Insurance Companies and Systemic Risk

The systemic risk posed by insurance companies is something that I’ve never been entirely clear about. I know it’s an enormous issue for large insurers who want to avoid additional oversight by the Federal Reserve. I’m well aware of the usual defense, which is that insurers are not subject to bank runs because their obligations are, in large measure, pre-funded by policyholder premiums, and policyholders must pay a price in order to stop paying premiums. But this has never seemed entirely convincing to me, because some insurers are enormous players in the financial markets, and the nature of systemic risk seems to be that it can arise in unusual places.

So I find very helpful Dan and Steven Schwarcz’s new paper discussing the ins and outs of systemic risk and insurance. Because it’s written for a law review audience, it covers all the basics, so you can follow it even if you know little about insurance. They cover the usual arguments for why insurers do not pose a threat to the financial system, but then posit a number of reasons for why they could pose such a threat.

A big reason is that insurers make up a large proportion of the buy side, especially for particular markets—owning, for example, one-third of all investment-grade bonds. Furthermore, insurers tend to concentrate their purchases within certain types of securities that provide them with regulatory benefits (sound familiar?)—such as the structured products that promised higher yield while providing the investment-grade ratings that insurers needed. The big fear is that large numbers of insurers could be forced to dump similar securities at the same time, causing prices to fall and harming other types of financial institutions. This may seem unlikely, since insurers only have to make cash payouts when insurable events occur (houses burn down, people die). But insurers have to meet capital requirements just like banks, so falling asset values will require them to adjust their balance sheets.

Another major problem is that it’s not clear that insurers are prepared for those insurable events. For example, insurers are not prepared for a global pandemic, just like they weren’t prepared for large-scale terrorist attacks prior to September 11, 2001.

Finally (and I’m skipping several factors), it’s possible that entire segments of the insurance industry are under-reserving for certain types of risks. This stems from the usual cause: companies compete for market share, and the way to win share is to charge lower prices, and the way to charge lower prices is to underestimate risk. This is all good in the short term, resulting in larger bonuses, and bad in the long term, when the risk actually materializes. Yet it seems that insurance regulators are shifting to “a process of principles-based reserving (‘PBR’), which would grant insurers substantial discretion to set their own reserves based on internal models of their future exposures.” For even a casual observer of the last financial crisis, this sounds like the system is taking on a large amount of model risk and regulatory competency risk, and we know how that story ended last time.

Schwarcz and Schwarcz conclude that the federal government should play a larger role in monitoring systemic risk in the insurance industry, which will make them just about the least popular people in most insurance circles. Given the downside risks, though, it seems like pretending that there’s no reason to worry about insurers is not a good long-term strategy.

This piece is cross-posted from The Baseline Scenario with permission.