Irrational Economist_ Making Decisions in a Dangerous World - Erwann Michel-Kerjan [90]
SOME PARALLELS WITH THE CATASTROPHE RISK MARKET
Before I turn to some direct evidence on this, I want to draw some parallels with the catastrophe risk market. This market has in the past witnessed dislocations that are similar to that of the 2008-2009 credit crisis. Yet it is in many ways a far less complicated and more easily interpreted marketplace.
In fact, the “cat” marketplace lends itself to a readily understood analogy, and there are several reasons for this. First, natural disasters are exogenous. In the case of the credit crisis, we know that the housing and commodity price bubbles have burst. But rather than reflecting concern over what exactly caused it, the cat risk example suggests we can and should move on, and look at the effects on capital provided to financial intermediaries. Second, cat risks and the damages they cause are physical phenomena that can be scientifically modeled. The likelihood of a given dollar amount of hurricane damage is much more objective and transparent than the likelihood of a deep recession, or, more accurately, of AAA securities trading at only pennies on the dollar. We can objectively simulate storm frequency, severity, and trajectory based on our knowledge of physical systems. We cannot objectively simulate the “madness of crowds”—by which I mean shifts in sentiment, so-called animal spirits, or runs on banks, runs on markets, or runs on consumption. Finally, cat risk is not systematic—it is not correlated with the risks of major financial markets. This means that we have a good idea of what the “fair market” price of cat risk should be. Specifically, because cat risk is diversifiable, the fair-value premium for a reinsurance contract that incurs no losses 99 percent of the time and incurs losses up to a given limit 1 percent of the time is 1 percent. That is, fair-value premiums are just equal to expected losses (probability × consequences). And since expected cat losses can be reasonably accurately and scientifically modeled, fair-value premiums can consequently be readily observed. This valuation process is far simpler than its counterpart in the financial world; while the market may try, no one has any real knowledge of what most underlying financial securities are worth. The same applies to real estate: Who can say convincingly that houses were too cheap or too expensive (or fairly priced) this year? Indeed, it is the lack of reliable markers of fair valuation that allow for enormous swings in the values of financial securities. During the most extreme bubbles we’ve witnessed (and there have been many over the past decade alone), pundits have argued compellingly that both the bubble and the bubble-has-burst prices are fair.
So what happens in the catastrophe markets in the aftermath of an event? As with the current crisis, some observers have contended that immediately following a hurricane is the best time to sell insurance (and reinsurance) because it’s then that the demand for protection is high. Those who bought plenty of insurance see its value, and those who didn’t feel once-burned, and, going forward, twice shy. With demand high, prices should be high. This is just the demand-shock view. And there are those who argue that, after an event, capital is depleted in the relevant financial intermediaries—reinsurers, in the case of cat risk—so that supply is low. Both arguments suggest that prices should be high after the event. And indeed they are. However, the high prices in each argument have opposite implications for what happens to quantities of intermediary risk sharing supplied.
The behavior of cat reinsurance prices in the aftermath of events is shown in Figure 20.4 with respect to Hurricane Andrew in 1992 and in Table 20.1 with respect to the three large hurricanes of 2005. These data make several points. First, they demonstrate that, in the aftermath of such events, the price of reinsurance rose substantially