Is there such thing as climate insurance?

Late last month, this article in the New York Times (originally published in May) was recirculated by some popular blogs that generally never touch insurance (Boing Boing and Slashdot). (The article was also covered by more specialized blogs at the time of its release.)

Looking at the Boing Boing article in particular I noticed a phrase that has become a become a pet peeve of mine - "climate change insurance" which Boing Boing says "costs big bucks."

The reason I blanche at that phrase is that:

  1. "Climate change insurance" does not (really) exist; but, nevertheless,
  2. People refer to it all the time.

The first claim I discuss in depth in my dissertation. It's also something that my colleagues at GlobalAgRisk have published on.

For the sake of brevity, I'll just say that  the nature of climate change risk (which is distinct from the climate risk described by teleconnections like ENSO) makes it very difficult to reinsure or even put on a market.

First, changes in global surface temperature play out over decades or centuries, while reinsurance contracts are generally year to year (or in the case of cat bonds may stretch out to a few years). That makes it ill-suited to insurance markets, which demand huge premiums for trying up large pots of capital over a long time horizon.

Second, climate change is not a truly random process in the same way that say, earthquakes are. The upward drift in global surface temperatures looks increasingly certain and certain outcomes can't be insured - no one will sell life insurance to a man on death row. So, while you can insure against extremes in the weather systems impacted by climate change, climate change in its most basic form is "classically insurable" .

Finally, climate change still has poorly defined regional impacts. That makes it very difficult to translate into the types of real risks that motivate companies and individual to insure themselves. While there may be some exceptions (like at the Poles) to the extent we understand regional impacts of climate change, those impacts are filtered through the kinds of teleconnections I discuss in my dissertation.

For all those reasons, there is (as far as I know) no large reinsurer actually offering "climate change insurance". And yet, the phase, and the idea of insurance against global temperature rise, remain in circulation. At GlobalAgRisk, my colleagues and I saw/heard it repeatedly used by experts in the field of international development as well as by academic economists. 

Perhaps I can interest those eager insurance purchasers in some killer robot coverage

Introduction to the ILS market - the firms

Now that I'm done with my dissertation, I'm on the job market.

I know that I want to work...

  • ...in catastrophic risk markets; and,
  • transferring new types of risk (like ENSO).

I also have a preference for: 

    So, given those parameters, it makes sense to start my job search with the capital markets groups of major reinsurance brokers.

    To understand my options, I've graphed a network of Insurance-Linked Securities (ILS) deals since 2005 using the network visualization framework sigma.js. (You can see a large database of ILS deals here. My code is on github.) 

    To start, let's look at the full network of deals. Each cedent (generally an insurance company) is represented by a yellow dot. Each broker or service provider is represented by a green dot. The cedents (yellow dots) are linked to the brokers (green dots) if they have worked together on an ILS issue since 2005.

    The first thing to note is that there aren't many distinct communities. Towers Watson Capital Markets only worked with a little set of cedents, who, in turn, only worked through them. (See the small star pattern at the bottom of the graph.) Those deals focused on hurricane risk in Florida. There was also one small deal related to lottery payouts (see the tiny line in the upper right of the graph). But those were the sole exceptions to the rule that ILS is a close knit community of roughly 80 cedents working with a half dozen brokers (and another dozen service providers).

    Network of ILS deals since 2005 with broker/service provider nodes in green and cedent nodes in yellow.

    Network of ILS deals since 2005 with broker/service provider nodes in green and cedent nodes in yellow.

    Below, I zoom in closer on that giant connected component of the network. Here each broker's node is proportional to the number of ILS deals it was involved in. There are four dominant brokers:

    1. Aon Benfield Securities;
    2. Goldman Sachs;
    3. Swiss Re Capital Markets; and,
    4. Guy Carpenter Securities.

    Most cedents work with at least two of these largest brokers. That promiscuity gives the graph its hairball shape.

    Major connected component of ILS deal network

    Major connected component of ILS deal network

    Aon Benfield Securities helped 24 cedents issue ILS issues since 2005, the most of any broker. In the graph below, each edge is weighted by the total value of the issuance on which the cedent and the broker worked. So thicker lines mean a greater volume of issuance.

    Aon's portfolio is balanced. It worked with many cedents and no single cedent dominated its issuance (i.e. the lines are of roughly equal thickness).

    Network of deals involving Aon Benfield Securities

    Network of deals involving Aon Benfield Securities

    Goldman Sachs was the second most popular broker, working with 23 cedents. It's hard to see from this picture, but USAA is the thick line on the left. It accounted for an outsized portion of the issuance that Goldman touched.

    Network of deals involving Goldman Sachs (edge thickness indicates the sum of the values of deals from each cedent)

    Network of deals involving Goldman Sachs (edge thickness indicates the sum of the values of deals from each cedent)

    Swiss Re Capital Markets' portfolio of deals covering 20 cedents was less balanced than Aon or Goldman. It benefited greatly from its relationship with Swiss Re, represented here by the fire-hose like link at the bottom.   

    Network of deals involving Swiss Re Capital Markets  

    Network of deals involving Swiss Re Capital Markets

     

    Finally, Guy Carpenter Securities' portfolio of deals was also relatively balanced.
    Network of deals involving Guy Carpenter Securities

    Network of deals involving Guy Carpenter Securities

    In the coming days, I'll be posting more about this network and what it means for individual firms in the ILS market.