BRUCE KIRSCH: Commercial Mortgage-Backed Securities, or CMBS, are a relatively new financing innovation. And what they’ve done is they’ve helped to create a more liquid and more plentiful debt market for commercial real estate. The essence of a mortgage-backed security is bundling and packaging of multiple mortgages from various properties into a pool, and then issuing shares in that pool.
And so investors will purchase the shares seeking a targeted yield on their investment from the scheduled interest payments, and then return of their principal at the end of the term of issuance. And all of this, when it’s said and done, will blend into a final IRR by the end of the timeline.
Now, this is pretty complicated. And so it’s easy to misunderstand how this all works. What do you think is the biggest misconception about what CMBS is and how it functions?
PETER LINNEMAN: That you can destroy risk. You can’t destroy risk. The mortgages have underlying risk, both because each mortgage is backed by a specific property, and that property has risk relating to that, building related to that market therein, to that property type, and when that loan is due.
The other side of it is that you can’t change those by packaging them. Those risks exist. And there’s this mythology that somehow I can eliminate risk. I can’t eliminate risk. That risk exists.
I can redistribute that risk. I can say, Bruce, you take more of that risk, and I’ll take less. But the total amount of risk is still the same. So that’s the one big myth.
There’s probably a second huge myth, and it’s crushed the market in the late 2000s window. And that was that the risk of one mortgage is unrelated to the risk of another mortgage. So the way that these products were packaged and sold essentially assumed that if I put 10 mortgages into a pool, each of them had a risk of default that was unrelated to the other nine.
And the rationale went, well, one’s in an office building in St. Louis, and another is a shopping center in Tampa. The other is an apartment building in Boston, and the other is a warehouse in Seattle. How could those buildings have correlated risk? They aren’t competing with one another. Therefore, it’s not like one’s stealing tenants from the other, or the whole city goes down. Gee, the risks are uncorrelated.
Well, some of us wrote at the time, they are correlated. They’re highly correlated. They’re not correlated as a microphenomenon. Namely, the market’s fine, and one building’s losing tenants to another competitor. They’re not correlated in that sense. But they’re huge correlated in the macro sense. Namely, a bad economy, a bad US economy, will hit Tampa retail sales. It’ll hit Boston warehouse activity. It’ll hit St. Louis office activity, and it will hit apartments as well. It will hit everything.
When people are losing their jobs, they’re not buying as much at shopping centers. They’re not leasing as much office space. They’re not storing as many boxes in their warehouses. And they aren’t as aggressive as a renting apartment. So that the cash streams of all these properties have a general macroeconomic risk that’s highly correlated.
And secondly, they all have capital market risk. Namely, no one wants to lend. No one wants to invest. And therefore, when no one wants to lend and no one wants to invest, they’re all going to experience difficulty in their valuations and difficulty in their ruling over of loans at the same time, even though they’re different property types in different markets.
Now, these risks may not be perfectly correlated. St. Louis may not get hit quite as hard as Minneapolis or whatever, but they are highly correlated. And the risk models package these as if they were uncorrelated. And you may recall the phenomenon of flipping coins.
Imagine flipping 10 coins that are independent. You get a different kind of pattern of outcome than if you say, I flip 10 coins, but whatever one comes up, they all come up. The risk is very different in those two situations.
And these products have historically been priced and analyzed as if the risks are independent, when in fact, they’re highly correlated, driven by macro– not micro, but macroeconomic trends and macro capital market trends. So their micro property may be uncorrelated, but the overarching risk is highly correlated. That has been probably the biggest myth that’s gone out there.
BRUCE KIRSCH: I think that it’s not even that far-fetched to be able to see how they might be directly tied, though. Let’s say that you’ve got, in this pool, a commercial office building in Chicago. And in the same pool, you have an apartment building in New York. And the folks who are the tenants in the apartment building in New York, or some of them, work for the company that’s based in Chicago. And so while it might be a little far-fetched– well, what are the odds of that? Well, it does happen, right?
PETER LINNEMAN: Yeah, and when you think about retail, for example, it’s even more obvious that people who are buying stuff in Tampa, those companies whose goods are being bought are tenants in an office building in St. Louis. They’re tenants somewhere. And they’re storing their boxes somewhere in a distribution center. And their employees are living in apartments somewhere.
And so when you think of them that way, there is undoubted macro correlation. And in the United States economy, the world economy, not every place falls the same. But most places fall at the same time. Most places rise at the same time. Some rise more, some rise less, some fall more, some fall less. But that commonality of rising and falling means the risks are correlated.
