A capital charge puts Bermuda's $375 billion sidecar stack on trial

The NAIC's September National Meeting handed its Life Risk-Based Capital Working Group a single instruction—develop a capital charge on cessions to reinsurers outside reciprocal jurisdictions—and the instruction is written about a class of jurisdictions where the offshore life trade has been building. It replaces one instrument with another of a different kind: a checklist asked a ceding life insurer to confirm that the jurisdiction receiving its risk met a standard the regulators had set, while a charge asks the same insurer to hold dollars against the possibility that it did not, putting the counterparty's domicile inside the ceding company's own risk-based capital ratio.

A charge prices the reinsurer's domicile instead of auditing it, changing the arithmetic of every treaty that reaches outside the recognized list, and the target is the gap that cessions into unrecognized jurisdictions have always represented—the Cayman gap, in the shorthand this market uses—so putting a capital price on it is the substantive move. Jurisdiction risk has entered the RBC formula as an amount rather than a screening question, and the amount will do the sorting. Whatever the working group settles on, the direction is fixed: ceding to a jurisdiction without reciprocal recognition gets more expensive, and ceding to one that holds recognition does not.

The calibration is where the argument will be, and it is early: a charge set at a token level absorbs into the formula and changes nothing, while a charge levied on the wrong base—gross ceded reserves rather than the capital relief a treaty actually delivers—penalizes volume instead of risk. The memo asked for a charge and left the formula to the working group, and the distance between those two things is where ceding insurers will spend their attention.

Ceding insurers will run the arithmetic the way this market runs it: compare the charge against the relief the treaty produces, and re-cut attachments where the answer is unattractive. The treaties that survive will be the ones whose economics still work with the counterparty's domicile priced in, which is a different market from the one the sidecar trade was built in.

Bermuda's life and annuity sidecar complex is where the offshore business went, and Morningstar DBRS's figures show how far: $375 billion in sidecar liabilities, quadruple the level of four years earlier, with something quieter than the growth—no known recaptures anywhere in the book.

A recapture is the test this market runs on itself: when a ceding insurer pulls risk back from a reinsurer, it states a view about the price, the counterparty, or both, and the structure that existed on paper answers to the party that knows the underlying liabilities best. Four years of fourfold growth without a single one means a $375 billion stack has been assembled without passing that test.

There are innocent readings, and they deserve stating: the treaties may be attractive enough that no cedent wants out, and blocks of this duration rarely present a live recapture option early in a deal's life. The pool remains untested, and the working group is calibrating a charge against an asset base that has not yet produced its own evidence on whether the risk transfer holds in substance.

AM Best has been running the same inquiry from the ratings chair and is further along: the agency's read is that property-catastrophe rates are adequate but falling, and its method for under-collateralized casualty sidecars is a look-through that follows adverse development to the sponsor's balance sheet. Strip out the asset class and the principle stands on its own: below some level of collateral, the credit that matters belongs to whoever stands behind the vehicle. That is the relevant half of the view for the offshore life complex, because the test it describes—whether the collateral holds when development turns against the vehicle—is the test a life sidecar faces over a horizon measured in decades.

Axis Capital is what a completed answer looks like: AM Best affirmed the A rating and wrote its rationale around two items, the 2023 reserve strengthening and the $2.3 billion Cavello Bay loss portfolio transfer that ended the early-2020s volatility. Sponsors carrying their own development questions should read that sequence closely. The way to settle an argument about a tail is to move it off your balance sheet, absorb the charge, and let the strengthened position be the thing the rating agency writes about. Axis's cleanup happened before the review rather than during it, which is the only version that holds.

The public comparisons, meanwhile, are thinning: NEWGT Re, a Bermudian reinsurer, exited AM Best's rating process after an A- affirmation, leaving one fewer published view in a market where published views were always understood to sit behind collateral in importance. A charge that prices domicile and a method that looks through to the sponsor push the same way: the evidence that counts is held by whoever posts the collateral and, behind them, by the sponsor carrying the tail.

That points to where the offshore competition goes next. If recognition stops being the differentiator at the margin, because the charge funnels the same cessions toward the same short list of recognized jurisdictions, the differentiator becomes the quality of what a supervisor publishes about the entities it oversees. Bermuda's advantage has been the recognition itself; whether it stays an advantage depends on how much of the $375 billion can be described in terms a US regulator will accept when it builds the charge, which is the point at which the BMA's oversight of two Aegon units stops reading like a detail of a redomicile.

Aegon put a price on the structure from a different direction: the group is redomiciling its US holding company under US regulators, while $800 million of capital stays behind in Bermuda, where two of its units remain under the Bermuda Monetary Authority's supervision. The parent's supervisor changes; the subsidiaries' does not, and neither does the capital standing behind them.

If the offshore pitch is capital efficiency, and the growth of the sidecar complex suggests it is, the $800 million is what that efficiency costs to maintain: a balance sheet that is not an appendage of the American parent but a separate entity with its own supervisor and its own capital requirement. Capital that stays behind is capital doing work where the licenses are, and continued BMA supervision of the two units is how the work gets accounted for. What the redomicile establishes is that an offshore cession's cost does not end at the treaty; Aegon's figure is a legible example of the second line.

None of that is an argument against the decision: a group that keeps capital where its licenses are is doing what the structure requires, and BMA oversight is what makes the arrangement legible to everyone on the other side of the treaties. The charge for a cession and the cost of the structure that makes the cession possible are separate lines, and the RBC formula captures only one of them.

The line of business where the test arrives fastest may be the one getting the least scrutiny: ceded US health premium has more than tripled since 2016, to $203 billion, and the terms it was written on matter more than the total does. By structure the trade resembles the life sidecar business—a ceding carrier moves a long-dated obligation to a counterparty whose capacity to absorb adverse development depends on assets and collateral that take years to grade—and a market that triples in a decade signs a great many contracts before the experience that will judge them exists.

Two clocks are running on the life side as the year turns: the $22 billion Corebridge-Equitable tie-up has its approvals and a Dec. 31 close to make, consolidation of domestic balance sheets on a schedule every counterparty can see, while the NAIC charge is at the instruction stage, where nothing runs on a date certain, and the businesses it reaches are the ones whose economics depend on the parts of a term sheet nobody models until a cycle turns.

For ceding life insurers, the work before year-end is unglamorous and specific: know the collateral position behind every cession in the book, on both sides of the reciprocal line, in a form the company's own RBC calculation can consume, because the working group's calibration will be built from what cedents can document. For sidecar sponsors, the work is a repricing of the structure itself: if the sponsor's balance sheet is the effective backstop for a thinly collateralized vehicle, as AM Best's casualty method implies, then the cost of that backstop belongs in the deal economics now rather than in the next ratings conversation.

The burden has moved from showing that a jurisdiction qualifies to showing that the collateral inside it is adequate. Bermuda did the first years ago; the $375 billion has to answer the second on a record that, four years and a quadrupling into the build-out, contains no known recaptures. That is the base the working group will calibrate its charge against, and the first ceding company to take a block back will say more about the answer than any formula the group publishes.