A fund compares a 5-year AAA(SO) bond rated on a parent guarantee at 7.9% with a plain AA plus bond of an unrelated manufacturer at 8.1%. What does the structured obligation actually protect, and which would you own?

The (SO) rating protects you only as far as the guarantee pays in full and on time, and on these assumptions the AA plus is the better bond. A 6% issuer default chance times a 20% chance the guarantee fails gives 1.2%, above the AA plus bond's 1%, for 20 basis points less yield. After expected loss the AA plus earns 8.00% against 7.78%. Even a guarantee that never fails leaves the SO bond at 7.90%, so it needs a reason other than yield, and a guarantee that is unconditional, irrevocable and paid on a fixed timetable.

It borrows someone else's credit. A student with a thin income can rent a flat because a well-off parent signs as guarantor; the landlord is really relying on the parent. A structured obligationA bond whose rating rests on a credit enhancement outside the issuer, such as a guarantee, a pledge or an escrow of cash flows, shown by the SO suffix. rating says the bond's strength comes from an enhancement, here the parent's guarantee, not from Kesarigiri's own balance sheet. Kesarigiri alone is closer to BBB plus. So the question is never how strong the AAA is, but how certain it is that the parent pays, and pays on the due date.

In two places. The first is enforceability and timing. A guarantee that is conditional, needs board approvals at the time of payment, or has no mechanism for the trustee to invoke it days before a due date, can pay late; and in a bond fund even a one-day delay is a default, a downgrade and a mark-down of the NAV. The second is correlation. A parent that has guaranteed a subsidiary is usually stressed in the same years the subsidiary is, because group companies share lenders, markets and management. The parent's strength on the day of rating says little about its strength on the day of the call.

Put the two failure points into one number. The SO bond loses money only if Kesarigiri defaults and the guarantee fails: 6% x 20% = 1.2% over five years, against 1% for the AA plus. With a 50% loss, spread over five years, the expected loss is about 0.12% a year for the SO bond and 0.10% for the AA plus. So the SO bond pays 20 basis points less and carries slightly more expected loss. Even a guarantee that can never fail does not rescue it: with zero expected loss the SO bond still earns 7.90%, below the AA plus bond's 8.00% after its own expected loss. Turn it round: the two tie only if the AA plus bond's five-year default chance is about 3.2%, more than three times the assumed 1%. So the SO bond cannot be bought for its yield; the only case for it is that its risk is easier to judge, through a parent the fund already knows well.

Close with a conditional view, because the answer turns on documents, not ratings. Own the AA plus unless the guarantee passes four checks: it is unconditional and irrevocable; it covers interest and principal; there is a payment mechanism with fixed days for the trustee to invoke it before each due date; and the parent's total guarantees are small against its own net worth. If all four hold, the SO bond's real risk is close to the parent's and 7.9% is a fair price for a parent-grade credit, though still not a better one than the AA plus. If any fails, the fund is buying a BBB plus credit at a AAA yield.