Capital released offshore is capital borrowed

Over the past decade a growing share of US annuity reserves has been reinsured to companies in Bermuda and other offshore jurisdictions, many of them affiliated with asset managers or set up as sidecars funded by outside investors. The ceding insurer typically transfers a block of fixed or indexed annuities together with the assets supporting it, receives a ceding commission or a reduction in required capital, and reports a stronger capital ratio. The reinsurer holds the business under a different capital framework, often with a different view of what reserves are needed.

I have seen these transactions from both sides. At Kerrowan Life Reinsurance in Charlotte I priced the investment assumptions for annuity blocks that insurers wanted to cede. At Wrenfield I negotiated a funds-withheld agreement in 2018 on part of our fixed annuity block, and since then I have recommended against two further proposals. My view is not that offshore reinsurance is unsound. It is that a ceding company's board should treat the capital it releases as borrowed, because in an important sense it is.

What a cession actually buys

When we cede a block, the policyholders still hold contracts with Wrenfield. If the reinsurer cannot pay, the obligation remains ours. What we have bought is a promise from the reinsurer to reimburse us, secured by whatever collateral the agreement provides. The capital we release is capital we no longer hold against that block, replaced by reliance on someone else's balance sheet.

Economically that resembles borrowing. We receive capital capacity today, and in return we depend on the reinsurer's continued performance for as long as the block runs off, which for annuities can mean decades. The cost is the reinsurer's share of the spread, adjusted for any ceding commission, and it should be compared with what equity or debt would have cost. Sometimes reinsurance is clearly cheaper. Sometimes it looks cheaper only because the reinsurer holds less capital against the same risk, and the difference is a risk we have retained without measuring it.

Our own 2018 agreement shows the arithmetic. We ceded a share of $2.4 billion of fixed annuity reserves on a funds-withheld basis, and our capital ratio rose by 28 points. In return the reinsurer took a share of the spread on that block for its remaining life. When I compared that share with the coupon we would have paid on holding-company debt raising the same capital, the reinsurance was cheaper by roughly a third, which is why I recommended it. But the comparison held only because the assets never left our balance sheet. Had the same block gone into a trust under looser guidelines, the lower price would have bought us a weaker claim, and the board would have been comparing two things that were not alike.

Here the difference between capital frameworks matters. A reinsurer working under another jurisdiction's rules may calculate reserves on a different basis from ours and may be allowed to recognise assets or expected spreads in ways our statutory accounts do not. Neither framework is necessarily wrong. But part of the capital released to the ceding company is a product of that difference, and a board should know how large a part.

Read the collateral before the rating

A reinsurer's credit rating tells a board how likely it is to pay in ordinary conditions. The collateral tells it what happens in extraordinary ones. In funds-withheld and modified coinsurance structures, the assets remain on the ceding company's balance sheet, which offers the most protection. In coinsurance with a trust, the assets sit in a trust for the ceding company's benefit, and the protection depends on the trust's investment guidelines, on how its assets are valued and on how quickly a shortfall must be made good.

When I review a proposed cession, I spend more time on the trust agreement than on the reinsurer's financial statements. Which assets may the trust hold, and who values them? May the reinsurer substitute assets, and on what notice? Must the trust be overcollateralised, and by how much? In what circumstances can we recapture the business and take the assets back, and at what cost? US regulators have paid growing attention to these questions, including through asset adequacy testing that reaches reinsured business, and I regard that attention as overdue rather than burdensome.

Concentration is the second issue. If several US insurers cede to the same reinsurer, or to reinsurers whose assets one manager runs, their policyholders share exposure to the same investment decisions. No single ceding company can see that concentration in its own data, which is one reason I prefer structures in which our assets stay on our balance sheet and under our guidelines.

Six questions before the board approves

Any board considering an asset-intensive cession should, in my view, require answers to six questions in the paper itself.

First, what would the same capital cost if raised as equity or debt, and how does the all-in cost of the reinsurance compare? Second, how much of the capital release arises because the reinsurer holds a lower reserve or capital requirement than we would, and where would we stand if that difference disappeared? Third, what collateral secures the reinsurer's obligations, who values it and how often? Fourth, what are our recapture rights, and what would recapture cost in a stress? Fifth, is the reinsurer affiliated with any of our shareholders, managers or advisers, and if so, who negotiated on our side? Sixth, how concentrated is the reinsurer's book across other US cedants?

At Wrenfield those questions led us to decline a 2023 proposal that would have lifted our capital ratio by about 40 points, because the reinsurer would have held the assets in a trust whose guidelines allowed a far higher share of privately rated structured credit than our own. The capital would have looked excellent in our statements. The risk would still have been ours, held on terms we could not control.

None of this requires a board to become expert in reinsurance law. It requires the paper to be written by someone who expects these questions, and a committee that asks them every time, including when the capital ratio is comfortable and the proposal arrives as an opportunity rather than a necessity.

Reinsurance remains one of the most useful tools an annuity company has, and we will use it again. I ask only that it appear in the board's capital plan where borrowings appear, with a counterparty, a security package and a repayment profile, and not among the sources of capital the company has generated for itself.