Editore"s Note
Tilting at Windmills

Email Newsletter icon, E-mail Newsletter icon, Email List icon, E-mail List icon Sign up for Free News & Updates

November 1, 2007
By: Kevin Drum

A QUESTION FOR THE ROCKET SCIENTISTS....Every couple of weeks I take a whack at trying to understand what Collateralized Debt Obligations and Structured Investment Vehicles really are. My latest take is that they're basically ways to game the credit rating agencies (a game the agencies were none too eager to stop, as it turns out). Is that about right?

UPDATE: OK, here's what I mean by this. Warning: abysmal ignorance about to be displayed.

Basically, the whole selling point of a CDO is that its risk-adjusted return is higher than the risk-adjusted return of its underlying parts — i.e., that you've modeled the CDO's diversification or correlation or something better than Moody's has, and it will thus yield a bit more than you'd normally expect based on the CDO's rating.

Since the rated tranches are pretty straightforward, this mostly depends on convincing the world (and the rating agencies) that the small unrated tranche doesn't affect the riskiness of the rated tranches by very much. Not enough to affect their stated ratings, anyway.

But in the long run, this can't work. If you're right, then eventually everyone else's models will catch up and the yield of the securities in the unrated tranche will go down, thus eliminating the pooling advantage claimed by the CDO in the first place. If you're wrong, then the CDO goes bust. The game only continues if you can consistently value the unrated tranche more accurately than anyone else, including the rating agencies. But how many people can actually do that?

Kevin Drum 7:54 PM Permalink | Trackbacks | Comments (55)

Bookmark and Share
 
Comments

My latest take is that they're basically ways to game the credit rating agencies.

Gee, I thought the rating agencies pretty much gamed themselves on this one.

Posted by: bleh on November 1, 2007 at 7:55 PM | PERMALINK

Collateralized Debt Obligations and Structured Investment Vehicles are not inherently "ways to game the credit rating agencies" but they can be used to do so if not carefully regulated.

Posted by: Econobuzz on November 1, 2007 at 7:58 PM | PERMALINK

I'm not a rocket scientist, but econobuzz has it right, above. They are essentially debt obligations that are repackaged and sold as securities, and as such, add liquidity and flexibility to the monetary system. As for "gaming," organizations like Fannie Mae, Freddie Mac, and so on, wouldn't exist without CDOs. They are, and have been, useful tools in financing America's economy, particularly in the world of real estate. The problem we've been seeing with them is the result of greed, bad judgment, and panic, not the securities themselves.

Posted by: Martin Gale on November 1, 2007 at 8:08 PM | PERMALINK

They are securitization structures - basically a pool of debt instruments (often mortgages) packaged together in the form of a security that pays interest and principal (if you are lucky) to the party that purchases them. They are often "tranched", meaning one class may pay interest only (usually riskier) and some may pay a return of principal only.

I think your reference to "gaming the credit rating agencies" may refer to the fact that mortgages that don't meet FNMA, FHLMC or GNMA underwriting requirements (e.g. too high of loan-to-value ratios, inadequate collateral, etc.) can be put into a CDO, but not into the securities that the aforementioned govt.-sponsored agencies like Fannie or Freddie issue. These vehicles also allow the originators to remove the original loan or mortgage from their balance sheet, provided they get "true sale" treatment. Thus, the risk of default associated with the underlying mortgage is passed to the buyer of the security. Because they often carry a higher "coupon rate" or interest rate, CDOs look appealing to investors like Citibank or B of A, when in reality, they are highly risky due to the shaky nature of the original loans that make up the pool.

Hope that helps....

Posted by: The Conservative Deflator on November 1, 2007 at 8:15 PM | PERMALINK

Just my opinion:
1 A lot of managers of a lot of money were under pressure to produce results (high % returns). Either from baby boomers starting from what they wanted rather than from what is possible. Or because they were responsible for pension funds that had been underfunded and so were forced to find ways to raise the rate of return.
2 If you are already actually reasonably intelligently, then you can not raise your rate of return any higher without taking on more risk.
3 These non-transparent investment vehicles allowed you to earn very visibly higher returns but pretend you did not see the risk.
4 Until now.
In other words, I think institutional constraints made all parties willing to play this game.
BTW, the perception that our investment markets are uniquely transparent and fair is one of our largest competitive advantages. This may wind up doing significant long-term damage.

Posted by: jessica rooney on November 1, 2007 at 8:15 PM | PERMALINK

You should spend some time at Calculated Risk, especially the Compleat Ubernerd posts (here). They have a specific slant, but they are good at disclosing their biases, and to the best of my knowledge, they are accurate and complete.

Posted by: Mark on November 1, 2007 at 8:23 PM | PERMALINK

There have been a lot of discussions about this over at Mish's:

http://globaleconomicanalysis.blogspot.com/

Posted by: arteclectic on November 1, 2007 at 8:33 PM | PERMALINK

Kevin,

I think what was being gamed here was the algorithms of some of the "quants." These elaborate debt structures were constructed to make them look far less risky than they were. Look less risky to whom? Not whom, I think, but what - namely the algorithms used by some/most of the hedge funds.

Posted by: capitalistimperialistpig on November 1, 2007 at 8:37 PM | PERMALINK

Generally, SIV's are a scam. The basic idea is that you borrow money on short-term commercial paper markets (traditionally low-risk, therefore low-cost) and invest in packages of long-term collateral-backed loans (higher risk, therefore higher yield). You make your money on the spread between the money you borrowed and the money you loaned, and roll over the short-term commercial paper as needed. You have to be a really big bank to play this game, so that people trust you when you dip deep into short-term commercial financing.

The problem right now is that nobody really knows what the SIV's (with a paper value of about $400 billion) are actually worth, but we can be quite confident that it is less than 100% of the paper value. It might be a *lot* less. Banks are not required to provide any information about the underlying assets of their SIV vehicles. In effect, this little game has contaminated the entire $2 trillion short-term commercial paper market by injecting an unknown amount of bogus assets into the mix.

As a result, short-term commercial markets have seized up. There are two consequences, one potentially serious, one potentially catastrophic. First, all of those SIV's have to liquidate at firesale prices if they can't float on easy money. That's why Citicorp, which holds about a quarter of the SIV instruments, is frantically trolling for a bailout. Second, short-term commercial loans are the lubricant that keeps business transactions moving smoothly. Without easy and cheap access to short-term loans, commerce stops.

There's an interesting ethical issue here. Short-term commercial loans have traditionally been something of a low risk, high benefit public service. In effect, the banks that created this little game are playing investors in short-term commercial loans as suckers by increasing the underlying risk of that market without providing compensation, or even transparency. It's not illegal. It *is* wrong.

Posted by: Kevin Bell on November 1, 2007 at 8:41 PM | PERMALINK

Kevin:they're basically ways to game the credit rating agencies

The credit agencies are well paid for their ratings but the vehicles have been in use for a long time. This latest variation is a way to differentiate yield classes from the same underlying obligations.

Take the case of the sub-prime instruments. The customer, unable to access the regular mortgage market, is approved anyway because his/her/their individual mortgage is being bundled together with lots of other mortgages of similar credit profile. No investor would buy that bundle without substantial reserves (cushion).But that bundle of mortgages can be stripped into different pieces with different yields which is a CDO.

All of the different pieces are valued using the same mathematical model. That model uses historical payment data of rates of repayment, early payoffs, defaults etc. So, using the model, the IB creates classes which have different claims on the cash flow from the repayments.

The top layer or "A" class, which has the lowest yield, gets a fixed payment stream from the underlying mortgages. The rating agency looks at the model and the assumed stream of payments and, if they're satisfied the structure matches up with the historical data, issues a rating of let's say "BB+". That CDO can be priced at a yield equivalent to the market price for "BB+" debt. The lower layers are constructed and priced in similar, higher risk, higher yield fashion.

What went wrong was that the first year actual performance didn't look anything like the model on which the CDO was constructed. The market went bonkers because if the pricing model was wrong, badly, then what were the CDOs actually worth?

The rating agencies were looking at a model based on limited data since sub-prime mortgages really had no history. Traditional lenders simply wouldn't give a mortgage to somebody with those credit scores. What's neat about them is the product (the CDO) came first, the marketing of the "bad" credit customer came second.

So the rating agencies were wrong, but to a naif like Kevin, it must be premeditated evil. It's just business, Kevin, nothing personal.

Posted by: TJM on November 1, 2007 at 8:43 PM | PERMALINK

I'm not a rocket scientist either here, and perhaps I'm simply misunderstanding the question, but it seems to me that Martin Gale and econobuzz are both correct. CDOs are basically just a way to securitize debts that could not otherwise be sold efficiently in the market. (Otherwise you could assemble your own CDOs just buy buying specific debt portfolios).

They also provide access (via the tranches) to a specific risk level; however, to a certain they actually *rely* on the credit agencies' accuracy in order to properly divide or sub-divide each tranche. The point of unrated (i.e., equity) tranches isn't to provide a guaranteed return; it's to provide access to equity investments in the CDO (and like equity investments, if something goes wrong, equity holders get to absorb the loss first).

Like many investment products, they represent a calculated gamble, and sometimes you win and sometimes you lose. The fact, however, that you can only win at roulette less than half the time doesn't mean that you can't make money off of them in the short term - even if in the long term the deck is stacked against you.

Posted by: Geoff on November 1, 2007 at 8:50 PM | PERMALINK

So the rating agencies were wrong, but to a naif like Kevin, it must be premeditated evil. It's just business, Kevin, nothing personal.

Posted by: TJM

I think your analysis is correct but your conclusion may be a little too glib. It's not "just business," it's unregulated, speculative business that was (a) clearly counter to all common sense about the riskiness of subprime and unconventional loans and (b) ignored early warning signs that it would all end in a liquidity and credit crisis.

That's where the Ayn-Rand-Assholes like Greenspan failed to do their job. There were plenty of signs that this would all end the way it has.

I think the point is that markets -- and instruments designed to provide needed liquidity -- are not immune to speculation and manipulation, and that they need to be regulated.

Posted by: Econobuzz on November 1, 2007 at 8:58 PM | PERMALINK

CDOs are payment streams, not bonds. Through magical math, one can create a payment stream based on subprime mortgages that can have a AAA rating. However, these ratings aren't based on anything real; in fact, the ratings agencies pretty much admitted that the actual instruments were never evaluated. The evaluation is now happening ex post facto.

The models upon which these instruments are based, in general rely upon historical performance of a class of loans, not the wild west show that kicked in between 2004-2007 where almost every link in the origination chain "looked the other way" in order to collect fees. Add to that, no one really understands this stuff very well, and thus no one knows what it is worth.

All that would be bad enough, but some aggressive financial types loaded up the boat with such purchases, leveraging them wildly. Just a simple change in margin requirements can bring down the house of cards. When the margin calls hit, and the securities to be sold can't be valued reliably, any liquidity in the market will come at a discount that approaches 100 percent, assuming anyone will play.

As an earlier commenter urged, check out the CalculatedRisk blog (www.calculatedrisk.blogspot.com). Tanta's Ubernerd series should help educate.

Posted by: RickG on November 1, 2007 at 9:05 PM | PERMALINK

The inherent illogic in the CDOs backed by subprime loans recently was in thinking that broadly distributing risk so as to make high risk investments more attractive to investors has the same affect as actually reducing the default risk of those loans that made up the portfolio in the first place. Not that this was conscious on the part of the debt issuers. In their mind, they were simply pricing risk according to the risk tolerance as expressed in the market for these securities. If investors thought that a certain return was adequate, that tolerance essentially became a guideline driving the risk pricing all the way up to the rate sheets bursting through the inboxes of the fly-by-night mortgage brokers every morning. In the end, no one knew what the real risk was in these subprime loans. And, ominously for such large financial institutions as Citigroup (down 7% today on the NYSE), the real risk is still largely unknown.

Posted by: sporadic on November 1, 2007 at 9:08 PM | PERMALINK

Yes, it very much is just business for the IBs. They have clients looking for yields above a US Treasury or a bank CD. They created these vehicles to give the customer what they wanted, higher yields.

(a) clearly counter to all common sense about the riskiness of subprime and unconventional loans

That's just your opinion. Virtually everything the IBs do is unregulated and common sense has nothing to do with it. "Everybody" knew making loans to low credit scores was risky to some degree, that's why there are different classes. The lowest yield reflected the cushion needed, under the assumptions of the model, to make a BB+ piece of paper out of risky loans. According to the model, the loans were risky the tranche was not.

ignored early warning signs that it would all end in a liquidity and credit crisis.

Hindsight is always 20/20, hell, sometimes it's 20/10 but it's never 20/400. Amazing isn't it?

Posted by: TJM on November 1, 2007 at 9:16 PM | PERMALINK

> (a) clearly counter to all common sense
> about the riskiness of subprime and
> unconventional loans

As I have said over at DeLong's, IMHO a large part of this is the way the word "risk" was, in the 1990s, stripped of all connection to the common meaning of that word and overloaded (to use the C++ progammers' term) with 7 technical meanings that bore little connection to the fundamental meaning. Then people, some of whom knew what they were doing and some of whom did not, started using the symbol "risk" to sell financial instruments. Those who did know what they were doing knew full well that such use was at best obfuscatory and at worst fraudulent (those who didn't know what they were doing where just chumps) but they continued anyway as long as they were making money. I doubt any of them were surprised or hurt by the events of the last 3-6 months.

Cranky

Posted by: Cranky Observer on November 1, 2007 at 9:18 PM | PERMALINK

> That's just your opinion. Virtually everything
> the IBs do is unregulated and common sense has
> nothing to do with it. "Everybody" knew making
> loans to low credit scores was risky to some
> degree, that's why there are different classes.
> The lowest yield reflected the cushion needed,
> under the assumptions of the model,

This is the kind of baloney that borders on the fraudulent. The most glaring issue is that if everybody knew this (and of course they were the self-described smartest guys in the room) then the current situation could never have arisen. But in fact the risks were correlated (gee, that was never anticipated: even though it has happened in very housing downturn since the Babylonian times), the models were wrong, the risk levels were underestimated, the fundamentals were ignored, the "trees don't grow to the sky" principle (another one known since 10,000BC) was ignored, the bezzle was forgotten or swept under the rug...

But the people responsible for all this are either taking $129 million golden parachutes or looking for a bailout from Washington - which they will get.

Cranky

Posted by: Cranky Observer on November 1, 2007 at 9:25 PM | PERMALINK

I'm sorry, but on the whole Rocket Science is a lot simplier than financial instruments.

Beb

Posted by: beb on November 1, 2007 at 9:34 PM | PERMALINK

Virtually everything the IBs do is unregulated and common sense has nothing to do with it.

My point precisely.

According to the model, the loans were risky the tranche was not.

There is an inherent conflict of interest in these "models" that requires regulation.

Hindsight is always 20/20

Exactly. And what will we do next time? Rely again on models created by folks who stand to gain from underestimating risk?

This was a crisis manufactured on Wall Street. To suggest that it was an honest mistake by folks who didn't stand to gain is fucking ludicrous.

Those who ignore history are condemned to repeat it.

Posted by: Econobuzz on November 1, 2007 at 9:36 PM | PERMALINK

Great explanations, all. I like Kevin Bell's the best, if only because it has so many echoes in our recent history. I think it should be dubbed the Colin Powell Theorey (borrowing someone else's credibility for your own foolish and short-sighted, greedy, criminal purposes).

Posted by: jussumbody on November 1, 2007 at 9:40 PM | PERMALINK

All these financial shenanigans always happen during the Republican administration.

BCCI, Saving and Loan, the Michael Milken, etc. and now this.

The dotcom phenomenon does not qualify as it was qualitatively different.

Posted by: gregor on November 1, 2007 at 9:40 PM | PERMALINK

CDO's and SIV's are two totally different things. Let's start with CDO's (Not a professional here, but I have read quite a bit about this.).

The selling point of CDOs is the premium for top rated debt (AAA, Aaa+ or whatever the ratings agencies call it). Many institutional investors, ie pensions, are restricted in the ratings they are allowed to purchase, and many are restricted to top rated debt. As a result, top rated debt is always in higher demand than all others when just risk and reward are put together, so they demand a price premium. So, if you are rating on a scale of 1-10, debt that is rated 10 is sold as if it were more like 13. There is a huge incentive to get your debt in a top rated instrument. So, if you have a bunch of debt that is really around an 8, you slice off some and say it gets paid first. The ratings agency calls that a 10 and the rest is a 6. Normally, it would average out to an 8 for no better terms than you had got. But, if the 10 is sold for a price that's more like a 13, you end up with 13+6/2 and an average of 9.5, a huge bonus. There's nothing wrong with this, however due diligence was never done. The debt that was rated 10 may have been much closer to 5! Lately, the ratings agencies have done a ton of wholesale ratings downgrades. When the CDO's were downgraded, they lost the premium and had to be sold by the investors, and they are now worth a lot less than originally guessed, and nobody really knows how much they are worth.

SIVs are a bit more shady. What I have gathered is they are ways for banks to make investments off books, without affecting their banking reserve requirements. They must take on outside investors, otherwise it would have to go on the bank's books. However, they have been hit hard by the CDO and subprime crises. Many of the investors expect the banks to make good on the promises, in which case they would have to put in huge infusions of cash. Otherwise, the SIVs would have to be liquidated and would have no assets left and the banks would have lost credibility and access to funds.

What this all means? Short term gains of long term stability.

Posted by: Adam on November 1, 2007 at 10:05 PM | PERMALINK

> There's nothing wrong with this, however
> due diligence was never done.

Even assuming that the rating agencies were honest and the due diligence was performed and reported, if the risk between the two instruments is correlated then the exercise of "stripping" is meaningless.

Cranky

Posted by: Cranky Observer on November 1, 2007 at 10:14 PM | PERMALINK

Rocket science just looks easier because the rocket scientists know how to get it right.

Posted by: a cornellian on November 1, 2007 at 10:18 PM | PERMALINK

Cranky:
The point of the stripping is to get the premium of the top-rated debt.

Posted by: Adam on November 1, 2007 at 10:20 PM | PERMALINK

This is the kind of baloney that borders on the fraudulent

I'm sorry you feel that a description of a process is baloney. Is there something inaccurate? Is there an opinion in that description? No, I didn't think so. But that is your opinion.

if everybody knew this (and of course they were the self-described smartest guys in the room) then the current situation could never have arisen.
LBOs, REITs, you could name any vehicle created by Wall Street and point to a problem with it--- after the fact.

even though it has happened in very(sic) housing downturn... Lots of people said the prices couldn't keep up the pace, but only a few predicted a real problem. What you say is true enough but, again, it's 20/20 hindsight. Now if you're in Vegas or Phoenix or parts of Fla. good luck, but it's not as bad in every market. Real estate does go through cycles. REITs in the 70s, commercial and residential in the early 90s, sub-prime in the 00s.

the models were wrong No,duh. The repayment assumptions were wrong but the model was eminently saleable to sophisticated investors who were chasing yield.

the risk levels were underestimated You're on quite a hindsight roll, here, Cranky, let's see what else you know now.

looking for a bailout from Washington - which they will get. Really? Not at Bear Stearns, not at Merrill, not at Citi or UBS. Where is this bailout? The Fed putting liquidity in the market and lowering rates to keep the consumer driving the economy could, I suppose, be a bailout, but individual IBs?

Kevin Bell et al, see this:The U.S. commercial paper market saw its largest increase since the start of the credit crisis in the week to Wednesday, though a $9 billion decline in asset-backed paper outstanding testified to continued strains in money markets.

and:the exodus from the asset-backed segment of the market continued.

As Lord Gresham said, "bad money drives out good."

Posted by: TJM on November 1, 2007 at 10:32 PM | PERMALINK

My latest take is that they're basically ways to game the credit rating agencies.

Gee, I thought the rating agencies pretty much gamed themselves on this one. Posted by: bleh

The biggest problem (beyond the basic dishonesty) is that the credit rating agencies aren't really regulated. Rather than being as conservative as they should (and historically had been), they bought into the idea of an endless upward spiral in real estate prices and turned two blind eyes to the inherent shoddiness of some of the bundled debt. The subprime loans were, are no different than junk bonds and as part of packages should have lowered ratings. However, much of these crap CDOs were getting AAA ratings.

If Congress had any balls (which it doesn't) and it really cared about the health of the economy as a whole (which it doesn't as most memebers of Congress are independently wealthy or damned close), they'd be investigating the lenders, the "bundlers" and the debt rating agencies with the end result that most of them would be heavily fined and/or stripped of their abilities to practice their trades.

I'm not a rocket scientist, but econobuzz has it right, above. They are essentially debt obligations that are repackaged and sold as securities, and as such, add liquidity and flexibility to the monetary system. Posted by: Martin Gale

True. This is in great part of what makes home ownership rates so high in the U.S. However, in years past, banks and even mortgage companies didn't deal in absolutely crap loans. The ugly fact of the matter is that many of the people who "got caught up in" the evils of subprime had no business applying for a real estate loan of any kind.

Years ago I worked in mortgage banking (early 1980s). This was the first era of ARMs. However, ARMs then weren't predatory as they are now. You got in on low rate, and rolled the dice that when the adjustment came that you'd either be able to refinance to a decent fixed rate then or that the adjustment wouldn't be too high. The adjustments were market based, which is not what was happening over the last couple of years where the loan rate might shoot past what one would pay for a 30-year fix by anywhere form 5-10%.

Posted by: JeffII on November 1, 2007 at 11:05 PM | PERMALINK

The CORE problem, the one that underpins this whole crisis, is that practically all market participants miscalculated the default rate on the underlying assets of these instruments.

They did not do this simply because of greed or stupidity (although those two factors played a role like they do in every other aspect of life) but because they based their calculations on what the (admittedly recent) history told them.

If market participants had had a better understanding of the ultimate rates of default there would have been less issuance but we would have had no crisis. Is it reasonable to claim they SHOULD have known better? I don't think so, or at least not without plenty of caveats.

Talk of CDOs being Ponzi schemes or fundamentally flawed are wrong.

Posted by: Gabriel on November 1, 2007 at 11:30 PM | PERMALINK

Ummmmm . . . rockets run on Newton's Third law???

Posted by: pjcamp on November 1, 2007 at 11:41 PM | PERMALINK

The CORE problem, the one that underpins this whole crisis, is that practically all market participants miscalculated the default rate on the underlying assets of these instruments. Posted by: Gabriel

No. One of the "core" problems is that dishonest mortgage brokers didn't, don't give a shit about the credit worthiness of their borrowers because they make their money on the volume of loans they generate and the fees they earn from this. They aren't holding the paper, so they don't care.

Another "core" problem, as I wrote above, was, is the lack of regulatory oversight for the credit rating business and its dishonesty in giving too many CDOs ratings far higher than they deserved.

The final "core" problem is that the agents for these CDOs, knowing full well that these instruments were mostly crap and potentially disastrous investments, essentially lied about them to investors. These are the "core" problems.

The default rate was easy to calculate because borrowers who couldn't qualify for a fixed rate mortgage at what are still low rates were sure to default when their introductory rates doubled or even tripled, regardless of where the RE market went.

And, as I wrote above, there was an irrational (where have we heard that word before used to describe investments?) belief that real estate prices would continue to rise 5-20% a year, thus giving unqualified borrowers and speculators "instant" equity in properties for which many never put a penny down of their own money. I'd like to have sympathy for these people, but very rarely does it work out that anyone gets something of value for nothing.

Posted by: JeffII on November 1, 2007 at 11:47 PM | PERMALINK

Jeff,

If the default rates had been calculated correctly, if, say, market participants had realized they would be closer to 20% than to 10% the issues you mention (dishonest brokers, for example) would have been irrelevant, they would have simply been priced in.

Are you part of this world at all? Have you ever analyze a CDO for example?

Posted by: Gabriel on November 1, 2007 at 11:52 PM | PERMALINK

CDOs have been overused, and as people like Bill Fleckenstein, Jim Jubak, Scott Burns and others point out, they never were "marked to market."

Sorry, Econobuzz and agreers, they're about as good for the market, especially with relaxation of the asset standards on lending, as was the dereg of S&Ls back in the 1980s. Guess what? We have a similar situation today.

As for the ratings agencies like Moody's, what we have is similar to the scandal over financial advisors and their touts of stocks a few years back: too much of an incestuous relationship.

And, as for investment tools, folks at Citibank or BofA were supposed to know better, but let themselves get infected with the greed bug.

And, bottom line is, this is the flip side of subprime loans. If a debt can't stand to be sold by itself in the market, then it has no business being out there while being hidden by a fig leaf. And THAT, Kevin, is what CDOs ultimately are: a massive fig leaf. Or the emperor's imaginary clothes. Choose your metaphor.

Jessica Rooney: London is already rated as having more transparency than NYC. Many international investors consider it, not the Big Apple, to be the world's financial capital.

Kevin Bell is spot on about SIVs. TJM is about the "marked to model" rather than "marked to market" problem of CDOs.

JeffII: Many subprime borrowers were NOT first-time homebuyers. They were buying their second home of their life, or third, and got greedy for 4K square feet rather than 2,500. OR, they were persons buying a second or third home to own at once as an investment vehicle. People in this last group are just like people sinking 300K in the stock market. The market tightened and now you have a "call" on your mortgage and you're short. Tough shit.

Posted by: SocraticGadfly on November 1, 2007 at 11:57 PM | PERMALINK

Are you part of this world at all? Have you ever analyze a CDO for example? Posted by: Gabriel

This has nothing to do with CDOs per se, which have been around for at least 25 years. The issue was the lack of credit worthiness of the subprime borrowers, many of whom, regardless of where the RE market went, were never going to be able to afford the future adjusted mortgage rates short of the miracle of the their incomes doubling or tripling. Add to them all the the speculators, all the idiots hoping to "flip" just one more house, and you find us exactly where we are today.

Posted by: JeffII on November 2, 2007 at 12:01 AM | PERMALINK

Well, you didn't answer my question. How long CDOs have been around doesn't tell me how much you know about them. Rockets have been around for much longer but most people are not rocket scientists.

Of course the issue is the creditworthiness of the borrowers. That's what I wrote. The core issue in these instruments is to try to accurately measure that creditworthiness. And they didn't.

Remember that even know the vast majority of borrowers continue to pay back on time. So all it takes is comparatively small changes in the default rate to completely change the models.

Posted by: Gabriel on November 2, 2007 at 12:05 AM | PERMALINK

Many subprime borrowers were NOT first-time homebuyers. Posted by: SocraticGadfly

Never said they were. I said these were borrowers that didn't have the credit worthiness to handle a conventional loan, otherwise they wouldn't be defaulting today.

No matter how you all want to spin this problem, that's what it boils down to.

Posted by: JeffII on November 2, 2007 at 12:06 AM | PERMALINK

Of course the issue is the creditworthiness of the borrowers. That's what I wrote. Posted by: Gabriel

No you didn't. This is what you wrote and what I disagreed with.

The CORE problem, the one that underpins this whole crisis, is that practically all market participants miscalculated the default rate on the underlying assets of these instruments. Posted by: Gabriel

This is mumbo-jumbo about miscalculation, which any of these new "masters of the universe" that bundled these bad loans could have figured out on a decent HP calculator. And again, since the bulk of these subprime loans were made by mortgage brokers, they were never in danger of getting caught holding the bag as long as they sold the paper, which is the hold point of being a broker rather than a bank.

Posted by: JeffII on November 2, 2007 at 12:15 AM | PERMALINK

We need to ask an honest guy like Buffett what the word was on this mess.

Posted by: ferd on November 2, 2007 at 12:16 AM | PERMALINK

Jeff,

I ask again, do you know anything of this market, have you any training in this area?

You seem to be pretty confused by the terminology since you don't realize that calculating the default rates and knowing the creditworthiness ARE THE SAME THING. (sorry I shouted but it needs to be clear).

I know that it feels good to read some newspaper articles and spout that anyone with an "HP calculator" could have gotten it right but, sadly, that's not how the world works. Not the world in general and certainly not the world of securitization.

You keep mentioning the role of the mortgage brokers which would have been of no importance if we had known what the default rates would be. That's the tricky question, how much of this could have been predicted in advance and in what areas?

Posted by: Gabriel on November 2, 2007 at 12:23 AM | PERMALINK

To some extent you're correct but it's just as much a way to game yourself as the credit agencies. You can think of CDOs in much the same way as junk bonds. There the returns looked higher than the risk because there weren't many junk bonds until Drexel came along. Once you did it on a large scale the perceived advantages disappeared. In order to feed the system with saleable product, in this case CDOs, on a larger scale you have to reach and this forces you to make riskier loans because there aren't enough good loans.

It's really scale that is the problem. There's also a Ponzi scheme element to it in that you wouldn't make the loans in the first place if you can't sell them off. The originating lender is not willing to hold the loan.

Posted by: hrk on November 2, 2007 at 12:25 AM | PERMALINK

Drum: But in the long run, this can't work. If you're right, then eventually everyone else's models will catch up and the yield of the securities in the unrated tranche will go down, thus eliminating the pooling advantage claimed by the CDO in the first place. If you're wrong, then the CDO goes bust. The game only continues if you can consistently value the unrated tranche more accurately than anyone else, including the rating agencies. But how many people can actually do that?

I sincerely don't mean to be a jackass here, but you're still half operating on the assumption that the debt instruments themselves were the problem, and you're focusing on them. The problem was the greed and judgment of the lenders, the ignorance of the borrowers: the securitization of the debt was irrelevant, except to the extent that it facilitated (made it easier to raise) the loans, which extent is something I don't think anyone knows. I think once that point is understood, the whole topic goes away, or devolves into some conversation that is quite fascinating, I'm sure, to an actuary or some such, but deadly dull to the rest of us. But whether or not CDOs themselves are bad should never have been brought up. Are CDs bad? Are passbook savings accounts, or money market accounts bad? How about a bond fund that mixes low-rated debt into its portfolio to juice the yield? It's the same principle at work: pooling debt and re-selling it to investors.

Your question all along should have been, "Should people with bad credit histories be given access to risky loans?" And that is a question we answered long ago when we chose a market-based economic system.

Posted by: Martin Gale on November 2, 2007 at 1:00 AM | PERMALINK

Martin,

The problem wasn't making risky loans. The financial system does that all the time in almost every area one can imagine. The key question should have been "Are we comfortable that we are measuring the risk correctly?".

Posted by: Gabriel on November 2, 2007 at 1:08 AM | PERMALINK

This is the beginning of the post capitalist era.We don't put 'in god we trust' on the $ for nothing.I fear another 'terrorist attack' is imminent.

Posted by: paper tiger on November 2, 2007 at 3:10 AM | PERMALINK

Sorry, Econobuzz and agreers, they're about as good for the market, especially with relaxation of the asset standards on lending, as was the dereg of S&Ls back in the 1980s.

Posted by: SocraticGadfly

Once again, they are financial instruments just as S&Ls are financial institutions. Both serve a market function. Both can be abused if not regulated.

In non-rocket-science terms, if the fox is left in charge of the chicken coop, expect the worse. The market will always sort it all out in the long run, but a lot of innocent folks will be hurt in the process.

Drawing the conclusion (not yours) that millions of innocent folks have to suffer because an honest, unavoidable "mistake" was made in valuing risk, by players who had nothing to gain by deliberate undervaluation, is ludicrous.

And, as I'm sure you will agree, blaming the victims who were enticed into loans with terms they ultimately could not afford is equally ludicrous.

Posted by: Econobuzz on November 2, 2007 at 6:50 AM | PERMALINK

And again, since the bulk of these subprime loans were made by mortgage brokers, they were never in danger of getting caught holding the bag as long as they sold the paper, which is the hold(sic) point of being a broker rather than a bank.

The mortgage brokers came after the product was created and the traditional banks had little to do with it. They're regulated and, except for ones like Citi, BofA and Wachovia, banks don't have the distribution channels of the IBs.

The clients wanted yield et voila!, a CDO was created, financial calculations were done (diversifying through volume rather than by type of asset) and now what was needed was a pipeline that consisted of making the loans (mortgage brokers) bundling the loans (IBs or proxies) and selling the strips.

Mortgage brokers are enablers, that's all. So, instead of conventional product, they attracted the "no credit, no problem") customer and off they went.

It's not the mortgage loan that matters, it's the stream of cash from that mortgage that matters. The model that was built to represent that stream of sub-prime loans was analogous to the conventional market but had no real history except loss on default.

Mortgage brokers never hold conventional loans either, that's why they're brokers. The sub-prime product was an expansion of an existing line of business which because of volume attracted all the brokers currently dropping like flies.

The default rate was easy to calculate

Well, that's your opinion but it's not true. Plus,it's only one part of the equation. Those "junk bonds" that Drexel (Mike Milken) sold turned out to be worth pretty close to what he said they would be over time (why do you think Lewis Ranieri is considered a genius today?). It's not just whether the loans default but what the collateral is worth if they do. A change in the velocity (turnover) of the mortgage market can have more of an effect than rates of default and that history, at least over the last 15 years, was that even if the loan defaulted, the house could be quickly sold (less than a year). The sale would reduce the tenor and the yield but not the value.

The reason for the crisis is that the default rate factor was wrong and more importantly, the slower volume of sales changes the repayment calculation and the drop in prices changes the value calculation.

TJM is about the "marked to model" rather than "marked to market" problem of CDOs.

Uh, not even close. The model referenced was the valuation model which was needed to sell the rating agency on the value of the pieces, the different CDOs; the mark to market depends on a calculated value, in this case the value of the CDO after the foundation assumptions are proven to be wrong. There's a crisis because no one really knows what these CDOs are worth. The pricing model is clearly wrong, but there is no other model to take its place. You can't just look at each loan value it and add up the pieces.

This is exactly what occurred with the "junk bond" problem. No one knew what those (corporate) loans were worth until several years passed and if you go back and look, the value was much better than was expected at the time of the crisis.

The reason for today's crisis is that nobody, and I mean nobody, knows what these CDOs are worth because there is no model.

Posted by: TJM on November 2, 2007 at 7:46 AM | PERMALINK

>> the risk levels were underestimated

> You're on quite a hindsight roll,
> here, Cranky, let's see what else you know now.

I wasn't being paid $35 million/year for my superior financial foresight, but strangely enough when both my mortgage broker and Alan Greenspan put great pressure on me to go with a balloon ARM I chose instead to stick with an inferior 20 year fixed. At that time people such as yourself were literally pointing at and ridiculing those who took a conservative approach to financing their living quarters, predicting all sorts of dooms and humiliating financial opportunity costs for those who didn't jump on the bandwagon.

As far as I can see the financial geeks dig themselves in deeper with every post on this thread. They deploy increasingly detailed and sophisticated arguments to justify their position that everything they did was correct and that the severely negative outcome that resulted from their actions was the fault of their customers, the government, stupid people, the stars - anything but their own lack of ability to see what was actually happening. Even though the consequences had been predicted by many disinterested but clearheaded analysts for at least 3 years. Now where else have I seen this pattern...

Cranky

Posted by: Cranky Observer on November 2, 2007 at 8:04 AM | PERMALINK

Ah, yes, the "I'm out of facts, let's argue based on personal decisions" mode. Just because I said you had perfect hindsight, which is true of all of us by the by, don't read what I wrote through your lens of "well, harumph, I could see what was coming and, harumph, I took the more prudent path." Too funny, great pressure? Really? To take a mortgage?

Of course at the same time you fail to point out where I said something wrong (not something self-evident, I do that all the time) but you fail to point out where I said I think this was the greatest product ever. Or, in fact, where I said I approved. That's likely because I didn't. I laid out how the product was created, how it works and why it failed.

You then, having already leaped to detail your vast hindsight, take what I said as a defense of what happened. What I said was, it's just business. There's no grand strategy other than coming up with a way to make money. I understand you think that's all dirty and crude and is, naturlich, done to deliberately crush the poor with the jackboot of capitalism. That's a world view not an analytical tool.

If you think "the Street" set out to crush the poor folk by losing billions, you're crazy not cranky. The traders made money, sure, but if they knew it would cost them billions, I don't think they would have done it. You sem to believe they would.

Just as an aside, I have a 30 year fixed and have done so all 3 times I bought a house. Oddly, I didn't feel any pressure to do that at all. It's too bad more people weren't as conservative (I mean that only in the fiscal sense) as you.

Posted by: TJM on November 2, 2007 at 8:59 AM | PERMALINK

TJM,
The Oxford English Dictionary(tm) called. They want to use your 8:59 as an example in the "irony" entry.

Cranky

Posted by: Cranky Observer on November 2, 2007 at 10:03 AM | PERMALINK

Here's a good explanation:

http://accruedint.blogspot.com/2007/03/how-does-cdo-work.html

Posted by: John on November 2, 2007 at 10:38 AM | PERMALINK

Yeah, thought that's all you had.

Posted by: TJM on November 2, 2007 at 10:41 AM | PERMALINK

My original statement:

>> But in fact the risks were correlated
>> (gee, that was never anticipated: even
>> though it has happened in very [every]
>> housing downturn since the Babylonian times),

TJM's paraphrase of my statement and response to the paraphrase:

> ots of people said the prices couldn't
> keep up the pace, but only a few predicted
> a real problem. What you say is true enough
> but, again, it's 20/20 hindsight. Now if
> you're in Vegas or Phoenix or parts of
> Fla. good luck, but it's not as bad in
> every market. Real estate does go through
> cycles.

TJM's paraphrase is much weaker than my original statement, and ignores entirely the fundamental question of whether these "strips and bundles" actually stripped the risks of the correlation - which seems to me to be a bit of an important question. But even with the weaker formulation TJM admits that my point is correct but fails to answer it in any meaningful way. If you are saying that there are always going to be business cycles and that the eventual crash is unforeseeable that is fine - but you are then essentially admitting econbuzz's point that the modern financial markets require intense heavy-handed outside regulation. And, I would argue, very high rates of personal taxation since the people raking off the huge sums aren't really adding any value to the economy but are just lucky enough to be on the roller-coaster at the right time (or they might even have an incentive to trigger off damage themselves - there's that bezzle again).

But TJM says "that all [I] got". Sure. What financial institution do you work for? I need to know where best not to do business.

Cranky

Posted by: Cranky Observer on November 2, 2007 at 10:58 AM | PERMALINK

Talk about missing the forest for the trees. This whole converstion is about nothing. the real issue isn't miscalculations of risk. These sub-prime loans have beeen made for years and the risk factors are well known. The real story if anybody cares is why are so many people defaulting.Riddle me this. Perhaps the economy for the people that hvae taken out these loans is shaky at best.

Posted by: Gandalf on November 2, 2007 at 11:00 AM | PERMALINK

Econobuzz, I disagree that it was an "honest mistake."

TJM, I think you're on, is on the right path, though using different language than me. CDOs were artifically priced without being placed on the market; I called it "marked to model" vs. "marked to market," rather than your take of "no model." I think what we're both saying, in different ways, is that their pricing was done without real-world input to guide it. The "marked to model" vs. "marked to market" angle is what Fleckenstein and Jubak use on MSN, and, because they're so good at explaining this issue, I'll stick with their interpretation, especially if you still consider yourself to not be on the same page as me.

JeffII, no, you didn't say how many subprime borrowers were first-time buyers vs. non-first timers. Nonetheless, your post implied a degree of sympathy for all subprime buyers that is unwarranted, hence my comment.

That was very willful and not an "honest mistake." I think you're being kind of naive. Some economic tools have more inherent potential to game the system, at least the way it's currently structured, than do others.

Posted by: SocraticGadfly on November 2, 2007 at 11:02 AM | PERMALINK

Gandalf,

If the risk factors were trully well known the expected defaults would have been measured correctly. They weren't.

Posted by: Gabriel on November 2, 2007 at 11:38 AM | PERMALINK
Basically, the whole selling point of a CDO is that its risk-adjusted return is higher than the risk-adjusted return of its underlying parts — i.e., that you've modeled the CDO's diversification or correlation or something better than Moody's has, and it will thus yield a bit more than you'd normally expect based on the CDO's rating.

Won't the risk-adjusted-return of any lump of randomly selected investments each having a non-negligible diversifiable risk virtually always be better than the risk-adjusted return of the underlying parts? Isn't that pretty fundamental to investment theory?

Posted by: cmdicely on November 2, 2007 at 4:51 PM | PERMALINK

Econobuzz, I disagree that it was an "honest mistake."

Posted by: SocraticGadfly

I wrote that to suggest that it was an honest mistake was ludicrous. I will have to be clearer next time. I think we agree about all this.

Posted by: Econobuzz on November 2, 2007 at 6:32 PM | PERMALINK




 

 

Read Jonathan Rowe remembrance and articles
Email Newsletter icon, E-mail Newsletter icon, Email List icon, E-mail List icon Sign up for Free News & Updates

Advertise in WM



buy from Amazon and
support the Monthly