The SEC case against Goldman Sachs is bewildering in its complexity to outsiders. It is very hard to know what actually happened without being familiar with the technical world of synthetic CDOs. Ambitious readers can go here and here to get a flavor of the complexity.
But some clarity is possible. John Paulson expected housing prices to go down. The ABACUS vehicle at the heart of the transaction allowed him to profit when they did go down. Who was on the other side of the transaction? Who lost the $900 million that Paulson collected?
Goldman had trouble selling the long side of the deal. Paulson was able to bet about a billion dollars that the housing market would go down by buying insurance against parts of ABACUS, insurance Paulson would collect if the housing market and therefore the underlying mortgages and therefore the underlying bonds of ABACUS plummeted, which they ultimately did.
Goldman told buyers of the package that ACA Management had chosen those underlying bonds. ACA Capital, another part of ACA, offered the insurance against losses. But ACA was only A-rated. (And ultimately grossly undercapitalized to keep all their promises.) So Goldman (I think, or maybe it was ACA) paid ABN Amro, a Dutch bank, to stand behind ACA’s insurance. For that promise, ABN Amro received $7 million. Alas, when the housing market went south, they ended up having to pay $841 million to Goldman and then John Paulson.
For reasons not clear to anyone that I can find, Goldman was left insuring a small piece of the losses, which is where the $90 million loss to Goldman comes in.
I may not have all of the above exactly right. What is more interesting to me given my claim that creditor rescue by government is the true cause of the crisis, is the rest of the story. From a CBS Marketwatch report:
Within months, the notes were “nearly worthless,” and IKB [which had purchased $150 million worth of ABACUS] had lost almost all of its investment. Most of this $150 million was ultimately paid to Paulson & Co. in a series of transactions between Goldman and the hedge-fund firm, the SEC explained in its suit.
ABN Amro assumed the credit risk tied to the highest-quality parts of the Abacus CDO, through derivative contracts known as credit default swaps, or CDS, the SEC said.
In late 2007, ABN Amro was acquired by a group of banks led by Royal Bank of Scotland.
The CDS contracts were unwound in August 2007, through a payment of $841 million from Royal Bank of Scotland to Goldman. Most of this money was subsequently paid by Goldman to Paulson & Co., the SEC said.
Royal Bank of Scotland was one of the hardest-hit banks in during 2008′s credit crisis, so much so that the U.K. government was forced to eventually become majority owner of the bank late that year. To date, RBS has received 45.5 billion pounds from British taxpayers.
IKB was similarly affected. It avoided collapse after the German government-owned KfW Bankengruppe stepped in with capital that ultimately gave it 91% of the bank. The rescue of IKB cost KfW roughly 8.5 billion euros, according to reports.
So the ultimate guarantor of Paulson’s play were the taxpayers of Germany and the UK. Is this any way to allocate capital or run a political system?
The other interesting twist is that as I mention in my narrative of the crisis, Gambling with Other People’s Money (link coming very soon), the Chief Risk Officer of the Royal Bank of Scotland, was and is, Riccardo Rebonato. Rebonato wrote a superb primer on risk management, The Plight of the Fortune Tellers, just before the crisis hit. In that book he explains how risk management is not a science and how Value at Risk (VaR) doesn’t really capture what is going on. I think he knew his bank was playing with fire. And yet the bank kept lighting matches. It suggests that the suits refused to listen to the geek, either because they felt they knew better or because they counted on government rescue while collecting their bonuses along the way.
BTW, I interviewed Rebonato for EconTalk. He attributed the failures of his bank and others to myopia and errors in judgment and attributing nothing to moral hazard and incentive asymmetries between management and creditors. But as an employee of RBS, and speaking on the record, it’s not surprising that that was his story. I wonder if that’s what he really thinks.
Share
Podcast RSS Feed
Full EconTalk Text
{ 40 comments }
For reasons not clear to anyone that I can find, Goldman was left insuring a small piece of the losses, which is where the $90 million loss to Goldman comes in.
It is not entirely clear that Goldman wasn't simply taking a small piece of the other side of Paulson's trade.
If there weren't enough people who believed the housing market is only going up, Paulson would not have been able to short the size that he did.
Credit spreads were shrinking. Risk could be spread worldwide. Nobody really knew where the limits were.
People severely under appreciate just how hard risk management is when you're not managing with hindsight. Whenever you take a position, you're playing with fire to some extent and when you have a complex portfolio to hedge, you're playing with a hotter fire.
It is not entirely clear that Goldman wasn't simply taking a small piece of the other side of Paulson's trade.
That is what I've read as well — that Goldman went long on the deal as did, obviously, IKB. If so, that means that Goldman is essentially being accused of defrauding themselves.
People who make these kinds of bets — for that is precisely what this investment was — know full well that they can only make the bet because there are other people willing to make the opposite bet.
Don’t you love the way government created an asset bubble, then popped it — and then imposed mark-to-market accounting rules to make the collapse as rapid and deep as possible?
I'm a huge fan of mark to market accounting – which was not imposed after the bubble burst but was the accounting rule for well before that. It is much more accurate and gives people much less room to mark the book inaccurately.
The alternative is mark to make believe. I don't have to spell out the dangers of that. Of course, when assets are illiquid, there's mighty little difference between the two accounting methods. Still, there is a difference and I'm willing to bet that these banks are now carrying overmarked assets on their books. That's going to be a problem down the road.
Members of congress are far too stupid to understand the mark to make believe vs. mark to market accounting arguments. They couldn't possibly have orchestrated the a more rapid collapse if they wanted to.
I know a few people who worked at interest rate derivatives traders and CDO traders during this time who hedged or traded many of these products. It may be hard to believe now, but back then, people really really DID believe their own bullshit.
Keeping in mind, of course, that it isn't the team of traders who execute the trades who are making the risk management decisions. Those decisions are made higher up.
Perhaps you should take the time to read “The big short”.
The book, not the reviews.
And when you talk money, it should be correct. Paulson never collected 900 BILLION USD, contrary to your post.
And by the way, Prof. Roberts, all what you have been writing about the crisis until now, has been a textbook case of confirmation bias. You are obsessed with proving that “government” and “expected government bail-outs” are the only and single cause of all of it. No matter how strong the evidence in other directions, you simply keep on ignoring it.
Right, mark-to-market was in place before the bubble burst and was suspended a few months after Lehman bankruptcy.
The problem with mark-to-market is that is doesn't reflect the value of an asset when the market doesn't really exist.
A more appropriate method would be to discount expected cash flows, since DCF analysis gives the smallest errors, when consistently applied. Together with the DCF analysis, the entity should give detailed discussion of all the assumptions it made. Then, investors could decided whether or not to hold the entity's securities.
Any chance you'll get Michael Lewis on EconTalk?
Lewis essentially leaves Fannie & Freddie & the CRA & Greenspan & Chris Dodd & Barnie Frank & the problem of moral hazard completely out of his story.
I'd like to someone to ask why.
Does anyone know what percentage of all the the subprime mortgages created and sold Freddie & Fannie owned or guaranteed?
For those who need to conduct their own research and don't have time to read every book that comes out, here's Anthony Randazzo's summary of Mr. Lewis book: http://reason.org/news/show/coming-up-short
The more I hear about this the more I see the comparison between Goldman and other like firms as mainly market makers looking for volume to generate revenue much like a bookie lives off the vig! I would like to see someone say to those self serving hypocrites in congress that they were just complying with the laws that congress passed and it's not his or her problem that such horrific unintended consequences occurred due to congressional incompetence!
I'm a huge fan of mark to market accounting – which was not imposed after the bubble burst but was the accounting rule for well before that.
My understanding is that historic mark-to-market rules had an exception for assets that were to be held to maturity, but that FASB rules changed that in 2007, in the aftermath of the Enron scandal, forcing everyone to value securities “at market” even when there was no intention of selling the security involved (or no market for the security at all. Is that not true?
In any event, I am not a fan of government-imposed accounting rules. The government should enforce the laws against fraud, but I see no justification for government dictating things like how assets should be valued on a balance sheet. Assuming the valuation method is not misrepresented some how, I see no role for government here.
Thanks for finding the typo. I've corrected it.
I do have a bias. I don't hide it. When my essay comes out you can see what I've ignored or exaggerated. My essay, like almost every other “explanation” of the crisis is an ex post narrative. It is very difficult to know which if any of these narratives is correct. It is almost impossible to confirm any of these narratives to the exclusion of others. My goal was to learn and to teach. So you should read it with many grains of salt. I hope you learn something anyway…
Russ,
Classy reply.
I hope that by reading your essay I can gain some insight into how to show the furor after Lehman's bankruptcy was due to regime uncertainty and not due to Lehman's actual bankruptcy. Meaning: the feds bailed out Bear, and then changed the game by not bailing out Lehman. Obvious to me, but hard to show.
I have a few paragraphs on this. Essay should be available on line tomorrow. I'll blog on it and link to it.
Mark to market is certainly best but what do you do when banks are making loans where no real secondary market exists? Isn't that the case for a good portion of the assets we are talking about?
“It is very hard to know what actually happened without being familiar with the technical world of synthetic CDOs.”
That's the problem. It's clear these complex porducts exist simply to set up a casino style finance where the banks ultimatly come out on top.
Nassim Taleb and many other respected economist suggest these products should be banned or at least isolated from direct benifit of cheap Treasury dollars and government gaurantees.
They rich bankers realized they couldn't make money with these things isolated to their own accounts and traded amongst each other. They needed to have rules changesd to allow them to infest homeowners equity and hard working peoples 401K's… and ideally for the bankers the Social Security Fund. Fortunatly that nuttery was denied.
Banker products needs labeling at the very least. That way reasonably informed people can avoid trans fats and CDO's from affecting their health.
When consistently applied is the problem. DCF is only as good as your assumptions.
The bottom line is that illiquid securities are very hard to price regardless of method. Virtually nobody is going to go assumption by assumption when looking at a firm's holdings – assuming the details of the holdings are actually revealed.
I'm half way through it. INCREDIBLE! There is nothing respectable about making a living on Wall Street the way these people do. They are some of the most despicable opportunist. It's a shame to see so much rewards going to such an unproductive… no make that counter productive aspect of the economy.
I should say that as an investor, I favour mark to market. The holding period is an arbitrary rule. Intentions change. As an investor, all I care about is what the next guy is willing to pay for my assets.
I also don't see a role for government. If a company really wants investors it will meet investor demand for disclosure, etc. That's the way it works in private investment pools including hedge funds. Investors never scream when they're making money. They only start pointing fingers if they lose, demanding government somehow punish the guy running the firm. And that's how we get government intervention. It doesn't help. I think we have a case in point here.
Yes, that is the case. There is no ideal way to price illiquid assets. That's the real problem.
Because the guy doesn't really understand how the instruments work. He was a French Lit major or something and then he spent a very short time on Solomon's trading floor as a SALES guy. He's very week on the economics and finance.
That's my guess based on my reading of everything from “Liar's Poker” to his current thoughts on the meltdown. His narrative is like swiss cheese
From what you've posted in the past, you exaggerate the assumption that firms assumed they would be bailed out. This is the outcome, if even considered, that firms wanted to avoid most because of the unknowns.
When risk managers built their models, they assigned an extremely low probability to the housing market collapsing on a national level. The rating companies assigned a zero probability. It had never happened before and one of the weaknesses of risk models is that they rely on what happened in the past.
Now, after what's happened, risk models assign a much larger probability to bailouts and housing prices going up and down together. Your heavy emphasis on bailouts for the non-GSE's is probably much more correct for today than the height of the housing mania.
Of course, with hundreds of thousands of people making daily decisions around the world, it's going to be impossible to decipher conclusively what really happened.
The only thing that is not in dispute is that bailouts greatly increased moral hazard. A thing the financial reform bill increases yet again. It's basically the screw the taxpayer bill.
I don't think they were marking their credit default swaps to market accurately or objectively before the collapse. The few guys buying CDS on CDOs were bidding through the settlement price for months. That's how they figured out that the guys they thought were just brokering the trade actually had the other side, or at least some of the other side. If there had been central clearing and something approaching an exchange, at least recording price information, the market wouldn't have grown as large because the slap down would have been sooner i.e. before Magnetar got really involved. That's what I gather anyway.
“It's clear these complex porducts exist simply to set up a casino style finance where the banks ultimatly come out on top.”
Except IKB, a bank, did not come out on top
Maybe you just mean “investment banks”.
Somewhere in the testimony before Congress, I wish someone would be honest enough to say “Guess why we were creating all these crazy AAA-rated securities? Because Basel bank regulations called for AAA-rated securities, the SEC-recognized rating agencies rated these securities AAA, so the Banks came to us desperate for them.”
I should add that every investment is a casino bet. Even a savings account is a risk (if the whole country melts down enough, even FDIC can run out of money). And frankly, no one in government was saying in the early 2000's “Hey banks, watch out, the home mortgage market is about the implode” instead government was saying “More homeowners every year, this is good, we should make it even easier for people to get loans.”
My thoughts exactly, Lewis is a storyteller, not a finance major.
Not someone to take seriously for narrative.
Your intellectual honesty is what keeps me coming back to the podcast and the website despite the bias which I usually don't share. I'm looking forward to your narrative. My own feeling is that moral hazard is a bigger factor for those insulated from its social impact ie creditors and shareholders. I suspect that the behaviour of management in the firms that are bailed out directly is pretty much explained by mis-aligned compensation. A bailout may save the firm/creditors/shareholders, but its unlikely to leave management with position and reputation intact.
When risk managers built their models, they assigned an extremely low probability to the housing market collapsing on a national level. The rating companies assigned a zero probability. It had never happened before and one of the weaknesses of risk models is that they rely on what happened in the past.
And there was no incentive for them to believe or behave otherwise, especially when Fannie & Freddie were ultimately backstopping these things, full faith & guarantee, etc.
That sounds like a naïve gamble, but in the end, it was true.
And we now find ourselves in the reverse-insano world of Dear Leader spewing flatly false crap like “This bill absolutely will put an end to bailouts.” If anything, bailout likelihood increases with the “reform” proposals, with said likelihood directly determined by proximity to Congress and political influence.
The SEC case against Goldman Sachs is bewildering in its complexity to outsiders.
This case actually isn’t that complex, though it involves some bewildering and arcane terminology known only to trading desks and IB types.
The spectacle of Carl Levin twisting in a net of his own ignorant, imprecise verbiage today was entertaining, though the dread that set in as the questioning progressed (nothing good can come of this witch hunt, if it is indeed real, which is unlikely) outweighed any fleeting amusement.
None of these hopelessly dim-witted notaries public (the Republican Senator from Oklahoma didn’t know that PM commonly refers to Portfolio Manager) have any capacity to digest hedging dynamics, or even simple offsetting arithmetic (large long positions generally require large short positions to offset them).
Sorry, no time to go further here.
mismarked books…now THERE'S something new on Wall Street, eh
That problem will never entirely disappear
exchanges have their benefits. However, there's a big downside to standardizing fixed income derivatives and forcing them onto exchanges.
Fixed income is more complex to hedge than equities where all you're hedging is price moves. Structured products can be created to hedge exactly the risks the securities holder wants to hedge and a lot of fixed income products aren't that standard themselves. So, if you standardize the hedges, you're taking away a lot of optionality. Equity derivatives were a hell of a lot easier to standardize.
Personally, I don't think forcing these contracts onto exchanges will provide the protection people think it will. Wall Street is always well ahead of the SEC and the politicians. They'll simply churn out more poorly constructed and more opaque derivative products that nobody else will understand until it's too late. It's already happening and that's how we got those horrible ultra long and ultra short ETFs. The SEC, while tasked with regulating them, doesn't understand them. About 9 months ago, the SEC showed up at a colleague's firm (a market maker) and asked the managing members to explain the ultras to them. And the Ultras trade on exchanges. Regulated by the SEC.
How sad is it that intellectual honesty is in such short supply?
Well said.
I agree that a standardized contract tracking prices of RMBS CDO's wouldn't have been sufficient for Wall Street but if that product existed it would have helped in pricing the specific CDO's and synthetics. But what I have in mind is more like a title company where all trades would be reported so open interest, price movement and a rough estimate of risk could be calculated. Participants in that market could use the standardized contract as an imperfect hedge much as funds use the various standardized interest rate products to hedge risk along the yield curve.
Paulson had to be counting on a bailout. He had to know the risk exposure to the insurers would overwhelm them when the market tanked. I simply can not believe that the people shorting the housing market were stupid enough to overlook this. The rest of them, were probably that stupid.
Agreed. Some standardized contracts would remove opaqueness, making structured products easier to price and mark. As long as OTC structured products are still allowed (they exist in equities despite the standardized options contracts trading on exchanges), an exchange would be a benefit.
Banks don't want the exchanges, of course. If the market is more transparent, it'll be more efficient, reducing their edge.
Not that I agree with Michael Lewis, but he is not a French Lit Major. He was a history major in his undergrad. Lewis also has a Masters degree in Economics from London School of economics.
For investing to be like a casino, the number of aces in blackjack would have to increase as the players became more confident the casino odds were increasingly in their favor because the casino could payout twice then three then four then five times the amount bet. Then when confidence in winning changed to fear of losing, the deck would contain only nines and face cards.
What Paulson did was hire Goldman to find players to sit with them and play with the deck Paulson stacked. All the other players lost and Paulson won and while Goldman got a bit tip from Paulson, Goldman as house lost as well. But another section of Goldman was counting cards because they could see the cards the Goldman dealer held.
Here's the more important question: What was Paulson paid FOR?
What value did Paulson contribute to society?
We know the value of a casino, and it isn't creating wealth, producing consumer goods or capital, or improving the welfare of society. A casino is at best entertainment, entertainment that stimulates primitive parts of the brain that drive the reward from finding food.
Paulson seems to provide the same social good of Bernie Madoff, but legally, thanks to Goldman agreeing to round up players who would play with Goldman's dealer using the deck Paulson stacked. As the Goldman casino has no gaming rules prohibiting stacked decks, Goldman's violation is failing to tell the players the deck was stacked by Paulson and that Paulson was the player to their right. Goldman's defense is “these are sophisticated gamblers who know casinos are allowed use the stacked deck of another Goldman client, and that Goldman has examined the deck beforehand and may decide to join the play.
That's a valid point, but economics is not finance, or trading.
Economics helps in understanding market action, but it does not tell you about hedging a position, or the dynamics of that hedge. It also doesn't tell you about risk management (too much).
As Methinks observes, Lewis was in the sales area of the trading desk, which is very different from the prop trading and even agency desks of large financial firms. He only heard parts of conversations with counterparties, and likely had little or no knowledge of firm-level position risk and mitigation schemes. (see David Viniar's testimony yesterday)
Would the $90mm loss to Goldman be like a self insured (ie–deductible) for the policy?
That's a valid point, but economics is not finance, or trading.
Economics helps in understanding market action, but it does not tell you about hedging a position, or the dynamics of that hedge. It also doesn't tell you about risk management (too much).
As Methinks observes, Lewis was in the sales area of the trading desk, which is very different from the prop trading and even agency desks of large financial firms. He only heard parts of conversations with counterparties, and likely had little or no knowledge of firm-level position risk and mitigation schemes. (see David Viniar's testimony yesterday)
Would the $90mm loss to Goldman be like a self insured (ie–deductible) for the policy?
Comments on this entry are closed.