What Risk Transfer Actually Transfers

Risk is never destroyed. It is moved to a party with a different view of it, a different capital cost against it, or a diversified book that makes it cheaper to hold.

Three mechanisms, three effects

Practitioners use "risk transfer" as a single term for arrangements that behave quite differently once tested. It is worth separating them.

Insurance responds to a defined event with an indemnity, subject to proof of loss. The insured must generally demonstrate that the event occurred, that it falls within the wording, and that the loss follows from it. Payment is therefore reliable but conditional and, in most cases, not immediate. Insurance is well suited to physical, operational and defined contingent exposures.

Derivatives respond to the movement of a reference price or index, irrespective of whether the holder suffered a loss. There is no proof-of-loss requirement and no adjustment process; settlement follows the formula. This makes derivatives fast and certain, and introduces basis risk — the possibility that the reference moves differently from the exposure it was bought to offset.

Guarantees and sureties substitute one credit for another. The obligation itself does not change; the party standing behind it does. The instrument is only as good as the guarantor, and its value to a lender is precisely the difference between the guarantor's credit standing and the obligor's.

What remains after transfer

Every transfer leaves residue, and disciplined structuring names it in advance. Four categories recur.

• Basis risk — the hedge and the exposure do not move together, leaving a gap that widens precisely when markets dislocate.
• Timing risk — the cover responds, but later than the obligation falls due, producing a liquidity failure inside an economically covered position.
• Counterparty risk — the promise is good, the promisor is not, and the exposure returns without warning.
• Definitional risk — the event that occurs is adjacent to, but not the same as, the event described in the wording.

None of these is a reason to avoid transfer. They are the reason transfer must be engineered rather than purchased.

Matching instrument to exposure

The selection rule is straightforward in principle. Where the exposure is to a defined, verifiable event, insurance is usually the efficient instrument. Where the exposure is to a continuous price or rate, a derivative is usually the efficient instrument. Where the exposure is to the performance of an identified party, a guarantee or surety is usually the efficient instrument. Where the exposure is a combination of the three — as it is in most real projects — the answer is a combination, and the difficulty lies in the joins.

The joins are where structures fail. A construction programme protected by a contractor performance bond, a delay-in-start-up policy and an interest rate hedge is well covered on paper, but the three instruments have different trigger definitions, different notice periods and different measurement dates. A single event can fall between them. The technical work is not in selecting the instruments; it is in aligning their definitions so that no gap exists between them.

Why the sponsor should not assemble this alone

Assembling a coherent transfer programme requires access to insurance capacity, derivative markets and guarantee providers simultaneously, and the ability to negotiate wordings across all three so that they interlock. Brokers operate in one market, banks in another, and neither is incentivised to police the seams between them. The sponsor is left holding the residual risk that nobody was engaged to identify.

A properly constructed programme is designed backwards from the risks the capital providers refuse to price, not forwards from the products the market happens to be selling.

A risk you have transferred is not a risk you have understood. The understanding has to come first.