Who's At Fault for Too Much Risk?

I appreciated Steve Waldman’s article at his excellent blog Interfluidity, which was also posted at Naked Capitalism.  I have a slightly different take on the topic, which I expressed in the comments section of each blog:

Steve, I think we had two, maybe three things go on here. First, the “originate to sell” model failed because basic underwriting was not done well. The incentives against failure were not left with the originator, i.e., having to hold onto a large equity piece.

If the underwriting had been done well, the next problem would be weak financing structures on the part of the certificate buyers. Many were leveraged higher than prudent, even on “super seniors.”

Finally, the servicing models are often flawed. There has to be adequate pay for servicing and special servicing, or else loss mitigation efforts will be poor.

Risks were taken and avoided, but many of the seemingly avoided risks come back when the one guaranteeing the avoid risk cannot perform.

It is true that there was a lot of demand for AAA assets, but there was also a lot of demand for mezzanine and subordinated assets out of complex debt structures.  Within all investor classes, there was a hunger for excess yield, whether it was a little extra at the AAA level, or a lot extra for subordinated and equity levels.

The demand for AAA assets, whether senior or super-senior, was often driven by leveraged investors, seeking to profit from being able to arbitrage the AAA securities versus their funding rate.  Safe assets were turned into unsafe assets by the added leverage.

And in this sense, the rating agencies are culpable, because they let the concept of a AAA, which means capable of surviving a depression, drift to a lesser standard.  They trusted simple mathematical models, and did not spend enough time on the quality of underwriting.  Of course, that takes time, and profits for the rating agencies comes from cramming as many deals out the door as they can.  That is, if you don’t care about your franchise.

When I was a mortgage bond manager, I spent time on any deal, even at the AAA level, by asking, “Who has skin in the game?”  If they were credible underwriters, I had greater comfort, but if the originator was selling and retaining little exposure to the outcome, I did not tend to buy.

Deal structure cannot make up for bad underwriting, usually.  Lousy assets lead to lousy returns for everyone in the capital structure.  I have owned AAA assets that have gone into default.  In every case, lousy underwriting of the original debts, not a bad economy, was the cause of the problem.

The entire period 2004-2007 was characterized by low spreads, as a hunger for yield depressed yields on newly issued corporate and structured debts.  Now we are facing the true value of those debt promises.

Of course after too much risk is taken, the regulators come along and say, “We must tame this!  No more excessive risk taking by investment banks because that leads to systemic risk.”  jck at Alea pokes at the recent efforts to do so, and if you look at the comments, I agree.

It is impossible to separate the desire for high returns from high risk-taking.  Having been a risk manager inside insurance companies, I read with some sympathy this article from The Economist.  Substitute actuary for risk manager, and marketer for trader, and the same situation plays out in insurance companies every day.

The only place that I have worked in that solved the problem bonused both marketers and actuaries on the same formula, offering slightly more reward from sales to marketers, and risk-adjusted profits to actuaries.  It got both sides on the same page, because they were compensated similarly.  I told the head of that division that he was fortunate to have business-minded actuaries.  He choked on his drink when I suggested that we were cheap for what he was getting.

My view is that investors did take too much risk 2004-2007.  They did it in many ways, by not underwriting properly, by levering up too much, by not servicing properly.  We are paying for that now.

Comments

[fatcat:]

You can argue all you want about what went wrong with the banking system,but all comes down to one point..all the rules of banking were thrown out the window over the years: know your customer,know your collateral,keep your ratios correct.If you make a home loan for more than 3 times annual family income,trouble will follow...

2008 Aug 08 11:32 AM

[dlaw:]

Oh, this is just nonsense!

The problem was not compensation structuring, it was the total lack of structure in the market from the theory up.

First of all, the statistical methods were bad. Market players insisted that the market is described by a Gaussian statistics when that has long ago been disproved.

Per the Economist article, this leads people who should know better to buy the story that the recent past is mathematical prologue.

Again, per the Economist article, professionals may *ask* themselves where liquidity problems might come from but their models for liquidity risk are clearly inadequate and generally based on the "last war".

The biggest problem - made clear from the "blame the lying borrower" narratives followed by soft-soap "analyses" like this - was self-deception and just plain deception.

The fact was, is and will always be that a traded market needs to have highly-regulated quality assurance underneath it to make sure the traded inventory maintains quality. Otherwise we get a "Market For Lemons", which is exactly what happened.

Once the charlatans figured out that the risk managers were dependent on poor models, they gamed those models. Bankers and originators did in two steps. First, they shut information out of the market. At the lowbrow end, they radically increased issuance of poorly documented loans. Why the hell - BTW - did you people THINK they were doing that? Did you not see the significance of huge volumes of loan product coming onto the market that were like used cars with "broken" odometers? What did you THINK was happening there? A couple fellows got a Nobel prize for their work on "lemon" markets and information asymmetry. Did you not have economists available to you?

At the "highbrow" end, bankers and derivatives writers shut out information by creating product that had more and more complicated relationships to the underlying inventory (mortgages). This is an absolutely standard element of most confidence games, by the way.

The second element in the confidence game was the canard. Once bankers and writers had moved attention away from the actual, underlying inventory (mortgage underwriting standards) , they substituted the judgment of stupid rating agencies - apparently rejects from the more highly-paid bank risk-management departments - and then wrapped the crap up in the gold foil of so-called bond "insurance" from the so-called "bond insurers" - apparently rejects from the rating agencies.

This was the kind of totally predictable duplicity that enters into an unregulated market because IT IS PROFITABLE TO LIE. Again, some fellows got a Nobel prize for proving that.

2008 Aug 09 03:16 PM