When I was a young actuary, say in the early 90s, my boss came to me, and gave me an unrequested lesson. ?He said something to the effect of:

Most pricing actuaries make assumptions. ?Well, I test assumptions. ?That involves checking how actual results are coming in?expected, but in the early phases of a new product, you are living under the law of small numbers — you don’t have enough data to be statistically credible. ?You should still do the statistical analysis, but I take it one step further.

I pull the first 10-20?claim files and look at the cause for the claim. ?If the qualitative causes are not chance events, but are indications that the business is being sold improperly to those who know they are close to death (or disability) and evade the limited underwriting of the group coverage, that means the group is low quality, and the program should be discontinued, or severely modified.

He then told me about some credit life insurance that the company was offering through two well known, prestigious banks, and how the deaths were coming in from non-random causes: AIDS, Cancer, Drowned in the Hudson River, Murder, etc.

He fought to get that insurance line shut down, and it took him five years, as the line manager argued there was not enough experience. ?The line manager tried to get my boss fired, and finally, the line manager was fired. ?But if the company had listened early they would have lost $10M. ?As it was, they lost several hundreds of millions of dollars.

Now, most of my readers don’t care much about insurance, but this tale is meant to illustrate that reason for losses matters as much, and sometimes more than the absolute amount lost. ?Now to illustrate this for a different and perhaps more timely reason:

Wups, wups,?wups,?wups, pop, Pop, POP, Yaaaaaauuughhhh!

Maybe I am growing up a little, but I am trying to have better titles for my articles. ?The subheading above would have been my title. ?But let me explain what it means:

The credit cycle tends to be like this: in the bull phase, a long period (4-7 years) with few defaults and low loss severity followed by a bear phase, a shorter period (1-3 years) with high defaults and high loss severity. ?This is a phenomenon where history may not repeat exactly, but it will rhyme very well.

In the bull phase of the credit cycle there are a few defaults, but when you analyze the defaults, they occur for reasons unrelated to the economy as a whole. ?What do the failures look like? ?Fraud (think Enron), bad business plans from a megalomanic (think Reliance Insurance, ACH, Southmark, etc.) , a sudden shift in relative prices (think Energy Future Holdings), etc. ?Bad banking — think Continental Illinois in 1984.

In the bull phase, companies that fail would fail in any environment. ?But now let’s talk about the transition between the bull and bear phase — that is the “pop, Pop, POP.”

As the credit cycle shifts, a few companies fail that are closely related to the crisis that will come. ?They are your early warning. ?Think of the subprime lenders under stress in 2007, or the failure of Bear Stearns in early 2008. ?Think of LTCM in 1998, or the life insurers that came under stress for writing too many GICs [Guaranteed Investment Contracts] in the late 80s and invested the money in commercial mortgages.

As the cycle moves on defaults become more closely related to the financial economy as a whole. ?Fed policy is tight, and a bunch of things blow up that borrowed too much money short term. ?This is when the correlated failures happen:

• Banks, mortgage insurers, and overly leveraged homeowners default 2008-2011.
• Dot-coms fail because they can’t pay their vendor finance.
• Mexico and the mortgage markets blow up in 1994.
• Commercial mortgages blow up in the early 90s.
• LDC loans blow up in the early 80s.

To the Present

The present is always confusing. ?I get it right more often than most, but not by a large margin. ?We have companies threatening to fail in China?and?Portugal, but I don’t see much systemic lending risk in the US yet, aside from what is leftover from the last crisis.

It is worth noting that deleveraging has occurred more in word than in deed over the last five years. ?Yes, debt has traveled from public to private hands, but that only defers the problems, as governments will either have to inflate, tax more, or default to deal with the additional debts.

I am not trying to sound the alarm here. ?I am trying to tell you to be ready. ?During the intermediate phase between bull and bear, the weakest companies fail from unrecognized systemic risk. ?Personally, I think I have heard the first ‘pop.” ?It is coming from nations that did not delever, and that may suffer further if the bad debts overwhelm the banking systems.

Are you ready for the bear phase of the credit cycle? ?Screen your portfolios, and look for weak names that will not survive a general panic where only the best names can get credit.