Skip to content

FRM Exam Part I · External and Internal Credit Ratings

Internal Rating Systems and Rating Philosophy for FRM Part 1

Updated 11 October 2026 · Fact-checked

An internal rating system is a bank's own process for assigning borrowers or facilities to grades that map to PD, LGD and EAD. Banks build it from scorecards, expert judgement or both. To solve questions, identify the borrower type, the model input, and the Basel IRB parameter involved.

Understand Internal Rating Systems and Rating Philosophy

A bank cannot rely only on external agency ratings. Most borrowers, such as mid-sized firms and retail customers, are unrated. So banks build internal rating systems: they assign each borrower to a grade, and each grade has an estimated probability of default.

There are two main building blocks. A statistical model or scorecard uses data such as financial ratios, payment history and sector to produce a score. It is consistent and testable, and it needs enough default data. Expert judgement uses credit officers' experience. It is useful for large corporates, banks, sovereigns and low-default portfolios where data are thin, but it can be inconsistent and biased. Most banks combine the two: a model score is produced, then an analyst may override it within limits, and overrides are documented and monitored.

Under the Basel Internal Ratings-Based (IRB) approaches, a bank estimates risk parameters itself. PD is the probability of default over one year. LGD is the share of exposure lost if default occurs, after recoveries. EAD is the amount owed at the time of default, including drawn amounts and expected drawdown of undrawn commitments. In the foundation IRB approach the bank estimates PD, while LGD and the credit conversion factors used to get EAD are supervisory values. In the advanced IRB approach the bank estimates PD, LGD, EAD and, usually, maturity. Regulatory capital then comes from supervisory formulas using these inputs.

Rating philosophy describes how ratings respond to the economic cycle. Point-in-time (PIT) ratings reflect current conditions, so borrowers migrate across grades with the cycle and borrower PDs move with current conditions. Through-the-cycle (TTC) ratings assess creditworthiness over a full economic cycle, often assuming stress-like conditions, rather than current conditions. Ratings and the PD attached to each grade are therefore relatively stable, while realized default rates still vary with the cycle. Banks often sit between the two (hybrid). PIT gives timely risk signals but more volatile capital; TTC gives steadier capital but slower warning.

A good system must be validated: it should rank-order risk (discriminatory power), produce accurate PDs (calibration), and be stable and well governed. Grades should be neither too few nor concentrated in one grade.

Key formulas to remember

Expected loss
EL = PD × LGD × EAD
Use PD as a decimal over a one-year horizon. EL is the average loss, not the capital buffer.
LGD and recovery rate
LGD = 1 − Recovery rate
Recovery is measured as a share of exposure at default, usually after costs.
EAD with undrawn commitment
EAD = Drawn + CCF × Undrawn
CCF is the credit conversion factor, the expected share of the undrawn limit drawn by default.
Foundation vs advanced IRB
F-IRB: bank estimates PD. A-IRB: bank estimates PD, LGD, EAD and, usually, maturity
In F-IRB, LGD and the CCFs used for EAD are supervisory values, and maturity is also set by the supervisor.
Rating philosophy
PIT = responds to current conditions. TTC = through the cycle, more stable
PIT grades migrate more with the cycle; TTC grades migrate less.

How to solve Internal Rating Systems and Rating Philosophy questions

Most questions test either a calculation with PD, LGD and EAD or a concept about how the system is built and behaves. Use the same routine.

  1. 1Read what is asked: a number (EL, EAD, LGD) or a concept (model vs judgement, PIT vs TTC, F-IRB vs A-IRB).
  2. 2For a calculation, list the inputs and check units: PD in decimal, LGD as a percentage of exposure, EAD in currency.
  3. 3If undrawn limits appear, compute EAD = drawn + CCF × undrawn before anything else.
  4. 4If recovery is given, convert it to LGD = 1 − recovery.
  5. 5Compute EL = PD × LGD × EAD and check the size is sensible.
  6. 6For concept questions, match the portfolio to the method: ample data suggests scorecards; low-default or large borrowers suggest judgement.
  7. 7For philosophy questions, ask whether the rating reacts to the current cycle (PIT) or ignores it (TTC), then link to capital volatility.
  8. 8Eliminate options that confuse EL with unexpected loss, or that give the bank estimation rights not allowed under F-IRB.

Quickest way: Three-check shortcut

When to use it: Use when time is short and options look similar.

  1. Check which IRB approach is named, as it tells you who estimates LGD and EAD.
  2. Compute EAD first, then multiply by PD and LGD.
  3. For philosophy, remember PIT means volatile ratings and capital; TTC means stable ratings and slower reaction.

Common mistakes in Internal Rating Systems and Rating Philosophy

  • Using the full undrawn limit as EAD, or ignoring it

    Students forget that undrawn commitments are only partly drawn at default.

    Fix: Use EAD = drawn + CCF × undrawn.

  • Entering LGD as the recovery rate

    Both are percentages and appear together in questions.

    Fix: Always compute LGD = 1 − recovery before using it.

  • Saying banks estimate all parameters under foundation IRB

    The names IRB and internal suggest full freedom.

    Fix: Remember F-IRB: the bank estimates PD, while LGD and the CCFs used for EAD are supervisory. A-IRB: the bank estimates PD, LGD, EAD and, usually, maturity.

  • Treating TTC ratings as always giving a constant PD

    Students confuse stable grades with fixed default rates.

    Fix: TTC grades and the PD attached to each grade are relatively stable, but realized default rates still vary with the cycle. Under PIT, borrower PDs move and borrowers migrate across grades.

  • Calling expert judgement unreliable and models always superior

    Models sound objective.

    Fix: Models need data and can miss information; judgement fills gaps but needs controls. Most banks blend both.

  • Confusing expected loss with capital

    EL is the headline IRB formula.

    Fix: EL is expected and is covered by pricing and provisions; capital covers unexpected loss.

Worked examples

Example 1

A bank grants a corporate a €10 million loan, fully drawn, with a one-year PD of 2% and an expected recovery of 60%. Calculate the expected loss.

Show the solution
  1. EAD = €10 million, since the loan is fully drawn.
  2. LGD = 1 − 0.60 = 0.40.
  3. EL = PD × LGD × EAD = 0.02 × 0.40 × 10,000,000.
  4. EL = 0.008 × 10,000,000 = €80,000.

Answer: €80,000

Example 2

A borrower has a USD 4 million drawn loan and a USD 6 million undrawn committed line. The CCF is 50%, PD is 1.5% and LGD is 45%. Compute the expected loss.

Show the solution
  1. EAD = drawn + CCF × undrawn = 4 + 0.50 × 6 = USD 7 million.
  2. EL = PD × LGD × EAD = 0.015 × 0.45 × 7,000,000.
  3. 0.015 × 0.45 = 0.00675.
  4. 0.00675 × 7,000,000 = USD 47,250.

Answer: USD 47,250

Exam tips

  • Expect short calculations of EL with a twist: undrawn limits with a CCF or recovery given instead of LGD.
  • Know who estimates what under F-IRB versus A-IRB; this is a frequent concept question.
  • For PIT versus TTC, link the choice to capital volatility and early warning.
  • Be ready to say when expert judgement beats scorecards: low-default portfolios and large, complex borrowers.
  • Read whether PD is annual; multi-year horizons need adjustment.

Practice questions from External and Internal Credit Ratings

Internal Rating Systems and Rating Philosophy in other exams

The same ground in other exams, if you are preparing for more than one or want another angle on it.

Internal Rating Systems and Rating Philosophy: frequently asked questions

How do banks develop internal credit ratings?

They gather borrower data, build a scorecard or model, add qualitative expert judgement, and map the result to rating grades. Each grade is calibrated to a PD. The system is then validated and monitored.

What is the difference between foundation and advanced IRB?

Under foundation IRB the bank estimates PD, while LGD and the CCFs used for EAD are supervisory values. Under advanced IRB the bank estimates PD, LGD, EAD and, usually, maturity, subject to supervisory approval.

Expert judgement or statistical model: which is better?

Neither is always better. Models are consistent and testable but need data. Judgement helps where defaults are rare, but it can be biased. Banks usually combine them with controlled overrides.

What is the difference between point-in-time and through-the-cycle ratings?

Point-in-time ratings reflect current conditions and change quickly with the cycle. Through-the-cycle ratings look past temporary conditions and are more stable.