Comment response

January 23, 2009

Comment on

Friday, January 23, 2009

Why Sequencing Is So Important In A Good Blogpost

Ran across an interesting piece over here [], that benefited heavily from being read on a rainy morning. The first paragraph of it is largely nonsense, asserting a massive conspiracy to defraud unsophisticated investors by hiding the true risks inherent in bonds backed by subprime originations. The centerpiece of it is an unfortunate chicken salad / chicken shit analogy that would make for a semi-amusing soundbite if it wasn’t completely inanalogous to the way structured finance works. Where the blogger’s hopped the track is that his analysis is based on some sort of one to one correspondence between the performance of a MBS bond and the underlying assets in the bond. There is none, by design. A MBS bond is structured in multiple tranches, each with different yields and different priority positions on access to cash flows flowing into the trust. This subordination, along with initial overcollateralization and excess interest spread, are the reasons that you can get a AAA rating on a structured subprime deal. Now keep in mind, it’s not the entirety of the deal that gets rated AAA. There’s a good chunk of the deal lower in the collateralization foodchain that’s going to get rated at mezzanine and sub-investment grade levels, and the ratio of these is higher on a subprime deal than it is on agency debt, as it should be. And any investor who’s buying these assets in the marketplace knows this.

To return to our chicken salad / chicken shit image, investors knew they were getting chicken shit. They also knew that the expected cash flows from a chicken shit sandwich were lower than a chicken salad sandwich, and thus the subordination levels for the chicken shit sandwich bonds were much higher than they would have been for the chicken salad bonds.

There was a massive failure in risk assessment across the entirety of structured finance without a doubt, but in no way is structured finance in general to blame for it, and in no way were large institutional investors defrauded on the credit composition of the trust’s collateral. There’s a case to be made that happened with respect to mezzanine CDOs, but certainly not normal subprime mortgage-backeds.

And this is why sequencing is so important in a good blogpost. Because if it wasn’t raining and a lousy time to go for a run I would have pitched on this piece after the first paragraph. Which would have been a shame, because starting around the fifth paragraph it gets very very good.

I began in this business as a subprime originations underwriter in the early 1990s, as the non-agency marketplace evolved away from finance company dominance to a wholesale securitization model. But they were still around at the beginning of my career, and I’ve put people into 65% LTV, 19% interest / 10 point loans at Beneficial, Avco, Transamerica in my day, and was trained by people that lent in branch offices and made loans knowing that they, personally, were going to be responsible for collecting them. My friend Johnny Rod once got his picture in the Sacramento Bee pushing a wheelchair down a ramp for an elderly borrower who he had to evict pursuant to foreclosure. Do that once and you learn a thing or two about underwriting that people stopped teaching years ago.

And for as wrong as the blogger was about the evils of securitizing subprime loans due to the fraud committed upon unsuspecting investors, he is completely right in the idea that a huge disadvantage of the securitization model as an exit strategy for subprime loans is the relative ineffectiveness of centralized servicing and collections for impaired borrowers. You’ll save a lot more borrowers with physically engaged collectors that work in the area of their borrowers than you will with collection activity coming from a phone farm in Bangalore.

But you don’t need to turn back the clock to 1985 to do this. You can maintain a subprime securitization exit strategy that keeps rates and points below usury levels and combine that with a much more robust (and labor-intensive, and capital-intensive) servicing strategy. Your servicing costs go up significantly, but not by 1000 basis points.

Some solid analysis here that I haven’t seen discussed too many places. I’m glad I was able to stick with the post long enough to get to it.
Posted by FC at 9:06 AM
BWS reples:
FC: Thank you for your comments regarding my post.
And you are correct in your comments regarding tranched CDOs.  You are also correct in clarifying that the ratings fraud and defrauding of investors was primarily associated with tranched CDOs rather than more standard MBS (securities composed of mortgages sharing like characteristics).  While you found the chicken salad analogy “unfortunate”, please understand that my comments and explanations are not exclusively directed to a finance-oriented audience.  Mine is an anti-statist blog, not exclusively a finance blog.  My goal, therefore, is to shave away as much complexity as possible from the mechanics of finance while still getting the general point across.  Some felt that I found the right balance.  You did not, and I respect your pointing it out in a well-written and thoughtful comment.  Regarding the structure of the post, I’ll take that into consideration in future entries and always appreciate comments of style and structure.
Regarding your final comments, it seems we at least agree that the ‘high-touch’, local servicing model is more efficient for sub-prime borrowers.  One way to look at pricing is that a consumer finance borrower was paying 12% for the right to have his loan restructured twice a year when he couldn’t make his payment.  As you evidently have experience in consumer finance, you know what I mean.  Another way to look at pricing, and more in line with the theme of this blog, is that it should determined by the market.  While you suggest the costs of this model should not add 1,000 basis points to the ‘price’ of the loan, I would argue that the market should determine what price is acceptable.  This price, in an ideally free market, will fall within the mid-point of 1) what the consumer will pay, and 2) what is necessary to keep the business profitable enough to sustain operations and continue to provide their service to those demanding it.  One thing I think we can both agree on is that, while some may find it unpalatable, demand for subprime loans will not go away any time soon.


2 Responses to “Comment response”

  1. George Matkov Says:

    FC’s analysis may be correct in the facts as he presents them but he conveniently skips, or is blissfully ignorant of, the fact that between these tranches and the customer is a sales person. I can tell you from my experiences in the corporate world, the sales people are paid for sales and they work very hard to accentuate the positive (high returns for low risk) and willfully ignore the possible negatives (high returns are inevitable accompanied by high risk). Stick to your guns good buddy.


  2. Thanks GM. You can bet I’ll be sticking to my guns as we explore a wide range of future issues.

    That said, I’m always open to feedback and questions- positive, negative, or neutral. Good for the debate. And these debates must take place if we are going to rise out of this mess.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: