Last week, in partnership with NPR's "Planet Money," we published an investigation about how Wall Street banks created fake demand in the run-up to the housing meltdown, increasing their bonuses -- and ultimately making the crisis worse. We asked for your questions, and now here are our reporters' answers.
Q. As a bank, aside from the short-term gain of bonuses, why would you invest in your own CDOs? If your aim is to get rid of the liabilities and manage risk, why would you buy CDOs based on your CDO? Wouldn't you want to get rid of your liability altogether?
A. It was a catastrophic error. There were several reasons why they would do it. They wanted to complete deals, to keep their mortgage securities assembly lines going. So if they had to take a bit of the CDOs to keep their deals flowing and the year-end bonuses coming, they would. (Or, more precisely, they would "sell" it to another part of the bank.)
In some cases, they were doing the proverbial "picking up nickels in front of the steamroller" trade. They would retain the top part of the CDO and hedge it, sometimes with a fragile insurer. They would make a little bit of money because the CDO threw off more money than it cost to hedge. But when the CDOs failed, the insurers (like MBIA and Ambac) collapsed and the banks were stuck with the virtually worthless CDOs.
Q. What balance sheet benefit resulted when one bank's CDO bought another CDO's lower-valued tranches?
A. If one bank bought a piece of another bank's CDO, then that was truly off the first bank's books.
However, in the late stages of the CDO boom, CDOs were essentially swapping pieces with each other in apparent quid-pro-quos.
Here's a hypothetical: Merrill sponsors "CDO Jesse" and Citigroup sponsors "CDO Jake." Jesse buys a mezzanine (or middle) slice of Jake and Jake buys a piece of Jesse.
Since each CDO owns something that owns itself, it has a less diversified set of assets. Those two CDOs are inherently weaker than they would have been.
If Merrill had retained the top portion -- or super-senior -- of Jesse, and Citi had done the same with Jake, then both banks were more exposed to the risk of losses than they would have been if they had sold the pieces of the CDOs outright to outside investors.
Q. The crux of the article is that banks with unsold CDOs on their books put them into pools of bonds to back new CDOs. I just don't see what is nefarious about that, since the composition of the pool is disclosed to the CDO buyers, who are all professional investors, who can look out for themselves. And it's only natural that banks would try to get risky assets off their books. Isn't that what we want them to do? Wasn't the problem that some banks had too many risky assets on their books, not too few?
A. The problem was, as our piece points out, that the assets weren't really off the books. They only looked that way. When a bank pushed unsold assets into a new CDO, but retained the top 80 percent or so (called the super-senior), it was still exposed to the majority of the new CDO.
Since the new CDO was filled with lousy assets, it was more fragile. So the banks were on the hook for losses that they wouldn't have been if they had really sold the assets or made the deals more solid.
Further, to the idea of what investors knew or could have known: One surprise for us was that investors didn't necessarily know all the assets that the manager selected when they invested. Deals often closed without being fully completed. Investors would know that the remaining 10 or 20 percent would be filled with CDO pieces, but they wouldn't know which ones.
But to the larger point, you are right. Much of this was disclosed in the prospectuses, although buried in hundreds of pages of legalese and nonspecific caveats.
The parties in the transactions that might be in the most legal jeopardy are the CDO managers. They had a fiduciary responsibility to manage the CDO properly and that their role was accurately represented to the investor.
Q. So who's paying those fees? I sell you something and earn a fee and then you sell me something and earn a fee. Where's the net income?
A. The income generated by the CDO itself provides the fees for the manager and the bank. In that respect, it's a bit like mutual funds, where the fees get taken out of your returns. Normally, there was some actual, or as they say on Wall Street, real money, in the deal. This would be used to pay the fees.
Q. Weren't these CDOs rated by "independent" ratings agencies?
A. I think you already know the answer to this question! Like the managers, the rating agencies depended on the banks for their income. As one rating agency official told us, agencies couldn't say no to a deal and the banks knew it.
In devising their ratings on managers, the rating agencies chose to look at the wrong things. As one executive at a CDO manager told us, the agencies "did heavy, heavy due diligence on managers but they were looking for the wrong things: how you processed a ticket or how your surveillance systems worked," adding, "They didn't check whether you were buying good bonds."
Q. Why wasn't this self-dealing fraud?
A. In some cases, it may have been. That's up to the SEC -- which is aggressively investigating the CDO business and, especially, banks' relationships to managers -- and ultimately to the courts.
However, some of the questionable behavior may have been perfectly legal. For instance, banks had agreements to own the assets that CDO managers selected before those assets were placed into the CDO. That gave the banks the right to veto managers' selections. As our story showed, in late 2006, banks began to routinely use this veto. Was this perfectly appropriate behavior? It may have been, at least on the banks' side.
The managers, however, have fiduciary duties. They cannot misrepresent that they were selecting assets if indeed they weren't. And they cannot do things that aren't good for the CDO. That could be illegal.
But these cases are challenging. The subject matter is complicated. The investigations are resource-intensive and time-consuming. And the legal disclosures on the deals themselves were extensive.
