Tom Adams: The Myth of “Insatiable” Investor Demand for CDOs

By Tom Adams, an attorney and former monoline executive

One of the ongoing myths of the financial crisis is that investor demand was what motivated the creation of so many bad securities. Banks, journalists and academics have all described the period prior to the crisis as a period of “insatiable investor demand” for things like subprime mortgages and CDOs. According to the conventional wisdom as retold by Michael Lewis and countless others, this mania-level investor demand is what caused banks and lenders to bundle up mortgage loans and related bonds as fast as they could.

As Yves Smith has noted on a number of occasions, the notion of “insatiable investor demand” is nonsense. Real cash investors were not the ones demanding more subprime loans and the bonds that they got packaged into. The demand was artificial and a form of Ponzi scheme. It came, overwhelmingly, from CDOs.

CDOs were transactions that bought up low rated tranches from mortgage backed securities. Issuers and bankers of mortgage backed securities came to rely on CDOs as the buyer of the low rated MBS bonds almost exclusively by 2006. As soon as the mortgage bonds were assembled, they knew with certainty that a CDO would buy them because the purchase had been lined up, frequently with companies or units effectively controlled by the lead investment bank, even before the loans were made. The issuance of a CDO was a great money making scheme for the banks, adding further to the demand for MBS bonds, because they would finance the MBS that were intended for the upcoming CDO in warehouse lines with the CDO manager. Creating CDOs for the purpose of buying mortgage bonds was an essential part of the banker toolkit in this era.

Even if the mortgage bonds were being sold in faux transactions to captive or friendly “buyers” in the form of CDOs, one would think that at least the market would determine the price of the CDOs when they were sold. In this way, presumably, the cold light of reality would trickle through to the underlying mortgage bonds and mortgage loans. However, if you thought this, you would be wrong.

Real cash investors were not demanding CDO bonds, no matter what the price. In fact, a significant amount of demand for CDO bonds (especially lower rated and hard to sell CDO bonds) came from other CDOs. A substantial portion of the collateral in so called high grade (amazing name, isn’t it?) CDOs was recently issued CDO bonds from other transactions. Many of these deals had up to 40% of mini CDO squareds. So-called mezzanine CDOs (the sort that consisted heavily of BBB rated tranches) generally contained 10% of other, low rated CDo tranches.

In addition, an even bigger portion of the so-called insatiable investor demand came from the banks themselves who had been tasked with selling the bonds. In particular, banks such as UBS, other European banks, and Citigroup, devised a strategy to purchase for themselves the senior most AAA rated portions of the CDOs. They claim that this was an attractive investment strategy. It was also useful as a way of continuing to feed the machine with “buying” for more subprime mortgage bonds.

The only CDO “investor” group that was somewhat independent of the CDO sausage-making factory was correlation traders, who constructed trades that approximated being long one CDO tranche and short another. These buyers were credit-indifferent and most suffered sizeable losses.

Without the demand from other CDOs and banks buying their own CDOs, the insatiable investor demand for subprime bonds would have vanished. The people at the tops of these institutions were in a position to know this but they either ignored it, because the fees from keeping the dance going were too great, or they were too foolish to realize it.

Thus, the myth of investor demand is exposed as nothing more than a fabrication to provide cover for the irresponsible behavior of the bankers.

More and more evidence is surfacing that proves out this story. For example, the Financial Crisis Inquiry Commission elicited some gems last week. At the end of this article from Bloomberg, Nestor Dominguez, the former co-head of Citigroup’s CDO business, is quoted as saying that he continues to believe CDOs provided a valuable service in meeting investor demand for the product. But just a few lines earlier in the article he admits that Citigroup itself was a huge buyer of CDOs, responsible for a good portion of their bailout needs, and thus a major contributor to non arms length “demand” for otherwise unnecessary bonds.

Yesterday, Bloomberg provided further compelling evidence of where the “demand” for CDOs, came from: Citigroup’s liquidity puts. According to the Bloomberg story, the Financial Crisis Inquiry Commission may be making the case that Citigroup used $14 billion worth of liquidity puts to aid in the purchase of CDOs by Citigroup sponsored commercial paper programs. Basically, Citigroup “sold” the CDOs to investors in the commercial paper but, because investors likely balked at the risk in the CDOs, Citigroup provided liquidity puts to the investors. These puts allowed investors to “put” the bonds back to Citigroup if they declined substantially in value. Which they did, causing Citigroup to buy back about $25 billion of CDOs at about a third of their face value.

Why did Citigroup provide these liquidity puts for the CDOs? To help get the bonds sold in a market that would have otherwise been saturated and unwilling to buy any more of the CDOs.

Prior to 2006, AIG and the bond insurers provided guarantees on a large portion of the AAA rated CDO bonds backed by MBS. This insurance helped investors get comfortable with the credit risk in CDOs (we all know how that turned out for AIG and the other insurers). Based on calculations I’ve made from publicly available transaction data, over 65% of the mortgage related CDOs issued between 2003 and 2005 had insurance provided on the senior bonds by AIG and the bond insurers. By 2007, thanks to AIG’s pull back from the CDO market and the size limitations of the bond insurers, this percentage fell to approximately 25%. Citigroup’s liquidity puts were basically a form of bond insurance and thus a way to give investors comfort that they didn’t have to worry about credit. By “selling” the CDOs to the commercial paper programs, subject to a liquidity put, Citigroup completely fabricated demand for CDOs or, to put it another way, became the insurer in AIG’s absence.

The fact that fewer and fewer CDO deals were being insured should have been a sign to Citigroup that people were no longer comfortable with the credit risk of the bonds. Instead, Citigroup and other banks continued to ramp up the volume by “insuring” it themselves. Citigroup, which was one of the largest issuers of asset backed commercial paper, was probably one of the worst offenders for this tactic (hence their massive bailout by the government) but other banks, such as UBS, Calyon, Societe Generale and others, likely used similar tactics.

The reason Citigroup and others were so eager to keep the CDO machine going (and growing) was because the deals were so lucrative for the bank management and its employees. The fees in the CDO business were among the highest in banking and there were all sorts of additional ways of generating fees and income using the CDO technology and regulatory arbitrage. So the bankers were highly motivated to find creative ways to find new “buyers” for CDOs so they could reap huge personal rewards – even mid-level CDO bank staffers were lavished with seven figure compensation packages and the top brass made multiples of that. Of course, it’s easy to make money if you assume that the underlying deals have no credit risk.

Citigroup officials continued to insist last week that they were meeting the insatiable demand of investors when they created more CDOs. Various industry apologists continue to perpetuate the myth of the exuberant buyer of CDOs, mortgage backed securities and sub-prime mortgage loans. In truth, by 2006, there were virtually no natural buyers for CDOs. This “demand” was a complete farce and if the demand for the CDOs was a farce, then the demand for risky MBS and the mortgage loans was a farce too.

The CDO bankers at Citigroup (and the other banks) who pushed their deals into the commercial paper programs, while calling it a “sale”, the bank managers who misreported these risks to the regulators and accountants, and the government officials who fueled the reckless policies with low rates and a total absence of regulatory scrutiny should all be doing the perp walk ( ). The entire mortgage related CDO business was a sham. However, it had a tremendously damaging impact on the economy by grossly distorting the mortgage market. In many ways, the CDO business came to resemble a Madoff like Ponzi scheme – new bonds were made to satisfy the “demand” of the CDS short sellers (i.e. Magnetar and company) and the CDO salesmen at the banks who had found the ultimate suckers to dump the bonds on – their own bank and, eventually, the taxpayers.

Addendum 4:00 AM, 4/14 (by Yves): Tom’s post does not include a point discussed more than occasionally among the Naked Capitalism team that has been working on CDOs, namely that they were also sold to unsophisticated institutions, generally foreign (the long-standing expression on Wall Street is “X was sold, not bought”). See former credit derivatives salesman Tetsuya Ishikawa’s book, How I Caused the Credit Crunch, for details.

Print Friendly, PDF & Email


  1. Gepay

    I was wondering how the big boys got caught. Usually the insiders know when to get out. AIG got out in 2006 and that was still too late. In a rational financial system the job of Wall Street and their like would be to distribute the surplus of the real economy to those who could best use it to make life better for people. the jobs of regular banks would be to lend money to businesses and households. It is clear that the financial system – Goldman sachs JP Morgan Citi etc is out of control. Congress has been bought. the regulators are asleep or castrated. the financial meltdown has made it all too clear that their main occupation is skimming off bigger and bigger slices of the surplus the real economy creates with complicated and arcane financial gimmicks. It is indeed one big legal Ponzi scheme. Just because it is legal doesn’t make it right.

  2. shargash

    I just finished reading “A Bubble That Broke the World”, by G. Garrett, written in 1932, and that is EXACTLY what the bankers said about the stocks and foreign government bonds they were peddling in the late 20s. “We were just grocers,” they said, fulfilling insatiable investor demand. Some of them used that exact phrase.

    1. Daniel de Paris

      “A Bubble That Broke the World” is a must-read.

      Once read, you’ll never tackle the financial press in the same way. I did in 2007 and I’m glad I did AT THAT TIME. The only issue here is that we should expect a Chine G.G. The chances are limited. Andy Xie is doing his best though.

      Thanks Garet… and Xie :)

  3. Hubert

    Great Analysis. But that would still leave IKB and some German Landesbanks as idiot buyers. And HRE ? Landesbanks have counted some 250 billion (€ or $, I forgot) in toxic US asseets. That is some demand…. And some Japanese bought too.
    Do you have any numbers there ?

  4. VenusVictrix

    You state that “the myth of investor demand is exposed as nothing more than a fabrication to provide cover for the irresponsible behavior of the bankers”.

    I think you are being too delicate. What they’re really covering up is the outright fraud involved in the entire process, from origination all the way through to the creation of the CDO. The investment banks could not sell this stuff without subjecting themselves to lawsuits and criminal complaints for fraud. Much better and safer for them to sweep their dirty little secrets into a black box VIE, or SPE, and make their profits on the Credit Default Swaps which reveal nothing about the so-called reference assets, except that they were crap. Meantime, the commercial paper aspect was what gave them leverage over our entire financial system.

    And didn’t Goldman do something similar with regards to using CP facilities with their Cayman Islands CDOs?

    Anyway, great piece. I’ve long believed the supposed “demand” for these crappy assets was a myth, and you’ve credibly supported your argument.

  5. fresno dan

    Great post over at the Baseline Scenario

    Thanks to Naked Capitalism and other blogs I am finally beginning to get some glimmerings of what happened and what the issues are.

    Financial blogs truly perform a great public service in educating the public in depth about these activities.

    Finally “The reason Citigroup and others were so eager to keep the CDO machine going (and growing) was because the deals were so lucrative for the bank management and its employees.” There’s your problem – I am so old and decrepit I remember the great complaint about banks was an unwillingness to lend. Well, financial “inovation” solved that problem – mega, mega bonuses to bankers (which they keep) irrespective of the outcome of the loans.

    1. Yves Smith Post author

      I’m perplexed at your comment, this is precisely what the post is addressing. The “insatiable demand” notion is an urban legend.

      Starting in mid-2005, CDOs (ABS CDOS, we are not talking about CLOs, which technically are CDOs) were increasingly constructed with a significant CDS component or entirely synthetic. Dealers claimed those CDOs were faster to launch and were the economic equivalent of cash CDOs.

      However, many traditional AAA buyers either were not permitted to fund or were cautious about taking down a synthetic exposure (in a hybrid deal, the higher tranches were typically pure synthetics, the lower cash bonds). The dealers themselves were the main AAA tranche funders. Sometimes these exposures were partially or fully hedged via monoline guarantees; they were also placed in SIVs, which had liquidity backstops from their sponsors and therefore were not as off balance sheet as believed.

      The so-called inner tranches (AA and below rated tranches) were often rolled into first gen CDOs as Tom described above, as well as higher order CDOs (sometimes junior AAA tranches were also rolled into other CDOs). They were also used by correlation traders, who were arbiing various tranches versus each other (note that there was almost no market for shorting a CDO tranche directly, but it could be approximated using CDS and other instruments).

      The inner tranches were also sold aggressively, with the prototypical target a not very sophisticated foreign firm.

      1. Boo-urns

        I don’t scour the web as much as you do, but when I read “insatiable demand” or the like, I tend to assume they’re referring to the Aaa rated tranches. And I haven’t seen anything to dispel the notion that there was insatiable demand for Aaa-rated securities from investors, which drove the increasingly synthetic creation of more Aaa-rated bonds.

        The point is, I respectfully submit that you’re comparing apples and oranges here. You are correct that there was very little demand for the lower-rated tranches, but that’s not what drove the creation of CDOs. It was the “insatiable demand” for what Gary Gorton calls “informationally insensitive” Aaa-rated debt that drove the entire process.

        1. Yves Smith Post author


          I am sorry if I sound a bit tart, but did you read the comment I made earlier in this thread? By 2006, ABS CDO creation has shifted in large measure to CDOs that had a significant synthetic component (note that market size/composition stats make it hard to isolate hybrids from cash and pure synthetic CDOs, and from what we can tell, there are classification issues in most of the overviews).

          So in a hybrid deal, the top tranche would typically be synthetic, and hence could not be repoed.

          Moreover, to the extent there were “cash” buyers for CDOs because the top tranches were cash bonds, not synthetics, your argument re repo actually supports our thesis. As Gorton explains in gory detail, the main driver of the need for repo was as collateral for derivative positions, particularly credit defaults swaps. The dealers themselves had the biggest CDS positions. Hence the dealers themselves would have far and away the biggest demand for repo as collateral.

          But the big driver of the retention of the AAA tranches was flattering treatment of capital requirements (minimal) which meant retaining the top rates tranches was more attractive for most players than selling them (note this was due to how internal management information systems incentivized business units on a P&L basis to “free up capital” or otherwise use it sparingly).

          The reason we belabor this point is that the industry has sought to excuse the conduct that led to bailouts by contending that they were merely acting as middlemen, intermediating between various customer groups. But the dealers themselves, via their own retention of super senior tranches, use of supposedly off balance sheet vehicles that turned out to be less than arm’s length, their own correlation trades, and use of other CDOs as the “buyer” of CDO tranches, were far and away the biggest customers for CDOs.

          1. Mindrayge

            Indeed. When you examine the constructs of the various CDOs from the prospectus that are available there is no doubt that the purpose of such vehicles was not a response to insatiable demand. The conflicts of interest sections – though you do have to read nearly the entire several hundred pages to truly understand those – are staggering. That is even for managed CDOs that, at least initially, had less than 5 percent of assets in the form of tranches from other CDOs in them.

            And it doesn’t stop there. What precluded TCW (part of SocGen) when acting as a collateral manager from buying tranches of other CDOs where SocGen subsidiaries or affiliates held those notes? Nothing.

            One another note, the NY Times article on Lehman’s Fenway is just the tip of the iceberg. That was a case where Lehman was trying to use the same assets as collateral against more than one borrowing. They weren’t the only ones doing that – all of the players were. Who knows how many such obfuscated transactions (via firms like Hudson Castle) were running through the commercial paper markets and repo markets where the lender unknowingly really had no collateral since the asset was already encumbered.

            Further, people tend to miss the verbiage in various prospectus where you find that rather than just say entity A is selling asset X to the CDO/MBS you see something on the order of “to sell or otherwise convey”. Some of these resulted in assets that were effectively loaned rather than sold. If that ultimately ends up in a CDO where the collateral manager could then use that borrowed asset as collateral against a commercial paper or repo borrowing of their own. And, of course, there were the curious CDO provisions where the collateral manager could lend assets provided that the borrower pledged 104 percent of the asset value in cash or treasuries. Why would anyone borrow an illiquid asset with their most liquid assets pledged as collateral? I haven’t been able to wrap my head around that yet. But perhaps you or Tom know why those provisions were in place and how and why they were used.

            In seeing things like the NY Times article and Tom’s post I am more and more convinced that the AIG bailout and TARP along with the BofA and Citi guarantees was nothing more than stuffing real collateral to back commercial paper and repo borrowings through maturity. In retrospect the events of September 2008 look no different than any other check kiting scheme unraveling. And the Lehman fiasco seems more and more like an attempt to isolate the money market mess to a single event – Lehman’s collapse – rather than allow it to fall on all of the players with blood on their hands.

        2. csissoko

          When I hear the phrase “insatiable demand”, I think: well then what is preventing supply and demand from reaching equilibrium?

          I don’t think anybody (especially the financial industry) has a good understanding of how derivatives (i.e. synthetics) interact with equilibrium in financial markets. But phrases like “insatiable demand” and “insufficient supply of MBS” are strong indicators that something was going very, very wrong in these markets.

          More here:

      2. MarcoPolo

        I’m lost.  Maybe the marketing terminology “synthetic CDO” dosen’t mean what it should until you understand “… CDOs … were increasingly constructed with a significant CDS component…” !!!  And I’m slow and I’ve got to stop to think about which side of that swap would be included. Not the insured (buy) side. No point in that.  So if I own one of those “synthetic CDO’s” what do I get?  A high risk investment that pays a nice (?) return and I’ve sold insurance against it blowing up?  Huh?  I suppose the cash that came off something like that should look attractive. But if it blows up I lose my investment and have to pay somebody else for his?  Am I missing something?  It would seem to me if you could sell two of those things that would qualify as “insatiable demand”.  Though I really doubt if I understand.     

        1. Yves Smith Post author

          This isn’t a marketing terminology, the deals were different, although the marketing types at dealers argued they were the same.

          It might help to remember that a CDO is a mini bank, with an asset side of the balance sheet (the assets that go into it) and a liability side (the various tranches).

          With a cash CDO, the assets are all real instruments, mainly bonds (tranches of residential mortgage bonds) although you might see a smattering of whole loans too. So the CDO whacks up the cash flows (interest and principal payments) into tranches (super senior through equity). Each of these tranches is ALSO a bond, which can be confusing.

          So investors in cash CDOs pay cash (hard dollars) to buy a bond (an instrument that pays interest and hopefully returns the principal).

          In a synthetic CDO, the assets are ONLY credit default swaps. So the tranches are “synthetic”. That means the “investors” do not invest in the traditional sense, but play a role akin to an insurer. They receive their participation in the “premiums” paid on the credit defaults swaps in the CDO. They also stand ready to make payments to the CDO if the credit defaults swaps in the CDO require payments (this is a simplification, but gets the basic idea across).

          Does that clear things up?

    2. craazyman

      I had the same question. But I am the dumbest guy in the room, I admit. These narratives often read as jargon riddled shop talk between high finance cognescenti. For those of us who don’t get paid to follow this stuff full time, it can get convoluted with all the short hand references and acronyms. It’s easy for full-time pros to lose sight of how slippery it all is.

      Somebody, somewhere downstream from the banksters supplied the funds that supported these bonds — whether synthetic or cash — and therefore “bought” the underlying toxic exposures, it seems to me, with the hope of profiting. In this respect, the banksters were middle men satisfying demand, as Sechel observes.

      It seems to me that the toxic exposures were like rat meat ground up — in all sorts of shell game ways — with allegedly prime Angus beef to make hamburgers, which were eaten by the downstream investors through a variety of investment vehicles. And were even held and eaten by the banksters themselves in some cases.

      The banksters argue that the downstream investors drove the demand for hamburgers. But the argument in this and in Yves posts on the issue is that the demand for hamburgers was created by the chefs who knowingly used rat meat and then sold the burgers in the dining room as Prime Angus burgers, more or less.

      One perspective on this is “How could the diners NOT KNOW they were eating rat meat??? I mean, look at all the warning signs evident to anyone with eyes as the bubble grew! If you deny you didn’t know, then you LIE!”

      My sympathies are with the perspective that the chefs knew it was rat meat and were morally if not criminally negligent. If this wasn’t against the law, it should have been.

      The bankster perspective is “How could the chefs have known this was rat meat? After all, they were just cooking with ground up meat that looked like beef at the time. Who can predict the future? We can’t judge based on 20/20 hindsight! We were all fooled by the bubble, but we were honestly fooled.”

      This seems lame to me, as a justification.

      This is about as far as I’ve gotten with this stuff. But I do know what all the acronyms mean and I know what a synthetic exposure is and I know how a CDS works. I just don’t know how it’s all put together in terms of the mechanics of creating the investment vehicles in a way the penetrate the veil of the shell game in a step by step 9th grade explanation. And anything can be explained well enough for a 9th grader to understand. I’m convinced of that.

      1. Mindrayge

        In the CDO world most of the time, if you were to look at a prospectus, what you would find is that the actual assets that would be in the CDO were not actually purchased yet and only a set of “rules” of what kinds of assets would be purchased were given. Sometimes the prospectus would already have a complete asset load and would mention a website or the ability to obtain a CD ROM where the assets are listed and the prospectus from the MBS, etc. that asset came from included. Other times they went out the door only partially loaded. The managed CDOs had collateral manager that could buy and sell assets, thus changing the actual assets owned by the CDO from the initial set the investor made their investment decision on.

        Investors were literally buying blind. What kinds of investors do such things? Well, the first thing is that the only way they could get US investors was via rule 144a since CDOs were unregistered securities. Rule 144a limits the purchase of such investments to “qualified investors”. Such investors have to have $100 million in liquidity. They also can’t be pension fund managers or 401K managers. The second thing about the kinds of investors that would buy a CDO tranches notes blind is that they all likely had some relationship with the issuing entity or show up in the conflict of interest section of the prospectus as one of the participants in the construction or management of the CDO or is one of the affiliates or part of some partnership amongst the various players. They likely knew or had influence in just what assets would get pulled into the CDO or would be sold from the CDO. Further, the investor could buy into the CDO and use that asset in their own fixed income fund that, in turn, they could sell to pension, 401k, IRA, gram and gramps, etc.

        I am not convinced of the supposition that the original CDO investors were largely unsophisticated foreign investors. More than likely they were offshore subsidiaries or affiliates of US firms that bought notes from these CDOs and turned them into assets of yet other vehicles that were then pulled in as registered securities and dumped on US investors from pension funds to municipalities to individuals.

        One unrelated point. These CDOs existed in tax havens, often the Cayman Islands or the Channel Islands and thus paid no US taxes. However they were bailed out (via AIG) with US taxpayer dollars. Should have never happened.

        1. VenusVictrix

          “I am not convinced of the supposition that the original CDO investors were largely unsophisticated foreign investors. More than likely they were offshore subsidiaries or affiliates of US firms that bought notes from these CDOs and turned them into assets of yet other vehicles that were then pulled in as registered securities and dumped on US investors from pension funds to municipalities to individuals”

          I think you are absolutely correct.

          1. Tom

            I’ve been following the cases around the globe on CDO litigation and, other than a few isolated cases I just haven’t seen that many involving unsophisticated pension funds or municipalities. Considering the collapse in values of the senior bonds from the CDOs (let alone the junior tranches), it seems likely there would be more noise by now had such investments been made.

            While the data for the cash issuance of ABS CDOs is remarkably inconsistent, our estimates indicate that a substantial portion of the bonds, especially the senior bonds, ended up remaining at the banks, in their “hedged” trading books or in their off balance sheet vehicles. Rather than disperse risk, the CDOs acted as a tool to concentrate otherwise illiquid risk in a remarkably small number of institutions.
            This is what why the banks’ defense, as employed by Citigroup last week, that they had no idea that they were so exposed or that this meltdown would hit CDOs is so offensive. In CDO transactions, banks had a tremendous information advantage over investors – they knew the loans, the prices, the parties, the motives, and the documents, and outside investors got the bare minimum of information. Despite this, the banks were the ones who ended up holding the bag on so many of these bonds. Real investors couldn’t get comfortable with even the AAA risks in the CDOs, particularly at the offered yields, but the banks ignored this and scooped the bonds up themselves.
            The absence of real third party investors for the CDO bonds, and on down the line to the mortgage bonds and mortgage loans, should have been reflected in the prices for the bonds, but it wasn’t.
            The rapid growth in the “vertically integrated businesses” at the banks allowed them to originate mortgage loans, issue MBS, buy MBS and package them into CDOs (and repackage some of those CDOs into other CDOs), and then buy the senior bonds into some unit – it was an amazing fee generating machine at every step.

    3. fresno dan

      Although it is better to be silent and be thought an idiot, than to speak and remove all doubt, I would make a suggestion that maybe would help people (me and a few other idiots) and that is some actual examples of the prices of these “transactions,” “premiums” and “services”.

      I can’t help but think that much of this was uneconomic (although if people buy something it can be argued that it is “economic” – transactions between entities) – like when Sears sold appliance “maintenace” that cost 10$ on a 20$ toaster.

      What I just find hard to fathom is the economics of bond insurance. If you believe market theory (or propaganda) wouldn’t any difference in bond risk already be priced into the interest rate? Or are bond buyers like consumers who buy toaster insurance at Sears? By buying bond “insurance” you have more fees, more counterparty risk, and more confusion – you must be losing yield with more middle men. Wouldn’t you get the same amount of safety simply by diversifying and getting some muni’s and treasuries? Its kinda like someone selling a stock selection method – if it really works, why is the author making chump change selling books, when he could make billions selecting stocks?

      Does it simply boil down to people thinking they could get higher yield with lower risk because the “insurance” was a good deal (something for nothing)?
      Is there some regulatory arbitrage I am missing? I guess if we had millions of people feeling they had to buy houses, I guess there must of been lots of people who felt they had to buy “hedged” or “insured” or “guaranteed” bonds.

      1. craazyman

        I love that opener, Fres.:)

        This one from Mark Twain is a gem (and probably already well known around here too). . .

        “Reader, suppose you were an idiot. And suppose you were a member of Congress. But I repeat myself.” -Mark Twain

      2. Mindrayge

        You aren’t missing anything. The more hands something passes through the less there is to go around. There is no possible way for any of this to work unless all of the participants were buying the asset with borrowed money on extremely short terms at very low interest and constantly rolling that borrowing over as inflation alone would destroy the return over 30 years. And beyond that it is obvious the players had put themselves into a situation where there wouldn’t be enough money to go around and rather than let the shit hit the fan they ramped the Ponzi scheme up even further and faster right into the collapse.

        But when one considers that a 125,000 mortgage at 7 percent will produce somewhere around $170,000 in interest, front-loaded, you have to wonder why banks would be willing to throw that kind of return to others when they are quite capable of arbitraging interest rates and inflation on their own.

        As an investor, it would seem there would be less risk and a better return had you simply bought the 10-year treasury instead of the mortgages – especially when it was clear the housing market was over priced.

      3. VenusVictrix

        “What I just find hard to fathom is the economics of bond insurance”

        Dan, I think that’s a great comment, and I agree.

        You are right that the risk “should” have been built into the rate, but lenders had to offer the low, “teaser” rates in order to get consumers to bite. On top of that, lenders were refinancing the loans so quickly that it’s really hard for me to believe that ANYONE made any significant return from these bonds even before the market collapsed.

        In fact, the economics and risk-reward ratio make so little sense to me that I am totally convinced that the CDS story is a hoax. It’s merely a cover-up to cloak what has been called a back-door bailout to the investment banks via AIG.

        Consider that AIG has a private mortgage insurance subsidiary: United Guaranty. This company has been around since 1963. Private mortgage insurance is the traditional means of insuring the lender against a borrower default for loans exceeding 80% LTV, written to standard, conventional underwriting guidelines.

        The loans that AIG supposedly guaranteed with CDSs were loans that were not eligible for private mortgage insurance, because of the low credit standards and/or lack of full documentation – and AIG knew that. In addition, private mortgage insurance costs significantly more than what AIG was supposedly selling the CDSs for – and yet the CDSs guaranteed the entire loan balance, whereas private mortgage insurance only covered at most 20% – 30% of the loan amount.

        So in other words, we are lead to believe that AIG would take on more risk, for less compensation, and guaranty a higher payout than anything they would take on through the fully-capitalized mortgage insurance unit. And they did it in such an astronomical amount that it put all of the corporation’s equity at-risk.

        You are correct, that this makes no economic sense.

  6. Kid Dynamite

    tom, in your final paragraph detailing who should be doing a perp walk, you left out one essential group: the ratings agencies who failed (fraudulently?) at their jobs more miserably than anyone else in the entire crisis.

    1. VenusVictrix

      Kid, hopefully the CT AG will succeed in having the ratings agencies charged and convicted.

      See my comment below, in response to Tim. I reference the recently-filed complaints against S&P and Moody’s.

  7. Yearning to Learn

    Perhaps I’m splitting hairs, but I’m not sure that I can totally agree with the wording of this article.

    To be sure, there was little so-called “natural” end-user investor demand for these products. however, there was a very high demand for the products in the banking system itself.

    The financial system has gotten itself into a self-trading loop, where it simply trades assets and pseudo-assets back and forth among itself, at ever higher and higher prices (at least in the past). it worked until it didn’t.

    Thus, there WAS insatiable demand, just not by the public at large.

    The fact that the end-users often demanded “insurance” (puts or a CDS or whatnot) doesn’t change the fact that there was still demand for the product. For instance: many people will only buy a car if they can get a 36,000 mile warranty on that new car… which is effectively insurance, right? The insurance bolsters demand, true, but there is still demand.

    Thus, I would agree that there was little demand from routine end-user investors. But there was large demand from the financial system itself (hedge funds, quant funds, banks, non-bank entities, etc). Of course during these hearings the banking executives try to obscure this fact.

    I will note as well that the same thing is happening right now in the equities and commodities markets IMO. the banks and hedgies are simply trading back and forth among themselves but the retail investor has vanished. thus we get higher and higher equity/commodity valuations on smaller and smaller volume.

    sadly (or not), it will work until it doesn’t

    I hope my point made sense as I’m not as good a writer or as knowledgeable as many here

    1. Mindrayge

      Yep. Eventually the incestuous trading will come to a stop if they can’t find suckers to come in and take those positions.

    1. VenusVictrix

      I would caution about some of the author’s conclusions, which aren’t supported by the data she gathered. For example, she asserts that the ratings agencies made “mistakes” in over-rating billions of dollars of CDO traunches. She explains this “mistake” at the end of her paper by suggesting it resulted from use of automation and inaccurate inputs. However it seems intuitive to me that prior to assuming a “mistake” was made, one should consider the extent to which that mistake might have been influenced by financial incentives, and one of her own sources referenced in the report (Roger Lowenstein) repeatedly refers to the perverse incentive created by the issuer-pays model for obtaining the ratings. Barnett-Hart errs by ignoring this potential conflict of interest.

      Though not available at the time she compiled her report, the recent complaints filed against Moody’s and S&P by Connecticut AG Richard Blumenthal provide very strong evidence in support of a claim that the ratings agencies knowingly and willfully inflated the ratings assigned to these CDO traunches.

      You can retrieve copies of them here:

      But I’ll shortcut to what I find to be the most damning evidence against these agencies.

      According to the complaint filed against S&P (McGraw Hill), the models used to establish the ratings were based on very limited data developed prior to 2001, and then they were never updated after 2001. In fact, S&P actually went through the process of DEVELOPING new models with updated information, but management refused to implement the changes. In spite of the huge explosion in RMBS activity after 2001, and the fact that new lending standards consistently degraded the credit quality of loan originations, they continued to use an outdated model with irrelevant data. (see pp 33 – 35)
      You can hardly call that a “mistake”.

      Those ratings were meaningless – and they knew it.

      In the Moody’s complaint it is revealed that although they did update their ratings models as the market changed in the early 2000s, as they started losing market share to S&P, they decided to “amend” their models to more closely mirror S&P’s ratings process. It is clear they did this to preserve market share!

      Obviously they weren’t as concerned about establishing a valid rating for the instruments as they were about making money off the process. (see pp 38 – 40)

      In defense of Barnett-Hart, Yves calls the paper a serious effort in spite of some shortcomings (like not asking serious questions about the relevance of the data), and I think if you are careful about questioning her conclusions, the data and statistical analysis are probably a good resource if you really want to learn more about the CDO market.

      However I think the CT AG’s complaints are much more readable, and they are pretty interesting if you want the real dirt.

  8. AK

    The article is definitely not complete.

    I saw a small article some time ago (it would be difficult to find it now for me I guess) that some people did CDO “arbitrage” — go long with one tranche and short with other one from the same CDO.

    So, when CDO was downgraded for example their positions were increasing in value.

    I don’t know if Yves explored those trading strategies before or not.

    1. Yves Smith Post author

      Please re-read the post, since the point you claim was overlooked is in fact addressed. The post contains this statement:

      The only CDO “investor” group that was somewhat independent of the CDO sausage-making factory was correlation traders, who constructed trades that approximated being long one CDO tranche and short another. These buyers were credit-indifferent and most suffered sizeable losses.

      This is probably more flattering to the dealers than it ought to be. The dealers themselves were major correlation traders (it is hard to know their market share vs. that of independent hedge funds pursuing similar strategies).

      As far as we can tell, direct shorts on CDO tranches were extremely rare. The short position on a CDO tranche was approximated using credit default swaps on mortgage bonds (often the same CDS used to create synthetic or hybrid CDOs) and other instruments.

      1. AK

        I guess the idea, which I referred to, was rather related to so-called “default correlation.”

        I will try to find that article later if I have time.

  9. Tim

    For the layman, the point is this ?

    The ultimate demand for CDOs were the “banks” who kept them.
    Therefore, they can’t act like they didn’t know and are “guilty”.

    1. EmilianoZ

      I’m a layperson and that is also the impression that I get.

      Towards the end of “Econned”, Yves explains why bank executives liked to keep CDOs on their books. I don’t have “Econned” with me at the moment so it’s from memory. My understanding is that those executives were gaming their bonus system. In the gains section that determine the bonuses, those guys could write the CDOs as if they had already paid off. In theory of course it would take years for a CDO to pay off in full. In practice they never did because they were crappy. But those bank execs have been paid as if those CDOs had been brilliant successes and as far as I know there have been no clawbacks.

      So, that is a very strange way to compute bonuses but that’s how it’s done according to “Econned”.

    1. tedb

      Dennis, It would be helpful if your comment was more complete and then if someone more knowledgeable than I would respond. I am not sure what you are referring to when you say that the government “coerced” banks into making unwise loans. The idea that the Community Reinvestment Act (CRA), for example, was a major cause of the crisis has been debunked because, IIRC, loans made by the banks subject to its provisions were subject to real lending standards and as a result have performed better. (Janet Yellen of the San Francisco Fed wrote a paper on this.) Many (not all, of course)of the bogus loans that led to the crisis were made by lending institutions not subject to normal government regulation, i.e., lending institutions that are not banks at all.
      Also, the word “bank” is confusing, at least to me. The banks referred to in this blog article are the big investment banks, and the government did not force them to create CDOs. These big investment banks, this post alleges, are the ones who created an illegitimate demand for bad loans that would never satisfy reasonable lending criteria.

  10. Dennis

    It seems that there is one more step back on this, The government, since the mid-90’s, was coercing banks to lend the original money to people that couldn’t afford it. That the banks, at least initially, looked for some way to lay off that risk seems reasonable (hence, tranches of mixed mortgages with the riskiest mixed in with good ones). The fact that a game like that, especially with criminally lax regulation and ever more exotic playoffs,eventually took off into the one described in the article shouldn’t be a surprise.

  11. ericv

    I wonder whether this would be RICO for criminal conspiracy. After all, the sales process didn’t bother to tell the buyers that Magnetar had negative postions in their own CDOs. Hardly full or material disclosure.

  12. MarkS

    As James Chanos remarked in an interview with Charlie Rose on Monday (… Lehman’s origination, purchase, and off-balance-sheet accounting of CDOs amounted to fraud. The immediate victims were Lehman shareholders… Fees from CDO origination ended up in employee and management pockets immediately, while the risks in the long term positions were absorbed by the stockholders and enabled by fictitious valuations. For others on Wall Street playing a similar game, the losses will be absorbed by everyone else in the economy via the increase in the FED’s balance sheet.

  13. najdorf

    I’m skeptical of this post due its almost complete lack of numbers. I know that banks retained a lot of CDO exposure, but part of the idea of CDOs was that when you securitized assets it was easier to distribute them. I also can think of many examples of unsophisticated investors getting stuffed with structured products in the run-up to the crash – for instance, UBS had some in their retail mutual funds but never really talked about what they were. There was also a lot of SIV paper in money market funds, for instance at Bank of America’s Columbia, and these were SIVs that were not backstopped by a bank – when the SIVs assets went bad and they stopped paying the CP, Bank of America had to infuse capital into the money-market fund to prevent losses. Insurers who weren’t involved in the CDO business also took significant losses on CDO portfolios where they were just trying to make a spread over Treasuries.

    If the claim is that retail investors weren’t calling up their brokers and buying CDOs, sure. But retail and institutional money wound up getting a lot of CDO risk distributed to them as banks tried to off-load what they could.

Comments are closed.