Siwoti Sunday: Blodget’s bizarre Goldman apologia

By Felix Salmon
April 18, 2010
Henry Blodget delivered his verdict: both Stewart and her broker, Peter Bacanovic, were not guilty on all counts. But at least he disclosed that he was prejudiced and that he “wouldn't mind seeing the government get egg on its face”:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

After covering the Martha Stewart trial for Slate, Henry Blodget delivered his verdict: both Stewart and her broker, Peter Bacanovic, were not guilty on all counts. But at least he disclosed that he was prejudiced and that he “wouldn’t mind seeing the government get egg on its face”:

My potential biases are so salient that I wouldn’t have made it through the first 10 minutes of jury selection. As a result of my own experiences as a defendant, for example, as well as a decade of working on Wall Street, I have a “rooting bias”: I regard corporate America as my home team and the government as the visitors.

Now Blodget is back at it, rooting for corporate America in the SEC case against Goldman Sachs. But while he at least attempted to play it down the middle in his Slate dispatches, he now seems to have gone simply bonkers: his case for the defense is so weak that one suspects it was written more as a pageview magnet than as a sincere reflection of his own beliefs.

Blodget kicks off with what he calls “some important background to keep in mind”. The problem is that Bloget’s “important background” seems to have been lifted from a fictional planet: it certainly doesn’t reflect reality.

For starters, in BlodgetWorld, this CDO was structured in the wake of numerous failed attempts to bet against the housing market. He talks about the “huge losses” which were suffered by “dozens of other investors who bet against the housing market from 2003-2007”, and goes on to reiterate:

Plenty of firms had been betting on the collapse of the housing market for years, and they’d all been wrong.

In 2007, the housing market had not yet collapsed, and everyone who had bet on it collapsing had lost huge amounts of money, gone bankrupt, and/or otherwise been rendered fools.

I have no idea what Blodget is talking about here, and I don’t think he has much of a clue either, because he names no names and provides no links to back up his assertion. People went bankrupt betting on the collapse of the housing market? Who?

The fact is that for most of recorded history it has been pretty much impossible to short houses. At the end of the housing bubble, Robert Shiller tried to create a market in housing futures, but it failed, plagued by low volumes and illiquidity. And until John Paulson came along, it was pretty much impossible to short mortgages, too.

In general, it’s extremely difficult to short any bonds other than Treasuries. That’s one of the reasons that credit default swaps became so popular so quickly: they allowed people to go short credit instruments in a way that was pretty much impossible before. As the housing bubble grew and the quality of subprime mortgages sank, various fund managers, including Paulson, started asking the likes of Goldman Sachs to sell them credit protection on subprime loans. Eventually, as the market in subprime CDS expanded, the ABX index was created, and entire synthetic CDOs started to be constructed out of these derivatives. But there was no such thing as subprime CDS in 2003, or even in 2005, which means that no one was shorting those bonds back then. Let alone going bankrupt doing so.

But if Blodget is wrong about the people who were short this market, he’s laughably, spectacularly wrong about the institutions which went long.

So much money had been made betting on further appreciation of the housing market, meanwhile, that investors were DESPERATE for vehicles that allowed them to make these bets in a more efficient fashion. That’s why the buyers of Goldman’s CDO bought the CDO: They thought housing prices were headed higher, and they wanted to make a killing on it.

This is one of the silliest things I’ve seen in a very long time. The decision to buy into Abacus was not an attempt “to make a killing” by “betting on further appreciation of the housing market”. To the contrary, it was an attempt to lock in a modest yield pick-up over Treasury bonds, in return for accepting a very high degree of illiquidity. That’s why Abacus was carefully structured to get a triple-A credit rating: the buyers of the instrument wanted no risk at all with regards to their payment stream.

Buying triple-A mortgage-backed securities and collecting coupon payments on them is not making money by betting on house prices going up, any more than buying Treasury bonds and collecting coupon payments is making money by betting on tax revenues going up. The whole reason for the overcollateralization and the waterfall structures and the equity tranche and the built-in diversification and all the other bells and whistles which so impressed the ratings agencies was to try to make it impossible that investors in the triple-A part of the structure could lose money. If the investors wanted housing-market risk, they would have bought the equity tranche. They didn’t, and they didn’t. Instead, they wanted to put their money in a structure where their principal was safe, but where they didn’t need to pay the liquidity premium seen in the Treasury market — since they were long-term investors and had no particular need for a liquid investment.

Blodget then continues:

As Goldman has observed, with CDOs like the one in question, there is ALWAYS a short side and a long side: The buyers of the CDO knew that someone was going to be betting against them.

This is true, but also misleading, because the central point of the SEC’s suit is that Paulson was represented to ACA as ACA’s friend and coinvestor in the deal, while in fact they were ACA’s enemy, taking the opposite side of the trade.

But let’s back up a bit and look at the Abacus structure from the point of view of ACA and IKB. Simplifying a little, there were basically three parts to it: equity, bonds, and super-seniors. The bonds were the product of the magic of securitization: triple-A-rated instruments which were inherently safe themselves even if they were made up of risky bits and pieces. The investors buying the credit protection on the underlying securities might get paid out here or there, but never in sufficient quantities to endanger the senior cashflows.

Then there was the equity part of the deal: that was the bit where speculation might be going on. An aggressive hedge fund like Paulson would take the long side, betting that it could make a killing as the underlying mortgages avoided default. (Remember that ACA thought Paulson was long the equity tranche.) Some other hedge fund, or perhaps somebody who was already long housing and wanted to hedge that position, would take the opposite side of the trade, protecting themselves against housing defaults or maybe even making a speculative bet that those defaults would be quite numerous.

In any event, the CDO was designed, in the eyes of ACA and IKB, to confine speculative activity to the equity tranche. Equity investors can lose money — or make a lot. And if they lose, then some unknown short will have gained. The bond investors, by contrast, were the boring ones, clipping coupons and sleeping well at night because they didn’t need to worry about losing their principal.

Finally, there was the super-senior tranche, also known in some circles as the “quadruple-A” tranche. This tranche was so safe, and so low-yielding, that it was never even sold: it was, in the parlance of structured finance, “unfunded”. The risk there was entirely theoretical, and was often laid off onto an insurance company which took in a very low premium in order to make sure all the books balanced.

In reality, however, something very different was going on. It’s a little unclear, but it looks very much as though the equity tranche here was unfunded: Paulson certainly didn’t want it, and Goldman never even tried to find someone so bullish on the subprime housing sector that they would buy it. That’s what Blodget is referring to when he says that “Goldman retained an ownership stake in the CDO and lost money on it”. While Goldman was leading ACA to believe that Paulson was speculating on house prices going up, in fact Paulson was betting that they would go down so much that the entire CDO would be wiped out — not just the equity tranche but also the safe triple-A bits. Even the super-safe unfunded super-senior tranches were at risk.

Goldman knew full well what Paulson was seeing, and deliberately kept its client ACA in the dark as to what Paulson’s motivations were, and what the risks were in the structure. While Goldman was trying to persuade the ratings agencies that the bond part of the structure was perfectly safe, it was also listening to Paulson explain why in fact the structure could blow up spectacularly. Neither ACA nor the ratings agencies ever suspected the true reason for the structure’s existence — because Goldman never told them what it was.

Which brings us to the meat of Blodget’s argument:

If Paulson had had control over which securities were selected for the CDO, this would OBVIOUSLY be fraud: Paulson wanted BAD bonds in the CDO, not good ones. The buyers of the CDO, meanwhile, wanted GOOD bonds. That would be a direct conflict of interest that should obviously have been disclosed.


Paulson did NOT have control over which securities were selected for the CDO.

This seems to me to be tantamount to an admission of defeat. If Paulson controlled which securities were selected for the CDO, says Blodget, then that’s not only fraud but it’s OBVIOUSLY fraud.

Blodget then goes on to explain why he thinks that Paulson did not have such control, but he’s pathetically unconvincing.

The firm that DID have control over which securities were selected, ACA, was a highly sophisticated firm that analyzed securities like this for a living. It had FULL CONTROL over which securities were included in the CDO. We know this because, of the 123 bonds that Paulson proposed for the CDO, ACA only included 55 of them. In other words, ACA dinged more than half of the bonds Paulson wanted in the CDO, presumably because they did not meet ACA’s quality hurdle.

Let’s remember here that in the end there were 90 securities in the CDO. Of those 90, it seems that 55 were chosen by Paulson. In other words, more than 60% of the securities in the CDO were picked, essentially, out of a stacked deck. It didn’t matter which securities ACA chose; Paulson had come up with his longlist of 123 securities precisely because all of them were particularly toxic. That’s a material fact which, if ACA had known it, would surely have sufficed to get them to exit the deal entirely.

But it’s actually worse than that: the fact is that Paulson not only proposed 55 of the 90 securities, but also had veto power over the other 35, and signed off on all of them. The timeline is complicated, and stretches from January 9 to February 26 of 2007, with quite a lot of back-and-forth between ACA, Goldman, and Paulson. And in fact it’s not entirely clear that exactly 55 of the final 90 securities were on the initial Paulson longlist. But what is clear is that whenever ACA proposed adding any new securities, Paulson could and did exclude those it didn’t like from the final structure: on February 5, for instance, it deleted eight of ACA’s names from consideration. And in the end it declared itself happy with every single name in the structure.

Given all that, it makes no sense to say that ACA had full control over which securities were included. Really, it was Paulson which had that control: ACA was confined to tinkering at the margins, and only to the degree that Paulson was OK with their alterations. If and when ACA tried to include securities Paulson didn’t like, Paulson simply removed those securities from the deal. So no, ACA did not have full control over anything.

Blodget tries to paint Paulson as merely “influencing” the contents of the CDO structure, in much the same way that PR flacks, the media, or stock screeners might influence decisions. But none of those influences had veto power over the outcome, or were proactively involved in putting the deal together, or, for that matter, paid Goldman Sachs $15 million in order to get exactly what they wanted from the structure. (It’s worth noting here that Alan Dershowitz has said that Paulson “could easily have been charged with conspiracy to defraud” in this matter.)

At this point, Blodget starts just making stuff up.

ACA, furthermore, did not just pick the securities. It BOUGHT THE CDO. ACA’s parent invested more than $900 million in the CDO.

Er, no. ACA’s parent did not “invest” money in the CDO: it did not take $900 million in cash, and give it to Goldman Sachs in return for securities. The parent that Blodget is referring to here is ACA Capital, the insurance company which insured the super-senior tranche of the CDO. From the complaint:

On or about May 31, 2007, ACA Capital sold protection or “wrapped” the $909 million super senior tranche of ABACUS 2007-AC1, meaning that it assumed the credit risk associated with that portion of the capital structure via a CDS in exchange for premium payments of approximately 50 basis points per year.

ACA Capital no more “invested” $900 million in the CDO than an insurance company is investing $1 million if it starts accepting insurance premiums on a $1 million house. Yet Blodget says that ACA “bet almost $1 billion”. He’s wrong. ACA certainly made a very bad decision when it agreed to write this particular policy. But it never took the best part of a billion dollars and wagered it. Maybe it should have done: the discipline of seeing $909 million disappear out the door would certainly have stopped ACA from writing this policy.

Blodget then says something very interesting:

Goldman argues that the Abacus process was the same process used to construct all of these securities, not just at Goldman–in other words, it was a standard industry practice. A transaction sponsor (in this case Paulson), paid for the CDO to be created, the bank created it and sold it, someone went long and someone went short, and everyone knew all of that.

Has Blodget been talking to Goldman? Because there’s nothing in the official Goldman statement along these lines. Probably because it isn’t true. The fact is that standard industry practice was for the transaction sponsor to take the equity tranche, precisely because that helped to align the interests of the sponsor with those of the CDO manager. In fact, I challenge Blodget to come up with a single other CDO where the sponsor was not an equity investor. This was not a standard deal, by any stretch of the imagination.

Blodget concludes with a conditional statement: “the SEC may criminalize this lack of disclosure in hindsight,” he writes, “but if it was a standard industry practice at the time, Goldman likely has a solid defense.” That’s a very big if. And it’s going to be very hard indeed for Goldman to make the case that it was standard industry practice for a speculative short to be given veto power over a CDO manager’s picks, without the CDO manager knowing of the speculator’s position or motives. Henry might be rooting for his Goldman Sachs home team here, but if this is their best defense, it’s looking decidedly rocky.


Comments are closed.