Felix Salmon

Goldman’s Abacus fee

Felix Salmon
Apr 22, 2010 15:09 UTC

A reader writes in with an extremely good question — does anybody know the answer?

Paulson reportedly paid GS $15 million to construct the transaction. Was this a standard practice for GS to charge an explicit fee outside of the transaction from one party to arrange a transaction versus charging a fee within the deal to all parties to the transaction? It is also a reasonable question to ask if this was the going rate.

This was I think the only Abacus deal which was done on behalf of an outside sponsor, rather than being created in-house at Goldman Sachs. In the other deals, presumably Goldman paid itself a fee by getting money from the cash investors in the deal — the people on the long side. It makes sense that banks would extract fees from the long side of the deal because that’s where the money is: the long side is made up investors who put up a large amount of cash upfront, in return for getting a stream of insurance payments in the future. It’s a lot easier for a bank to skim off a small percentage of that cash as a fee than it is to ask the people paying insurance premiums to pay the bank an extra fee for the privilege.

More generally, was $15 million a normal fee for a deal of this size? The total cash invested was only I think about $200 million, which implies that Goldman’s fee was a whopping 7.5% of the cash size of the deal. Is that the kind of money that synthetic/hybrid CDO arrangers normally took on these things? It seems high to me.

As economists love to say, incentives matter, and if you have unusual incentives, either in the way your fee is structured or in its sheer size, then that surely increases the chances of an unusual outcome or non-normal behavior. So it would be great to see what the fee structure was on other, similar deals.


The $15m is in proportion to some deals I saw first hand.

Posted by vk9141 | Report as abusive

Did ACA know Paulson was going short?

Felix Salmon
Apr 22, 2010 00:56 UTC

Steve Liesman has got his hands on Paolo Pellegrini’s testimony to the SEC, and you can see why the Goldman case seems to be building up to the conclusion that ACA did know Paulson was going to short the deal:

In one part of Pellegrini’s testimony, a government official asked him: “Did you tell (Schwartz) that you were interested in taking a short position in Abacus?”

“Yes, that was the purpose of the meeting,” Pellegrini responded.

“How did you explain that to her?” the government official said.

“That we wanted to buy protection on traunches of a synthetic RMBS portfolio.” Pellegrini said.

Contrast the SEC complaint:

Had ACA been aware that Paulson was taking a short position against the CDO, ACA would have been reluctant to allow Paulson to occupy an influential role in the selection of the reference portfolio because it would present serious reputational risk to ACA, which was in effect endorsing the reference portfolio.

This is fascinating stuff, not least because there doesn’t seem to be any hint of it in Paulson’s letter. But if Pellegrini’s testimony turns out to be reliable, it surely constitutes a simple disproof of the SEC statement.

Once again, we’re left desperately wanting to see the full testimony, rather than the little bit cherry-picked by Liesman, or his source. The SEC promises that we’ll see just that — “in court at the appropriate time”. Which could just be a way of dragging this whole thing out and keeping the pressure on Goldman to settle the case, or could be simple confidence that seen in context, Pellegrini’s testimony doesn’t really end up being all that powerful.

If this does go all the way to court, the room is going to be packed on the days when Schwartz and Pellegrini testify. But in the meantime, it would be great if an enterprising journalist somewhere could get one or other of them on the record. It might clear up a very great deal of confusion.

Update: Also note the comment below from Mustard, who looked carefully at the dates involved:

Pellegrini likely did not know during his January 27th, 2007 meeting in Jackson Hole with ACA (paragraph 31 of complaint) that Goldman in its January 10, 2007 email to ACA (paragraph 47) had mislead ACA into reasonably believing that Paulson was taking the high-risk equity tranche of the synthetic CDO.

Thus Pellegrini’s statement to ACA that Paulson “wanted to buy protection on traunches of a synthetic RMBS portfolio” would have in his mind implied a short position, but to ACA – after the misrepresentation by Goldman – the statement would have suggested a hedge on Paulson’s supposedly intended high-risk equity position, not a naked short.


Goldman could put their whole rigamarole on YouTube, but it would still look like an assisted suicide by them and Paulson on ACA. Only question remains whether ACA was a willing victim, and what motivated them (or anybody else now, for that matter) to go on supping the lethal GoldAids.

Posted by HBC | Report as abusive

Revisiting rent vs buy

Felix Salmon
Apr 21, 2010 20:39 UTC

David Leonhardt today revisits the great rent vs buy debate — something I was talking about on All Things Considered just this weekend. He’s also updated the excellent NYT rent vs buy calculator, which he uses to agree with Dean Baker that the point at which it makes more sense to buy than to rent is probably closer to 15 than to 20, if you’re looking at the ratio of house price to alternative annual rent.

This all makes sense to me — but at the same time I’d still be tread warily if I were buying a place right now. For one thing, it’s worth avoiding the housing market between now and the end of the month, as would-be home buyers scramble to sign a contract before the federal tax credit of up to $8,000 expires. That’s bound to push prices up a little at the margin, and there’s no point getting caught up in the frenzy. If you need the tax credit to justify buying a house, then you shouldn’t buy that house.

More generally, Leonhardt seems to assume we’ve had all that we’re going to get, when it comes to house-price corrections. Check out his assumptions in his blog entry:

Among other things, I assumed rents and house prices would both increase 2 percent a year — along with inflation — and a mortgage rate of 5.25 percent.

The mortgage rate is reasonable — indeed, it might even be lower than that right now, although it does seem to be bouncing around quite violently, for reasons I don’t understand. But if you plug in zero appreciation for both rents and house prices, then in Leonhardt’s example you never get into positive territory, and level off with a loss of more than $5,000 on the house.

It’s hard to imagine a world where rents and house prices never go up — but we might well be entering just such a world. I rehearsed many of the arguments a couple of weeks ago, and I shan’t repeat them here — interest rates are low and rising; the government is artificially propping up the market; prices need to continue to fall just to revert to their long-term mean. But when it comes to rent vs buy calculations, the point is that you can no longer safely assume that rents will always go up. They might not have risen as much as prices during the bubble, but they did outpace inflation, and it does make conceptual sense that if house prices aren’t rising then there’s no real reason for rents to do so either.

My feeling is that house prices are going to fall in real terms — that they won’t keep pace with inflation — for the best part of this decade. And in that environment, buying a house is really not that smart, especially given the opportunity costs associated with doing so. Leonhardt’s calculator puts those costs, for his hypothetical $384,000 house with a 20% down payment, at $3,830 a year — 5% of the down payment. But in reality the opportunity costs are much greater than just the return that you would get on that money if you invested it. Renting gives you much more agility and mobility than if you tie yourself down by buying a place which might be quite hard to sell in a pinch. And having $76,800 in cash can come in extremely handy in many circumstances. For one thing, if say you’re laid off, it can pay the rent for a months if not years; if you’ve already used that money for a down payment, then you can’t use it again to pay the mortgage on your house.

So while Leonhardt is right that in much of the country price-to-rent ratios have come down to their historical mean, that’s not necessarily a good reason to buy. It would be more accurate to say that one of the many good reasons not to buy has gone away.


Excellent discussion.
Although the NYT Calculator and most of this discussion seems to be dealing with the US market, I have tried to apply it to Europe.
Very few European markets are doing well. And except for very few major cities, it is almost always cheaper (read well, I’m not saying ‘better’) to rent.
Especially for 60+ semi-retired and restless, travel-hungry people like myself.
Say I do have the EUR 300.000 in cash needed to pay for a very nice apartment in Lisbon. Why on earth would I freeze that money, when I can also rent the same, or a similar apartment for EUR 1100/1200 per month? I’ll make some interest on secure investments and saving accounts.
In Winter I could go and stay 3 months in a 5* hotel on the Turkish Riviera for an average EUR30 per day.
Mind you, that includes my round-trip flight, transfers, a double suite for single use, all meals and drinks, internet and daily cleaning.
Egypt would be another option, when things were less stormy.
Friends and family can use my apartment in Lisbon when I’m not there.

What I read from this debate is that most people defending Home Ownership, do so for emotional reasons. They either have just bought one, or are trying to sell their house, which isn’t going too well.
There’s nobody honest enough to say: “Sh@t, you’re right. I wish we could sell and rent instead”.

In the old days you bought a house for your family, close enough to work. Many made their career within the company and never left.
These days are different. Companies move, and will get rid of you when you do not move with them. Universities have no space, so you have to go and get your degree at the other side of the country. Or somewhere else in Europe.
Imagine you are stuck with a mortgaged house you can’t sell. What do you do?

I doubt very much that the average m2 prices of houses in Europe will increase in this decade. Those countries where the interest rate on mortgages were tax-deductible (Netherlands, e.g.) are gradually abolishing this option.
The result is that people are getting worried, and start putting their houses on the market.
Few are buying, because that deduction was one of the FEW reasons their monthly costs were similar and only slightly higher than rental.
A dark picture ……

My tip: when you see a wonderful home, or an apartment in a wonderful area, try to make a good rental deal. I bet your purchasing price will be lower next year!!

Posted by Maria2011 | Report as abusive

Zuckerman defends Paulson

Felix Salmon
Apr 21, 2010 19:09 UTC

Bess Levin has a Q&A today with John Paulson’s biographer Greg Zuckerman:

Do you think even if Goldman had disclosed what the S.E.C. says it should have, regarding Paulson’s role, the investors would’ve made the same decision on it?

Yeah, I don’t think many investors would have had a second thought about taking the other side of a trade of John Paulson’s back in 2006 or early 2007. He was seen as a tourist investor dabbling in real estate, and some people thought he was out of his league—even Goldman Sachs thought he was out of his league. Josh Birnbaum, a top trader at Goldman, sat across from Paulson in his office and warned him about what he was doing.

Goldman is all about—or saying it’s all about—the clients being number one, and now because of all this, people, investors, are questioning that commitment. But Paulson was a Goldman client. So they were still working in one client’s best interest. Sometimes you have to play favorites, right?

Yeah, I mean, you could argue that if John Paulson had lost a billion dollars on this deal, that he should have complained that he’s a client of Goldman Sachs, and he wasn’t warned that a sophisticated, huge German bank which does credit analysis on a daily basis was on the other side of his transaction.

I’m beginning to think that we really need to hear from the people at ACA who were involved in the transaction — and ideally look at contemporaneous emails from them as well. It’s entirely within the realm of possibility that they thought Paulson was an idiot to short this stuff, and that they reckoned they were making lots of money off him by taking the other side of the trade.

But that’s not what the SEC alleges, and it’s worth remembering that the SEC has been looking into this deal since the summer of 2008, and got 8 million pages of documents on it from Goldman Sachs alone. The SEC says clearly and unambiguously that Goldman misled ACA into believing that Paulson was long, rather than short. So I do wonder where this idea is coming from that ACA actually knew Paulson was short.

Because if what Zuckerman is saying is true — if it wouldn’t have made the slightest bit of difference to inform ACA that Paulson was short — then why on earth didn’t they do that? Hell, why didn’t Paulson do that, if Goldman didn’t want to? This suit would develop a gaping hole and would probably never have been brought if Pellegrini or anybody at Paulson had simply mentioned, in the meeting with ACA in Jackson Hole, that they intended to fund the entire structure. The fact that they didn’t is prima facie evidence that they thought that such disclosure would have made a difference after all.

As for the argument that Paulson could have complained about Goldman’s actions if he’d ended up losing money, that’s simply ridiculous and I don’t understand why Zuckerman is giving it any credence at all. This was a reverse inquiry: the whole deal was Paulson’s idea. Goldman just made it reality, and it genuinely didn’t make the slightest difference to Paulson who was on the other side of the trade.

And yes, it did make a difference to ACA who was on the other side of the trade, if that person was Paulson in particular — precisely because it was always Paulson’s deal, and Paulson had enormous control over what securities went into it. It’s one thing entering into a wager with someone about an outcome neither of you can influence — it’s something else entirely to enter into a wager with someone when the outcome is even partially under your counterparty’s control. If ACA knew that it was betting against Paulson, then Paulson’s role in picking the contents of the deal would have looked very different indeed.

Goldman, of course, had a duty to be honest about the deal; the only reason that Paulson isn’t being charged here is that he had no such duty. But they both lied to ACA, if lies of omission count as lies. And the attempted defenses of Paulson are far from convincing.


I am still so confused how the guy who writes a book about Paulson’s greedy guts deals says Paulson is innocent of all wrong-doing and defends Goldman and says the SEC case is weak, so many media/people believe that they must not be guilty so it is now over?

Pelligrini says he told ACA that Paulson was shorting. But at any time was Paulson misrepresented as an independent third party that designed the portfolio? it looks as though Paulson is also defending themselves by defending Goldman.

Just who did Pelligrini tell. Just the one woman? and what did he say and what written proof is there of this or was it divulged over a few beer? What constitutes the telling if there is no documentation to the fact? Is this the way ‘highly sophisticated traders’ mean to do business?

It seems that the media is backing away from pressuring Goldman, but with 8 million pages of crap to go through, there is evidence of wrong-doing that triggered the charges. And being there were more then a few similar deals made other then Abacus, how many others were as unethical.

There were billions of dollars given in bonuses to brokers for money that was ‘foreseen’ but not yet ‘realized’, so doesn’t Goldman have a fiduciary responsibility to its shareholders for having given bonuses for cooked books?

I guess if being an ‘investment’ bookie is legal, then crafty accounting is also OK. Goldman said it wasn’t crooked to it’s shareholders Apr 7, and said it didn’t bet against its own clients, then I guess it must be a great monolithic untouchable that should be left alone to do business as usual.

And all of the fellow firms (God bless all the institutions doing God’s work!)say Goldman should just cut their losses and take a big hit and pay the fines, even though they are innocent. Sadly, it’s because they know they will be next.

Good lord, if nothing else, isn’t his another wake up call that reform is necessary?

Posted by hsvkitty | Report as abusive

The Paulson letter

Felix Salmon
Apr 21, 2010 15:06 UTC

Both the WSJ and the FT have got their hands on a letter from John Paulson to his investors, but infuriatingly neither of them have seen fit to share it with their readers. So we’re left with a few snippets of direct quotes, and we have to simply guess at the overall tone of the note.

Here’s the WSJ:

Mr. Paulson sent a letter to investors Tuesday night saying that in 2007 his firm wasn’t seen as an experienced mortgage investor, and that “many of the most sophisticated investors in the world” were “more than willing to bet against us.”

And here’s the FT:

“Paulson was transparent and open regarding its concerns about the mortgage market which were driven by analysis of publicly available data,” the letter states…

In his letter, Mr Paulson took pains to cite the SEC complaint that states that while he suggested 123 securities to be included in Abacus, ACA, the firm that acted as collateral manager and chose the mortgage securities to be included, ultimately accepted 55 and rejected 68. “All our dealings were through arm’s length transactions with experienced counterparties who had opposing views,” the letter adds.

Without seeing the actual letter, I’m going to have to conclude that it reads largely as a set of Goldman Sachs talking points, which is interesting: the fact that the SEC has charged Goldman but not Paulson gave Paulson an opportunity to try to distance himself from the alleged fraud. But instead he seems to be lining up with Goldman here — a stance which he might regret, if the SEC makes its fraud charges stick.

Once again, there seems to be a lot of hints and nudges in the direction that ACA knew that Paulson was short and willingly took the opposite side of his trade. Which sophisticated investors is Paulson talking about, who were more than willing to bet against him, if not ACA and IKB? Especially seeing as how the “transparent and open” Paulson had meetings with ACA in the run-up to the Abacus deal.

And once again we see the odd Goldman defense that ACA rejected a chunk of Paulson’s picks, with the unhelpful implication that if it hadn’t done so, then disclosure might have been warranted. At that point the question becomes just how many Paulson picks ACA needs to reject before you don’t need disclosure any more, and I don’t think that Goldman is going to like the answer to that question.

Paulson is now one of the largest hedge-fund managers in the world, largely because of all the people who piled into his funds after he made lots of money shorting mortgages. Those people aren’t particularly loyal to him: they haven’t been with him all that long. And if they think he’s starting to smell a little toxic, they’ll probably start looking for other places to put their money. Once upon a time, investors probably liked the fact that Paulson was so aggressive in his bets that he would do things like hire former SEC chairman Harvey Pitt, who then lobbied against proposed moves which would allow underwriters to modify mortgages and thereby help homeowners. Nowadays, however, that kind of activity is increasingly the kind of thing that investors want to disassociate themselves from.

So while I doubt Paulson will get a massive wave of redemption letters next Friday, I do think that his grip on his AUM is becoming increasingly tenuous. In turn, that might lead him to take a few more risks in an attempt to boost his returns and keep his investors that way.


A broker to any transaction, has responsibilities to BOTH parties, not just the ‘client’. Paulson spins that this is somehow ‘legal’ because fees were paid and someone else did the selling.

The essential crime is in designing an investment to fail in the first place. Defacto, if it sells, the party that is long has not been adequately informed.

In another world it is as if an insurance company discovers that one need not merely place bets against certain aircraft designs, but rather designs something that clearly will never fly in the first place. That’s easy. What’s needed is an unstoppable sales agent to sell it. (GS).

Posted by firstbridge | Report as abusive


Felix Salmon
Apr 21, 2010 07:26 UTC

What on earth is Bloomberg doing hiring Ben Stein to write for them? — Bloomberg

JPMChase sleaziness, implying that PIN debit transactions are less safe — American Banker

Did Palm really say that Tourre told ACA that Paulson was shorting the structure? I didn’t take him as saying that — Telegraph


You could ask the same thing about Amity Shlaes, who left the Financial Times shortly after becoming a prominent apologist for the Bush Administration’s response to Hurricane Katrina.

Or Kevin “Dow 36,000″ Hassett.

Bloomberg is becoming the retirement home for some of Wall Street’s most inane commentators. It was very frustrating to me when I worked there, as the company maintains high ethical standards on the reporting side. To then read these commentaries, published with the same little (Bloomberg) dateline, in the same typeface, as if they had gone through the same scrutiny as our reported, fact-checked, exceedingly neutral-voiced stories — there were definitely times when it set me off.

Posted by Setty | Report as abusive

Revisiting the Magnetar Trade

Felix Salmon
Apr 21, 2010 01:45 UTC

In the wake of the SEC charges against Goldman Sachs, a lot of people have been wondering whether Magnetar, or its bankers, might be next up in the SEC’s crosshairs. Moe Tkacik has been looking into Magnetar for a couple of months now, and I spent a lot of time on IM with her today, learning just what it is about Magnetar which seems to drive people crazy — to the point at which they’ll happily spend seven months on a journalistic investigation of the company.

I’m not going to even attempt to get into all the nooks and crannies and conspiracy theories which surround Magnetar, but I think, after talking to Moe at some length, that I have a much better grasp of the very big picture than I did after reading the ProPublica report.

The story begins in 2005, when Greg Lippmann of Deutsche Bank gets ISDA to create some standardized language for credit default swaps on subprime mortgages. That was the spark that lit the fire — and one of the first people to realize the enormity of the potential conflagration was Alec Litowitz of Magnetar. Lots of hedge funds had the bright idea of shorting mortgages: you can read Michael Lewis’s book on many of them, or Greg Zuckerman’s on the biggest one of all, Paulson & Co.

But that was a simple trade. Someone like Michael Burry or John Paulson or Andrew Lahde would go up to Deutsche Bank or Goldman Sachs, and say “hi, I want to buy credit protection on the BBB-rated tranches of subprime RMBS”. And then those banks would have to find someone else willing to take the other side of the trade, which wasn’t always very easy. And the market never really got very big or important.

But then Litowitz hit on the idea of a mezzanine subprime hybrid CDO — and that was a real game-changer.

First it’s worth explaining exactly what such a strange beast is. A CDO is just a collection of fixed-income instruments, aggregated and tranched. In this animal, all the instruments are “mezzanine” — which means they carried the very lowest investment-grade rating, triple-B. They were also all subprime. And “hybrid” means that the CDO was a mixture of cash bonds and artificial credit default swaps, normally in a ratio of roughly 4 CDSs for every bond.

There were subprime synthetic CDOs before Litowitz came along — Greg Lippmann had created some — but they were generally linked to the broad ABX index, which meant that you couldn’t construct them with anything like the level of specificity and granularity that Litowitz was looking for. And Litowitz’s insight was that the ratings agencies’ models didn’t look at loan-level data on the contents of CDOs, they just looked at the ratings of the contents of CDOs. So you could fill a CDO up with all manner of subprime NINJA nuclear waste, and it would look to the ratings agencies exactly the same as if it held much safer BBB tranches of prime fixed-rate mortgages.

Once the ratings agencies blessed the structure with their magic-fairy-dust triple-A ratings, there was always someone willing to buy it at a modest pickup over Treasuries or Libor. That was the nature of bond investors during the Great Moderation: they simply didn’t have the time or the inclination to investigate the contents of every triple-A bond they were shown. Instead, they were busy dreaming up clever structures of their own: if they funded themselves short in the ABCP market, and invested the proceeds long in the subprime CDO market, they could make a fortune. And they could always get funding in the ABCP market because their collateral had the precious triple-A rating. In many ways it didn’t really matter what was inside it, just so long as the rating was there and it ended up paying off in the end.

Magnetar’s subprime CDOs started appearing in the spring of 2006, as the first cracks were beginning to appear in the housing market, and continued through the summer of 2007. The likes of Michael Burry had already put on their short positions at this point, and were just waiting for the inevitable market plunge and their handsome paydays, as the value of their default protection soared.

But Litowitz threw a weird spanner in their works. He kept on buying up cash mortgage bonds to put into his CDOs, which kept that market high. And he bought up the toxic equity tranche of the CDOs too, which no one in their right mind wanted to touch with a bargepole. (Although, scandalously, naive public pension funds were buying equity in these things too, right up until the end.) And he was creating these CDOs in such enormous volumes that even with him providing a huge amount of the short interest, there was still a lot more to go round, to the likes of Burry and Paulson, who would get offered the opportunity to buy protection at lower rates than they were used to.

That’s a mixed blessing for a hedge fund manager like Burry. On the one hand, he wanted all the protection he could lay his hands on, the cheaper the better. But on the other hand, he had to mark his holdings to market, and the price of protection was going down rather than up. And so his investors started getting very antsy indeed: not only was he paying out millions in insurance premiums, but the value of his insurance was falling.

Meanwhile, on the long side of things, the fact that subprime CDS prices were staying cheap, even after the real-world subprime housing market had started falling apart at the seams, only served to confirm in the minds of bond investors and monoline insurers that the models were right and that there was nothing to worry about. So they kept on buying those triple-A bonds, even when they were made up of nothing but triple-B dreck. After Magnetar was happy to buy the equity, which was specifically designed to insulate the bondholders from any harm.

In a narrow sense, then, the Magnetar Trade did indeed involve capital-structure arbitrage of custom-designed synthetic CDOs, as they say it did. Litowitz saw that there was a mispricing in the market, and that insurance was being sold too cheaply, so he created as many vehicles as he could just so that he could buy insurance on them.

The bigger picture, however, was that Magnetar’s vehicle production line helped to perpetuate the bubble, and almost certainly reassured not only bond investors but even the likes of Hank Paulson and Ben Bernanke that the problems in the subprime mortgage market were confined to a few mortgage companies, and wouldn’t have knock-on effects in the financial system. After all, the market knew exactly what was going on in California, and didn’t seem to be worried in the slightest!

Here’s part of what Moe IMed me today:

Magnetar sponsored $40 billion worth of the worst of the worst CDO deals, thus propping up the whole damn market, and Paulson did $5 billion, at the end, only after (I imagine) he caught onto Magnetar’s trade. In the end Paulson looks guiltier because he made more money. BUT. What Magnetar did was much more akin to what the investment banks do every day, which is spin rampant-conflicts-of-interest into megasurefire profits.

In Magnetar’s case, however, they only had their investors to answer to, so they have not had to bother making up some mendacious bullshit explanation of their role in the broader economy such as “efficient allocation of capital”.

From a systemic perspective, Magnetar had a much bigger effect — and a much worse effect — than Paulson. That’s what makes people like Moe and Yves Smith angry. (Much of what Moe learned she got from Yves and her book.) Magnetar was in many ways the engine which was responsible for many of the worst losses from New York to Dusseldorf. Those losses didn’t directly become Magnetar profits, because Magnetar was long equity and generally hedged in a way that Paulson and Burry weren’t. But they did end up helping to cause the biggest recession in living memory.

Update: Just to clarify, both ProPublica and Smith have done amazing work uncovering the Magnetar Trade and its sheer enormity. It’s a really hard story to crack, and they did an enormous amount of heavy lifting and deserve everybody’s gratitude. ProPublica, especially, is a case study in how to put source documents online and lay complicated things out in a wonderfully transparent manner — while at the same time being accessible enough for This American Life on NPR. I  didn’t mean to denigrate their excellent reporting in any way, but it came out that way and I apologize for that.


Felix, I don’t think the question here is the terminology of the most senior tranche, but whether it would have been possible to maintain the profitability of CDOs in the absence of an unfunded (and therefore very low cost) most senior tranche. I think that cash investors were demanding higher yields than unfunded “investors” and that there’s no way that all the deals that took place with unfunded super senior could have been issued as cash deals.

Posted by csissoko | Report as abusive

Goldman’s unwanted super-senior position

Felix Salmon
Apr 21, 2010 00:05 UTC

Kate Kelly has the obvious answer to the question of what on earth Goldman Sachs was doing with that 45-50% super-senior tranche:

“The core of the SEC’s case is the allegation that one employee misled two professional investors by failing to disclose the role of another market participant in a transaction,” said Goldman CEO Lloyd Blankfein said in a recent message to employees. The firm “assumed risk in the deal, and we lost money,” he added.

But Goldman invested the money only because sales of the deal didn’t play out as planned, forcing Goldman to step up with its own money, people familiar with the matter say.

This is surely exactly right. We’ve already seen that it took Goldman a full five weeks after the deal closed just to wrap the 45% of the super-senior tranche that it did manage to get off its books — and it did so at the end of May 2007, when the mortgage market was hitting three-month highs and there was a lot of hope in the market that the early-2007 crash in the ABX index had been a short-lived aberration.

Here’s my theory of what happened: Goldman intended to wrap the super-senior tranche with ACA more or less at exactly the same time as the deal closed. But it couldn’t come to terms with ACA on the question of posting collateral: ACA, which only had a single-A rating even then, wanted a Berkshire Hathaway-style deal whereby it didn’t need to post collateral on a day-to-day basis and just needed to pay out once there was a credit event. That wasn’t acceptable to Goldman’s risk managers, which spent a good amount of time putting a deal together whereby ABN Amro would buy the wrap from ACA, and in turn would sell insurance to Goldman while posting collateral as necessary.

But the price rose when Goldman had to go through ABN Amro, and so Goldman ended up offloading only the top 45% of the 50% super-senior tranche. The rest it kept on its own books as a trading position, looking to sell that unwanted long position on an opportunistic basis.

And then Goldman ran out of time: before it could sell the position, the structure had started imploding, and Goldman was stuck with $90+ million of losses.

Which now Blankfein and Goldman are trying to turn into some kind of virtue.

This is one of those things a bit like the argument saying that ACA rejected half of Paulson’s picks: making the argument involves making a subtle implication that the case would be stronger were these facts not true. Is Goldman saying that the SEC’s case would be stronger if ACA had not rejected any of Paulson’s picks? Is it saying that the SEC’s case would be stronger if Goldman had managed to wrap the whole super-senior tranche and had ended up making a profit on the deal? Because it’s pretty obvious that Goldman would have been perfectly happy with the outcome in both those cases, and still wouldn’t have disclosed anything more than it did.

Of course, you can’t sue a bank for not disclosing information in a hypothetical deal which never actually happened. But the legal case is only half the point, and not even the more important half. The ethical case is what matters, and Goldman’s doing a very bad job of answering that one.


What’s your source on that, Felix? I’ve been trying to work out the who-held-what-when questions since Monday, and I’ve yet to see anything compelling to suggest that Paulson managed to short (i.e. buy protection on) the *entire* capital structure (0-100) of the ABACUS 2007-AC1 reference portfolio.

Paulson probably would have loved to have purchased protection on the whole thing, but Goldman wouldn’t likely sell it to them unless it had someone on the other side on whom to offload the risk. From what I can glean from reports/allegations/defense docs/bloomberg, the ABACUS deal only had $192M of securities issued (to only IKB for $150M and ACA for $42M). I believe these securities represent what is effectively a 34.4%-45% tranche on a $1.818B reference portfolio. Incremental to this (but never passing-through the ABACUS legal entity) there was a supersenior (50-100) trade between GS and ACA/ABN to the tune of $909B. (Note: this is where I get $1.818B = $909Bx2). Of course, Goldman sold all of that protection (and a little bit more) to Paulson.

If you’re wondering where all the lower tranches (the 0-34% first-loss risks) went, stop wondering. If my suspicious are correct, it never existed. The lower tranches referred to in the pitchbook never found a way to market because Goldman couldn’t find any investors. Now, mechanically, there are a number of ways to accomplish this feat; one way is for Goldman to enter into a single-tranche bespoke basket default swap facing the ABACUS SPV (in the first instance, but this was ultimately backed-to-backed with Paulson). A single-tranche bespoke basket default swap is basically a customized tranche (e.g. 34.4% to 45%) of an unfunded synthetic CDO all wrapped up in an ISDA-friendly derivative contract. It’s a bilateral derivative agreement with a specified reference portfolio (usually it just points to the deal docs for the notes) and attachment/detachment points.

Posted by Sandrew | Report as abusive

How the SEC cracks down on unethical behavior

Felix Salmon
Apr 20, 2010 17:25 UTC

Emanuel Derman has a fantastic two-line blog entry on the SEC/Goldman affair, which I can’t really help but quote in full:

The architects of the bailout have been trying to cure insolvency by treating it as illiquidity.

The SEC may be trying to cure unethical behavior by treating it as illegality.

This is also known as “how people behave when the only tool they have is a hammer”. Central banks can inject liquidity much more easily than finance ministries can spend money. And the lawyered-up SEC, if it finds a deal it considers odious, will go to great lengths to find a way in which that deal is illegal. Once they’ve done that, Goldman’s lawyers at Sullivan & Cromwell will go to equally great lengths and start quoting City of Monroe Employees Ret. System v. Bridgestone Corp, along with lots of other prior cases, in support of their argument.

I don’t doubt for a minute that Goldman’s behavior was more unethical than it was illegal. Goldmanites never stop talking about how they always put clients first, but because the U.S. has a rules-based rather than a principles-based regulatory system, that’s not an explicit regulatory requirement. One thing that the SEC has already done, in filing its complaint and making it public, is reveal that at least one banker — “Fab” Fabrice Tourre — does not fit that conception at all. Rather than treat every client with the utmost respect and transparency, he favored the sponsor of the Abacus deal, John Paulson, who was paying Goldman $15 million to put it together. And he blithely talked about a 0-9% equity tranche in emails to ACA, when no such tranche even existed.

Here’s the cunning thing about the SEC filing: they could settle with Goldman for $1 tomorrow, and Goldman’s reputation would still be tarnished for years to come. Here’s Allan Sloan:

I don’t much care about the legality of what Goldman allegedly did, because something doesn’t have to be illegal to be wrong. And almost everything about the Abacus 2007-AC1 “synthetic collateralized debt obligation” deal was wrong…

The SEC case seems more than a little weak legally, but tars Goldman as amoral at best, immoral at worst. It will take years for Goldman to erase the stain to its reputation, if it ever does.

If I were a betting man, I’d put money on Goldman prevailing in court, should the SEC charges go to trial. But in the court of public opinion, Goldman has already lost. And given the current state of things, that’s the court that matters.

This helps to explain why Goldman leads off its public self-defense by characterizing this deal as “a single transaction in the face of an extensive record”; general counsel Greg Palm said something very similar on the conference call today. Everybody at Goldman is hyper-aware of the degree to which a single mistake can ruin a bank’s reputation, and also of the degree to which that reputation is singlehandedly responsible for bringing enormous amounts of money into the bank. If you’re a journalist, no matter how good you are, you get better if you start working for the NYT or WSJ: your calls are returned that much more quickly, you’re that much more likely to be chosen as the outlet for leaks, and so on.

Similarly, if you’re a banker, you get better on your first day working for Goldman Sachs. Your calls get returned, you get better access to clients, and so on and so forth. The revelations in the SEC’s suit will hurt Goldman’s revenues, and Goldman wants to signal to the markets that they’re not at all typical of how it normally does business — while at the same time maintaining that they’re not even well-grounded in the first place.

It also helps explain why the SEC didn’t give Goldman the opportunity to settle the charges. The real damage to Goldman has already been done, and is much greater than any fine it might end up having to pay the SEC. (Which won’t be large, given that the firm lost money on the trade.) It’s an interesting case of the SEC using its rules-based structure to impose damages on firms who violate a hypothetical principles-based regime. It’s not what the SEC is necessarily meant to do, but that doesn’t mean it isn’t extremely effective.


In the court of public opinion, Goldman is already toast:

“73% Say It’s Likely Goldman Sachs Committed Fraud”
http://www.rasmussenreports.com/public_c ontent/business/general_business/april_2 010/73_say_it_s_likely_goldman_sachs_com mitted_fraud

Posted by DanHess | Report as abusive