies to inclusion of events in the loss data used for the loss component.
Data history required
10 years of data; minimum 5 years may be accepted at first implementation
With less than 5 years of good data, the bank uses BIC only (ILM = 1).

How to solve Operational Risk Loss Data and Governance Requirements questions

Use this order for any question on SMA loss data, the loss component or related governance rules.

  1. 1Identify the bucket: BI up to €1bn means ILM = 1 and loss data does not change capital.
  2. 2Compute BIC by slicing BI: 12% up to €1bn, 15% from €1bn to €30bn, 18% above that.
  3. 3Work out annual losses: use net losses (after recoveries) above the threshold, averaged over ten years.
  4. 4Multiply the average annual loss by 15 to get LC.
  5. 5Compute the ratio LC ÷ BIC, raise it to the power 0.8, add (e − 1), and take the natural log to get ILM.
  6. 6Multiply BIC by ILM for capital. Multiply by 12.5 only if the question asks for RWA.
  7. 7For governance or data questions, check data history length, threshold, net versus gross, and disclosure.
  8. 8Sanity check: LC above BIC must give ILM above 1; LC below BIC must give ILM below 1.

Quickest way: Direction check, then calculate

When to use it: Use when options are far apart or the question only asks whether capital rises or falls.

  1. Compare LC with BIC: LC = 15 × average annual loss.
  2. If LC = BIC, ILM = 1 and capital = BIC. This is the anchor point.
  3. If LC is above BIC, capital exceeds BIC. If below, capital is lower. Often this removes two options.
  4. Only compute the power and log if options are close.
  5. For rule questions, recall: 10 years, 5-year transition, €20,000, €100,000 for bucket 2, ILM = 1 for BI up to €1bn.

Common mistakes in Operational Risk Loss Data and Governance Requirements

  • Using gross losses instead of net losses in the loss component

    Candidates see a large headline loss and forget recoveries.

    Fix: The SMA uses losses net of recoveries. Subtract recoveries before averaging.

  • Forgetting the factor 15 when computing LC

    Candidates treat the average annual loss as LC directly.

    Fix: LC = 15 × average annual loss. Always multiply before comparing with BIC.

  • Applying the BIC rate to the whole BI

    It looks like one flat rate.

    Fix: Marginal coefficients apply slice by slice. For BI of €5bn: 12% on the first €1bn and 15% on the next €4bn.

  • Applying the ILM formula to a bank with BI up to €1bn

    Candidates apply the formula mechanically.

    Fix: Those banks have ILM = 1 by design. Capital = BIC.

  • Mixing up the data history rules

    Ten years, five years and the thresholds are all numbers that look alike.

    Fix: Ten years is the standard. Five years is a transitional minimum. Below five years, use BIC only. €20,000 is the standard threshold. €100,000 is an option for bucket 2.

  • Thinking loss data is disclosed only when it affects capital

    Candidates link disclosure to use in the formula.

    Fix: Banks disclose their loss history even where ILM = 1. The point is transparency, and the data also feeds risk management.

Worked examples

Example 1

A bank has a Business Indicator of €5 billion and an average annual net operational risk loss of €60 million over the past ten years (above threshold). Compute the SMA operational risk capital.

Show the solution
  1. BI is above €1bn, so ILM applies.
  2. BIC = 12% × €1,000m + 15% × €4,000m = €120m + €600m = €720m.
  3. LC = 15 × €60m = €900m.
  4. LC ÷ BIC = 900 ÷ 720 = 1.25.
  5. 1.25^0.8 = exp(0.8 × ln 1.25) = exp(0.8 × 0.22314) = exp(0.17851) ≈ 1.1954.
  6. e − 1 = 1.71828, so the term inside the log = 1.71828 + 1.1954 = 2.9137.
  7. ILM = ln(2.9137) ≈ 1.069.
  8. Capital = 720 × 1.069 ≈ €770m.

Answer: About €770 million. ILM is slightly above 1 because LC exceeds BIC.

Example 2

A bank has BI of €800 million and an average annual net loss of €50 million over ten years. What is its SMA operational risk capital? A) €96m B) €114m C) €750m D) €60m

Show the solution
  1. BI is €800m, which is below €1bn, so the bank is in the lowest bucket.
  2. For these banks ILM = 1, whatever the loss history.
  3. BIC = 12% × €800m = €96m.
  4. Capital = BIC × 1 = €96m.
  5. The loss component (15 × 50 = €750m) is a distractor. It does not enter the capital for this bank.

Answer: A) €96 million.

Exam tips

  • Start every numerical question by checking BI against €1bn. It can save the whole calculation.
  • Memorise the anchor: LC = BIC gives ILM = 1. Use it to eliminate options.
  • Watch for traps: gross versus net losses, the missing factor 15, and the flat BIC rate.
  • For rule questions, learn each number with its purpose: ten years, five-year transition, €20,000, €100,000 for bucket 2.
  • Expect scenario questions on data quality: missing events, wrong dates or unreviewed data should be linked to governance and disclosure failings.

Practice questions from Basel III: Finalising Post-crisis Reforms

Operational Risk Loss Data and Governance Requirements: frequently asked questions

How is the loss component calculated in Basel III operational risk?

LC is 15 times the average annual operational risk loss over the previous ten years. Losses are counted net of recoveries and above the loss data threshold. LC is then compared with BIC to find the ILM.

What is the loss data threshold under the SMA?

The standard minimum threshold is €20,000. Supervisors may allow banks in the middle size bucket to use €100,000. Check the exact wording in your readings, as this is a national discretion.

How many years of loss data does a bank need?

Ten years of good-quality data are expected. On first implementation, supervisors may accept at least five years. A bank without five years of good data uses the BIC alone, so ILM = 1.

Do small banks need to collect loss data?

Yes. Banks with BI up to €1 billion have ILM = 1, so losses do not change their capital. They still need loss data for risk management, and the disclosure expectations still apply.