Readers of this blog may know that we broke story in our book ECONNED of the role that the hedge fund Magnetar played in increasing the severity of the subprime bubble through its program of hybrid CDOs (meaning composed of actual tranches of subprime bonds plus credit default swaps).
To recap: Magnetar embarked on an unheard-of program of CDO creation to enable it to take a no-lose bet. It sponsored CDOs by funding the equity tranche, the riskiest layer that received high interest payments, typically 18-20%. It used the cash flow from the equity to sponsor an even larger short position, using the instruments in the CDO. It balanced its exposures so it would show a thin profit as long as the CDO performed, and a much larger return when it failed. Sponsorship of the equity trance most importantly gave Magnetar influence over the parameters of the deals (such as the ability to reject certain exposures). As an industry source said:
At their peak, Magnetar was *THE* driver of RMBS [residential mortgage backed security] CDO issuance. The size of their “Constellation” program was the most amazing thing I’ve seen in my entire career. . . .
Magnetar’s idea was that CDOs were destined for long term failure—that the leverage on leverage based on cr*p assets made the BBB tranches long-term zeros. And, they realized that while most other hedge funds were content shorting the BBB tranches from subprime RMBS, shorting BBB tranches from
RMBS CDOs was a much more slam dunk of a trade. The commentary is right . . . without someone willing to fund the equity of a CDO there was no way to get one done. So, Magnetar made the logical leap . . . they’d fund the equity necessary to create the structures and then short a multiple of the bonds their equity money had allowed to be created.
The gravy was that the equity was typically good for one or two VERY HEFTY cashflow distributions—i.e., these structures went terrifically bad, but it usually took a little while from a timing perspective for that to happen. So, their carry cost of the shorts was offset by the one or two equity payments. After that, their upfront costs were covered and they would own the 100 point options for free.
Magnetar made A TON of money . . . I’d expect every bit as much as Paulson [a hedge fund manager who earned $15 billion shorting subprime mortgages in 2007].
The discussion of the Magnetar program in ECONNED includes:
• How a seemingly small amount of BBB tranches from subprime bonds used in Magnetar’s CDOs, had a devastating impact on the subprime market. Consistently conservative analyses indicate that in the peak years of 2006 and early 2007, Magnetar’s program drove the demand for roughly 35% of subprime bonds. Industry sources have estimated that the number to be as high as 50% to 75%.
• Magnetar’s trade was imitated by other hedge funds and dealers, further increasing the systemic impact.
• The deals were mostly hybrids, typically with 20% cash bonds and 80% credit default swaps; why this structure was advantageous for Magnetar.
• The links between the demand for CDOs and the “negative basis trade” that was arguably a widespread form of bonus fraud. (When a AAA instrument was insured by an AAA guarantor, internal reports typically treated it as if all the expected income in future years was discounted to the present. As we know now, in the overwhelming majority of cases, bonuses were paid on income that was never earned. This mechanism was THE reason many banks would up holding so much AAA CDO inventory – it was more lucrative for the traders to retain and “hedge” it than sell it.)
• The masquerade of almost entirely BBB subprime risk as AAA suppressed CDS spreads on BBB subprime bonds, the most actively traded tranche of the original bonds. Via arbitrage, this also influenced subprime bond pricing, which in turn lowered the yield on the loans themselves. In this manner, the mezzanine CDO market directly influenced spreads in the subprime housing market.
The Wall Street Journal tonight reports that the SEC is investigating Magnetar:
Investigators carrying out the SEC’s broad examination into these securities, known as collateralized-debt obligations, are holding face-to-face meetings with Wall Street executives about deals involving hedge fund Magnetar Capital, according to people familiar with the matter….
Critics have said that the actions of players such as Magnetar helped fuel the boom in mortgage-linked securities after cracks had begun to show in the subprime housing market. That subsequently worsened the financial crisis when Wall Street banks incurred heavy losses from CDOs they bought or were unable to sell.
Magnetar offset its risk by betting against the default of these securities using instruments called credit default swaps. Magnetar ended up making big profits when these CDOs collapsed, while the investors in the supposedly safer parts of the security suffered big losses.
Among other things, investigators want to know how the assets that were put into the CDOs were valued at the time, what the terms of the deal were, what triggers were put in place to determine whether investors would incur losses and at what point did firms involved in the deal bet against the assets in the CDO, people familiar with the matter say.
Yves here. Note that the subsequent expose on Magnetar by ProPublica disclosed that many of Magnetar’s deals were “triggerless”, a particularly bad feature for the AAA investors. Normally, CDOs had “triggers” built in to change how the cash flows in a CDO were distributed if the deal started to suffer credit losses. They were analogous to the way the body responds to hypothermia, by restricting blood flow to the extremities to preserve the brain and vital organs. In CDOs, the triggers would restrict payments to the equity and lower-rated tranches. Eliminating them worked to the advantage of Magnetar, since it would keep getting good payouts even as the deal was starting to crater.
Back to the Journal:
People familiar with Magnetar’s deals say the firm actually played a role in how the securities were put together, encouraging these independent managers to pick mortgage assets with certain characteristics to put in them.
Magnetar told its investors in April that it “did not control asset selection in CDOs in which it participated” but said it “often communicated” with the bank and asset managers when the deals were being put together.
Yves here. Magnetar’s defense is technically accurate but spurious. Yes, an asset manager picked the assets that went into the deal, but those choices had to be satisfactory to Magnetar. Everyone knew that if the equity sponsor didn’t like the assets chosen, it would withdraw, and that would kill the deal. To the Journal again:
Magnetar rejected the characterization of its deals as bets on a collapse of the mortgage market. “It was not a ‘bet’ that any CDO, any group of CDOs, or the housing or mortgage markets as a whole would fail either in the short term or long term,” the firm said in its letter.
Yves here. Again legalistic. No, Magnetar was not betting on the housing market, it was betting on certain highly correlated BBB subprime exposures. And it can claim it wasn’t betting “against” the CDO, since it had a long as well as a short position in them.
I hope the SEC is able to see through Magnetar’s dubious claims. It’s time that the role of the sponsors of toxic CDOs were brought to justice.
Bases on my belief that greed is by definition insatiable; I would suggest that if the investigators really want to solve and prosecute these cases. They should look for the secondary and tertiary layers. I am certain they exist.
There may be evidence showing that financial executives accually pumped/pimped the story of the mortgage meltdown.
There will somewhere be links to other benefit streams, both personal and corporate, where gains were pocketed from the collapse of the CDOs. There will probably be short positions on the companies most affected and possibly even some type of second layer of default swaps??
Questioning the managers will only result in years spent in circular testimony and obstruction, where everyone is covering mutual butts.
Egads. The whole construction of the investment structures in the CDO scheme and Magnitar’s part in it, is enough to make my head hurt, just trying to follow the constructs and layers of it. Now you are implying there may have been yet another level of more sophisticated gaming on top of this? I hope you are wrong.
Rex; I’m not suggesting more complication. Actually, since greed is basic to some human’s nature, I suggest that the smoking gun will probably be found in personal bank accounts. At some point certain individuals involved and having foreknowlege of the situation, will have decided that there was no reason to share the wealth into corporate coffers. They will have made side bets which would only benefit one or two people. If these transactions can be exposed the people who benefited will be prosecutable. Once we can address the crime on an individual basis the perpetrator is much less protected by the financial industry. This should crash the house of cards. Nobody will personally take the heat for the whole mess.
At my most generous, I see the Obama administration handling this with kid gloves to attempt a continuation of the status quo. If so they are naive. The economy is complex and requires large degrees of trust to allow it’s opperation. Once the common people realize that the “King has no new clothes”, they may ‘go along to get along’, but the little shadow of doubt will, over time reduce economic activity until it stalls. Far better to have dealt with the problem at the start.
Also thanks to Yves for her part on putting pressure on our government to do the right thing. Lets hope they follow through and some people go to jail.
Small point in this bigger discussion, but…
Yves mentions, “the subsequent expose on Magnetar by ProPublica”
I found it annoying, at the time, when other blogs started mentioning ProPublica as the source for their information about Magnetar, yet I had recently read Econned which told me about Magnetar before the ProPublica piece, especially considering the book’s writing and publication delays.
Props to you Yves. Keep up the good work.
Thanks! If you want to help, please send an e-mail to Wikipedia (ECONNED is mentioned nowhere on its page on Magnetar) and journalists when you see stories about Magnetar that mention ProPublica and not ECONNED. This has been a source of enormous frustration to me. It’s as if what we’ve done has been airbrushed out of existence.
What’s the reason for the cash/swap mix?
Why not 100 per cent swaps?
It appears to have been to get monolines to guarantee their deals. No AAA rated monoline would insure a pure synthetic (ACA was not AAA rated). Having a monoline insure the top AAA tranche(s) took the dealer out of that part ofhis risk position, so he didn’t need to find an “investor” for them (the bulk of the par value of the deal) if a monoline took the other side (note the monoline typically wouldn’t insure all the AAA tranches, there would normally be 4-5). Finding an equity investor was the hardest part of placing a CDO, but finding an substantial AAA investor was the next big hurdle. Dealers would usually line up the equity tranche and find an investor for a substantial part, if not all, of the AAA tranches, before they really geared up to put the deal together.
The monoline wanted to see at least 35% of the liability side consist of cash (people how had paid real money), not unfunded commitments. They wanted to see “real money investors” on the hook.
While you would often see some funding in a pure synthetic (the liability side “investors” putting up a portion of their commitment in cash, generally the investors in the lower-rated tranches), it would fall well short of the 35% de facto requirement of the monolines. Conversely, investors would pay hard dollars for CDOs that consisted entirely of cash bonds.
It took roughly 20% cash bonds in the asset side for investors to put up 35% via cash payments (purchases of notes, as opposed to agreeing to pay in the event of losses on CDS). The Magnetar structure was considered “innovative” and was copied by hedge funds and dealers.
We have a typical Magnetar deal structure and a discussion of how the CDO was put together in an appendix in ECONNED, “How to Short Subprime in Large Quantities”.
I bought ECONNED but haven’t made my way to the appendix yet.
“It took roughly 20% cash bonds in the asset side for investors to put up 35% on the asset side.”
I think you meant 35 % on the liability side. But what’s confusing me is – doesn’t a 35 % funded liability position create a 35 % cash asset collateral position anyway? i.e. even if the 35 % funding is not used to buy cash mortgage assets, doesn’t that funding go toward buying lower risk cash asset collateral on top of the swaps on the asset side?
If so, I don’t see why investors wouldn’t be indifferent to the nature of the cash assets – unless it was an amorphous queasy feeling about 100 per cent derivatives on the asset side per se. But I don’t see where a substantial difference in credit risk comes into play.
Yes, you are right I meant liability side, have corrected that comment, but (which is confusing) only the “funded” portion of the synthetic is kept as a reserve account.
Look at Appendix II, there is a balance sheet. You’ll see which tranches were issues as notes (cash payment) and the assets side (the reserves are the “funded” portion of the CDS).
Thank you and congrats, Yves! This is your scoop … and coup. May it lead to more, and even light a torch under our AWOL DOJ. Thankfully, it appears the SEC has found enlightment in your book, ECONNED and perhaps Naked Cap. It’s time to make it part of Prez Obama’s PDB reading. Now that would offer real hope.
You starkly exposed Magnetar and Wall Street cronies as malevolent malefators—corrupt gamblers intent on self-enrichment at the premeditated expense of others. This terrorism and defilement of the Earth must not, cannot continue.
“Where there is no vision, the people perish” (Solomon’s Proverbs 29:18)
Too many people and too much creation is perishing today (with great heartache), so we need renewed vision badly. But that is so conspicuously absent in current American leadership (political, financial, or journalistic). So it is blogs like NC, not captive to asylum management, with sharp analysis, well formed ideas, and rigorous debate, that offer our best hope to re-awaken our vision.
I happen to agree that this kind of information should be included with Obama’s PDB. And that failure to clean up this mess and clear out the rot will silently doom the long term economic vitality of the US.
My understanding of human nature, particularly in the US, is that people don’t function as well in situations where they are forced to pay 20+% to banksters for short term debt, while Big Fish are left untouched. It’s not human nature to work harder in order to benefit oligarchs.
And Magnetar is oligarchy built on computer code, physics (astronomy), and asocial gamesmanship. Because economics is a fundamentally social activity, failure to sharply rein in and penalize misconduct leads to the dissolution of that game.
The problem with Magnetar is that it appears to have missrepresented the CDO’s it sold. Penalties for that and beyond that I’m uncertain.
The trade itself, create a failure, insure it and then collect the insurance may well be a criminal act. But then, recognizing that the housing markets in several locales were in a bubble is a very interesting proposition. Is/was there a viable short against the housing market bubble other than the CDO and its companion the CDS?
What should have been the appropriate method of correction of the over priced housing market? In that correction, who should have sufferred losses?
I don’t think we’ll ever address those issues.
Although the SEC is investigating Magnetar, it is not clear any charges will be filed. The reason is Magnetar merely “sponsored” the CDOs. It didn’t sell the CDOs, it didn’t prepare or approve any of the sales materials. It merely bought the equity tranche, used the power it gained to influence the deal structure, and took a short position bigger than its equity stake.
The party that should be liable (aside from the investment banker if he made misleading statements in the sales process) is the asset manager, who is supposed to watch out for the interests of all the investors, not just the equity buyer. But the asset managers were the shallowest pockets in the room. The industry cliche was it was a guy or two with a Bloomberg terminal.
As this presentation notes (see p. 8) it’s hard to catch a thief when everything is legal.
Thanks for that. Very interesting and it would be amusing were it not that the occurence is such a sorry affair.
I’m reading your book and find it to be well written and a rich resource. A companion or following book is possible, are you considering a discussion as to the failure of the global fiat currencies?
My view is that the CDO/CDS trade on the housing markets is a socially valid trade to the extent that it served to assist in the correction of the house price bubble.
Having asserted that I’m very cognizant that a great many people with no direct interest in the housing market bubble were very seriously hurt.
In the specific, however it is very important to recognize that their injury is the result of fraud. Recording Repo105 as a pure sale is at best a misrepresentation. Creating Abacus as an off-shore entity is profoundly instructive. A CDO represents a claim against the first loss income streams. How the hell do you get a AAA rating against that which has a contingent value of zero, it was not matter of wheteher it was a matter of when. In fact how the hell do the rating agancies determine their rating for products for which they have no experience data?
For most who were injured by this financial failure this is a fraud not directly incurred but indirectly thru the mechanism of a fractional reserve banking system.
As I examine the events and preceded and followed, I find that the erosion in purchasing power of the dollar lies at the very core of our current and continuing economic distress.
Debunking efficient markets and the related nonsense is quite worthwhile, but what about the fallacy of fiat currencies and under capitalized financial institutions and 300 years of contract law. Fractional reserve banking is, to me, a very curious institution in the sense that what is the point of holding reserves that have no inherent value. Effectively you are reserving nothing, puzzles the hell out of me.
Or is this merely the stage filled with sound and fury and no significance.
I think that unless you can prove that the asset manager is in cahoots with Magnetar there is no legal case.
“Everyone knew that if the equity sponsor didn’t like the assets chosen, it would withdraw, and that would kill the deal.” That is true but the other side is that it makes no sense that asset managers want CDOs they are associated with to fail.
I also have read other books that say banks retained AAA CDOs for short term lending collateral in the repro market.
I as well you want justice, but where was the injustice you claim – be specific.
Too much mob mentality and rage will obscure meaningful solutions to preventing another financial crisis.
Moose; I’m not connected to this other than, like everyone else (an innocent bystander).
However, I know human nature. The proof exisits somewhere. Also, the money flowed into many pockets. It took a lot of money to repeal Glass Steagall and replace it with the much looser Gramm Leach Bliley.
Somewhere there is a template describing what an “acceptable” CDO looked like. Somewhere there is a directive to lend money to certain unqualified borrowers. A chain of evidence needs to be built linking these assets to the CDO structure.
You say that ‘Asset managers wouldn’t want CDOs in their portfolio to fail’. That may be true in most cases. However, if they were paid enough to compensate for the embassasment, some would certainly be accomadative.
Another point to remember is that this was structured like an inverted pyramid. The “bad mortgages” were not that large compared to the monetary structure they supported. The subprime provided the cashflow which made a very boring asset worth keeping. When the mortages collapsed the CDO was nolonger worth keeping and changed from an asset to a liability. Thus there is a glut of CDOs on the market and a lot of people holding assets which are worthless because there is no yeild.
These are not unsubstantiated allegations. My sources include people who worked on the correlation trading desks and were putting together Magnetar’s deals. Their intent was understood by the dealer employees who spoke to me (in fact, that’s why they came forward, they understood what had happened and were not happy about it). If employees at the dealers who contacted me (more than one) knew what was afoot, it seems implausible that this was not widely understood.
I don’t understand your point re repo. So? As I detail in ECONNED, AAA collateral, in particular AAA tranches of ABS CDOs, were used as repo, with what now look like astonishingly thin haircuts. The reason the AAA tranches were retained in the vast majority of cases was because the correlation/CDO desk (the name of the desk handling CDOs varied somewhat by firm) employees were gaming the bank’s bonus system. At most firms, if you hedged an AAA instrument with a guarantee from a third party (it usually but not always had to be an AAA counterparty, and it seldom had to be for the full value of the exposure, in the case of UBS, for instance, a mere 4% would do), the traders would then be able to discount the future income from the position, less the hedge costs, and book it as income on their P&L. It turns out this future income was never realized, and instead spectacular losses resulted, but the bonuses were nevertheless paid on fictive earnings.
The credit risk of the deal was not the trading desk’s problem, it was the bank’s treasury’s problem once the hedge was put on.
The asset managers were typically very small shops who were very handsomely paid to do very little, and were sometimes recent employees of banks who were significant feeders of deals to them. In a subset of those cases, they were even co-located (as in on the premises of a dealer). In those cases, their independence is suspect. Moreover, the asset managers were dependent on the investment banks for their warehouse lines. Does that sound independent to you? From a legal standpoint, perhaps, but from a practical standpoint?
So your position is only a suspicion that the asset manager was working with Magnetar to make the CDO fail and you may be right. Who is the asset manager and what do you suggest we do to make asset management truly independent besides “hanging those varmonts”.
As for mentioning repros,it is my understanding that the whole borrowing money scheme for the banks was based on AAA tranched notes used as collateral for lower borrowing rates. We now know that the crisis was due to this collateral losing its worthiness, so why would banks want to see non-synthetic CDOs to fail. In fact their failure is what lead to the banks failure until the government intervened. And in my opinion the intervention was more for propping up the insurance industry and the banks got lucky.More should be investigated as to the insurance companies’ role in the crisis and how to make sure that there is separation between monoline operation and the normal realm of insurance business.
You really need to read ECONNED, I treat these issues at length.
First, this wasn’t just Magnetar. The lack of independence of the asset managers was pervasive. Look at Harding, which was effectively captive to Merrill. Harding’s role as providing window dressing to Merrill even merits discussion in Michael Lewis’ The Big Short (a book I’m generally not keen about and have debunked here) and even he points out that asset managers generally were not “independent” in any practical sense. Or look at the Senate hearings on Goldman, where Goldman in e-mails that were released makes clear that ACA, the asset manager on the Abacus deal under investigation, was selected precisely because it would do what it was told. ACA was actually not one of the “two guys and a Bloomberg terminal” sort of asset manager, but it was deemed suitable because it would not buck what Goldman wanted to happen.
As for the repo, you are treating the banks as if there was a centralized intelligence running them. Not the case. The banks (again, as I discuss in ECONNED) have a serious command and control problem, they are basically hostage to the “producers” as in the business unit managers.
Due to internal management information systems which sought to allocate capital to various business units and reward activities that took less capital, hedging AAA paper looked hugely attractive to traders, they could and did book eight figure profits when they did so.
The risk managers, the only check on this process (as I again describe in ECONNED) were weak politically. And do you think senior management knew bupkis about the risks of CDOs? The only guys who understood were the guys on the CDO desks, and the saying on Wall Street was “IBG, YBG”, “I’ll be gone, you’ll be gone” as in when the trade blew up, it would be someone else’s problem.
This was institutionalized looting. I’ve set it forth long form in ECONNED. It’s disturbing that no journalist or official investigator has wanted to turn over these rocks. This was central to the crisis.
As to the monolines, they were stand alone companies, not part of bigger insurers. They had only two lines of business (despite their names). They had been muni bonds guarantors originally, and then diversified into insuring structured credit products. BTW, the ratings agencies pressured them to stay in that market to maintain “diversified” earnings. So as much as the monolines now look pretty stupid for having guaranteed CDOs, it isn’t clear they had much choice. Had they withdrawn from the market because they didn’t like the deals, the ratings agencies would have downgraded them for being less diversified. Their business model depended on having an AAA rating; a downgrade was a death sentence.
Yves, your frustration well justified. A month or two ago Frank Rich wrote a story about Goldman/Magnetar and referenced the ProPublica piece. I read Rich’s op-ed shortly after it was posted on the NYT website (on Saturday night) and immediately submitted a comment — from a bar — saying it was your scoop. Nothing. I think Frank Rich is a great writer, but it was really disappointing. Oh well…
Thanks for your efforts. The more people that remind journalists of the oversight, the better the odds of changing perceptions.
I’m thinking its going to be a long slog reminding the world this is your scoop, but you have the stamina, tenacity and charm for it. Fact remains its yours.
So next phase is going to have to be some more excellent coverage of the story.
This part caught my attention:
Among other things, investigators want to know how the assets that were put into the CDOs were valued at the time,
The SEC has GS lined up as the misrepresentation test case. I expect they should be looking for the next big fish test case.
If the SEC is now looking at how the assets were valued perhaps they’re looking to make a case for collusion.(or if they’re not maybe there’s a story worth writing from that angle).
The beauty in investigating Magnetar is that Magnetar swam with all the big fish, so any investigation of Magnetar is going to involve each of them. And as you’ve pointed out, they used pretty much the same template over and over again.
If the asset mangers were captive ( and given what we already know about i.e the Norma clowns that’s not really in doubt) and the asset pricing was determined by the captors, then it should be possible to demonstrate that the dubious valuation methods used by the captors allowed them to offload mispriced assets. The asset managers did what they were told to do, so if the Merrils were providing crap valuations, maybe that could be considered crimminal.
It could be simple enough to compare marks in the assets on the deals vs internal marks at the BDs. That info should become available through discovery, I think.
This story is no where near the end.
Thanks for your kind words and continued interest.
Agreed 100% there was collusion, and the role of the asset manager appears to be the focal point. They weren’t terribly independent to begin with, particularly when you factor in that they got their credit lines from investment banks.
Our thesis (but we haven’t found a smoking gun) is that the asset managers took down big chunks of the liability structure of RMBS. This guarantees less than arm’s length pricing.
See this post:
So does Obama admit that Rahm Emanuel is an enemy of the state, a financial terrorist, and that the USA is under siege from within?
Thank you for your excellent work. I will make sure people understand this is your work.
Please keep digging, I am learning more than I ever wanted to know about this subject, but it’s become necessary. Everybody needs to understand that this was not a case of a business deal gone bad due to extraneous circumstances, but fraud.
I have managed money for over forty years. I was shown a CMBS, a RMBS and a CDO. passed on all three. The applied cap rates for the CMBS were sub 6%, nice portfolios but the cap rates were too low. The RMBS contained largely sub-prime paper while the interest rate was 300 bps below what the implicit risk should have returned. The CDO was clearly a loser in that the income streams were reliant on first loss tranches of other bonds.
Who the hell buys a first loss tranche on high risk paper. Gotta be stupid. Now buy a CDO and a CDS against it and you an interesting trade. It’s not an investment, it’s a trade.
Were it not for the bailout of AIG, I suspect that GS and MS would have failed. That’s a story that has yet to be aired. Recall that GS restated their annual report going from fiscal to calendar and effectively lost December which might have otherwise been the month in which they would have been forced into the bankruptcy court.
It’s no wonder that GS wants an extension on it’s fraud case. But then will anything really happen there, after all, DOJ’s given Casano a pass. 44,000 CDS deals and he did no wrong?