Cat bonds wouldn’t have helped the Philippines

By Felix Salmon
November 13, 2013
might well have been the biggest and strongest storm in recorded history, with wind speeds exceeding 200mph and hurricane-force winds extending more than 50 miles from the storm's eye.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Super Typhoon Haiyan might well have been the biggest and strongest storm in recorded history, with wind speeds exceeding 200mph and hurricane-force winds extending more than 50 miles from the storm’s eye. Moody’s estimates that half of the Philippines’ sugar cane crop has been destroyed, along with a third of its rice-growing fields. Most devastatingly, thousands of people were killed by the storm. In other words, Haiyan is the very model of a modern environmental catastrophe.

At the same time, however, Haiyan is not a particularly devastating financial catastrophe. For all that the afflicted areas had the bad fortune to get hit just as the storm was at its peak strength, they were a long way from Manila, the commercial heart of the Philippines. The loss estimates of about $14 billion are large, but not crazily so given that the Philippines generally suffers about $5 billion per year in storm damage. And the IMF’s mission chief in Manila, Rachel van Elkan, says that the economic prospects for the country are just as rosy as they were. After all, all the money coming in to the country to help rebuild the devastated areas will end up making a positive contribution to the country’s GDP.

Like most natural disasters, then, this one is not a huge economic disaster. And while all developing countries can make good use of financial inflows, it’s not clear that the Philippines needs money right now more than it normally does. The national accounts are strong, and Haiyan won’t hurt them significantly. Which means it’s probably no big deal that all those talks back in 2011, about the Philippines issuing some kind of catastrophe bond, never amounted to anything. If the Philippines had issued a cat bond back then, the value would have been for some tiny fraction of $14 billion, and even if it paid out in the wake of Haiyan, the money wouldn’t actually help a great deal — after all, Haiyan is already causing a huge influx of capital into the country.

More to the point, there’s no good reason to believe that a cat bond would have paid out in the wake of Haiyan. Cat bonds tend to pay out only under certain narrowly-prescribed conditions, and those conditions would probably have drawn the geographical area to be protected much more narrowly than the entire Philippine archipelago. (Remember that after the Japanese earthquake, most Japanese cat bonds did not pay out. “Typically, for a cat bond to trigger, you need a bull’s-eye to be hit instead of a general shot in the right direction,” explained JP Morgan’s Tom Keatinge at the time.)

Haiyan, then, far from demonstrating the need for catastrophe bonds, actually does the opposite: it’s a pretty clear example of a case in which cat bonds would have been superfluous.

Still, Haiyan does raise an interesting question: if it’s one of many cases in which a country didn’t have insurance it didn’t need, then what’s with all the press, all of a sudden, about how cat bonds have become a big thing? Bloomberg, last month, had a story about how cat bonds constitute “the biggest change to the reinsurance sector’s capital structure in the last 20 years”, in the words of one quoted expert; Matthew Klein followed up with a column worrying about how such a phenomenon “might lead to overbuilding in risky areas and laxer enforcement of building safety codes”. And the Economist, in an article headlined “Perilous paper”, uses words like “frothy”, and warns that cat bonds might cause a “man-made disaster” which could be “just as frightening” as a natural disaster.

It’s all very odd, seeing as how the cat-bond market, in reality, is going pretty much nowhere. In 2007, total issuance of cat bonds was $7 billion; this year, that number is finally going to be surpassed for the first time, with $8 billion in new cat bonds being issued. In both cases, cat-bond issuance is still just a drop in the total insurance bucket: AIG alone writes more property and casualty insurance than that every quarter. Global net written premiums, meanwhile, are running at more than $450 billion per year: it’s hard to believe that such a huge industry is likely to be significantly disrupted by the tiny cat-bond market.

There are deeper reasons why the insurance industry need not feel threatened by cat bonds: insurance protects against losses, while cat bonds don’t. A cat bond pays out a set amount, or doesn’t, regardless of the losses incurred: it’s a binary thing, which is triggered by a specific event. And the fact is that there’s precious little demand out there for such a product. I don’t want insurance which will pay out a certain amount if I fall off my bicycle while riding down 9th Avenue during the hours between 5pm and 7pm. I want insurance which will pay out if I wind up incurring significant losses after falling off my bicycle. Cat bonds can do the former, but they can’t do the latter: in jargon, they cover parametric risk rather than indemnity risk.

So, please, enough of the concern-trolling about what might happen if cat bonds take over the world. They haven’t, and they won’t. They’re an idea whose time will never come, even if they exert a weird perpetual fascination on a certain breed of financial journalist.

Comments
9 comments so far

“then what’s with all the press, all of a sudden, about how cat bonds have become a big thing?”

Historically, a surge of interest in catastrophe bonds tends to coincide pretty well with “a biggish player in the market has built up some underwriting exposure it’s not happy with and is looking around for someone to lift it out”.

Posted by dsquared | Report as abusive

Actually most catastrophe bonds are not parametric in nature or trigger. In fact the majority issued now are indemnity, so linked directly to the issuers loss experience. The second most widely used trigger structure is one which is linked to the insured losses of the industry.

In fact, just 2 of the 31 cat bonds issued this year are parametric.

In the case of cat bond for developing nations, then yes parametric is likely most appropriate. However, to say that a cat bond payout would not have helped is strange, as how could some extra disaster risk financing not have helped?

Yes, structurally a cat bond for the Philippines would have to overcome many issues to ensure broad coverage. Something that needed work and may have prevented the issuance of one before. But, it most certainly would have been a useful, and much-needed right now, source of extra post-disaster contingent capital in my opinion.

Posted by steveevans | Report as abusive

“Cat bonds can do the former, but they can’t do the latter: in jargon, they cover parametric risk rather than indemnity risk.”
In fact a substantial proportion of cat bonds provide cover on an indemnity basis and the figure is on the rise.
I think the issue around creating a Philippines cat bond would be more about the ability to model the various risks eg from typhoon, flood, storm surge given the level of data available. This is typically the issue with cat bonds covering non-US regions.

Posted by KarinaWhalley | Report as abusive

“*modern* environmental catastrophe”

What on earth work is “modern” supposed to be doing there? You don’t think people were killed and fields destroyed by typhoons 100 or 1000 years ago? Its sloppy thinking/writing trying to make things seem nearer and more au courant than they are.

As for why cat bonds are getting press? Companies are always trying to drum up press for their new products and journalists are too stupid to call them on it.

Posted by QCIC | Report as abusive

Cat bonds aren’t like ordinary insurance, they are capital structure tools for re-insurers. They aren’t written to cover human catastrophes, but for very specific risks that are punitively expensive for the insurance industry to hold. Risks like a Florida hurricane or California earthquake that could wipe out the reinsurance industry. Compare them to hybrid bank debt–like equity when the issuer needs it, like a bond in good times.

Posted by rogierswierstra | Report as abusive

Cat bonds aren’t like ordinary insurance, they are capital structure tools for re-insurers. They aren’t written to cover human catastrophes, but for very specific risks that are punitively expensive for the insurance industry to hold. Risks like a Florida hurricane or California earthquake that could wipe out the reinsurance industry. Compare them to hybrid bank debt–like equity when the issuer needs it, like a bond in good times.

Posted by rogierswierstra | Report as abusive

Cat bonds aren’t like ordinary insurance, they are capital structure tools for re-insurers. They aren’t written to cover human catastrophes, but for very specific risks that are punitively expensive for the insurance industry to hold. Risks like a Florida hurricane or California earthquake that could wipe out the reinsurance industry. Compare them to hybrid bank debt–like equity when the issuer needs it, like a bond in good times.

(Hope this isn’t posted multiple times)

Posted by rogierswierstra | Report as abusive

Cat bonds aren’t like ordinary insurance, they are capital structure tools for re-insurers. They aren’t written to cover human catastrophes, but for very specific risks that are punitively expensive for the insurance industry to hold. Risks like a Florida hurricane or California earthquake that could wipe out the reinsurance industry. Compare them to hybrid bank debt–like equity when the issuer needs it most, like a bond to buyers in good times.

Posted by rogierswierstra | Report as abusive

QCIC – I think it’s a (fairly strange) reference to modern major general

Posted by eisegesis | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/