Last week some guys at Credit Suisse got in a bit of trouble for marking some structured credit assets at fantasy valuations in order to appease their bosses and get better bonuses. I commented that this was a bit different from the run of CDO-fraud cases we’ve seen: the CS guys seemed to think they had a better chance of passing off fake-o valuations on their bosses than they would on the market, and so rather than dumping those assets on unsuspecting clients, they held on to them and hoped/lied.

They might have chosen differently had they met Bobby L. Hayes:

BOBBY L. HAYES, an engineering entrepreneur in Incline Village, Nev., used to trust financial institutions. This is the story of why he no longer does.

Short version of the story is: Because he got stuffed with the equity tranche of a CLO that had bought loans at par when they were really worth 95, making his equity tranche worth zero.*

This is sort of Version 2.0 of the usual synthetic CDO fraud/fraudishness where it’s like Party A wants to bet Thing X will go down, Party B wants to bet that things of Class XYZ will go up, Bank C convinces Party B to go long Thing X because it’s in Class XYZ, Thing X and the rest of Class XYZ go down, and Party B is shocked, shocked, to find GAMBLING against Thing X (really “gambling AGAINST Thing X” which is even dumber) and it’s all sort of woolly-headed. Here, the claim is not that the investors were deceived about who was on the other side of the trade, but about the market value of the underlying securities. That sounds worse. But I’m not so sure it is.

What happened here is that a CLO manager bought commercial loans to package into a collateralized loan obligation, holding them in a warehouse facility until the CLO was complete. The manager bought $400mm face value of loans for a bit over par, and sold CLO interests for a total of about $403mm. The CLO’s private placement memo basically says “we’ll buy the loans from the warehouse at the acquisition price, which will be no more than 100.3%”; it does not say “and that will happen to coincide with the market price.” Elsewhere the PPM says that the collateral manager will in the future do arms-length-y things to buy loans fairly, but it doesn’t say that about the warehoused loans. And it appears, per some people who looked at it, that those loans were “worth” around 95, based on (apparently) changes in the trading level of the relevant index between the purchase time and the CLO issue date. So if the CLO was carrying them at 100.3, it had an immediate loss when it was sold to the public. And that loss ate up the BB-rated Class E tranche and most of the BBB-rated Class D tranche in this capital structure:

And that in turn ate up Bobby L. Hayes and his trust in financial institutions:

Mr. Hayes said he told his broker that he didn’t want to take risks with the money that went into the investment, a collateralized loan obligation known as Lyon Capital Management VII that was issued in July 2007. “I was a trusting client, and it was like a bad dream,” Mr. Hayes said. “I had a lot more assets in the bank, and it was unfathomable to me that they would deliberately do this to even a small depositor.” …

MR. HAYES had been sold the riskiest piece of the loan pool, known as the equity or E tranche. Because of the way losses are distributed in these instruments, the loans in the pool had to decline only by one-half of 1 percent before Mr. Hayes’s investment would be wiped out. The entire security was liquidated at a loss of around $75 million about 16 months after it was sold.

There is a lot that’s suspicious here, starting with “I didn’t want to take any risk … so I bought half of the BB rated tranche paying Libor + 500bps that was subordinated to paper with a face amount greater than the face amount of underlying loans.” I mean, maybe that is what happened – he told his broker “no risk” and the broker just put him in maximum risk anyway – but no risk rarely means L+500.

More generally, I am not a huge fan of the narrative of:
(1) banks had some bad things;
(2) they packaged those things in confusing ways with the intention of tricking clients;
(3) it worked, shifting bad things to clients at off-market prices.
There are a number of reasons for doubting that narrative, including the fact that banks ended up with a whole lot of bad things themselves, and the fact that clients can mostly read and do their own homework and there’s no reason that putting $380mm worth of loans into a box and calling the box a CLO would make hedge funds any more likely to pay $400mm for that box.

There are specific reasons for doubting it here. For instance, you could doubt that this CLO was created to help BofA offload its bad shit, because the underlying loans weren’t BofA’s – they were originated in various places and acquired in the market by the CLO. How do you know that? Well, BofA will tell you, if you call them up and ask them. The PPM says that the underlying loans were owned by the CLO, not BofA, though it also says that BofA may have provided warehouse financing. They probably did, which might give them some incentive to move the thing along – I suspect the CLO manager was not super well capitalized and so BofA probably had some risk in getting these loans out.

More importantly, you know this wasn’t exactly BofA dumping its bad loans because of how loans work if you’re BofA. Specifically banks like BofA don’t mark their commercial loans to market: you put on a loan at 100, you carry it at 100, minus some reserve, as long as it performs. The loans here were performing – their market value had just (allegedly) dropped on some index-adjusted basis. So unlike the Credit Suisse guys who were marking their trading books to fantasy because that got them a nice bonus, BofA bankers had no (personal, current-P&L-driven) need to offload loans just because their market value had dropped – they wouldn’t even notice that market value. Nobody was paying them based on the market value of their loans, so they had no incentive to conceal it when it went south.

But that’s not really the point – even if BofA was misvaluing a third-party CLO manager’s loans in selling them to its clients, that still looks bad. And I can’t really tell if they were. The case was decided by arbitration so the decision doesn’t tell you much of what’s going on, but the arbitrators seem to have believed that the loans were worth $380mm and BofA lied to Mr. Hayes. This does not, though, seem to be a case of actively traded loans where you could look at screen prices that day and see what they were worth – rather, BofA thought they were worth $400mm based on purchase price, performance, ratings, whatever, while Hayes’s experts thought they were worth $380mm based on where a comparable index traded. (In fact, they were “worth” way less than either number in the sense that there was worse performance in the future and the market value ended up being like $325mm when the CLOs were terminated, but no one knew that at the time.)

Without unambiguous trading data in the specific loans, I guess they were worth what a buyer would pay for them. And there was a buyer who paid for them and it was – the CLO investors. And there’s no particular claim about any actual deception at all. BofA said there were loans, they had a certain rating, future loans would be bought at market prices, previous loans would be valued at acquisition price – nothing at all was said about the market values of the existing loans. The investors could have diligenced to their heart’s content. Except that the PPM didn’t identify the underlying loans. So they couldn’t have.

This all seems very strange and there’s fault to share. Hayes probably shouldn’t have bought equity in unidentified loans carried at their purchase price from months ago. BofA maybe should have done more to assure him that those loans were worth their purchase price. But you could compare this transaction to the other way a retail-ish investor, like Hayes, could have bought equity in a set of not-marked-to-market commercial loans. He could have bought equity in a bank. BofA, say. Which held loans on its books at acquisition prices, not mark-to-market prices.

The CLO at issue here didn’t just slice up and tranche out some commercial loans. It also did a subtler kind of magic. It took loans that were trading on the market as mark-to-market assets and transformed them into acquisition-price, held-to-maturity looking assets. Just for long enough to sell them to investors, like Hayes, who didn’t really know the difference.

A Wipeout That Didn’t Have to Happen [NYT]

Did BoA’s 2007 CLOs Defraud Investors? [Securities Litigation & Consulting Group]

* Nah, kidding. Not zero. If Thing Y pays off (Value of Thing X – $100), and Thing X has a possible value in three years of between $0 and $102 and is currently trading at $95, what is Thing Y worth? Zero is in any case the wrong answer.