CLO Equity: Underwriting the Arbitrage

At SOFR+325bp on assets (SOFR = 4.30%):

• Interest income: $500m x (S + 3.25%) = 500S + $16.25m
• Interest expense: 310(S+1.20%) + 60(S+1.80%) + 30(S+2.50%) + 20(S+4.00%) = 420S + $6.35m
• Net: 80S + $9.90m = $3.44m + $9.90m = $13.34m (the equity earns SOFR on its own $80m notional because only 84% of assets are debt-funded)
• Fees: 0.50% x $500m = $2.50m, plus $0.50m = $3.00m
• Pre-loss equity cash flow: $10.34m, or 12.9% on $80m

Deduct expected credit losses: 2% defaults x 30% loss given default = 0.6% of $500m = $3.00m. Equity absorbs first loss, so $7.34m, or 9.2%. Note losses do not arrive smoothly - they arrive in clusters.

With spreads 125bp tighter: income falls by $6.25m. Pre-loss equity cash flow is $4.09m (5.1%); after the same $3.00m of expected losses it is $1.09m, or 1.4%.

The conclusion is the answer, not the arithmetic: the manager's 12-14% depends on spread levels it does not control and on a reinvestment period that will reprice the entire portfolio.

The projection is an arbitrage forecast, and every input to the arbitrage is a risk.

• Reinvestment risk. Assets are 2-year reinvestable; liabilities are fixed. If spreads tighten, the manager must reinvest at lower yields or hold cash - and both compress the equity return. The manager assumes today's spreads persist, which is the most favourable assumption in the document.
• Default correlation and timing. A benign 2% average default rate hides path dependency: a cluster in years 1-2 hits the equity before the OC tests trap cash, permanently impairing the return.
• OC/IC tests. Ask where the coverage test triggers sit relative to current levels and how much cushion exists before cash is diverted from equity to amortize the senior notes.
• Manager quality - the largest unhedged variable. A first-time CLO manager has no track record of trading around downgrades, managing a failedOC test, or working out defaults. Ask for the team's loan workout experience and the deal's replacement/eligibility criteria.
• Tranching/risk retention. Confirm the manager retains the required horizontal interest and that the equity being sold to the pension is not the retained slice being recycled.
• Valuation. The tranche is illiquid; the pension will mark it with dealer quotes in stressed periods, so reported volatility will exceed what a hold-to-maturity IRR implies.

None of these makes the trade uninvestable - but they mean the 12-14% should be haircut, not taken as the base case.

No - this is a classification error with real consequences for risk budgeting.

• Economic exposure. CLO equity is the first-loss residual on a leveraged loan portfolio. Its return is levered credit spread income plus the residual of the arbitrage, and its loss behaviour in a default cycle is equity-like, not investment-grade-like. Booking it in IG credit understates both the beta and the drawdown.
• Correlation. The plan's IG credit sleeve is likely meant to diversify equity risk. CLO equity is positively correlated with credit spreads and defaults - roughly the same factor exposure as high yield and equities, magnified by leverage.
• Mark-to-market behaviour. In a spread widening episode the tranche will be quoted far below amortized cost while the IG sleeve is expected to hold value. That mismatch produces the wrong governance response (forced selling at the bottom to stay within the sleeve's risk limits).
• Correct treatment. Classify it as opportunistic/private credit or as an equity-risk allocation, size it against the plan's loss-absorbing capital, and apply an equity-like stress test (e.g. 4-5% cumulative defaults at 30% LGD) before approving $40m. If that test shows the tranche going to zero, that is the risk being taken - it should be visible on the risk report.