I don't particularly hang out with insurance and I assume that if you do today's big angry report on captive reinsurance, by Benjamin Lawsky and his New York State Department of Financial Services,1 doesn't come as much of a surprise. Still it's fun for me because of its overlap with some of my favorite financial transactions. Like, here is the basic idea of reinsurance:
You've got a company (a New York State insurance company) that has insured some risks, and is required to hold reserves and be capitalized against those risks, and that seems unfair to that company, so it goes and finds another entity (a Cayman Islands reinsurance company) with less stringent capital requirements, and it buys reinsurance from that company, reducing its own capital requirements (it has no more exposure to the reinsured risk!) without 1-for-1 increasing the reinsurer's capital requirements (better regulatory environment!), so there's a positive-sum transaction.
This in a nutshell describes all capital arbitrage transactions in banking: a big bank has a risk that brings it capital requirements, it goes and finds a pension fund or hedge fund (or insurance company!) that is willing to bear that risk, the transfer of risk from bank to fund reduces aggregate capital requirements, and a trade is made. This is a sort of "shadow banking," and in fact Lawsky's report borrows the term as "shadow insurance" to describe mechanically similar reinsurance transactions.
The fun part is what Lawsky gets mad at, which is a trade that goes something like this:
The technical term for that trade is "nothing." Nothing happened! You bought insurance from someone, and then they bought it back from you. Maybe you paid yourself some fees but you reduced no risk. But if you did it the right way you reduced your capital requirements. Where "the right way" means "a technically correct way that Benjamin Lawsky nonetheless is pretty sure is the wrong way." And, I mean, he's got a point:
As part of its investigation, DFS uncovered that 17 New York-based insurers used some form of parental guarantee to support collateral arrangements in reinsurance transactions. Those shadow insurance transactions together totaled more than $48 billion.
The simplest way to do this is that (1) the reinsurer (normally a captive reinsurer subsidiary of the insurer itself, but never mind that) posts a letter of credit from a bank to collateralize its reinsurance contract, and (2) the parent insurer guarantees the letter of credit. So:
- The reinsurer promises to pay anything the parent owes on the reinsured claims
- The LOC bank promises to pay anything the reinsurer owes on its reinsurance treaty
- The parent promises to pay the bank anything the reinsurer owes on the LOC.
For perfect circularity. Pretty good huh? There are variants of this, such as those using a "naked parental guarantee" that cuts out the LOC and the bank entirely, achieving the circularity in fewer steps though at the cost of greater obviousness,2 or just conditional letters of credit where the bank's guarantee doesn't mean much and so neither does the reinsurance. In any case the basic idea is for the insurer to keep the risk itself and just lower its capital requirements through an exercise in paper-shuffling and box-checking.
Connoisseurs will notice that this is a theme in bank capital arbitrage too. For instance there's the lovely trade that Credit Suisse did where they bought CDS from Guggenheim Partners and then entered into a non-recourse credit agreement where they agreed to pay any amounts Guggenheim owed under the CDS contract. So if CS had any losses, Guggenheim would cover them out of money that CS gave to Guggenheim. I applauded this as an aesthetic triumph, though Basel capital regulators disagreed, but in any case I guess a conclusion from this reinsurance report is that it wasn't entirely original to Credit Suisse. Insurance companies have been doing roughly the same thing roughly forever.3
This all looks a little silly when you diagram it but it's not just a case of financial-industry creativity bowling over regulatory stupidity. The core argument for allowing these trades is the same in both cases: that regulatory requirements don't match economic reality. From the Journal's article on the report:
MetLife maintains that it holds more than sufficient reserves to pay claims on its policies and has cited its reinsurance subsidiaries as a cost-effective way of addressing what it and many other publicly traded insurers contend are excessively conservative reserving requirements for certain insurance products. ...
Many insurers and state regulators have long supported a plan for the National Association of Insurance Commissioners, a group of state regulators that sets solvency standards for adoption by states, to overhaul the rules for how life insurers set up their claims reserves, potentially eliminating the need for the captives. New York opposes a move away from the existing formula-based methodology.
Which is not a terrible argument, when it's of the form "if we can find a prudent market participant who will insure this thing at arm's length and capitalize it less than we have to, we should be able to do that." I mean, market participants have been wrong before, but so have capital regulators, whatever, why assume a priori that the regulators are more right than the shadow bankers? It does seem to break down a little, though, when you're selling the insurance to yourself.
New York Sees Trouble in ‘Captive’ Reinsurers [WSJ]
Governor Cuomo Announces Investigation Uncovers Billions of Dollars in Hidden Shadow Insurance Risk that Could Threaten Policyholder and Taxpayer Interests [NYS, sic!]
Shining a Light on Shadow Insurance: A Little-known Loophole That Puts Insurance Policyholders and Taxpayers at Greater Risk [NYS DFS]
1.Who have been busy - apparently they're also cooking up something with AIG-FP. Remember them?
2.And more limited use:
Through a “naked parental guarantee,” a captive insurance subsidiary engaging in shadow insurance does not even bother to obtain a letter of credit — conditional or otherwise — as collateral. It simply promises that its parent company would cover potential losses, without identifying any specific, dedicated resources to pay for them. While New York does not allow insurers to back insurance claims with naked parental guarantees, other states allow such arrangements.
3.Also they're buying the reinsurance from their own subsidiaries, which makes it all a bit sillier / more impressive.