Skip to content

FRM Exam Part II · An Introduction to Securitisation

Securitisation Role in the Subprime Crisis: Risks and Ratings

Updated 11 October 2026 · Fact-checked

Securitisation spread subprime mortgage risk through the originate-to-distribute model. Weak lender incentives (moral hazard), investor information gaps (adverse selection), over-reliance on ratings, and underestimated default correlation in tranches made senior CDO tranches far riskier than their AAA ratings implied. Solve questions by finding which failure the scenario describes.

Understand Risks, Ratings and Lessons from the Subprime Crisis

Securitisation pools loans, such as mortgages, and sells claims on the cash flows as tranches. Losses hit the junior tranches first. Senior tranches are protected by subordination and can earn high ratings even when the underlying loans are weak.

In the originate-to-distribute (OTD) model, a lender makes loans and sells them on instead of holding them. This creates moral hazard. The originator keeps little of the loss, so it has less reason to screen borrowers carefully. Underwriting standards fell, and subprime and low-documentation loans grew. Investors faced adverse selection. The seller knows loan quality better than the buyer, so buyers risk receiving the worse loans.

Ratings made things worse. Many investors relied on agency ratings instead of their own analysis. Agencies were paid by the issuers they rated, which creates a conflict of interest. Ratings also described default probability and said little about price volatility, liquidity or how the tranche behaves in a systemic downturn. A AAA structured tranche and a AAA corporate bond were not equivalent in risk.

Correlation and model risk were central. Tranche risk depends heavily on default correlation. Models often used the Gaussian copula with a correlation that was too low and assumed it was stable. Housing prices fell across regions together, so defaults clustered. Correlation rises in stress, so senior tranches took losses that models treated as near impossible. Short data history and no prior national house-price decline added to the error.

Resecuritisation compounded the problem. Mezzanine tranches of ABS were pooled into CDOs, and a large share was rated AAA. These tranches were highly sensitive to the same housing factor, so one shock hit many layers at once. Complexity, opacity and leverage then spread the losses to banks, conduits and structured investment vehicles, and funding and liquidity stress followed.

Key formulas to remember

Tranche loss allocation
Tranche loss = min(max(Pool loss − Attachment point, 0), Detachment point − Attachment point)
Losses hit the lowest tranche first. Express pool loss and the points in the same units, as a percentage of pool or in currency.
Tranche thickness
Thickness = Detachment point − Attachment point
A thin tranche can lose its whole principal on a small change in pool loss.
Correlation effect on tranches
Higher default correlation → equity tranche risk falls; senior tranche risk rises
A general result for typical pools. Mezzanine tranches can move either way depending on structure.
Retained-interest principle
Originator retains part of the credit risk → stronger incentive to screen
This is the logic behind risk retention rules. It reduces moral hazard but does not remove it.

How to solve Risks, Ratings and Lessons from the Subprime Crisis questions

Use this method for any question on securitisation failures, ratings or the crisis.

  1. 1Identify the stage in the chain: originator, arranger, rating agency, investor or the funding market.
  2. 2Name the failure: moral hazard, adverse selection, rating reliance, correlation or model risk, or opacity and leverage.
  3. 3Check who bears the loss and who makes the decision. A gap between the two points to moral hazard.
  4. 4If numbers are given, apply attachment and detachment points to compute tranche loss.
  5. 5For correlation questions, decide the direction: higher correlation raises senior tranche risk and lowers equity tranche risk.
  6. 6Match the fix to the failure: risk retention, own credit analysis, stress testing, better data and disclosure.
  7. 7Eliminate options that overstate, such as 'AAA means no risk' or 'ratings measure liquidity'.

Quickest way: Failure-to-cause matching

When to use it: For conceptual MCQs where the stem describes a behaviour or a loss and you must name the cause or remedy.

  1. Underwriting got weaker after loans were sold: moral hazard.
  2. Seller knows more than buyer about quality: adverse selection.
  3. Investors skipped own analysis and trusted a rating: rating reliance.
  4. Senior tranches lost despite a high rating after a broad housing fall: correlation underestimated, model risk.
  5. Losses across layers of a CDO of ABS: resecuritisation concentration.
  6. Pick the answer that fits the stem most specifically.

Common mistakes in Risks, Ratings and Lessons from the Subprime Crisis

  • Treating moral hazard and adverse selection as the same thing.

    Both are information problems in the OTD model.

    Fix: Moral hazard is changed behaviour after risk is transferred (weak screening). Adverse selection is a hidden-quality problem before the trade (buyer gets worse loans).

  • Saying higher correlation lowers risk for all tranches.

    Students recall diversification and apply it to every tranche.

    Fix: Higher correlation fattens the tail of pool losses. Senior tranche risk rises, while equity tranche risk usually falls.

  • Equating a AAA structured rating with a AAA corporate rating.

    Both use the same letters.

    Fix: Structured tranches carry far more systematic and model risk, so they can fall sharply in a downturn while the rating covers only default expectation.

  • Blaming only rating agencies.

    Ratings are the most visible failure.

    Fix: Name the full chain: originators, arrangers, agencies, investors and regulators who relied on ratings for capital.

  • Computing tranche loss from the pool loss without the attachment point.

    Students allocate losses pro rata.

    Fix: Subtract the attachment point first, then cap at the tranche thickness.

Worked examples

Example 1

A CDO pool of $500 million has a mezzanine tranche attaching at 5% and detaching at 15%. Pool losses are 11%. Find the tranche loss in dollars and as a percentage of the tranche.

Show the solution
  1. Thickness = 15% − 5% = 10% of $500 million = $50 million.
  2. Pool loss above the attachment point = 11% − 5% = 6%.
  3. 6% of $500 million = $30 million.
  4. Loss is below the 10% thickness, so no cap applies.
  5. Percentage of tranche = $30 million ÷ $50 million = 60%.

Answer: The mezzanine tranche loses $30 million, which is 60% of its principal.

Example 2

A bank sells all the mortgages it originates within weeks, and its default rates rise after it relaxes income verification. Which concept best explains this, and which measure best addresses it?
A) Adverse selection; higher ratings
B) Moral hazard; risk retention by the originator
C) Model risk; a lower correlation input
D) Liquidity risk; larger tranche thickness

Show the solution
  1. The bank sells loans quickly, so it bears little of the loss.
  2. It then weakens screening. That is changed behaviour after risk transfer, which is moral hazard.
  3. The remedy is to restore the originator's stake in the loss through retaining part of the risk.
  4. Option A names the wrong concept and ratings do not fix incentives.
  5. Options C and D do not address underwriting incentives.

Answer: B) Moral hazard; risk retention by the originator.

Exam tips

  • Match the scenario to the failure first. Most conceptual questions test one clear cause.
  • Always apply attachment and detachment points in order, and cap at tranche thickness.
  • Remember the direction of correlation effects: senior risk rises, equity risk usually falls.
  • Beware absolute words such as 'always' or 'eliminates' in options on ratings and risk retention.
  • Link ratings to their limits: they measure default risk, not liquidity or price volatility.

Practice questions from An Introduction to Securitisation

Risks, Ratings and Lessons from the Subprime Crisis: frequently asked questions

How did securitisation contribute to the subprime crisis?

It let lenders pass on loans, which weakened underwriting. Complex tranches and resecuritisation hid risk, and investors relied on ratings. When house prices fell together, losses reached senior tranches and spread across the financial system.

Why did CDO ratings fail in 2007-09?

Models used low and stable default correlation, short data histories and assumed no broad national house-price fall. Agencies also faced issuer-pay conflicts. Mezzanine ABS tranches pooled into CDOs were all exposed to the same housing factor.

What is moral hazard in the originate-to-distribute model?

The originator sells loans and keeps little of the loss, so it has less reason to screen borrowers. Weaker underwriting follows. Risk retention requirements aim to restore that incentive.

How does correlation risk affect securitisation tranches?

Higher default correlation makes joint defaults more likely. That raises the loss risk of senior tranches and usually lowers the risk of the equity tranche. Correlation tends to rise in stress, so models using low values understate senior risk.