Insurer Credit Exposure in Investment Portfolios

Why Investment Portfolios Create Credit Exposure

Insurers invest premiums and capital in assets intended to support future obligations. Bonds, loans, and other debt instruments create a direct credit exposure: the issuer may fail to make contractual payments, and the insurer may recover less than the amount owed. A portfolio can therefore be exposed to credit losses even when its assets are held to support long-term liabilities.

The key is to distinguish credit loss from a change in an asset’s market price. A bond’s market value can fall because interest rates rise, even if the issuer remains able to pay. Conversely, an issuer can default and cause a credit loss even if the bond’s market price had not previously signaled that outcome. The calculations below isolate credit loss; they do not model interest-rate valuation changes.

Definition: An insurer has investment credit exposure when it holds a claim against an issuer or borrower and may lose value if that party does not meet its contractual obligations. Relevant holdings commonly include corporate and government bonds, private-placement debt, commercial paper, and loans. In each case, the specific issuer, amount exposed, and expected recovery matter.

The discussion is aligned with the CAS Exam 9 (Risk Management for Actuaries) syllabus. Earlier material on recovery rates, exposure at default, expected credit loss, and portfolio aggregation provides the basic measures used here; the emphasis is on applying them to insurer investment holdings.

Classifying Debt Holdings

An asset is a credit exposure because of the contractual claim it represents, not simply because it appears in an investment account. A corporate bond is a claim on its issuer. A loan is a claim on its borrower. Commercial paper is a short-term debt claim on its issuer. A government bond is also a debt claim; the relevant credit risk depends on the specific government and the possibility of nonpayment. An insurer may hold multiple instruments issued by the same borrower, so it is useful to identify both the instrument and the ultimate obligor.

The form of the instrument affects how exposure and recovery are assessed. Bond terms specify principal, interest, and payment dates. A loan’s outstanding balance may decline as principal is repaid, or increase if additional borrowing is permitted and drawn. Secured and unsecured claims may have different recovery prospects. For instruments with a complex payment structure, the contractual priority and the assets supporting payment can affect the amount ultimately recovered.

Holding	Credit claim	Exposure questions
Corporate bond	Contractual payments from the bond issuer	What amount would be owed at default, and what recovery is expected?
Loan	Repayment obligation of the borrower	What balance is expected to be outstanding at default?
Commercial paper	Short-term payment obligation of the issuer	Will the issuer have funds to pay at maturity, and what is recoverable if it does not?
Government debt	Payments due from a government issuer	Which government is the obligor, and what recovery assumption is appropriate?

A fund that invests in debt may give the insurer indirect exposure to several issuers. If information about the underlying holdings is available, identifying those issuers helps reveal the actual credit exposures. If it is not available, the insurer should not treat the fund as though its underlying issuer risks were known precisely.

Exposure, Recovery, and Credit Loss

For a single issuer over a specified horizon, let $EAD$ denote the exposure at default: the amount owed to the insurer if default occurs at the relevant time. Let $RR$ denote the recovery rate, the proportion of that exposure ultimately recovered under the stated assumptions. The loss given default is the unrecovered proportion. If recovery is measured as a fraction of $EAD$, and recovery costs and timing effects are ignored, then $LGD=1-RR$.

Definition: Under the simple proportional-recovery model, loss if default occurs is $EAD \times LGD$, where $LGD=1-RR$. If the probability of default over the same horizon is $PD$, expected credit loss is $PD \times EAD \times LGD$. This formula assumes the exposure, default probability, and recovery rate refer to the same issuer and horizon, and that the stated recovery rate is appropriate for the exposure being measured.

Exposure at default is not automatically the original amount invested or the asset’s current market value. For a loan, it may be the outstanding balance at the default date. For a bond, it may include principal and amounts contractually owed under the chosen exposure definition. The analyst must state which amount is being used. Similarly, recovery should be measured consistently with that exposure: applying a recovery percentage to an unrelated amount can materially misstate the loss.

The simple formula also abstracts from recovery timing and costs. If recoveries arrive later or incur costs, a more detailed model may reflect their present value or net amount. Unless such adjustments are specified, the examples use a recovery rate applied directly to exposure at default.

Why the Expected-Loss Formula Holds

The expected-loss result follows by conditioning on whether default occurs within the horizon. Assume the exposure and recovery rate are fixed, default causes loss $EAD(1-RR)$, and no credit loss occurs if there is no default. The loss random variable is therefore:

L= \begin{cases} EAD(1-RR), & \text{if default occurs},\\ 0, & \text{if default does not occur}. \end{cases}

The loss takes its nonzero value with probability $PD$. Taking its expectation gives:

E[L]=PD \times EAD(1-RR)=PD \times EAD \times LGD.

This derivation is for the stated single-exposure model. It does not imply that actual losses are certain to equal expected loss; the realized outcome may be zero or may be substantially larger. Nor does summing expected losses across exposures require the defaults to be independent: expectation is additive. Dependence matters for the distribution of total losses and their tail outcomes, not for adding these expected values.

Worked Examples

Worked Example: Expected Loss on Two Bond Holdings

A hypothetical insurer holds a municipal bond with an exposure at default of $600,000 and a corporate note with an exposure at default of $400,000. Over the same one-year horizon, assume the municipal issuer has a default probability of 0.8% and a recovery rate of 60%; assume the corporate issuer has a default probability of 2.0% and a recovery rate of 35%. Calculate each exposure’s loss if default occurs and the portfolio’s expected credit loss. Ignore recovery timing and costs.

Method. Calculate each loss given default as exposure times one minus the recovery rate, then multiply by the default probability to obtain expected loss.

For the municipal bond, $LGD=1-0.60=0.40$. Its loss if default occurs is:

\$600{,}000 \times 0.40 = \$240{,}000.

Its expected credit loss is $0.008 \times \$240{,}000=\$1{,}920$.

For the corporate note, $LGD=1-0.35=0.65$. Its loss if default occurs is:

\$400{,}000 \times 0.65 = \$260{,}000.

Its expected credit loss is $0.020 \times \$260{,}000=\$5{,}200$. Add the two expected losses: $\$1{,}920+\$5{,}200=\$7{,}120$.

Check. Each expected loss is less than its corresponding loss if default occurs, as each default probability is less than 1. The portfolio expected credit loss is $7,120. A common mistake is to apply the recovery rate as though it were the loss rate; the correct loss rate is $1-RR$.

Worked Example: A Loan Balance at Default

A hypothetical insurer makes a $5,000,000 business loan. Before the one-year horizon ends, the borrower repays $800,000 of principal. Assume another $100,000 of accrued interest is included in the amount owed at default, so $EAD=\$5,000,000-\$800,000+\$100,000$. The recovery rate on that total exposure is 35%, and the one-year default probability is 2.2%. Calculate the loss if default occurs and expected credit loss. Ignore costs and recovery timing.

Method. First determine the amount owed at the default date. Then apply the recovery rate to that exposure, and weight the resulting loss by the default probability.

The balance after principal repayment is $\$5,000,000-\$800,000=\$4,200,000$. Including accrued interest gives:

EAD=\$4{,}200{,}000+\$100{,}000=\$4{,}300{,}000.

The loss given default rate is $1-0.35=0.65$, so the loss if default occurs is:

\$4{,}300{,}000 \times 0.65=\$2{,}795{,}000.

Expected credit loss is:

0.022 \times \$2{,}795{,}000=\$61{,}490.

Check. Expected loss is less than the default loss, as expected for a 2.2% default probability. The expected credit loss is $61,490. A common mistake is to use the original $5,000,000 principal as the exposure despite the stated repayment, or to omit the accrued interest expressly included in $EAD$.

Worked Example: Different Recoveries for Claims on One Issuer

A hypothetical insurer holds two debt claims on the same borrower: a bond with $EAD=\$1,000,000$ and a 50% recovery rate, and a secured loan with $EAD=\$800,000$ and a 75% recovery rate. Assume the issuer’s default probability over the horizon is 2.5%. Calculate the total loss if the issuer defaults and the expected credit loss. Ignore recovery timing and costs.

Method. Because default by this borrower affects both claims, calculate the loss on each claim using its own recovery rate, then add the losses. Do not average the two recovery rates without weighting them by exposure.

For the bond, $LGD=1-0.50=0.50$, so its default loss is $\$1,000,000 \times 0.50=\$500,000$. For the loan, $LGD=1-0.75=0.25$, so its default loss is $\$800,000 \times 0.25=\$200,000$.

The total loss if the issuer defaults is $\$500,000+\$200,000=\$700,000$. Expected credit loss is:

0.025 \times \$700{,}000=\$17{,}500.

As a check, total exposure is $\$1,000,000+\$800,000=\$1,800,000$, and total recovery in the default scenario is $\$500,000+\$600,000=\$1,100,000$. The unrecovered amount is $\$1,800,000-\$1,100,000=\$700,000$, agreeing with the claim-by-claim calculation. The expected credit loss is $17,500.

Common Pitfalls and Portfolio Interpretation

An insurer’s investments should be reviewed by both instrument and issuer. Two different securities can create exposure to the same borrower, while securities with similar labels can have different default and recovery characteristics. When a borrower defaults, claims against it may not share the same recovery rate; contractual priority and security can matter. Use the recovery assumption for each claim when one is provided.

It is also important to keep credit loss separate from market-value movement. Suppose a bond’s price falls because market interest rates rise. That decline is a valuation change, not automatically an issuer-default loss. Likewise, an expected credit loss estimate is not a forecast that exactly that dollar amount will be lost. It is a probability-weighted average under the stated assumptions.

Do not substitute face value automatically for $EAD$. Use the exposure amount specified for the default horizon.
Do not confuse recovery with loss. If 40% is recovered, the unrecovered fraction is 60%.
Do not use an unweighted average recovery rate when claims have different exposure amounts. Calculate recoveries or losses claim by claim, then aggregate.
Do not add a market-price decline to a default loss unless the question explicitly defines a combined loss measure that includes both.

Key takeaway: For insurer debt investments, identify the obligor and the amount exposed at default, apply the recovery assumption to that exposure, and distinguish the resulting default loss from ordinary market-value changes. Expected credit loss is the default loss weighted by the matching horizon’s default probability.

Check Your Understanding

For each hypothetical exposure, use the stated recovery and default assumptions. Ignore recovery timing and costs unless a question says otherwise.

A hypothetical insurer holds a bond with $EAD=\$300,000$, a recovery rate of 55%, and a one-year default probability of 0.6%. Find the loss if default occurs and the expected credit loss. (Answer: Default loss $135,000; expected credit loss $810.)
A hypothetical loan has an exposure at default of $2,400,000, recovery of 40%, and a matching-horizon default probability of 1.25%. Calculate expected credit loss. (Answer: $18,000.)
An insurer has two claims on one hypothetical issuer: a $500,000 bond with 20% recovery and a $1,000,000 loan with 60% recovery. If the issuer’s default probability is 3%, calculate total default loss and expected credit loss. (Answer: Default loss $800,000; expected credit loss $24,000.)
A hypothetical bond has $750,000 of principal and $15,000 of accrued interest included in $EAD$. Its recovery rate is 35% and its default probability is 2.4%. Calculate the exposure and expected credit loss. (Answer: $EAD=\$765,000$; expected credit loss $11,934.)
A hypothetical bond’s market value declines from $1,140,000 to $1,080,000 as interest rates change. For a separate credit-loss calculation, use $EAD=\$1,200,000$, recovery of 45%, and default probability of 1.8%. Calculate default loss and expected credit loss, and state whether the $60,000 market-value decline is itself part of that credit-loss calculation. (Answer: Default loss $660,000; expected credit loss $11,880; the $60,000 market-value decline is not included in the stated credit-loss calculation.)

What You'll Learn