What are CDOs and how did they contribute to the Global Financial Crisis?

John Kerschner: So, CDOs. What’s a CDO, a collateralized debt obligation? This gets a little more complicated. I know it sounds a lot like CLOs. But a CDO, the idea was, trying to create a new type of security out of kind of the pieces of securitization that existed back then.

So basically, you could take a little bit of ABS [asset-backed securities], a little bit of CMBS [commercial mortgage-backed securities], a little bit of RMBS [residential mortgage-backed securities] and put it in a structure, and obviously you get cash flows from all those securities and then divide it up into different ratings.

The reason why it was attractive, or it worked, was because the rating agencies, really their models rest on diversification to a large extent. So, the idea being, if I buy some ABS and CMBS and RMBS, and they’re all different types and geographically spread out, that if one part of the country has a problem or one asset class has a problem, the others won’t. So, you can basically create a security that requires less what we call credit enhancement, or support, by having this very diverse pool of assets. It’s just how the rating agency models work.

So that’s what CDOs first were. They kind of eventually became very RMBS heavy and pretty much just all subprime mortgages. So, if you think about it, you know, if you’re taking something that we found out was worth very little in a lot of these subprime mortgages and package them together, how could you get securities that were rated AAA off that? That is the gist to understand, right. You’re taking something that seems very fraught with a lot of risk and then creating a AAA security from that.

The reason for that is because the rating agencies didn’t have a lot of data because subprime mortgages didn’t exist before really the year 2000. And every model out there for home price appreciation assumed that housing would not go negative on a nationwide basis. And why is that? Because it had not happened since the Great Depression in the 1930s.

So there have been regional downturns – northeast in the early ‘90s and actually California in the early ‘90s – but nationwide, home price appreciation had never gone negative. So back then, if you looked at the models, it would say: OK, these are the losses you can expect from these mortgages if home price appreciation is 10%. And these are losses you can expect, a little bit higher, if it’s 7%. And a little bit higher if it’s 3%. And a little bit higher if it’s flat. And that’s where the model would shut off. No one ever was trying to forecast what would happen if home price appreciation was negative, because it had never happened. People just assumed it wasn’t going to happen. And yet, in basically 2006 to 2008, home price appreciation was -32% nationwide.

So that was the big problem. The rating agencies didn’t have the data. They basically calibrated their models on data they did have, which never showed home price appreciation going negative. Everybody just assumed that housing was going to continue to go up, probably not the same rate as it had been, but as long it was positive, a lot of these affordability products and subprime mortgages still would have been OK. But when kind of the roof caved in, and home price appreciation not just went negative but severely negative, it all collapsed on itself.

How did synthetic CDOs compound the problem?

Kerschner: Going back to the CDO market, why that was so important, [was] because it added leverage to the system. So, if you think about it, there are a lot of people that wanted to buy houses back in the early 2000s. They had bad credit scores. They could go to some of these New Century’s or Option Ones, kind of these mortgage lenders that focused on that area, and they could get a loan.

But if you think about it, there’s only so many subprime borrowers out there. I mean, there’s lots, but it’s a finite amount, right? And the demand for these securities was so high because money managers started printing CDOs. So, they would go out and buy a few hundred million of these subprime mortgages and then, again, put a structure around it, get a rating agency to rate it, and issue a CDO that was backed by the subprime mortgages.

That was one layer of leverage. But then they started creating synthetic CDOs. So, what is a synthetic CDO? It basically means you don’t have a cash bond, you have a credit default swap contract. So, let’s just say there’s a cash bond over here, bond 123, and that bond, you know, goes into a CDO. But different banks could look at that bond and say, OK, we’re going to create a credit default swap off that. So, what all that means is the contract, whoever owns the credit default swap will get the same cash flows from that cash bond. And if the bond defaults, whoever sold that credit default swap contract will then get a payment because it defaulted. And so, you could have one bond and really an unlimited number of credit default swaps off of that. If you think about back then, there were probably a couple hundred billion issued of cash bonds, but then there were multiple hundreds of billions more than that from synthetic CDOs.

Doing a CDO takes a lot of time and effort and work. You’ve got to spend months buying all these cash bonds and doing the analysis on them and then putting them in the framework and getting the rating agency to rate it. With a synthetic CDO, you could almost do it overnight. Like if you had a list of bond names that you liked, you could go to a rating agency in the street and just say, “We want to do a CDO with these.” It’s not quite overnight, but within weeks, literally, and issue a synthetic CDO. So, it caused a product that had inherent leverage, and put leverage on leverage, and then it was really infinite how much you could issue off that.

Where were you leading up to the Global Financial Crisis?

Kerschner: I started my investment management career in the mid ‘90s, came out of business school, joined a very small money manager, about $30 billion at the time, called Smith Breeden. They were a mortgage specialist. So, I worked on agency mortgages for about five or six years. And then I was getting itchy for a new responsibility, and they gave me the ABS/RMBS group. The firm was not doing anything in ABS and RMBS, and so they said, “Here Kerschner, you go have a stab at it, and good luck.”

And so given what happened with RMBS, non-agency I’m talking, from really the year 2000 to the year 2008, it went from almost zero – not quite zero, but almost zero – to $2.7 trillion. So, what was the ABS/RMBS group very much became the non-agency RMBS group.

I left Smith Breeden in the end of 2006, so for almost 7 years I was very much in the thick of it and went to a commercial real estate hedge fund after that. So, throughout the GFC, I was living the stress of kind of investing in commercial real estate globally. But yeah, it taught me a lot of things and I made it through and [I’m] glad to try to educate people here today.