The tiny cat-bond market
BusinessWeek‘s article on catastrophe bonds’ performance in the wake of the Japanese earthquake reveals, I think, a lot of the miscomprehensions about this market.
The article starts off by bemoaning the fact that “the cat bond market won’t be of much help in covering Japan-related insurance losses,” partly because most of the Japan-related cat bonds were written to cover a catastrophe in Tokyo rather than anywhere in Honshu. It continues:
Insurance companies show no signs of abandoning the cat bonds, even though the market failed to deliver a big payout for the Japan disaster. Reinsurers may need to issue new securities to cover future losses in Japan. Insurers, sure to face higher premiums from their reinsurers, may do the same. Swiss Re, the world’s second-biggest reinsurer, sold $95 million of zero-coupon catastrophe bonds on Mar. 30 through its Sector Re V unit containing loss triggers that include another earthquake in Japan. There is also strong demand from pension funds, which may push up issuance of cat bonds to $6 billion or $7 billion this year, vs. $5 billion in 2010, according to Axa Investment Managers.
I’m going to pass over the repeated incantation of the word “losses” here for another day — suffice, for the time being, to say that payouts aren’t losses. Instead, it’s worth looking at the numbers scattered around the article, which reveals that cat bonds will pay out $300 million as a result of the Japanese earthquake. That’s 18% of the $1.7 billion in Japan-focused cat bond market, and 2.4% of the $12.5 billion in global cat bonds outstanding.
Neither of those numbers seems small to me. Yes, the total insurance payout in the wake of the earthquake and tsuanmi is going to be very large — somewhere in the $20 billion to $30 billion range. But it’s worth putting those numbers in perspective. The catastrophe-insurance business paid out some $43 billion in total in 2010, and about $27 billion in 2009. And net written premiums are running at a rate of about $425 billion a year. Those premiums have to cover property damage even when it isn’t caused by a natural catastrophe, of course. But at the end of last year, cat broker Napco described conditions in the insurance market as “soft,”,\ and said that “the market is now so well capitalized that catastrophe losses would have to total more than $50 billion” before anything much changed in that regard.
The purpose of cat bonds, of course, is to help reinsurers absorb the costs of truly massive insured catastrophes — ones they can’t easily afford to pay out on their own. And while the Japanese earthquake and tsunami caused a lot of economic damage, it was still within the bounds of what insurers reasonably expect to see in any given year on a global basis. So if 2.4% of cat bonds are paying out as a result of the catastrophe, that seems about right to me — insured losses are, after all, a fraction of what they would be if the earthquake had hit Tokyo, or if a hurricane hit Miami.
And in the context of those $425 billion a year in insurance premiums, cat-bond issuance of $6 billion or $7 billion this year is a drop in the bucket. I’m reminded of the optimism I saw three years ago at the Milken Global Conference, after a record $7 billion of cat bonds were issued: everybody agreed that the number was still tiny, but there was a lot of hope that it would rise fast. Instead, it fell dramatically in 2008 and then again in 2009; after a modest recovery in 2010, the best hope for 2011 was that more bonds would be issued than were actually maturing.
The fact is that catastrophe bonds are the capital-markets security of the future, and they always will be: insurers will always accept lower returns on their capital than the kind of ROI that hedge-fund cat-bond buyers are looking for. And as I mentioned back in 2008, beyond that there’s a fundamental, endemic reason why cat bonds won’t take off: the difference between parametric risk and indemnity risk.
Bond investors want to pay out based on science: magnitude of earthquakes as measured by the modified Mercalli scale; hurricanes as measured by wind speed and the like. Insurers and insured, by contrast, want their payouts based on losses. The basis risk between the two is large: Everybody can think of large losses from relatively small events, or small losses from relatively large events. And it’s not easy how that basis risk can be reduced. Unless and until that gap can be bridged, catastrophe insurance will remain the domain of the insurance industry. And the bond markets will only be involved at the very margin.