Ed Noonan is frustrated. He’s chief executive officer of Validus Holdings, which manages an investment portfolio of insurance-linked securities, among other things. These are more frequently referred to as “catastrophe bonds”; in the event of a natural catastrophe, they offer a return but lose their principal to pay for insurance losses. Validus is being undercut in this market by fund managers who are, according to Noonan, no longer behaving rationally. From an earnings call last week in which he discussed the market for catastrophe bonds covering hurricane risk in Florida:

We’ve always viewed the ILS [insurance-linked securities] managers as behaving rationally. I can’t honestly say that’s what we’re seeing in Florida right now. I mean we have large ILS managers who are simply saying ‘whatever they quote, we’ll put out multi-hundred million dollar lines at 10 percent less.’ And I think it’s not a good model.

He might just be frustrated that everyone else got in on his specialized game. Pension funds, in particular, have moved into catastrophe bonds. But if Noonan is right, other fund managers are becoming irrationally exuberant about a high-yield risk that’s fundamentally unlike any of their other investments.

Pension funds and a novel risk: What could go wrong?

The first catastrophe bond was floated in 1996. Although it looks like a bond, as an investment an insurance-linked security demands a completely different set of tools. Some of these securities are tied to industry losses, which means an investor has to understand how insurers write their own policies, what kinds of houses are covered, and where they are. This is not impossible to do, but it requires highly specialized skill.

And investors in catastrophe bonds have to understand the underlying risk; understanding an earthquake or a storm can require the skills of a geologist or meteorologist. Models that can predict the frequency of hurricanes and the losses they might cause are useful. They are also flawed and require skepticism and understanding. Again, that’s possible—and specialized.

When Validus got into the market in 2005, catastrophe bonds were a niche product. The company hired academics to evaluate the risks it was taking on, and it built a structure around understanding the specific risks of catastrophe bonds as an asset class.

On Monday morning, however, the Financial Times reported that sales of catastrophe bonds hit a record level in the first quarter of 2014. The FT story offered a straightforward theme: Because they’re such a novel risk, catastrophe bonds offer a high yield, and high yields are hard to find right now. Catastrophe bonds used to be the hedge, since they were uncorrelated with financial markets risk. Now they’re the investment.

Which brings us to Ed Noonan’s appraisal of the market, quoted above. If he’s right, the people who buy insurance-linked securities for funds are pricing with the short-term forgetful optimism of the market, not the long-term performance of the underlying risk. As reinsurers have long known, this is a dangerous position to get into. Ordinarily, after a massive, once-in-a-decade catastrophe, a couple of reinsurers fail because they underpriced their risks. The ones that endure can raise their prices again.

In an exhaustive piece last December, Bloomberg’s Oliver Suess pointed out that many of the new investors in the insurance market are pension funds, which now provide about 85 percent of the funding for insurance-linked securities. If the pension funds have hired PhD meteorologists and geologists, and if they’ve hired away actuaries from reinsurance companies to understand insurance portfolios, then this is a smart move. If not, they can rely on models. Should be fine, right?

Before it's here, it's on the Bloomberg Terminal. LEARN MORE