Tutorials › CAS Exam 9 - Risk Management for Actuaries › Deriving Cat Bond Principal Write-Downs

CAS Exam 9 · Securitization and Cat Bonds · Tutorial 577 of 995

Deriving Cat Bond Principal Write-Downs

Learn to convert an eligible catastrophe loss into a capped principal write-down and calculate investors’ remaining principal.

Advanced 11 min read

What You'll Learn

  • Define attachment, exhaustion, layer width, and initial tranche principal
  • Derive the proportional principal write-down from the fraction of the layer affected
  • Calculate write-downs and remaining principal below, within, and above the layer
  • Distinguish a layer’s covered loss from the bond principal reduction
  • Identify boundary errors and incorrect use of the gross loss amount

From Loss Layer to Principal Write-Down

A catastrophe bond tranche transfers a defined layer of eligible catastrophe loss. The attachment point identifies where the tranche begins to absorb loss; the exhaustion point identifies where its loss-bearing layer is fully used. To determine investors’ remaining principal, the eligible loss must be located relative to both points, then translated into a principal reduction.

This tutorial assumes the contractual loss amount has already been determined. It is a deterministic calculation for a realized loss, not a model of the probability of that loss. The amount \(X\) below means the eligible loss measure specified by the bond contract, in the same units as the attachment and exhaustion points. It should not be read as gross physical catastrophe damage unless the contract defines it that way.

Definition: Let \(A\) be a tranche’s attachment point, \(E\) its exhaustion point, and \(C\) its initial principal. Assume \(0 \leq A < E\) and \(C>0\). The layer width is \(E-A\). For an eligible loss amount \(X \geq 0\), the tranche loss within the layer is the portion of \(X\) above attachment, capped at the layer width. Under a proportional principal write-down, the principal reduction is the same fraction of \(C\) as the fraction of the layer affected.

This proportional convention means that a loss halfway through the layer reduces half the initial principal, regardless of whether the principal amount happens to equal the layer width. The principal reduction and the layer’s loss amount are therefore not automatically the same dollar amount.

$$ L(X)=\min\!\bigl(\max(X-A,0),\,E-A\bigr) $$

Here \(L(X)\) is the loss allocated to the tranche layer. The write-down \(W(X)\) and remaining principal \(R(X)\) are

$$ W(X)=C\frac{L(X)}{E-A} =C\min\!\left(1,\max\!\left(0,\frac{X-A}{E-A}\right)\right), \qquad R(X)=C-W(X). $$

The first expression makes the steps explicit: calculate the amount of loss in the layer, divide by the layer width to get the affected fraction, and apply that fraction to initial principal. The second expression writes the same calculation in one line, with the minimum and maximum enforcing the cap and floor.

Why the Proportional Formula Holds

The write-down is defined to vary in direct proportion to the layer loss while the eligible loss is between attachment and exhaustion. At \(X=A\), the layer has incurred no loss and the write-down is zero. At \(X=E\), the layer is fully affected and the write-down is all of \(C\). Linear interpolation between these two endpoints gives the formula.

$$ \frac{W(X)-0}{C-0} = \frac{X-A}{E-A} \quad \Longrightarrow \quad W(X)=C\frac{X-A}{E-A}, \qquad A<X<E. $$

For \(X\leq A\), the tranche has not attached, so the write-down is zero. For \(X\geq E\), the tranche is exhausted, so the write-down cannot exceed \(C\). Combining the interior formula with these boundary conditions gives the full piecewise result:

$$ W(X)= \begin{cases} 0, & 0\leq X\leq A,\\[4pt] C\dfrac{X-A}{E-A}, & A<X<E,\\[8pt] C, & X\geq E, \end{cases} \qquad R(X)= \begin{cases} C, & 0\leq X\leq A,\\[4pt] C\left(1-\dfrac{X-A}{E-A}\right), & A<X<E,\\[8pt] 0, & X\geq E. \end{cases} $$

At the exact boundary \(X=A\), remaining principal is \(C\); at \(X=E\), it is zero. The formulas agree at those boundaries, so there is no jump in principal at attachment or exhaustion. If \(C=E-A\), the dollar write-down within the layer equals \(X-A\). If \(C\neq E-A\), the write-down scales that layer loss by \(C/(E-A)\).

Calculating the Write-Down in Three Positions

The following examples use hypothetical contract terms and realized eligible losses. Amounts are stated in millions of dollars. No discounting, interest, mortality, or probability distribution is involved: the task is to apply the tranche’s specified proportional write-down to a given loss amount. Each example assumes no reinstatement or other adjustment to principal.

Worked Example: Loss Below Attachment

A hypothetical catastrophe bond has initial principal \(C=\$24\) million, attachment \(A=\$80\) million, and exhaustion \(E=\$120\) million. The eligible loss amount is \(X=\$65\) million. Calculate the tranche write-down and remaining principal. Use the capped-layer method.

Step 1: Calculate the layer width. The width is \(E-A=\$120-\$80=\$40\) million.

Step 2: Locate the loss. Since \(X=\$65\) million is below the \(\$80\) million attachment, \(X-A=\$65-\$80=-\$15\) million. The maximum with zero is zero, so the layer loss is \(L(X)=\min(0,\$40)=\$0\).

Step 3: Calculate the write-down. The affected fraction is \(\$0/\$40=0\). Thus \(W(X)=\$24\text{ million}\times 0=\$0\).

Step 4: Calculate remaining principal and check. \(R(X)=C-W(X)=\$24-\$0=\$24\) million. The write-down plus remaining principal is \(\$0+\$24=\$24\) million, equal to initial principal.

Common mistake avoided: Dividing the gross loss \(\$65\) million by the exhaustion point \(\$120\) million would suggest a positive reduction, but that ignores attachment. Loss below \(A\) does not enter this tranche.

Worked Example: Partial Loss Within the Layer

A hypothetical bond tranche has \(C=\$30\) million, \(A=\$50\) million, and \(E=\$90\) million. Its eligible loss amount is \(X=\$74\) million. Find the loss allocated to the layer, the principal write-down, and the remaining principal. Use the proportional layer method.

Step 1: Find the layer width. \(E-A=\$90-\$50=\$40\) million.

Step 2: Find the layer loss. The amount above attachment is \(X-A=\$74-\$50=\$24\) million. Since \(\$24\) million is less than the \(\$40\) million layer width, the cap does not bind: \(L(X)=\$24\) million.

Step 3: Find the affected fraction. \(L(X)/(E-A)=\$24/\$40=0.60\). Thus 60% of the layer is affected.

Step 4: Apply that fraction to principal. \(W(X)=0.60\times \$30=\$18\) million.

Step 5: Find remaining principal and check. \(R(X)=\$30-\$18=\$12\) million. The write-down and remaining principal sum to the original \(\$30\) million. The loss is between attachment and exhaustion, and the remaining \(\$12\) million is 40% of initial principal, as expected.

Common mistake avoided: The tranche absorbs \(\$24\) million of layer loss, but its principal is not \(\$40\) million. The write-down is therefore not automatically \(\$24\) million; it is \(\$24/\$40\) of the \(\$30\) million principal.

Worked Example: Loss Above Exhaustion

A hypothetical tranche has initial principal \(C=\$18\) million, attachment \(A=\$125\) million, and exhaustion \(E=\$155\) million. The eligible loss is \(X=\$171\) million. Determine the layer loss, principal reduction, and remaining principal. Use the cap at exhaustion.

Step 1: Calculate the layer width. \(E-A=\$155-\$125=\$30\) million.

Step 2: Cap the loss allocated to the layer. The amount above attachment is \(\$171-\$125=\$46\) million, which exceeds the \(\$30\) million width. Therefore \(L(X)=\min(\$46,\$30)=\$30\) million.

Step 3: Calculate the affected fraction and write-down. The fraction is \(\$30/\$30=1\), so \(W(X)=\$18\times1=\$18\) million.

Step 4: Calculate remaining principal and check. \(R(X)=\$18-\$18=\$0\). The write-down equals the original principal, as it must at or above exhaustion. Loss beyond \(\$155\) million does not create a principal reduction greater than \(\$18\) million for this tranche.

Common mistake avoided: Using \((X-A)/(E-A)=\$46/\$30\) without capping gives a fraction greater than one and a write-down above initial principal. The layer is fully exhausted; the write-down is capped at \(C\).

Boundary Checks and Exam Traps

A reliable calculation checks both position and principal conservation. The loss amount must be compared with the attachment and exhaustion points before interpreting a ratio. Then confirm that \(0\leq W(X)\leq C\) and that \(R(X)=C-W(X)\), so \(0\leq R(X)\leq C\).

1
Identify the contract loss amount.
Use the eligible loss measure specified for the calculation, not an unrelated gross damage figure.
2
Compare loss with attachment and exhaustion.
Below attachment, the write-down is zero; at or above exhaustion, it is all initial principal.
3
Scale within the layer.
For an interior loss, divide loss above attachment by layer width, then multiply by initial principal.
4
Check the result.
Verify the write-down is between zero and \(C\), and subtract it from \(C\) to obtain remaining principal.

A frequent exam trap is to confuse three quantities: the total eligible loss \(X\), the amount of loss allocated to the tranche \(L(X)\), and the principal reduction \(W(X)\). In general, these are different. For example, losses below attachment are part of \(X\) but not \(L(X)\); and when principal differs from layer width, \(W(X)\) differs from \(L(X)\) even inside the layer. Another trap is to calculate the partial-layer fraction using \(X/E\). The correct fraction is \((X-A)/(E-A)\) when \(A<X<E\): both the loss before attachment and the layer width matter.

Key takeaway: For a proportional write-down, first cap the loss above attachment at the layer width \(E-A\). Divide that capped layer loss by \(E-A\), multiply by initial principal \(C\), and subtract the result from \(C\). The resulting principal reduction is zero below attachment, partial within the layer, and capped at all principal at exhaustion.

Check Your Understanding

For each hypothetical tranche, calculate the requested amount. Assume the stated loss is already the contractually eligible loss measure, all amounts are in millions of dollars, and there are no reinstatements or other principal adjustments.

  1. A tranche has \(C=16\), \(A=40\), and \(E=64\). Eligible loss is \(X=35\). Find write-down and remaining principal. (Answer: write-down \(0\); remaining principal \(16\).)
  2. A tranche has \(C=25\), \(A=70\), and \(E=110\). Eligible loss is \(X=94\). Find the layer loss and write-down. (Answer: layer loss \(24\); write-down \(15\).)
  3. A tranche has \(C=12\), \(A=30\), and \(E=50\). Eligible loss is \(X=49\). Find the affected fraction and remaining principal. (Answer: affected fraction \(0.95\); remaining principal \(0.6\).)
  4. A tranche has \(C=21\), \(A=90\), and \(E=120\). Eligible loss is exactly \(X=120\). Find write-down and remaining principal, and state which boundary applies. (Answer: write-down \(21\); remaining principal \(0\); exhaustion applies.)
  5. A tranche has \(C=28\), \(A=45\), and \(E=85\). Eligible loss is \(X=102\). A colleague computes the write-down as \(28(102-45)/(85-45)\). Correct the calculation and find remaining principal. (Answer: cap layer loss at \(40\); write-down \(28\); remaining principal \(0\).)
Clinical Biostats