The old saying is “better late than never,” but as we hope to demonstrate, the SEC is awfully late to take an interest in collateralized loan obligations. The problems it has gotten curious about now were discernible years ago. And the failure to take interest until now means that misbehavior that was discussed in the press during the crisis is almost certain to go unpunished, since the statute of limitations for securities law violations has passed.
By way of background, CLOs are a form of structured credit. Their constituent ingredient is leveraged loans, a fancy term for loans to companies that have a lot of debt once the financing is completed. Most leveraged loans are created in private equity acquisitions. The sponsor takes a pool of leveraged loans (100 is a representative number) and creates a series of securities from them. Just like residential mortgage backed bonds, the cash flows from interest and principal payments on the loans are paid out in a specified priority: AAA investors get first dibs, then when the’ve gotten their share, the funds are allocated to AA investors, and so on.
Technically, a collateralized loan obligation is a type of collateralized debt obligation, but the term “CLO” was well established, and no one in the media wanted to use the label “asset backed securities collateralized debt obligation” aka ABS CDO, as a moniker, so subprime-based CDOs have become “CDOs” and CLOs remain CLOs. But there are differences besides in asset type. CDOs were effectively resecuritizations, in that they took risky pieces of a previous structured credit deal (residential mortgage backed securities) that no one wanted (the BBB tranches) and sausage-like, bundled them with a little bit of other credits and ground them up again and sold them.
Due to the leverage-on-leverage and the lack of real diversification, they would fail catastrophically if they failed. By contrast, despite being based on risky loans, CLOs are a first generation structured credit, and hence natively less nasty.
But “less bad than a CDO” is hardly a high bar. The SEC is looking into two issues with CLOs. As the Wall Street Journal notes:
SEC investigators are looking at whether banks and companies are using the bond deals to hide certain risks illegally, said the people close to the probes. A number of likely cases in that area are in the pipeline, one of the people said.
Separately, the government has expanded an inquiry into how Wall Street banks sell the deals, the people added. The securities being examined aren’t traded on any exchanges or open platforms, and their prices are negotiated privately between buyers and sellers.
We’ll deal with the pricing issue first, since it’s more accessible and was an area of abuse during the crisis. Back to the Journal:
The SEC also expanded an investigation into whether a number of Wall Street banks are cheating clients by mispricing certain bond deals, according to people close to the investigation…
Bill Harrington, a structured-finance expert who formerly worked at ratings firm Moody’s Investors Service, a unit of Moody’s Corp., said an investigation made sense, given how nontransparent the market is. “In general, CLO prices can’t be easily checked or compared across sellers,” he said.
I actually saw an attempt at this sort of mispricing first hand. Before the storm hit, there had been a massive takeover boom, and banks still held a lot of unsold leveraged loan and CLO inventory when the markets froze. CLO losses got a lot less attention than subprime-related losses, since the latter were so much larger. Nevertheless, just paying attention to Bloomberg and Financial Times stories, you had a rough sense of where the market was and the trajectory. I happened to be waiting to meet someone in a bar, and I overheard a Goldman Sachs salesman trying to hawk CLOs to what I assumed was a wealth management client, as in an individual investor. He was telling them they were a great deal at 95 cents on the dollar. Even I knew that 95 was wildly overpriced and the current market prices were 90, tops, and anyone with an operating brain cell knew they were headed lower. Moreover, Goldman would probably dump its weaker positions on a guy like that.*
The other mispricing, which was occasionally reported in the financial media later in the crisis, was bank trading itty bitty trades with each other or friendly clients to establish a “market” price for marking their books. Where was the SEC then?
Let’s turn to the SEC’s other focus, again from the Journal:
Banks often use so-called structured-finance transactions to offset their own risks. For example, some banks have used derivatives to structure deals that transfer the risk of loan defaults off their books, without them having to move the actual loans. These transactions are favored by banks because moving the loans can be operationally cumbersome and may require the consent of borrowers.
Keep in mind that banks can have risk transfer fail even when it puts its exposures in an off balance sheet vehicle,** but that does not appear to be the primary focus of the SEC’s investigation.
The “derivatives to structure deals” refers to credit default swaps. Often banks will use synthetic CLOs, meaning ones composed of credit default swaps, to offset their position, while keeping the loans on their balance sheet.*** The problem there is what traders call basis risk, that the prices realized in an event of default on the CDS will not match the price the bank penciled out when it sold the deal.**** So the SEC is correct to be concerned. But again, this has been an issue from the very inception of using CDS to transfer loan book risk to investors, from the famed JP Morgan Bistro and less well known Swiss Bank GLacier Finance transactions in 1997. So the SEC is 17 years late. Charming.
And in the middle of the Wall Street Journal piece, the authors Jean Eaglesham and Katy Byrne unwittingly pass on SEC propaganda:
Before the crisis, Wall Street assembled and sold trillions of dollars of collateralized debt obligations, or CDOs, which are securities based on pools of mortgages and other debts that are sold to investors in slices of differing credit ratings and risks. CDOs have been the focus of numerous high-profile SEC actions since the 2008 meltdown, but the agency has no plans to bring more CDO cases related to the financial crisis, according to people close to the agency.
The SEC feels it brought all the CDO cases it could, and the agency also is bumping up against a statute of limitations in some cases, the people said.
The SEC filed a token one CDO case per major firm and got “cost of doing business” fines from the biggest single cause of the crisis. As we explained long form in ECONNED, it was heavily synthetic CDOs that allowed banks to create a multiple of real economy subprime risk, thus greatly multiplying those exposures, and turning what would otherwise have been a S&L level crisis into a global financial crisis. And why was the SEC not more aggressive? Because its head of enforcement, Robert Khuzami, had been general counsel for the Americas for Deutsche Bank during the crisis. Anyone who has a casual knowledge of the subprime saga, such as reading The Big Short, knows that patient zero of toxic CDOs was Deutsche Bank trader Greg Lippmann. In other words, a serious investigation of CDOs would have implicated Khuzami. No way was that going to happen.
So I wouldn’t hold my breath that the SEC is finally prepared to get tough, but I would be delighted to be proven wrong.
* If you doubt that this sort of behavior was common, read up on what JP Morgan did to one of its most important private clients, Len Blavatnik.
** Sponsors are use off balance sheet vehicles and structured credits to offload risks onto investors, and thus report any exposure as having been removed from their balance sheet when the deal is sold. But there is a proud history of supposedly off balance sheet vehicles proving not to be off balance sheet when the deals went bad. The famous example, which was covered well during the crisis, was structured investment vehicles, or SIVs. There were some variants in the structure, but they held longer-dated assets, including, natch, subprime exposures and were funded with comparatively short-term, significantly or completely reliant on the commercial paper market. When investors woke up to subprime risk in August 2007, the first big casualty was the asset backed commercial paper market, meaning commercial paper than funded SIVs. When the short term commercial paper funding SIVs matured and the SIV could not roll the CP, the choice was to let the SIV collapse or have the bank step in and fund it itself. Investors screamed at the banks that they had better not stick them with losses of a disorderly collapse, and banks found themselves funding them (at least until they could unwind them).
Similarly, credit card securitizations have been regularly bailed out by sponsors when they perform poorly. Basically, banks can’t afford to burn these investors too badly and expect to be able to continue to sell these deals.
***In case you wondered why this would be attractive, from “Understanding the Risk of Synthetic CDOs“:
**** This results from both the fact that the timing of defaults could differ from what the issuer had estimated, and that CDS are settled shortly after the event of default, while the losses on loans are determined through a much longer bankruptcy and/or restructuring process. The differences in losses realized on the CDS versus on the referenced credit (which is an entity, but the banks will be holding a specific loan or bond) are typically significant.