Before 2005, it had seemed that Property and Casualty business was a cyclical business, boom to bust, from the long periods without a disaster and many start-up prospering… to a major disaster, which would knock out some companies, sometimes rather big ones, that happened to be at the wrong place at the wrong time. Some other P&C companies would also be crippled, but not go under.

After the disaster, rates would go up, as the regulatory pressure would be to keep P/C companies underwriting, and the claims would justify higher rates. There also would be a competitive hole left from P/C going under or companies scaling back, either in areas or in total.

But as time progressed without any major calamity, the pendulum would swing back in the other direction.

To understand this cycle further, it helps to consider the Law of Large Numbers and its opposite, diversification. In more normal times much of the P&C claims are small independent events, that is the Law of Large numbers apply. During these times the risk is diminished by underwriting more policies. As you get more policies, the claims experience becomes much smoother and predictable. Like flipping a coin for tails many times you will approach 50% head or tails.

However, if claims are highly correlated as with a wide scale disaster like a bad hurricane season, then the law of large numbers does not apply. Rather than smooth and highly predictable, the p&c companies have concentrated claims all at once. Not only are the number of claims correlated, but the size of the claims also. More policies from even a diverse pool means they can experience a surge of claims. But if there is any concentration, which often is the case, as agents specialize by areas and companies… then the claims can be ruinous.

As we all know the 2005 Hurricane season left huge claims for the insurers to pick-up. However, rather than many going under, 2005 had less P&C companies go into receivership than the long term average P&C companies go under. This can be attributed to lessons learned, more diversification was needed than was naturally occurring. As competitors would often capture high concentrations of the business in various areas.

Further, business in the coastal cities were seen for the risk that it was… highly profitable for many years with a large claim season wiping out many of those profitable years. So companies learned to limit how much exposure they would take. They also analyzed their business for concentration risks. They would trade through reinsurance , risks to various coastal areas. Stop loss reinsurance became a way to smooth out any one area concentration risks.

Yet, risks management can only go so far, if the P&C company does not have a diverse enough lines of business,

Securitization of mortgages was supposed to work the same way. But of course now its clear that it didn't. It has been said that securitization did nothing for the investors, but shuffled the cards around so that everybody could hold the least regulatory capital as possible. That is the sum of the parts of required capital were significantly less than holding individual mortgages separately. However, that is revisionist history, the pooling and the tranching was supposed to give less risk through the law of large numbers and more diversification.

Of course individual home prices have long been vulnerable to foreclosures. Your home price is highly correlated with your neighbor's and your city and area's. If your neighbor forecloses, your city or area has high vacant/foreclosed homes your price would of course be affected. However, spread this risks out across the USA and suddenly it becomes much more predictable… The law of large numbers applies, even in prior recessionary downturns, the risks was spread so the result were more predictable.

Yet, recessions are not like hurricanes, they are caused by mis-allocation of capital.

So the stage is set. The regulators says we've got this conquered. For a large portion of these securities are completely safe. Will only require a minute amount of capital for this portion we call "highly rated"…. In fact if you want to jump through some hoops to set them up right where you'r proving your monitoring the interest rate risks, we won't require any capital for these off balance sheet Structured Investment Vehicles (infamous SIV's).

But the markets disagreed, they gave a healthy spread even as they narrowed to 20-30 bps over treasuries… The ratio of the spread to required capital for these highly rated RMBS was great. Everyone loved their ROC (Return On Capital).

Like those calm P&C claim years, everybody seemed to be making money. The more numbers you added made predicting those risks smoother. Even the recession of 2000-01, those higher foreclosures were spread more evenly and without large surprise. It appeared the regulators were right, they had conquered the risks. Everybody was accelerating…

This of course caused the '30 times leverage' we heard about. Nobody, could get enough of these… hence driving the spread down down down. This made the regulators even happier as home ownership was up, up, up.

What the regulators missed was the First law of investing, if everybody is doing the same trade everybody is wrong. If everybody is booking extraordinary profits… nobody is as wealthy as they think.

That is hurricanes cause real damage of wealth, mis-allocating capital causes book losses. This is true even in a 12 trillion dollar boring mortgage bond markets; the markets can over allocate capital.

Further, its clear, that like the chaotic path of a hurricane, there are tipping points. Over build too much, in too many places and suddenly the well will run dry. Like the path of a hurricane, those tipping points may appear clear after the fact. But given a few more butterfly's or home-builders here or a few less there and who knows what path it might have taken as the areas waters warmed and cooled here and there.

You may have heard of RESTART, ( a fed program to get mortgage underwriting going again)… They can talk about transparency and more due diligence on the RMBS buyer's part. But whats not being said, how much credit support these mortgages will need. And for the buyer how much extra capital will be needed. These questions however can't even begin to be addressed until the current batch of RMBS are sufficiently run-off the books.

The stress test, however, gives us some idea of the new quality levels and level of capital to be required, but the regulators just can't raise the minimum numbers explicitly without feeding the panic.

In short bubbles and panics may be as chaotic as a hurricane, but are man made by over allocating capital. Because of this, their devastation will affect the whole area that has been overcapitalized, not just a region.

Therefore, the only answer, is to diversify more not just throughout the housing markets, but throughout the markets themselves.

George Parkanyi comments:

Your point about over allocation of capital to the same correlated trades and everyone thinking they are more profitable than they are can be illustrated by one of the funnier scenes from the Jim Carrey movie Bruce Almighty. Bruce, who now has God's powers and responsibilities, is trying to answer everyone's prayers, and becomes overwhelmed. Frustrated, he decides to clean up his backlog by granting everything asked for. The next day, it turns out that several million people have won the New York State lottery, which sets off rioting because the shared winnings end up being about $17 a person.





Speak your mind


Resources & Links