A March 11th article in Bloomberg, “Catastrophe Prevention Drives Insurance Pitch to Clearinghouses” by Matthew Leising caught our eye. It reports on an effort to provide an additional layer of default protection for CCPs in the form of insurance. Lets take a look.
The idea is to create a backstop that could provide liquidity to a CCP should all the layers above it on the risk waterfall be eaten through. It sounds like an interesting and clever idea. From the article:
“…The insurers’ proposal could ensure clearinghouses don’t become insolvent during another crisis…”
“…The GSCA insurance would kick in if all those steps weren’t enough to keep losses from snowballing,…Clearinghouses would pay for the insurance with their own money, not member funds…”
What is the capacity to provide protection?
“…About 20 insurers were brought together by GCSA LLC, a New York-based underwriter, who declined to identify them. They are capable of offering $6 billion to $10 billion of protection to clearinghouses such as CME Group Inc. and LCH.Clearnet Group Ltd…”
OK. So by underwriting the last layer of default risk, it certainly makes the CCP more robust. This insurance is not without cost. Exactly how much is hard to tell – but it won’t be cheap and will certainly be passed on to clearing members in one way or another.
The article acknowledged that insurance companies writing the policy needed to avoid “wrong way risk”. In other words, this is the correlation risk that the insurance company could not pay on the default insurance because their own exposure to the derivatives market (or financial markets in general) has deteriorated their financial stability.
“…The GCSA consortium, which is working with broker Lockton Cos., is made up of insurers that aren’t connected to the derivatives market or the banks that serve as members of clearinghouses…That’s meant to lessen the concentration of risk in the event of a bank failing, he said…”
We wonder which insurance companies have the knowledge and financial capacity to do this kind of stuff are not already “connected to the derivative market”. Not many we expect. But this is more about risk absorbing capacity. Haven’t insurance companies like Warren Buffett’s Berkshire Hathaway taken these sorts of long tail risks before – when the price was right – and would probably be our first call (well, if we had their number).
But there is a problem here. If the insurance companies underwriting the risk are indeed not already exposured to the derivatives market, they will be after writing the policies. This increases inter-connectedness and hikes up systemic risk in the process. Is that necessarily a bad thing? Well, it has been known to be. We hope the insurance regulators are keeping an eye on this. $10 billion of CCP exposure might very well be digestible for the insurance industry without creating a concentrated risk problem. Not sure.
The other issue is CCP correlation. CCPs will, in all likelihood, all get into trouble at the same time. Insurers like to avoid correlated risk. This is one reason that private insurance companies don’t write flood insurance. A bad hurricane and the policies all get hit at the same time. And it is also the reason that only the government is willing to write these polices: they have the capacity to withstand massive correlated risk. There is a lesson in there somewhere.
The insurance companies could even securitize the risk wth catastrophe bonds (CAT). These bonds pay a nice rate of return provided that some listed catastrophe does not happen. For example a CAT bond might insure against European windstorms or US East Coast hurricanes. If the issuer suffered a loss exceeding a threshold (think of it like a deductible on a car insurance policy) the excess loss would be funded from a reduction in principal on the CAT bond. The cash raised in the CAT bond is held in escrow, isolating the insurance company performance risk. The risk ends up being fully funded, removing the “wrong way risk”.
CAT bonds are appealing to many investors because they are not correlated with other markets. But a CCP CAT bond would need to find investors that are not already substantially exposed to CCPs or financials who in turn are exposed. That might be a challenge.
We applaud the ingenuity behind this kind of thinking. The risk of a meltdown that could wipe out CCPs is remote, but finding a way to hedge that long tail risk (outside of lenders of last resort) is still hard. It would certainly be interesting to see what these policies cost as a way to size the value of the tail risk.