Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.
I think you are confusing the metaphor used by Mr. Liu with the Ponzi Scheme. He was refering more to a (demographic) age pyramid. Also, it is untrue to qualify the new credit derivatives as gambles. Gambles are the optional contracts (where the final pay off depends on decisions that one party may make later, hence they are often called "contingent claim" and not directly in the books of non-bank companies), or an arbitrary formula with conditions for the pay-off (e.g. "digital options" are pure gambles: e.g. you bet 100$ that the Apple Stock will hit 200$ by year's end... that or Jérôme's 100$/bl oil bet).

The CDS (credit default swaps) are actually just an insurance mechanism (you insure against failure of a debtor to repay). Only when you insure a risk that you do not have in your book, do you make a gamble. And it is not the primary goal.

The CDO (collateral debt obligations) are a risk and reward sharing agreement where everything is laid bare right at the start, no optional clauses, even if the risk-reward trade-offs are not homogenous between the tranches (tranche is slice, in french, I dunno why it is so, may be the first CDO's were "cooked" by a french chef). It is in concept no different than age-old practice of companies having preferential dividend shares, multiple-vote shares, golden shares, death-pill shares, etc, etc...

There are two Ponzi schemes in effect in the present crash, but they are not where you think. The first is simply the fractional reserve banking system itself. Techno just posted a diary which pretty much justifies it a mean of growth to get where we stand now (a mature industrial civilization), I doubt it will still be adequate for the post-peak contraction.

The second scheme is a classic ponzi bubble of selling an overvalued good for yet more money, to "some yet greater idiot": more and more bankers, brokers, realtors (all fee-earning), and yes even borrowers (who made capital gains refi after refi), where lured into a system of easy money that lasted until there was no more fools to bring in to keep the prices going up. And the last of the fools where simply those of the lesser quality (the subprime borrowers), and they were brought in as a last resort to keep the scheme running for another cycle because the guys before them didn't want to be the loosers.


by Pierre on Sat Aug 18th, 2007 at 04:23:43 PM EST
He was refering more to a (demographic) age pyramid.

He was? Where? I don't see any implication that demographics are involved.

It is in concept no different than age-old practice of companies having preferential dividend shares, multiple-vote shares, golden shares, death-pill shares, etc, etc...

The differences:

  1. The ratings were fixed. Supposedly AAA tranches were nothing of the sort - and both speculators and ratings companies knew this. The distinction between low and high risk tranches was largely fiction.

  2. Low rates keep the pot boiling deliberately. This was a deliberate political move across the entire market to inflate the national economy and make it appear bouyant, and not random market action by small companies.

  3. Unlike individual share action, there's a good chance the Fed is going to hit the markets with another dose of low-rate crack, while leaving the individual borrowers to lose their homes.
by ThatBritGuy (thatbritguy (at) googlemail.com) on Sun Aug 19th, 2007 at 12:25:39 PM EST
[ Parent ]

least risky tranches at the top and the more risky tranches with commensurate premiums towards the bottom

it's drawing loans like an age pyramid, except you don't sort by year of birth but by year of default (death). Early years bottom, already defaulted->equity tranche, future years (not yet defaulted) --> up in the senior tranches.

This extra value when siphoned off repeatedly from the overall mortgage pool required an ever larger base of sub-prime mortgages in the new pyramid shape, thus increasing the systemic risk further

Here actually he gets plain wrong, or he's trying to push a point to a laymen audience the wrong way. Once a given loan fund is built to be sliced in new CDO, no more loans are added to it. New loans -> new wagon of funds, tranches, etc... You don't feed new risky loans in a sold out fund.

The nexus that connects the CDO and subprime, is that CDOs enabled bad loans to be dumped on the market. There are not enough buyers for all bad loans if they are all rated junk. The CDO, by putting "watermarks", enables packagers to pretend most of the fund is safe and find buyers for it. So they siphoned Wall Street money into McMansions where realtors shoved illiterate borrowers. And everybody took a fee in the process. And to keep on the promises that borrowers could refi, they had to find more borrowers to keep the prices going up. The ponzi scheme is here, and not in the building of the tranches.

The differences:(snipped)

Totally agree. 1) The behavior of rating agencies was fraudulent, close to Arthur Andersen in the Enron case, and they will face consequences. But this is not intrinsic to CDO slicing, the greed and the low rates have turned US real estate financing into a giant fraud, and the fraud just used the fashionable instruments of the time
2) and 3) Yup, it's a political thing, and the politicians will use the RA as a fuse when they are equally guilty.


by Pierre on Sun Aug 19th, 2007 at 02:26:03 PM EST
[ Parent ]
but by year of default

So the tranching is done a posteriori, the loans are not pre-sorted?

Say you have a CDO with three loans and three tranches. One of the loans default now, it is automatically assigned to the last tranche owner who now have a foreclosed house, right?

by Laurent GUERBY on Mon Aug 20th, 2007 at 02:27:07 AM EST
[ Parent ]
CDOs don't transfer ownership of house : they are insurance contracts. The owner of the equity tranche pays a pre-decided amount of money while the first n% of the basket of loan takers default ; and indeed the tranches aren't pre-sorted, which is the whole point of CDOs : it was supposed that if you took a bunch of loans, even though they were bad, it was unlikely that more than 50% of them would default, thus AAA rating for the senior tranches.

The problem in valuing such CDOs being that in a crisis, like a bursting of the housing bubble, the probabilities of individual subprimes loan takers to default become highly correlated rather than independent events ; if one underestimated that correlation, one overvalued the senior tranche of the CDO, which is why some end up being hit unexpectedly, having to pay premiums they didn't think they'd have to.

Un roi sans divertissement est un homme plein de misères

by linca (antonin POINT lucas AROBASE gmail.com) on Mon Aug 20th, 2007 at 03:02:30 AM EST
[ Parent ]
Do you know what the pre-decided amount looks like relative to the estimated value of the houses?

Assuming the institution holding the loans bought CDO when defaults happens they get both the house and the CDO pre-decided payment, right?

by Laurent GUERBY on Mon Aug 20th, 2007 at 03:54:48 AM EST
[ Parent ]
The amounts are not directly related to the value of the houses. CDO funds are pretty large, they pool thousands of mortgages and the total debt before slicing will typically run into the billion $ (you need to amortize a lot of fixed bureaucratic costs). And then, defaults are cumulated by value of coupon and capital lost to determine what tranche they hit. The "watermarks" for the tranches can be surprisingly low. e.g. for corporate debt CDOs, the "equity" tranche ends at around 3%. Then the mezzanine tranche ends at 5-7% of defaults. And then you hit the senior tranches of various ratings, but in any case you'll be hitting (and eating) into AAA+ before 30-50% depending on the initial quality of the loans agregated. And this will definetely be hit with funds of fraudulent subprime loans.

by Pierre on Mon Aug 20th, 2007 at 04:01:26 AM EST
[ Parent ]
And then, defaults are cumulated by value of coupon and capital lost to determine what tranche they hit.

Ok so it's not purely time of default based, and capital lost plays a role according to you, so not exactly what I understood from linca.

I guess the devil is in the details of each CDO but it's hard to judge what's going on. My questions resolve around wether the holder of loan+CDO can make money in case of default or not (payment of CDO + payment from selling the house > original price of loan + original price of CDO).


by Laurent GUERBY on Mon Aug 20th, 2007 at 04:19:46 AM EST
[ Parent ]
My questions resolve around wether the holder of loan+CDO can make money in case of default or not

I don't think this is possible. Actually, the holder of a CDO tranche security will never get a house back. The fund has a contract with a realtor for day to day management of the pool of houses, and when there is a default, this manager will arrange for a refi settlement, a foreclosure, a rental, or an auction to try to make the best money out of it. But in any case, it is only money that is funneled back to the owner of the security. There is a default as soon as the coupon (interest payment) is not 100% of what is was meant to be (there can be partial payments) and/or the nominal has lost value (because it is already known that a house has been auctioned off for less than what had been loaned for it, and the borrower is bankrupt).

So for the owner of a risky tranche will never make a capital gain. In the extremely unlikely event that a defaulted house becomes fund property, is rented while the market is depressed, and may be sold at a gain much later, the excess money will be used to offset other losses in the fund. And if there was a total surplus (which would require the housing market to skyrocket next year), the remainder would go to the funds sponsors (a bank), not the CDO holder.

Because remember: a CDO belongs to the realm of "fixed income" securities, like a bond, and the nominal paid at maturity is agreed at the beginning. There is a coupon to pay a premium for a risk of default, but rarely a performance bonus !! this is in the realm of corporate/venture stock only ...

Note that a local boost in housing is not impossible: if a godzilla hurricane destroys all of florida and 10 million have to be relocated, funds that are geographically focused (on other areas) would benefit. But for the banking system as a whole, it's still a disaster, whatever happens now.


by Pierre on Mon Aug 20th, 2007 at 05:13:18 AM EST
[ Parent ]
Ok then the buyer of the CDO does own the loan and so the house in case of default: detail of property management is just outsourced and only cash flows go out. This is not what I understood from linca comment above.

Thanks for the precisions!

by Laurent GUERBY on Mon Aug 20th, 2007 at 06:31:50 AM EST
[ Parent ]
Liu says:
This is the structural un-sustainability of CDO securitization, irrespective of the state of the economy...
This sounds pretty specific.

I am still puzzled with what drove new CDO buyers to enter. There hadto be something promising - or was there no profit chain, only fear of loss?

It possible that the excitement with CDOs coincided with the first problems of subprime mortgages. Someone may have figured out that a way to get rid of "junk" packages (or to mitigate a subprime fallout temporarily) was to unwrap the pyramid potential of CDOs. On the other hand, I heard that CDOs were rarely traded... Was there only a frenzy of buying new CDO items?

by das monde on Sun Aug 19th, 2007 at 12:46:32 PM EST
[ Parent ]
CDO are not unsustainable. Liu should just be stating he doesn't like them, that's it. But they're not unsustainable by and of themselves. The reason institutions kept buying them is simply that they did not know what to do with the money that was raining on them (as the fed was minting so much... it's really a monetary spigot problem).

There is no pyramid to unravel in CDOs themselves, the pyramid is in housing. Just a debt-fueled asset bubble, and the repackaging makes it very hard to know who is hit (or hit worst, as everybody will be hurt eventually).

And it's not a subprime thing either. Supbrime just happens to be the first bubble to burst, but there is more than one bubble in the US today. Presently a large proportion of traded US corporate bonds are junk rated, and it's only the easy refi that kept them from defaulting. It will come next, in 2008, and it will be even bigger than subprime. And there will be CDOs too, and again, they are not the pyramid, they are just spaghetti on top of it ...


by Pierre on Sun Aug 19th, 2007 at 02:09:55 PM EST
[ Parent ]
I made a Google web search on "pyramid CDO", and I found confirming opinions:
CDOs, which are debt instruments, are built as a pyramid. The bonds are packaged into different tranches of securities. At the bottom is the equity tranche backed by the riskier debt or loan collateral. On the top of the pyramid are the "super senior tranche" first, followed by the triple-A rated tranches. Those senior tranches are backed by investment-grade debt or high-quality loans. In between are the mezzanine tranches, which offer a bit of a mix of the pros and cons of both extremes. Naturally, the equity tranche at the bottom of the pyramid is the one with the potential highest risk and reward, which is why it offers the highest yield. The senior tranches on top of the pyramid pay the lowest yield because they are safer, being backed by higher credit-quality assets and benefiting from the "cushion" offered by the equity tranche that gets hit by losses first in the event of credit downgrades affecting the collateral.

Typically hedge funds get involved in the CDO market for two main reasons. The first is to obtain returns. In such cases they act as CDO investors, buying almost always the rewarding equity tranche. The second reason will lead them to issue CDOs in order to obtain a new source of funding. In these cases they are the collateral managers, those who gather and buy the assets for the collateral pool. They obtain funds by issuing the notes sliced into the various tranches that are offered to the market. Hedge funds in such cases are no different from the corporations that issue bonds on the capital markets for their funding needs.

This is from Bloomberg.com, no less:

Bundling mortgages into asset-backed bonds and then agglutinating those bonds into collateralized debt obligations sliced into different flavors of risk always smacked of a sophisticated pyramid scheme. As the foundations crumble, even the apex of the CDO market is looking shaky.

This book has a big chapter on CDOs (No 20), and this chapter:

4. Liquidity, the Credit Pyramid and Market Data
4.1 Bond liquidity
4.2 The Credit Pyramid
4.3 Survey and engineered spread data
4.3.1 Survey data
4.3.2 Engineered data
4.4 Spread and rating

And here you can find some technical info about "Fortis Bank Pyramid CDO, a Euro 268mm ABS CDO". Is this an apt name for some financial product?

The following reference is more rhetorical, but still informative:

[All buyers] liked the extra yield this stuff provided and chose to ignore the dubious basis on which A credit ratings were obtained or the fact that these instruments were not traded and so could not be marked to market. They were worth what the issuers said they were worth - until they tried to sell them.

The sheer scale of these ponzi activities is indicated by the fact that in the first quarter of this year alone US$251 billion worth of CDOs were issued and including US$121 billion of credit default swaps - instruments by which banks sell risk to each other.

by das monde on Mon Aug 20th, 2007 at 05:17:09 AM EST
[ Parent ]
There was a pyramid scheme not because of the pyramid structuring of CDOs, which has nothing to do with Ponzi scheme, but because there may have been a bubble in CDO valuation ; Financial bubbles are when something is bought because of an expected increase in value may allow reselling at profit, and those may be described as Ponzi schemes.

The first link is wrong in its description of CDOs, BTW : the seniority of the upper tranches comes not because they are backed by safer securities, but because it is supposed they'll have to pay premiums only if all the backing debtors default, which is supposed to be very unlikely (and which may prove to be false in the case of subprime CDOs)

Un roi sans divertissement est un homme plein de misères

by linca (antonin POINT lucas AROBASE gmail.com) on Mon Aug 20th, 2007 at 05:39:33 AM EST
[ Parent ]
I do see a pyramid structuring of CDOs. I would not call it bubble - the CDOs were rarely traded. They were perhaps hardly tradable by design or expectation. As the last citation says,

"They were worth what the issuers said they were worth - until they tried to sell them."

by das monde on Mon Aug 20th, 2007 at 05:48:30 AM EST
[ Parent ]
The CDOs were traded ; that's how they were more effectively priced. Now that's there is not trust in CDOs, they can't be traded, and thus are losing their value.

A Ponzi scheme means essentially that you give the first entrants high returns with the money invested by later entrants ; this is not the scheme of CDOs at all : CDOs are insurance contracts.

Un roi sans divertissement est un homme plein de misères

by linca (antonin POINT lucas AROBASE gmail.com) on Mon Aug 20th, 2007 at 06:05:31 AM EST
[ Parent ]
I clarified my use of Ponzi/pyramid terminology in a parallel sub-thread. I do not refer to CDOs as Ponzi.

The lack of  CDO trading was especially emphasised at the moment of BNP Paribas crisis. CDO valuation was anything but effective. Banks and hedge funds would have normally sold junk CDOs away, but they were scared reveal the real value of CDOs; and the real value of CDOs was opaque because they were rarely traded. That was the reasoning so shortly ago.

by das monde on Mon Aug 20th, 2007 at 06:34:41 AM EST
[ Parent ]
CDOs weren't a full-Ponzi, but there was more than a hint of Ponzi about them.

I'm guessing the first few iterations of the high-risk tranches would have posted impressive returns, and this would have been enough to persuade hedge funds to funnel money at them, starting a mini-Ponzi stampede, which would in turn have made CDOs as a whole look more attractive - especially with unrealistic risk ratings.

All it's going to take is proof that someone somewhere moved money from Tranche A to pay off a high return on Tranche B in the hope that money would come in to cover Tranche A at some point, sooner or later, and you have a classic Ponzi.

I'll be surprised if no one tried this.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Aug 20th, 2007 at 07:49:58 AM EST
[ Parent ]
Trading dried up because of the crisis a few weeks ago ; before that, although the market wasn't extremely liquid, trading happened regularly.

And the 'real value' of anything is given by the market, anyway, in our utilitarist world.

Un roi sans divertissement est un homme plein de misères

by linca (antonin POINT lucas AROBASE gmail.com) on Mon Aug 20th, 2007 at 08:09:20 AM EST
[ Parent ]
I think I got the profit chain of the CDO game:

Each mortgage refinancing gave the old loan - and the CDOs covering them - immediate and safe profit, while shifting the risk to a new loan - and new CDOs.

Wall Street was probably betting on continuous refinancing. In that case subprime loans in CDOs were very attractive because they were "guaranteed"  to be refinanced soon, ending all risk.

Here we see a classical element of a pyramid scheme: profit of one item is based on a generation of new items.

Exponential branching appears to be missing on the mortgage level, though typically greater size of new loans alone may be generating a commulative effect. The "slicing and dicing" routine with CDOs probably make a pyramid "complete", through a mathematical model is welcome.

Pyramid elements are floating around, and they make the CDO scheme very unstable. Like in a straight pyramid scam, looks excitingly fine until a logically unavoidable limit is met. In the CDO case, the limit is met when the real estate inflation and refinancing tricks must stop. More generally, broad economic models conspiciously lack clear acknowledgment everything  of positive feedbacks, as Michael Hudson notes. We are left then with Ponzi/pyramid frazeology with every hurting runaway boom and bust.

by das monde on Tue Aug 21st, 2007 at 06:18:25 AM EST
[ Parent ]
Here we see a classical element of a pyramid scheme: profit of one item is based on a generation of new items.

... but the profit is not generated by the refinance of the loans ... indeed, the investment grade tranches are protected from pre-payment risk, as well as default risk.

What the refinance of the loans did was obscure the systematic risk faced by the holders of the higher return junk grade tranches.

I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.

by BruceMcF (agila61 at netscape dot net) on Tue Aug 21st, 2007 at 09:03:08 PM EST
[ Parent ]
It is indeed principally risk that is being transfered.

But it can be said that the profit is "in effect" generated by the refinance: the books are closed for the old loan and in parts of the related CDOs. Those parts in CDOs get a final profit margin, "at expense" of new CDO items.

Although the profit before refinance is supposed to be small (since interest is low until refinance is "forced" by high interest rates kicking in after a fixed time period), the tranching trick may give the lower CDOs higher than the nominal profit: If I understand right, the lower tranches are first to take any gain or pain. In this picture, it is the senior tranches that look like total suckers: they get meager profit in good times, and they are not really protected in bad times anyway. But I may need to learn (if I would dare) all details of CDO slicing to be sure.

by das monde on Wed Aug 22nd, 2007 at 05:16:55 AM EST
[ Parent ]
... of mine, and reserve banking is no Ponzi scheme.

That is, to be precise, reserve banking is not intrinsically a Ponzi scheme ... provided the credit contract created by the bank is sound, the credit money is sound.

What is required to prepare reserve banking for the post-peak loss of ongoing extensive growth, to be replaced at best by periodic waves of technology driven intensive growth, is a return to low and stable cash rates and regulation of banking operations to require prudent behavior.

Or, in other words, all those prudential regulatory schemes instituted in financial centers after the Great Depression, which have been progressively stripped away since the 60's.

I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.

by BruceMcF (agila61 at netscape dot net) on Sun Aug 19th, 2007 at 04:50:39 PM EST
[ Parent ]
Here is my way of disambiguating terms "Ponzi schemes", "pyramid schemes", etc.

A Ponzi scheme (or Ponzi financing, as in Minsky's theory) is a scam initiated by an entrepreneur seeking investment. Mr Ponzi himself did this that way: he promised and initially gave high return for investment, which attracted much more investment, so that for a while he was able to give high returns from the investment growth alone. But of course, he had top cut and hide at some time.

Ponzi financing may occur not only as a designed scam, but from enterpriser's desperation as well. I call the Russian GKO scheme because the initiative came from the Russian government (for whatever reason), not from investors.

The situation of "too much money chasing too few assets" has different flavor. Here the initiative comes from eager investors. The mild term for this run is "asset inflation". I would avoid to use "Ponzi" or "pyramid" terminology here. I do not actually know how to call this stronger.

In a narrow sense, a pyramid scheme is a scam with a defined structure (mathematically, an exponentially growing tree graph) of cash contributions and flow. Here there is no formal distinction between investors and enterprisers - everyone plays the same role if only the game can continue indefinitely. Ironically, the current situation with all investors borrowing and all entrepreneurs investing is very analogous in this respect.

More generally, to incorporate the cases like CDO (presumably) a pyramid scheme can be defined as any designed or emergent financial structure with something organized within an unsustainable pyramid "geometry", as Liu describes. That would be the general idea.

by das monde on Mon Aug 20th, 2007 at 05:40:41 AM EST
[ Parent ]
Sorry for hasty writing. I did not avoid some typos. To captivate most dubiuos errors:

The first paragraph should end:
But of course, he had to cut and hide at some time.

The last paragraph should better by typed as
More generally, to incorporate the cases like (presumably) CDO, a pyramid scheme can be defined as any designed or emergent financial structure with...

And the stronger characterizations of "asset inflation" can probably be "bubble" or "overheating".

by das monde on Mon Aug 20th, 2007 at 05:52:30 AM EST
[ Parent ]
... Ponzi financing is not just leverage ... its financing that collapse without an ongoing increase in the level of investment, which means, of course, it cannot be sustained in a steady state.

Prudential finance involves debt that can be serviced by the income from the financed activity even if income flows from the funded project fall short of projections ... how prudent is a matter of how bad things can turn and still service the finance.

The trap that CDO are heir too is, rather, a debt-farming shell game. If the top tranche of a CDO is considered an investment grade asset, and so allows participation in non-investment grade activity by those limited to holding investment grade assets, and the low risk premium = high price of the investment grade tranche(s) more than compensate sofr the high risk premium = low price of the junk grade tranche(s) ... then instead of the classical small town bank looking on the interest on mortgages (and small business loans) as its bread and butter, mortgages become a way to generate ongoing commissions on origination, selling the mortgages on in order to be able to originate more.

Of course, market pressure will invariably push the finance sector into non-prudential behavior until a series of events realises the systematic risk that has been lying dormant ... unless a regulatory authority imposes a market penalty for engaging in non-prudential behavior. And it is the reaching for those short term gains (ignoring their attendant long term pains) that has driven the ongoing stripping of prudential regulation from financial systems around the world.

I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.

by BruceMcF (agila61 at netscape dot net) on Tue Aug 21st, 2007 at 08:57:16 PM EST
[ Parent ]


Top Diaries

Occasional Series