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

March 17, 2008
By: Kevin Drum

IN WHICH I PUBLICLY MAKE THE STATE OF MY IGNORANCE MORE SPECIFIC....Like a lot of people, I've spent a fair amount of time over the past few months trying to understand the subprime debacle: CDOs, SIVs, mezzanine tranches, the housing bubble, diversification correlations, etc. etc. And some of this stuff I've finally figured out. Sort of. But there's one part that I still don't get, a piece of the puzzle where a card always seems to get palmed with no further explanation. For example, here's Jared Bernstein explaining what happened when the housing bubble burst:

Then the loans started to go bad, but because they were squirreled away who knows where, no one knew quite what to do. So the part of the economy that provides credit froze up, and without free-flowing credit, our economy can't expand.

I'm not trying to pick on Jared; his post just happened to be handy. But why is it that "the part of the economy that provides credit froze up"? There always seems to be some hand waving at that point, and aside from bits and pieces that I can't quite put together into a cohesive whole, there's never an explanation of why a meltdown in one particular (admittedly large) sector of the financial industry caused the entire commercial paper market to essentially freeze solid.

So consider this post a public plea for someone to explain that particular part of the puzzle in layman's terms. Why are the credit markets frozen?

And on a related note, I guess I have another question: is it really that hard to figure out the value of all the assets rolled up into the various CDOs and SIVs (the "who knows where" in Jared's post above) that caused this mess in the first place? I know it's not trivial, especially when the market is in freefall, but there are underlying assets behind all these vehicles. The proposition that nobody really knows what any of this stuff is worth just doesn't seem all that plausible. Anyone care to explain this too?

Kevin Drum 1:03 PM Permalink | Trackbacks | Comments (99)

Bookmark and Share

This doesn't really answer your question, but I found Nouriel Roubini's column today on the "shadow financial system" interesting and it may help you out. Got there via Krugman's blog.


Posted by: Chris on March 17, 2008 at 1:06 PM | PERMALINK

Kevin asks: Why are the credit markets frozen?

Counter-party risk.

Or 'Who's next?'

Posted by: MsNThrope on March 17, 2008 at 1:10 PM | PERMALINK

Simply put, the market has tolerated a large amount of FRAUD. They cannot reveal the true nature and scope of this crime without incriminating themselves and further destroying our confidence in Voodoo ecomonics.

Posted by: FredF on March 17, 2008 at 1:11 PM | PERMALINK

A too-simple answer to your second question is that houses are not selling, so the price discovery of housing hasn't happened yet. This is important because, as Calculated Risk has shown, no one knows how much of the country is going to be underwater, which affects the number of defaults. Since no one knows how many defaults there are going to be, no one knows what these securities are worth. It all will become clear when houses start selling again.

Posted by: House Whisperer on March 17, 2008 at 1:12 PM | PERMALINK

Why no one is figuring them out is obvious. The top dogs know enough to know that actually valuing the assets in the "who knows where" would lead to staggering losses even taking into account what is already happening. Thus:

1) They don't want to ever do it, because they are afraid.
2) They know they have to do it, but think spreading out the collapse longterm mitigates their pain more than an all-at-once crash.

Posted by: MNPundit on March 17, 2008 at 1:12 PM | PERMALINK

Simple solution:
Use public funds to bail out the banks.

Then bring back the 90% income tax bracket, make the fuckers who brought home the million dollar bonuses during the mortgage fraud feeding frenzy PAY for the bailouts. (and the war too!)

Posted by: osama_been_forgotten on March 17, 2008 at 1:15 PM | PERMALINK

I think my level of understanding is about where yours is, Kevin, so I probably shouldn't attempt to explain this bit, but what the heck.

The credit markets discovered that the risk levels associated with a bunch of financial instruments were wrong; there was a higher chance of default than they had thought. Without knowing that risk, it's nearly impossible to know the value of those instruments. Without a handle on how to value existing and prospective instruments, the markets are hesitant to use them to extend further credit.

Posted by: Shelby on March 17, 2008 at 1:16 PM | PERMALINK

What OBF said! But how likely is it?

Posted by: thersites on March 17, 2008 at 1:18 PM | PERMALINK

Perhaps the best thing to remember is that the credit market does not pull value out of thin air. Current capital is laid out in the expectation of future appreciation and realizable gains. Why would you continue to loan money if you thought that there was a very real chance that you would lose money? Let somebody else do the loaning. And once this (very realistic) attitude becomes widespread, credit tightens further, markets become more illiquid (i.e. no new cash to make new purchases), current assets (unsold houses, say) fall further in valuation, scaring the credit market further, and the circle keeps going around. End result, things are frozen solid.

Posted by: ppp on March 17, 2008 at 1:22 PM | PERMALINK

MNPundit is right, as part of the explanation... nobody WANTS to find out the "real value" of a lot of this crap.

Beyond that, is the original bundling, and folks like Moody's doing the original card-palming and saying, "We'll rate it all AAA!"

At that time, the Bear Stearns of the world knew much of their shit was worth less than what they were marketing it at; they didn't know then, and still don't, exactly how much less they're worth.

And, because of that, nobody wants to sell them and find out.

Posted by: SocraticGadfly on March 17, 2008 at 1:23 PM | PERMALINK


Let's say for a minute that your house in (Irvine?) was worth $750K in 2004 when you purchased it. The bank happily lent you $600K and you all watched as the value of your home increased to a cool $1M by late 2006. At which time, you did a refi for $800K, taking $200K for "other opportunities."

Today, that house of yours is back to $750K. You are probably underwater and so is your mortgage holder to some degree. But you only have to "book" that loss if you sell. So no one sells. Or if they have to sell, it is probably at a distressed price, increasing the loss to everyone. So the transaction market seizes up (leading to a lack of "price discovery" - thus no, they cannot really tell what things are "worth").

And so does the lending market as obviously no one of going to refi your home for the amount you owe. And the lender is stuck holding your loan because no one will buy it from him at a reasonable price. And this seizingup of the home loan market has spillover into all sorts of other parts of the credit markets, leading to a worrisome state of affairs indeed.

Layer on top of this all sorts of complex derivative transactions that supposedly "hedge" some part of someone's risk - and do so in good times. But in bad times, now you not only have to wonder what your underlying asset may finally fetch if it is indeed ever sold. But you have to wonder whether the other side to your derivatives deals (counterparty) is solvent or will go down like Bear Stearns.

This is pretty much how the Great Depression began and it is why the Fed (in the absence of any help from the Hoover, I mean Bush, Administration) is taking such historic steps, knowing full well they are bailing out some bad actors.

(And a special PS to all the corporation and banking hating readers of yours out there. Grow up.)

Posted by: martin on March 17, 2008 at 1:27 PM | PERMALINK

One last point. This crisis is entirely due to government policy, and is not just "one of those cyclical things" that happen from time to time. Some of the talking heads seem (for good reasons) to want to make this into just some bad storm that has knocked us around in a completely random and hence blameless way.

You can burst a bubble before it gets too huge, but you need to have a reality-based government in order for those tough-love measures to be implemented.

Posted by: ppp on March 17, 2008 at 1:27 PM | PERMALINK

Kevin, my understanding is that it comes down to leverage. Places like Bear Sterns were leveraged, say, 20:1. THis was great when everything was going up. Big profits! But if things drop 5% your 20:1 leverage is wiped out. As soon as the subprime crisis hit the banks and IBs realized that they had been wrong about the risks associated with their nice AAA bonds. And you only have to be a tiny bit wrong to go from big profits to huge losses. Do you want to loan money to BS knowing that they might have no assets to pay you back? They might be in the whole so bad that bankruptcy will end you chance of getting paid back?

Posted by: EmmaAnne on March 17, 2008 at 1:29 PM | PERMALINK

Kevin, I know you have a tech. background, so let's draw a tech bubble analogy. When the bubble burst in 2000/2001, large parts of the technology funding infrastructure "just froze up". This included VCs (who stopped investing) and investors/ibanks/IPOs (which stopped happening). There were plenty of great companies out there -- Google and Salesforce.com were both in existence, for example -- that simply had to wait and bide their time because it was impossible for them to efficiently and effectively demonstrate their value to a skeptical market. There were other potentially great companies that got wrecked by the freeze. And there was a bunch of tragically laughable crap.coms that got rightfully swept away. Think about how, if you were the Lord High Inspector of Tech Companies, you would have gone about evaluating and assessing what was worthwhile and what was not, if you'd suddenly been appointed to run a RTC-like entity for the whole tech sector. Think of the complexity and specificity of every particular company, its market, its patents, its code, its people, etc. A total nightmare - so basically we just let the market run and let companies die or live on the basis of their own capabilities.

The housing market is at least that complicated. What neighborhood, what condition (termites?) what are local salaries and local market conditions, what are the terms of the loan, is the income stable, etc., etc., are all (in the aggregate) at least as complicated as figuring out tech. companies would have been. While each individual transaction may be simpler, for tech there were perhaps two thousand discrete entities to assess, and this time, there are perhaps twenty million discrete entities to assess.

CP (commercial paper) is lent to and by people like GMAC, who sit on huge piles of mortgages and car loans; I used to work in GMAC's CP division in a past life, when they were moving between $1 billion and $2 billion per day. That velocity of capital was entirely driven by the perception that, when Citibank parked $150 million with GMAC for eleven days (I saw actual transactions like that all the time) in order to make a couple hundred thousand in interest, their $150 million was basically risk-free. As soon as it became possible that there were even a few risky loans in the giant pile of loans that is GMAC, things started to go south; and as significant potential risk emerged, suddenly the prospect of a few hundred thousand in interest vs. the risk of $150 million or more became silly. So it... just... stopped.

The liquidity crisis is entirely driven by the solvency crisis - I won't lend you money if I don't trust the value or stability of you and your assets. The analogy to what happened in tech in 2000/2001 is pretty exact... except this time the scale is probably 20X, and there is no further asset bubble to bail us all out.

Posted by: Ethan Stock on March 17, 2008 at 1:32 PM | PERMALINK

Shelby is basically correct but it's my understanding that CDOs and SIVs are main culprit here. These new financial instruments were supposed to reduce the risk but to do it they spread the risk around and suddenly assets that should be perfectly safe are now at risk thanks to being packaged with these bad mortgages. Also because the risk was supposedly much less (and no doubt plain old greed as well) the leveraging of assets has been obscene. I read somewhere last night that the average leveraging by hedge funds was around 14 to 1. It doesn't take much or problem for this leveraging to have a huge impact on a compaanies balance sheet and thus make lenders unwilling to lend any further.

Posted by: Jay on March 17, 2008 at 1:33 PM | PERMALINK

In lay terms, over simplified but tolerably close to the truth.

First, you have to remember that to a lender, a loan is an investment, an asset. Many of the entities making or buying these groups of loans used a great deal of leverage, putting up only two or three percent of their own capital and borrowing the rest.

Unfortunately, many of the loans were far more risky than the market originally thought they would be, and the value of these investments (to the extent that the value can even be calculated) went far below the limited capital the owners of the mortgage had invested in them, driving these owners into insolvency.

Not only the investors are hurt. The entities that financed these investments (who may also be investors in these loans themselves) are also underwater on many of these loans, and have to account for their anticipated losses by charges that reduce their capital.

Lenders can only make loans up to a limit set by their capital, so this loss of lender capital, and perhaps even more the loss of confidence in knowing if other lenders are solvent (ie - who will be the next Bear Sterns) has mean that lenders are increasingly unwilling to make the normal kinds of lending transactions that are essential to a functioning economy.

This is a stick figure version of part of what's going on, of course. Still, I hope it gives you a feel for the situation.

Posted by: Tom O on March 17, 2008 at 1:35 PM | PERMALINK

Most of the above explanations are headed in the right direction, when pointing to counterparty risk, but they also leave out a key component. The modern financial system is entirely dependent on short term financing - overnight (and other very near-term) lending between banks & non-bank financial institutions (hedge funds, SIVs, etc.). To gain access to this short term credit is relatively easy as long as you have some decent assets to post as collateral.

As soon as the markets started to decline for the first CDOs & other structured finance products, the whole class of securities became suspect. These things were so engineered & the actual state of the underlying loans so shaky, people became afraid to touch them. Once that happened, not only did other financial institutions not wish to purchase them outright, they became afraid of taking them as collateral as well (if you've already got exposure to this toxic stuff on your own books, why take on exposure to someone else's).

It's at this point that credit markets really began to freeze up, as assets once considered safe - i.e. AAA MBS/CDOs/ABS - and thus promptly accepted as collateral, could no longer be used as such. Suddenly, everyone was requiring higher quality assets or cash/cash-equivalents to lend to one another, even for very short term periods. No one really has enough other assets that they can pledge, and so short-term financing/credit markets seize up. Without this short-term financing, the basic business model on which all these non-bank institutions are founded, is undermined and they can no longer function.

Posted by: Tosh on March 17, 2008 at 1:36 PM | PERMALINK

IIRC, Bear Stearns was leveraged at 31:1.

But hey, that's no reason not to extend the friendly hand of taxpayer-funded corporate socialism to their buyers. After all, what's $30 billion, more or less, among political and financial titans?

Now, health care for the poor, well, that's another question. They need incentives to pull themselves up by their own bootstraps, and we can't be throwing good money after bad just to make life easier for a bunch of lazy bums.

Posted by: bleh on March 17, 2008 at 1:36 PM | PERMALINK

You can burst a bubble before it gets too huge, but you need to have a reality-based government in order for those tough-love measures to be implemented.

This time around, not only did the Fed not take away the punch bowl when the party got started, they brought in extra punch.

Posted by: Doc at the Radar Station on March 17, 2008 at 1:36 PM | PERMALINK

I made a comment on this at the end of Kevin's Bear Stearns post earlier today. When something doesn't make sense, it is usually because someone is doing something which we don't know about.

The someone is a collection of large financial institutions with low questionable loan exposure. The something is that they are creating a crisis by refusing to lend although they could. The reason for this is that they intend to acquire distressed assets at pennies on the dollar. The method is to first expand credit greatly; then to reduce their own exposure; and finally to reduce credit to create a crisis.

Perhaps I'm wrong but it is the best explanation I've heard so far and the only one that fits completely.

Posted by: margrave on March 17, 2008 at 1:40 PM | PERMALINK

35 trillion in credit derivatives as of early 2007--if they fail in the same proportion as housing mortgages, to which they're linked--every 3% loss would represent a loss of a trillion dollars.

Does that make you feel better?

Posted by: Neal on March 17, 2008 at 1:40 PM | PERMALINK

From Paul Craig Robersts over at Counterpunch last week:

"Reforms often do more harm than good. This is currently the case with the “mark-to-market” rule, which is imploding the US financial system by requiring financial institutions to value subprime mortgages at their current market values.

"This makes a big problem for balance sheets. These financial instruments became troubled prior to a market being established for them, as they were marketed direct from issuers to investors. Now that they are troubled and with their true values unknown, no one wants them. Their lack of liquidity assigns them a low value.

"The result is tremendous pressure on balance sheets. The plummeting value of subprime derivatives is pushing institutions that own them into insolvency, destroying their own stock values and forcing the financial institutions to sell untroubled liquid assets, thus resulting in an overall decline in the stock market.

"The solution is to suspend the mark-to-market rule. Instead, allow financial institutions to keep the troubled instruments at book value, or 85-90% of book value, until a market forms that can sort out values, and allow financial institutions to write down the subprime mortgages and other troubled instruments over time."

Posted by: Brojo on March 17, 2008 at 1:57 PM | PERMALINK

Remember that at the national economy level, "providing credit" means "buying debt obligations." Big Shitpile was so prevalent throughout CDOs that when things went south, no one wanted to touch anything that even remotely might contain a piece of it. So they stopped buying MBS securities, CDOs that contained any MBS component, and to a large extent, debt obligations issued by companies that might be exposed to MBS in any way (and this swept in most insurers, pension plans, I-banks, commercial banks, etc.).

And to make a long story short, a world where these sorts of debt obligations cannot be sold is a world where "credit has dried up."

Posted by: Joe on March 17, 2008 at 1:59 PM | PERMALINK

The passage Kevin Drum cites is indeed "hand waving"; it seems that Jared Bernstein was sketchy on some other aspects of the story.

1.When mortgages are securitized, the objective was to bundle comparable risks together as investment instruments. The problem is, the investment bankers who designed these instruments assumed that the risk of default is normally distributed; and while a single EXTREMELY low grade mortgage has, perhaps, a 10% risk of default in any single year of its lifetime, they supposed that this risk was distributed randomly throughout each bundle of securitized mortgage obligations (SMO's). It wasn't. Rather than offsetting each other, default risk either rises or falls as a group.

2. While these securities were initially regarded as excellent revenue streams, there is now a general recognition that not systematic method exists for calculating the real value of them. The junkiest of junk bonds does not have a unique, stable probability of default; and when portfolios of many institutions carry these instruments, the risk actually spreads to higher-grade investment instruments as well, since their value may well be contingent on the value of underlying junk bonds (or SMO's).

Hence, the real mystery is not "where" (i.e., in whose portfolios) the illiquid assets are, but where the contingent risk is. That's something that would be hard enough to assess even if an objective formula existed for estimating the new probability-weighted risk of default for each item. No such formula exists because everything depends on each other, and probability implies a random, uncorrelated event--like proton decay in a mass of uranium. This is more like the odds of a bank run spreading to other banks, or predicting the path of a forest fire.

3. An odd thing that economists have largely faild to discuss (other than Roubini & Setser, linked above) is the role of the nation's immense current account deficit (about $180 billion per quarter, or $700-730 billion per year); this is "offset" by a current account deficit that oscillates widely around an annualized mean of about $650-660 billion.

So the flow-through of assets is subject to an analogous risk linkage. If the demand for US assets declines, the dollar declines in value, which triggers a further decline both in foreign investment (i.e., demand for US assets) and intervention by the Fed to reduce interest rates, thereby reducing the income potential from dollar-denominated assets... leading to a further flight from the dollar to something like euro or yen.

Posted by: James R MacLean on March 17, 2008 at 2:00 PM | PERMALINK

Al Quida has already won! No Christian, Jew or Agnostic will want to have anything to do with compound interest when this depression is over. Just like fundamentalist Moslems.

Posted by: slanted tom on March 17, 2008 at 2:06 PM | PERMALINK

I'm confused by a slightly different aspect of this. The reason mortgage-based securities are/were considered safe is because there is a real asset, a house (or something similar) behind them. If you make a $1 million loan to a bad borrower, you can just grab the $1 million house and come out even. Now the houses may be worth a little less than the note, but they are still there and still worth a good percentage of that value. They haven't disintegrated or anything. There's a loss, but not a "fire sale for $2 per share" loss. So blaming the whole thing on housing prices seems a bit off. What am I missing?

Posted by: ArkPanda on March 17, 2008 at 2:13 PM | PERMALINK

First, the market hasn't exactly frozen, but things have certainly tightened, and we're in the perverse position where interest rates on assets (deposits, bonds, etc.) are falling, while interest rates on loans are increasing. But one thing is certain; the total amount of available credit is falling.

One of the major reasons for this is that the secondary market for mortgages (mortgage backed securities) has pretty much vanished. Blame Moody's, Standard & Poors, blame crooked mortgage brokers and appraisers, blame greedy homeowners. The fact is, no person in their right mind would buy this junk (please note the Fed is now buying it).

As to figuring out the value of these assets, the problems are manifold. First, in the case of many of these mortgages, (especially refi's) the collateral was overvalued to begin with; in some cases by a factor of 2 - 3X. Second, contrary to what everyone in the media seems to assume, the real estate market is LOCAL. There is no factor you could apply to adjust for market conditions. The housing market varies from neighborhood to neighborhood, not city to city or state to state. Third, these things are such a mess that it's not altogether clear who owns what (or what part of what). And Fourth, many if not most of these securities are in relative terms, too small to bother with individually. Given the high cost of actually doing the due diligence, relative to the value of the assets, it simply hasn't (and won't be) done.

Posted by: Dave Brown on March 17, 2008 at 2:16 PM | PERMALINK

You've probably gotten more explanation than you want, but here's my two cents: the reason it's hard to value the assets behind the CDOs is that they're payment streams - they represent the present value of future mortgage payments, payment obligations that were supposedly bundled together so in the AAA rated ones, the risk of nonpayment of any particular mortgage in the pool would be hedged by the presence of other mortgages with a different risk profile (if this sounds impossible, guess what, it probably is).

Now that the people who took out these mortgages can no longer expect their houses to be worth more than they paid (i.e., the housing buble burst), no-one knows if the mortgage holders are going to keep paying on these mortgages on houses they can't sell and maybe can't afford to keep.

The fact that a lot of these mortgages are going to reset from low teaser rates to 9% or more doesn't make it more likely that the mortgage holders will keep paying -- and the fact that you probably don't know which CDO holds what percentage of these mortgages doesn't help you to value the CDOs either.

And if you've been in the business of trading these CDOs to someone else, why, there's a certain temptation to see if you can't manage to leave them with that somebody else, while you keep the full-value price that you received for them back in July. Because from your point of view that would kind of solve your problem of trying to figure out if those people who took out more mortgage that they could afford will still pay....but the other side doesn't want to be stuck with them either. And so the market dries up.

This is a very simplified version, but I think this is what's going on.

Posted by: Diana on March 17, 2008 at 2:18 PM | PERMALINK

The solution is to suspend the mark-to-market rule. Instead, allow financial institutions to keep the troubled instruments at book value, or 85-90% of book value, until a market forms that can sort out values, and allow financial institutions to write down the subprime mortgages and other troubled instruments over time.

All this will do is keep companies from going under. It will do nothing to unfreeze credit. In fact, it will just prolong the credit freeze and make it worse. It did not work for Japan and it sure as hell will not work for us.

Posted by: on March 17, 2008 at 2:19 PM | PERMALINK

Basically, once all these crap loans were bundled into anything with a AAA rating, they became hot potatoes.

Hot Potato
Don't pick it up, pick it up, pick it up
Pass it on, pass it on, pass it on
You don't want to have it
when the big one comes.

Douglas Adams lives.

Posted by: mroberts on March 17, 2008 at 2:28 PM | PERMALINK

"Now the houses may be worth a little less than the note, but they are still there and still worth a good percentage of that value. They haven't disintegrated or anything. There's a loss, but not a "fire sale for $2 per share" loss. So blaming the whole thing on housing prices seems a bit off. What am I missing?"

Bear is essentially worthless because it got a piece of the upside, not the whole pot. Basically, it was packaging $108 in mortgage payments, selling it out for $100 at 6% interest, and pocketing $2 itself. When the mortgage payments really turn out to be $85, then the investors lose a crapload of money, and Bear is essentially insolvent.

You are right that the MBS are not valueless because there is, after all, a home there. However, it has become almost impossible to ascertain what that value is. Say a Miami condo was bought in 2005 for $1 million. Its value now is probably between $700,000 and $900,000, but that won't even be determined for six months to two years until the property is foreclosed upon and sold at auction. Plus there's an uncertainly in the costs of doing that. So essentially, you have an asset that may produce between $500,000 and $850,000 in nominal cash at some point in time over the next few years. How on earth do you value that? Multiply that by ten thousand, and you get some sense of why no one will touch a CDO right now.

Posted by: Joe on March 17, 2008 at 2:29 PM | PERMALINK

The Milgrom-Stokey theorem says that two rational risk averse agents will never make a purely speculative bet if each takes into account the other's willingness to bet when assessing the probability of winning. I might think it was extremely unlikely that Britney Spears would shave her head in public, but if you walked into a bar and offered to bet $50 that it had happened, I should decline. Similarly, if I'm considering buying some of your mezzanine tranche crud, I should suspect that the particular portion you're willing to sell is especially stinky.

Financial markets can overcome this if the risk is small and there are real gains from trade. The idea of tranching is to take a really complicated asset and divide it into a large safe part that can be held by many of the players with big money to invest, plus toxic waste handled by experts wearing lab coats and latex gloves. It's a nice concept, and I don't know anyone who claims to have a better way to finance home lending, but what people apparently didn't think about enough is that changes in the housing market and the macroeconomy can render "safe" stuff either risky or just worthless.

Posted by: Andy McLennan on March 17, 2008 at 2:40 PM | PERMALINK

For the record, my original post was just shorthand for many of the very thoughtful posts above.

A few other points (probably made already) re Kevin's query:
It's all about transparency and confidene at this point: if you don't know how exposed a borrower is, you're not going to lend to them. The nature of the securitization is that lenders stopped believing in the bond ratings, the valuations, the health of the balance sheets, etc...BS is the classic EG.

It's like if you and your friends wanted to go play ball on a local field. On the way, you pass a bunch of people with terrible poison ivy who were just playing there. They tell you, 'there's a patch of poison ivy somewhere in that big field.'

You don't know where, and you're spooked by the possibility of getting infected, so you go home and watch TV instead. I have a feeling Kevin might view this as an over-reaction, and it is, but that's what I meant when I said "Capitalism often overshoots"--both going up and going down.

Why don't we know precisely where the infected tranches (ie, the poison ivy) are? Basically, because most loans will be fine, and you never learn about the bad ones until they default. This is the long and arduous period of 'price discovery' that folks wrote about above.

That happens to take longer in a housing bubble burst than when the dot.com bubble burst. In that case, it was quick "mark-to-market"--you learned very quickly that your tech assets just flopped. But with home loans, it can take a long time for lenders to accept the reality of the situation--the fact that they're going to have to eat big losses.

Posted by: Jared Bernstein on March 17, 2008 at 2:44 PM | PERMALINK

Well; the real disaster isn't the loss of value of people's homes.

The real disaster is the loss of value of the American Dollar.

Those few who have hundreds of millions of them socked away in numbered accounts in barbados will be okay.

The rest of us?

We'll be lucky to keep our jobs; let alone demanding a raise so that we can maintain some semblence of a standard of living.

The lucky ones?:
The ones who got fed up with Bush and moved to Canada, and are now getting paid in Canadian dollars.

Posted by: osama_been_forgotten on March 17, 2008 at 2:49 PM | PERMALINK

Something like this just can't happen over night. Where and when would the first signs of this trouble have appeared in the housing/mortgage markets, and who would have seen them? When should one of our government employees have noticed a problem and taken action?

Posted by: ferd on March 17, 2008 at 2:51 PM | PERMALINK

To use yet another analogy - let's talk about cars.

I believe we all know that the "value" of a brand new vehicle drops suddenly the moment it's driven off the dealer lot. For this reason, many people don't ever buy brand new cars. Many still do, reasoning that over the life of the car, the value of their use of the vehicle will exceed that lost when they drive it off the lot.
Some folks buy 2 year old cars because they're almost new, and cost significantly less because they aren't NEW new.

If you know that the housing market is still declining, why buy now? You're just going to be paying more than it will be worth in 3, 6, 12, 36 months. And you'll wind up "upside-down" on your mortgage - unless of course, you can afford to pay cash. In which case you'd be rational to look at it as a negative ROI purchase - it may yet be worth it to you, if you like the house enough and plan to be in it long after the market stabilizes and prices return to where they were when you bought it.

That's an individual, though. And a single property.
Institutions aren't looking at selling or buying single loans - they're holding paper that bundles the mortgages(of discrete value) on hundreds, thousands or even more homes. Huge bucks.
And they know, KNOW, that the value of those homes is declining. And they likewise know that some percentage of those mortgages will be defaulted.
They also know that the rate of default is increasing, while the value of the properties is decreasing. And they don't know when that's going to stop.

Posted by: kenga on March 17, 2008 at 3:00 PM | PERMALINK

My understanding of this whole mess has been that a lot of supposedly smart people let themselves be gulled into believing in alchemy. Some people figured that if you mixed some really risky paper in with enough AAA paper you would be able to keep the AAA qualities in spite of the increased risk. Now, as the risky stuff is imploding, it is so thoroughly mixed in with AAA stuff that nobody really knows what risk is involved with any mortgage backed securities. The way it is spreading is, because nobody really knows how much exposure any particular institution has to this risky stuff, nobody is willing to extend credit in other areas. That is why Bear Stearns is such a BFD, the smart boys on Wall Street are crapping their pants because nobody has any confidence about what anything is worth today. I mean did Bear Stearns really F up so badly with the paper they are holding that their share price should be devalued from about $158/share in April of 2007 to $2/share in March of 2008? If the answer is no, JP Morgan will be riding high (and JP shares are on the rise), but if the answer is yes, the Bear Stearns buyout could drag JP Morgan down with it. Frightening to think about, no?

I don't think there is anyone in their right mind who would even entertain the thought that $2/share for Bear Stearns wasn't a completely risk free price to pay, but then nobody in their right mind was really considering the possibility, last year at this time, that we would ever face a liquidity crisis like the one we are now facing. In his blog, Paul Krugman described the disconnect:

What’s going on now, however, is beyond that: the “financial accelerator,” with deleveraging causing a credit crunch that forces further deleveraging, and now threatens to produce a sort of pancake collapse of the whole system, was not, I think, so widely foreseen. Certainly Nouriel Roubini was on the right track; I can claim to have described something quite a lot like what’s now happening, though I covered myself by making it a slightly jokey scenario rather than a straight prediction. But in Rubin’s defense, I don’t think many people saw how much the system itself would break down.

Posted by: majun on March 17, 2008 at 3:05 PM | PERMALINK

Perhaps Mr. Drum's ignorance, and the ignorance of most Americans, is the fault of the media we rely on for information.

The New York Times Company has struck a deal with a pair of hedge funds that want to shake up the company, giving the funds two seats on the board in order to avoid a proxy fight, the two sides announced Monday.

Posted by: Brojo on March 17, 2008 at 3:06 PM | PERMALINK

One other factor in drying up of credit: all these securities that can't be valued are assets that banks and entities used to make loans against. If they're banks, they have limits on what they can lend against their assets/capital. When the value of their capital shrinks, they have to call in loans and/or curtail issuing new ones. That adds up to credit shrinkage.

More than anything, this episode looks like the old 19th-century panics in the way it's playing out. 1837, 1857, 1893, somebody on the teevee mentioned 1907. Whatever sparked a run, the sequences of events were pretty much like this one-- credit contraction.

Posted by: Altoid on March 17, 2008 at 3:10 PM | PERMALINK

Al Quida has already won! No Christian, Jew or Agnostic will want to have anything to do with compound interest when this depression is over.

That's not Al Quida, that's Al Gebra

Posted by: optical weenie on March 17, 2008 at 3:17 PM | PERMALINK

Brojo: Roberts' idealistic solution won't work when people are actually making calls, unless you want the SEC to freeze a bunch of trading.

Martin: Get some ethics, you wanker, instead of calling for people who have reason to hate aspects of our current cheating economic system to "grow up."

Andy McLennon: Milgrom-Stokey? The market isn't rational. Even Greenspan, in his Financial Times interview, admitted there's room for, and need to be cognizant of, behavioral economics.

Kenga, et al: People who have certain tranches of MBS instruments, what are they going to do, take possession of fractiional portions of houses? These are houses, not mineral rights. And, none of these people want to sit on houses for five years, either.

We have not yet begun to see the clusterfuck.

Posted by: SocraticGadfly on March 17, 2008 at 3:19 PM | PERMALINK

Inflation/dollar in the crapper have already done their work.

The U.S. is no longer the world's largest economy; the Eurozone is:


And, if Big Ben is serious about a full-point rate cut, the Eurozone's going to be ahead of us for a long, long time.

Posted by: SocraticGadfly on March 17, 2008 at 3:21 PM | PERMALINK

I'm not near as smart as Kevin or the rest of you folks and after reading all of the above posts I'm not sure I understand anything at all. What does become clear is that we seem to be in real trouble but what's the bottom line? In other words, from a "local" standpoint what does this mean for your average schmuck who has a home they can afford, pays their mortgage and other obligations on time and is in most circumstances in good shape?

Posted by: Ciabnw on March 17, 2008 at 3:24 PM | PERMALINK

Ciabnw. Read just above your post, at mine.

It means inflation, inflation, inflation.

Or, more and more likely, stagflation, stagflation, stagflation.

That said, for other shakey investment banks, fewer sugar daddies are out there. Sovereign wealth funds have already indicated they don't want to make more unsecured investments, and Ben Bernanke would be publicly crucified on Pennsylvania Avenue if he extended guarantees to them.

Posted by: SocraticGadfly on March 17, 2008 at 3:36 PM | PERMALINK


It means that you and I are going to be paying for this for a long time, both in higher taxes, a weaker dollar (due to increased national debt) and higher credit and insurance costs.

Posted by: Dave Brown on March 17, 2008 at 3:39 PM | PERMALINK
The solution is to suspend the mark-to-market rule. Instead, allow financial institutions to keep the troubled instruments at book value, or 85-90% of book value, until a market forms that can sort out values, and allow financial institutions to write down the subprime mortgages and other troubled instruments over time.

All this will do is keep companies from going under. It will do nothing to unfreeze credit. In fact, it will just prolong the credit freeze and make it worse. It did not work for Japan and it sure as hell will not work for us.

The latter paragraph seems right. The credit market will freeze not because assets are worthless but because lenders are unable to value the collateral behind prospective loans. If the ability of a borrower to pay and the value of assets to be seized in the event of default can be assessed accurately, then lenders will be found. So long as uncertainty reigns supreme in the marketplace lenders will be absent. You only would want to eliminate the mark-to-market if you hope to lure some suckers into the marketplace; that pool is shrinking, however.

Posted by: rk on March 17, 2008 at 3:39 PM | PERMALINK

And a special PS to all the corporation and banking hating readers of yours out there. Grow up.

Let's see: a bunch of corporations did bank-type things but existed outside the normal system of regulation that bank-type financial institutions are subject to.

In exchange for operating under those regulations, the Fed does all sorts of things to back you up and bail you out in times of trouble.

Only now, the nonbank banks want to have it both ways: they've been immune from regulation, and it's gotten them in a world of hurt. So now they're asking for the Fed to bail them out - and the Fed's doing it.

And we're told that the mess they made isn't just their problem, but our problem too.

But the people who made money on the upswing aren't being asked to give it back, nor is anyone being asked to give up their golden parachutes for when the next wave of bad news hits. They keep getting rich, even as we pay.

What's not to hate?

Posted by: low-tech cyclist on March 17, 2008 at 3:51 PM | PERMALINK

Don't focus on the mortgage part of this problem -- we were always assured that that portion was relatively small and managable. The problem is all the other complex derivative stuff out there that the big boys were trading back and forth on a daily basis to generate outsize commissions. Turns out that one day after the subprime stuff got exposed, the number of people willing to trade for that other stuff started declining, then -- like musical chairs -- someone is left without a chair and is out of the game. Bear is just the latest and biggest loser to date. Why do we care -- because the banking system lent odles of money to the players to participate in the game. Now that the players are losing, the banking system is at risk. Greenspan still thinks that regulating hedge funds and their fellow travelers is counterproductive. Of course, he also supports baling them out to save the system. Nice game -- too bad we all can't play.

Posted by: steve on March 17, 2008 at 4:05 PM | PERMALINK


S = The Sales Value of a Mortgage (Expected present value of a single mortgage note)

V = (The discounted present value of all future payments on the note)

P = (percentage of future payments likely to be made on the note, modified by perception of riskiness (lower) or safety (higher).

Determining V is straight bond math. Determining P is for what you pay a loan officer and why we have credit bureaus and appraisers.

There also is a term in there for the expected liquidation value of the collateral at time of default, minus the costs of foreclosure and recovery. This is partial recovery of capital - we've lost some of the money we've lent if this happens.

The valuation of a mortgage note, therefore, is a highly individual, requiring much due diligence on the part of lenders and strong assurances on the part of the borrowers that the terms of the note will be fulfilled.

Further, the value of a single mortgage is much more subject to chance variations in localities and in people - regional or municipal downturns may reduce the ability of a borrower to make payments or a note liquidator to recover outstanding capital from a default.

Before the securitization of mortgages after WWII, this meant that mortgages were short-term and expensive and small. Short-term, to allow for volatility in the environment; expensive, because of the perceived riskiness of the instrument; small, because the market for more expensive properties is necessarily less liquid than that for cheaper properties. The market for mortgages was also much less liquid, because of the informational costs of managing diverse and highly individual and small instruments.

Then we have securitization of mortgages. This is where many thousands of mortgages are sold to a single entity and repackaged into a collateralized debt obligation, whose notes are further resold.

In other words, the lender who contracts the mortgage (buys the mortgage from the borrower) resells that mortgage to a third party for a discount. The third party combines that mortgage with thousands of other mortgages in a pool or mortgages. The stream of payments into that pool (from the original borrowers) becomes a salable asset, because the variability in the expected return on mortgages is much reduced when there are thousands of them. Reduced variability means more reliability. Reliable, predictable streams of income are very salable. Therefore, this stream of payments is sold again, in portions, as a collateralized debt obligation, to other investors.

This innovation has brought down borrowing costs and enabled borrowers to borrow more, achieving higher levels of leverage with the same level of assured safety as before.

The packaging of CDO's is highly complex. The principal and the interest stream can be split up and sold separately. Very risky and very safe securities can be combined to make medium risk.
Very risky and very safe securities can be sold separately for risk discounts and premiums. Practically, the ways that these obligations repay the lenders are unlimited - they can structure the payments any damn way they like.

This can serve to improve the predictability of the behavior of the underlying asset, or to obscure the likely potential future behavior of the asset. Buyers like predictability, and traders like volatility.

When the analysis is used for the forces of good, the predictability of the instrument is high, and its value is much less volatile. Shit is shit, and is treated with punitive interest rates and short maturities and onerous collateral requirements; and high quality is rewarded accordingly.

When analysis is used for the forces of evil, CDOs can be structured for maximum opacity - opacity superficially relieved by astute and energetic boosterism. A CDO may be structured so that a pool of extremely risky mortgages are split into different tranches (slices, with each slice containing a different part of the income stream). If the mortgages are divided into principal and interest payments, which are then recombined in varying proportion, an aggregately risky asset (the crappy mortgages), can divided into safer and riskier instruments. Principal repayments are generally more likely, because of the mechanism of default liquidation, so instruments that specify a high proportion of payments from the principal are relatively safer, thereby earning a higher bond rating. Interest repayments from risky mortgages are less likely, so bonds representing interest repayments will be correspondingly lower-rated, and bear, in compensation, higher rates of return.

Over the past 30 years this worked because the information going into mortgage origination was more reliable than not.

Now, the bubble. The social folkways of the mortgage security industry could not deal with two things:

1) the increased demand for mortgages in the face of rapidly rising asset values;


2) the moral hazard on the part of mortgage originators and bond-rating agencies to honestly represent the truth about their crap.

These combined to create an exquisitely bad effect.

There were too many borrowers and not enough loan officers, so loan officers were replaced by salesmen, computer programs and autistic guidelines that proved able to pump out greater volumes of mortgages with reduced aggregate informational reliability. These mortgages, represented as having historically expected reliability, were resold and repacked into CDOs with all of their purported reliable structuring, which were resold to suckers who were happy to get 4.5% when the federal funds rate sat at 1.5%.

The behavior of serially dumping misrepresented crap was pretty costless for four years, from the beginning of 2003 through the end of 2006, so mortgage originators took a tremendous and undeserved windfall when they sold mortgages for value premia related to true aggregate discount value (they fraudulently stole money from their hoodwinked CDO customers).

When more than a small proportion of individual mortgages in the aggregate pool of mortgages in a CDO behave at variance with expectation, the CDO behaves at variance with expectation. The revelation of insolvency naturally resolves at the riskier end of the spectrum, with the riskier interest-heavy or interest-only tranches of CDOs going insolvent due to higher levels of loan default and lower overall income, of which a higher proportion is principal repayment due to asset liquidation.

In many cases, during the early part of the boom, the rising prices masked the riskiness of marginal or subprime mortgages. A mortgaged property in an inflating market could be sold easily for more than its debt burden, so the extension of mortgages to marginal or subprime borrowers seemed relatively riskless. The retention by financial service companies and originators of CDOs of the securities representing these types of mortgages became a no brainer. A bank like Bear Stearns could retain the riskiest portion of its CDOs on book, purchased at a notional discount, and show extremely high profit growth as those "risky" instruments paid off in the short-term (tracking price increases in the broader real-estate bubble).

Revulsion against these instruments began in earnest at the start of 2007, and by last August, when Bear Stearns took that enormous write-down, risky mortgages, and the CDOs that were made of them, were unsalable. Further, they weren't realizing the income that they promised, so financial service companies that held this crap were caught in a liquidity squeeze.

Then the cash conveyor that funded the mortgages broke down, and the irrational and fraudulent expectations that undergirded the CDO industry were renounced, and vast numbers of discrete, individual mortgages began behaving at great variance with initial representation. This variatibility is highly unpredictable, and is at least partially dependent on the liquidity of the regional housing markets, which depends on easily available credit. The mortgage collateral for CDOs has become highly suspect, so CDOs are unsaleable unless at high premia. Likewise for new virgin mortgages. The purchasing agents for mortgages are starved of working capital, so they can't buy more mortgages, so mortgage originators create fewer numbers of new mortgages in much reduced volume, and thus a vicious cycle of revulsion, default, and illiquidity.

while a trillion or more dollars "worth" of crap still floated around.

Posted by: The New York City Math Teacher on March 17, 2008 at 4:24 PM | PERMALINK


In other words, from a "local" standpoint what does this mean for your average schmuck who has a home they can afford, pays their mortgage and other obligations on time and is in most circumstances in good shape?

If you have a marketable skill and you have no medical problems your standard of living will drop but you should be OK.

The price for gas will rise dramatically. That will push the price of other things up too. At best you'll have flat wages, at worst you'll lose your job and have to take a lower paying one if you can find one.

Buy a locking gas cap. Plant a garden. Own a gun. Start doing more 'real' work and you'll be OK.

Posted by: Tripp on March 17, 2008 at 4:29 PM | PERMALINK

The New York City Math Teacher,

Excellent. You should post here more often. And get a shorter handle - one that is easier to type (grin).

Posted by: Tripp on March 17, 2008 at 4:38 PM | PERMALINK

This comment thread is a remarkable mish-mash of knowledgeable and insightful analysis and bat sh*t crazy rantings. Never seen anything quite like it.

Posted by: Poster on March 17, 2008 at 4:57 PM | PERMALINK

I'm not an expert, but I think I've sussed out the answer to Kevin's first question, which is: "why [does] a meltdown in one particular (admittedly large) sector of the financial industry caused the entire commercial paper market to essentially freeze solid[?]" I'll try to express that answer in the vernacular:

The answer is that the lenders are also borrowers; they are (often) not lending their own money. They are lending somebody else's money, which they've borrowed using their own assets as collateral. Why? Because their own cash has long been leant to somebody else (nobody wants to hold cash, because it just sits there, rather than returning interest). The assets that serve as collateral are the loans they've already made; in theory, they can collect the money that other people owe them to pay off the loans they take out, or they can transfer the obligation to pay those loans to the people they borrowed money from. This sounds a bit insane, but it makes sense when you're talking about billions of dollars, because at that enormous volume a few decimal points of interest still adds up to some serious cash.

Okay, so that system has been working fine for a long time, since no lenders have had any trouble borrowing money from other lenders. However, when the value of the "asset" (like a mortgage-backed derivative) a lender wants to use as collateral on a loan becomes notoriously uncertain, it can no longer be used as collateral. No other lender will trust it, so they won't lend any money if that's how the loan would be secured. As a result, the lender owning the "asset" now has no way to turn that asset into cash that they can then lend to someone else. All the lenders fear finding themselves in a position where somebody else asks them to pay up, and they can't. I think that's a "liquidity crisis." That's why uncertainties in the value of mortgage-backed assets cause the credit market to grind to a halt.

The scary part, though, is what happens if the lender is already in a position where they can't pay up if someboday calls in a loan they've taken out. If that happens, they actually have to call in some of the loans they have made, and they don't know if those calls will yield any cash or not. As I understand it, that's when a "liquidity crisis" turns into a "soolvency crisis"--when the shit really hits the fan. If the ground levle loans (the mortgages themselves) aren't good, it looks like you end up with a big game of musical chairs, only the company without a chair at the end goes balls-up and billions or trillions of dolalrs in value evaporate.

Was Bear Stearns the first company to find itself standing? Probably. Was it the last? We have to hope so, but I kind of doubt it.

I'm an English professor, not an economist, but I'd be interested to hear from the economists in the audience whether this explanation makes sense or is somehow flawed.

Posted by: FearItself on March 17, 2008 at 4:58 PM | PERMALINK

Poster: come here often?

Posted by: Altoid on March 17, 2008 at 5:07 PM | PERMALINK

You have gotten some very good points made in several comments.

But one of the key points you have to remember is that markets do not clear instantly. Often there is a several day lag between the time you buy or sell a security and the change in ownership actually occurs.

So imagine that you want to buy a stock or other security from Bear sterns on Friday So you send them the money. but you will not receive ownership till next week. but if bear stearns goes into bankrupracy over the weekend your ownership goes into some sort of legal limbo. Bear now has your money, but you do not have the security. Yes, eventually when the lawyers get through with it you will your security. but for the time being you are out of your money and have nothing to show for it.

When you recognize that there is actually a risk that this could happen you just quit trading with Bear Stearns. If you want to buy a stock or sell a bond you go to Merrill Lynch of some other dealer.

So as far as Bear Stearns is concerned it is now frozen out of the market. But it makes its money by trading. consequently, it has no revenues to pay its salaries and other obligations and has no choice but to declare bankruprcy.

This is what people mean when they say markets freeze up. It can happen to a firm or to a certain type of security.

Posted by: spencer on March 17, 2008 at 5:16 PM | PERMALINK

Fear of deadbeats

Posted by: Luther on March 17, 2008 at 5:27 PM | PERMALINK

The failure of Carlyle Capital Corporation shows three mechanisms involved in the credit freeze. They described themselves as: "Carlyle Capital Corporation Limited invests in a diversified portfolio of fixed income assets including high-grade mortgages and credit products." And because they were highly leveraged they could offer 30% or higher returns. Less than 1 billion "owned" 20 billion of mortgages of which 19 billion was borrowed. However as the price of these mortgages dipped slightly, CCC stopped buying mortgages (extending credit) because they needed that 1 billion of capital to keep their lenders happy. About the same time, the institutions who had extended credit to Carlyle demanded that Carlyle find some more cash so that Carlyle's cash and mortgage pool would be able to repay the 19 billion that had been lent, i.e. they cut off Carlyle's credit. Shortly thereafter they began to make margin calls which CCC couldn't meet so it went bankrupt and the 1 billion of capital and the now less than 19 billion of mortgages were divided among the creditors who took a loss. Not only did they lose money, but they now had a pile of deteriorating mortgages instead of cash so their ability to make loans was reduced by about 18 billion dollars.

The first mechanism from this example is highly leveraged debt buyers hang onto cash in order to survive. The second mechanism is cash holders stop lending to anyone who looks risky. The third mechanism is cash lenders both take losses when they foreclose and wind up with illiquid assets instead of cash on the bad loans they have already made.

Posted by: zeno2vonnegut on March 17, 2008 at 5:27 PM | PERMALINK

Easy: wages have stagnated for a generation. The cost of living increased. People somehow believed they could buy and sell homes at a tremendous profit when the actual buyers could not afford such premium mortgages.

Moral: The best way to sell products that retain their value is to put money into wage earner's pockets. Politicians buy hookers. CEOs buy in France. Dubuque is on food stamps.

Posted by: Sparko on March 17, 2008 at 5:32 PM | PERMALINK

FearItself: I'm not an economist either, but here's how I explain it: start with the cash stream. Any cash stream can be "securitized," ie sold as the right to it or to parts of it. The right is what's then sold as a security, ie some form of financial paper.

There are old, established forms that banks and corporations issue and sell, and most of these are regulated. Then there are new forms created by outfits like the late, semi-lamented Bear Stearns that are largely unregulated, because brokers are treated differently than banks and bond-issuing corporations.

These newer forms combine many different streams from many different sources, as discussed by Math Teacher. These streams are often tapped at several removes from the original source-- it's paper issued on other paper issued on other paper, because a cash stream is an asset and there's no real limit to the number of generations of securitization it can be subject to.

If any doubt arises about the regularity of substantial portions of the cash streams, and/or if streams are so far away from sources that no one can tell whether they might be good or not, you have a decline in the value of the paper. It tends to spread like wildfire, as it did in the 19th century several times. They called them "panics" then; classic credit contractions.

The mortgage payments issue affects a lot of different kinds of paper because of the way the cash streams have been redirected through several generations of securities. When they say nobody knows what a security is worth, that's why-- nobody really knows where the cash stream is supposed to originate and whether it's reliable or not.

So the entire commercial paper market is affected. That means things our money-market funds have "invested" in, and paper that American and European banks have bought and use as part of their capital base. The Europeans wrote most of it off last summer and fall; our banks are slowly doing that, so we get a drip-drip of bad news about them, and their ability to lend money shrinks.

They and other credit issuers now don't want to lend without iron-clad guarantees of payment and iron-clad collateral, and who can blame them? Nobody now wants to buy paper without that same guarantee. So the credit freeze comes from both the issuing and the securities-buying sides.

Declining house prices affect the cash streams and the value of the securities because they reduce the value of the collateral and because they increase the likelihood of non-payment, hence of the cash stream not flowing. But that loss of confidence in the cash stream can't be confined to only one kind of security because so many generations of securities are built onto mortgage payments in fiendish combinations and because they're combined with other non-mortgage kinds of cash streams.

This is what Atrios so eloquently calls the "Big Shitpile." And why, to my understanding, he calls it that.

One thing the Fed is doing now is accepting all kinds of paper at face value and swapping it for gold-plated federal debt, in part so banks don't have to contract their credit. Federal debt is a solid asset that banks can lend on. Their problem is that they have to swap it back and have to have good assets on hand when they do it. They're all hoping the shitpile will be sorted out in 90 days, I guess.

Posted by: Altoid on March 17, 2008 at 5:41 PM | PERMALINK
Why are the credit markets frozen?
Actually that's not that hard to understand.

When you have wealth, it can sit there, with the risk that inflation will cause it to lose value, or you can invest it and get a return on the investment. If you can afford to lose it all, you buy stock and exchange the risk of total loss for the higher returns of stock. Otherwise you lend the money to people you are reasonably sure can pay you back and accept a lower return for the lower risk of default. The credit markets cater to that second type of investor, the one that won't accept the risk of stocks.

There is a risk-free investment, called treasury bonds. You will get a return that usually stays ahead of inflation. The rate of interest is considered the risk-free rate because you run NO RISK of DEFAULT. But it is a relatively low rate of interest.

If you want a higher return, then you have to accept some risk that you will not get your money back. Your rate of interest will then be the risk-free interest rate that you could get on a treasury bond, plus the estimated risk percentage that your borrower will default and not pay you back at all.

The rule is, if you want more return you have to run a greater level of risk of default. You calculate that level of risk and add it to the risk-free rate of return.

What happens if you can't calculate the added interest rate caused by the risk of default?

Then you don't lend to that borrower.

So the borrower says, "Hey! I have assets you can have if I don't pay the loan back?" If the assets are themselves reliable, then you lend to that borrower. The job of the rating agencies was to evaluate the reliability of those assets that could be pledged to get loans. It turns out that the bundles of mortgage loans had an unknown amount of sub-prime loans in them and the rating agencies were blowing smoke when they claimed to rate how reliable those subprime loans were and to what extent that they effected the overall batch of assets that were being put up as collateral for the loan.

So no one knows how to estimate the probability of default in the bundles of mortgages to be used as collateral, and the companies that want the loan were shown not to be able to repay the loans they already had.

The result? No sensible lender is going to lend money because they don't know if they can get it back, and they can't calculate the risk of default that has to be calculated to set the interest rate at which they WOULD lend based on that collateral.

Thus, the credit markets have simply frozen up. Better not to lend at all if you can't estimate the risk of default. If you could estimate the risk of default, you could set an interest rate high enough so that over large numbers of loans you would not lose your principle.

If the problem were limited to a few institutions, the credit markets could let them go and die. Only that part of the market would be frozen out. But now it is clear - every potential borrower carries a clear, but unestimatable risk of default. The rational interest rate to charge on such a loan cannot be set. Too much is unknown.

As a result, the risk of loss to default is higher by some unknown amount than the risk of not lending and losing return. So potential lenders choose the better risk - or the risk that is better understood - and don't lend.


Posted by: Rick B on March 17, 2008 at 6:43 PM | PERMALINK

Well; the real disaster isn't the loss of value of people's homes.

The real disaster is the loss of value of the American Dollar.

But the dollar is going down partly (largely?) because foreign money is not flowing in to mortgage the homes.

Posted by: Bob M on March 17, 2008 at 7:09 PM | PERMALINK


If you have enough money to lend you can spread your loans out to a lot of borrowers, so the risk of default on any one loan becomes even more predictable (law of big numbers. The same rule that makes insurance possible.)

As long as the risk of default is known, then you can borrow money to lend and take the difference in risk premium as even greater profit.

Bear Stearns thought they understood the risk of default, so they borrowed more money to lend. Anyone with that much money to lend has little competition, so the borrowers have to accept the relatively higher interest rate BS would have charged to riskier borrowers. If they didn't accept it, there was no one else large enough to lend to them.

Unfortunately, BS (like everyone else) miscalculated the risk of default and did not price their loans high enough. And in fact, no one in the market has enough information the properly price such loans. So BS's lenders stopped lending to them and even started to call back loans that had already been made. As Jared Bernstein says, that's a run on the bank.

BS then had to sell assets (loans) at fire sale prices to meet their needs for cash and avoid corporate default.

So it wasn't just the leverage that killed them. All banks use leverage. (That's why all banks are at risk of runs - and why the government insures bank acounts.) It was leverage that did not properly price the risk of default.

Posted by: Rick B on March 17, 2008 at 7:12 PM | PERMALINK

Kevin - You have identified a central puzzle in macroeconomics. What causes the credit allocation system to suddenly "freeze up"? More specifically, why does the capital market stop providing funds to profitable projects? It is exciting to see how precisely you put your finger on the problem!

I am a research macroeconomist who has thought about this issue for many years. I also teach corporate finance to undergraduates. The origins of market-wide illiquidity, and the reasons why firms place so much value on liquidity, are a great puzzle.

For various reasons, I am dissatisfied with conventional theories on this subject (such as Bernanke and Gertler's theory, for example). My thoughts are presented in an academic paper (Journal of Monetary Economics, Sept. 2003, pp. 1215-41), which shows that efficient allocation of funds to projects relies on information channels that may suffer persistent damage in a crisis. This theory still has a few loose ends, however, and I continue to view the "freezing up" problem as a great puzzle.

Posted by: Prof. G. Ramey on March 17, 2008 at 7:28 PM | PERMALINK


That alternative to mark-to-market is historical cost. That means that if you are dealing with a company that has bad loans on its books, you don't know it. Such loans build up unseen. Ultimately the failures will be much larger.

Mark-to-market will expose the bad loans sooner, so that there will be fewer built up. In this case the problem is not mark-to-market, but the fact that sub-prime loans were improperly valued by the market as a whole (meaning that everyone trusted the rating agencies who were being bribed for high ratings of poor securities, often because they did not understand what they were rating.)

Individual companies hate mark-to-market because it exposes their failures, but outside accountants know it is the preferable way of reporting. In particular, the management paying for the auditor hate having someone they are paying turn around and expose their failures to the world. Mark-to-market is unpopular with a lot of people, but that's because they don't want others to know what went wrong or when.

Transparency is preferable to secrecy, making mark-to-market preferable.

Posted by: Rick B on March 17, 2008 at 7:28 PM | PERMALINK

Comments to FearItself and Altoid - Altoid is basically alluding to the famous Modigliani-Miller Theorem: the value of securities is equal to the value of real assets that underlie them. If new information reveals that assets such as housing have been overvalued, then security valuations fall. Bankruptcies may reallocate ownership claims, but the overall value of securities is determined by the underlying assets.

The puzzle is, why should the revaluation of some existing assets cause the market to stop funding new profitable projects? Something about preexisting cash flows is complementary to making new investments. All theories about credit crunches rely on some "credit multiplier" or "propagation mechanism" that achieves the needed complementarity. None of the theories is really satisfactory in my view.

Posted by: Prof. G. Ramey on March 17, 2008 at 7:42 PM | PERMALINK

Rick B.Mark-to-market is unpopular with a lot of people, but that's because they don't want others to know what went wrong or when.

Nice theory but one dependent on believing that all markets are like the stock market. When, as in the case of the CDOs, there aren't trades, then value becomes a highly subjective issue.
"Bad" loans are a perfect example of why mark to market doesn't work. The market for these loans is neither wide nor deep. Buyers can try to argue a value far less than actual value and there may even be a sale of some piece of a loan. The OCC, for example, is fond of arguing that a sale of a loan or even a quote is a market and the bank should mark such a "bad" loan down to that trade or quote.
Taking such a tack is just lazy. The banks I know of did a rigorous analysis, every quarter, of their "bad" loans to establish such a value. The anomaly here is that the loans are in pieces spread in tranches and the only fact is that the models used to value them don't work. There isn't enough data, other than monthly collections, to properly value what's left.
Mark to market for so called "bad" loans is just passing more regulated money into unregulated hands.

Posted by: TJM on March 17, 2008 at 8:01 PM | PERMALINK

Prof. G. Ramey,

It's just a guess, but I'd suspect that almost all of Bear Stearns' business was buying securities using borrowed money. They weren't funding new profitable projects, they were slicing and dicing other people's projects.

The sub-prime mortgages got thrown into the pot and made the actual value of the securities that BS bought and sold unknowable because institutions that should have provided the risk data failed to do so. Since BS was so highly leveraged, their capital quickly disappeared, and since their assets cannot be reliably valued, their lenders not only quit lending but started calling in loans - in effect, a run on the bank. The problem is that as middlemen in the financial chain, they simply were not in the position to fund new profitable projects. Someone else was doing that, securitizing the result, and selling it to BS. And the system was not giving reliable data on the value of the profitable and unprofitable projects which are bundled together. Since the true value of such securities cannot be presently known with any reliability, no one is going to lend to BS. They have no capital and no collateral. Teh risk of default cannot be estimated.

Since this is a systemic problem, one not unique to BS, every investment bank that borrows money to buy and sell securities is facing a similar problem. That will be true until either the most leveraged of them disappear, or until the number, true value and impact of defaulted sub-prime mortgages can be determined.

As I say, that's just a guess, but I'll bet it's pretty close.

Posted by: Rick B on March 17, 2008 at 8:03 PM | PERMALINK

Excuse my ignorance, but it seems strange that home mortgages are such a large part of this system.

Is there a percentage of defaulted business-property mortgages (or other business loans) included in the mix?

Posted by: Tilli (Mojave Desert) on March 17, 2008 at 8:11 PM | PERMALINK

Prof. Ramey: I have two questions about what you've said. First about securities and underlying assets (I didn't know there was a theorem about that, but of course there had to be). What happens when you have a second and third generation of securities based on the same asset? Each takes its value as some multiple of the expected cash flow, presumably, but each generation is less able to get at the underlying assets, isn't it? That's my understanding of what happens when you get a chain of liquidations like happened in 1929, or more apropos for what's happening now, in the case of counter-signed IOUs and other notes that passed as money before the Civil War-- you'd try to collect payment from the last holder and then work your way up the chain of signatures, and usually you were SOL because nobody had the assets (or it cost too much to bother trying). That's how I think of these securities. Is that wrong?

Second, this has probably been used in some of the theories you mention and is somewhat discredited in history of science now, but if we accept that markets are psychological mechanisms, shouldn't we be thinking in Kuhnian paradigm shifts? As in, suddenly we have to re-examine all our premises because relying on them has just cost us big-time? Probably been done, but I thought I'd throw it out--

Posted by: Altoid on March 17, 2008 at 8:29 PM | PERMALINK

"This is pretty much how the Great Depression began and it is why the Fed (in the absence of any help from the Hoover, I mean Bush, Administration) is taking such historic steps, knowing full well they are bailing out some bad actors.

(And a special PS to all the corporation and banking hating readers of yours out there. Grow up.)
Posted by: martin on March 17, 2008

At least they had a Dust Bowl and agriculture failure excuse in those days. Today it's just all greed and some stupidity.

I wonder how many people in this tragicomedy need to go to jail for "justice" to prevail.

Posted by: MarkH on March 17, 2008 at 8:30 PM | PERMALINK


So your proposal is to keep known bad loans on the books at historical value or some arbitrary value close to it? That's worse than not being able to apply an accurate value.

You are removing and concealing the indicator of corporate risk from the financial statements so that the corporation can somehow secretly try to prevent their own stockholders and lenders from accurately valuing their company.

The difficulty of finding an accurate value for such securities does not justify hiding the fact that they are badly priced securities. That's the same problem as "smoothing earnings." You are prettying up the financial statements so that outsiders cannot properly evaluate the company doing the reporting.

Financial statements have to expose such problems, not conceal them. Otherwise comparisons between companies don't work.

If you can't convince a lender or stock buyer that a bad loan on your books is an asset, then don't ask your accountants to apply an arbitrary value to it and conceal the nature of the asset. That's not the purpose of audited financial statements. Transparency is. Difficulties in establishing an accurate value for an asset don't change that fact.

Posted by: Rick B on March 17, 2008 at 9:17 PM | PERMALINK

After reading the warnings about sub-prime mortgages and stories about the escalating home foreclosure rate over the past two years, I can only conclude that the credit market froze because the banks were waiting for an excuse to freeze it.

Maybe I've been listening to too much of Naomi Klein but I see the "Shock Dcotrine" in practice. I'm certainly no economist but this is how it appears to my eyes.

Just look at what has happened.

A few years ago Americans realized that they were drowning in debt and began declaring bankruptcy in record numbers. The banks then lobbied Congress and the bankruptcy laws were changed.

Americans then borrowed heavily against their home equity just to make ends meet. Some also transferred existing credit card debt to new cards offering low introductory rates. Many new home buyers (even those with good credit) entered into sub-prime mortgages that allowed them to purchase far too much home than they could reasonably afford. Credit flowed freely and Wall Street, foreign investors and the banks were making billions.

The banks were making billions in consumer penalty fees and steadly increasng credit card interest rates. Countrywide and Bear Stearns were well aware that they were sitting on tons of sub-prime mortgages and bad credit liabilities.

Now Americans with sub-prime mortgages cannot meet their payments when the rates adjust. Home property values have plummeted. Some Americans now owe more for their mortgage than the value of their home. Home foreclosures are at all time highs. And those who transferred credit balances to new cards can't meet their payments. American consumers have run out of money and assets to borrow against.

So after the banks had picked the American consumers pockets cleaned they froze credit. And when the banks froze credit, the politicians acknowledged "a possible recession", and gave The Fed an excuse to act. The bailouts began.

Bank of America bails out Countrywide, JP Morgan bails out Bear Stearns and The Fed lowers interest rates to make it easier for banks to borrow money from other banks.

If one didn't know better you'd think that the banking industry planned it.

Posted by: Pamela Lyn on March 17, 2008 at 9:30 PM | PERMALINK

I know this doesn't address the specific question Kevin raised (which has already been addressed at length, I think), but I remember seeing a great primer some months ago that described the secondary mortgage system as a whole in the form of a slide show that used a hydraulic metaphor. The payments from individual mortgages was represented as a pipe carrying a stream of "liquid" cash down into a larger stream, which ultimately debouched into "buckets" that represented the various resulting securities; each bucket was filled by the overflow of the bucket above. If the flow was reduced, the riskiest buckets (the lowest-rated securities) at the bottom went dry first. A trough at the bottom stood for the ultra-risky payments, which were then poured into still more buckets and repackaged (with artificially high ratings). I can't seem to find the link to this slideshow, but I thought it was a very accessible presentation of a complex idea. Does anyone recognize my description and have a link to that resource?

Posted by: FearItself on March 17, 2008 at 9:30 PM | PERMALINK

I sure don't see how a meltdown can cause a freeze up. That Melting is freezing is OK but to simultaneously say that down is up is going too far (up, down or sideways).

You better stop mixing metaphors or Paul will get you http://tinyurl.com/2784bf and http://tinyurl.com/2ef3rz

Posted by: Robert Waldmann on March 17, 2008 at 9:34 PM | PERMALINK

"My understanding of this whole mess has been that a lot of supposedly smart people let themselves be gulled into believing in alchemy."

That, plus Duncan Black's "Big Shitpile" analysis, describes perfectly why Kevin's quixotic quest to "understand" the situation is both hopeless and useless. Yeah, I took lots of finance classes and have the textbooks. Yeah, a lot of my classmates went to Wall St. and LaSalle St. and worked in that industry. What is going on right now is not covered in any of those textbooks. Criminal fraud textbooks maybe, combined with control systems engineering texts heavy on the feedback theory. But finance? Economics/economists? Those are the self-described "smartest guys in the room" **who created this situation**. I doubt they have any meaningful explanation.


Posted by: Cranky Observer on March 17, 2008 at 9:36 PM | PERMALINK

The value of the underlying assets can be worked out. That's not the problem. The real problem is that no-one is sure exactly who owns the mortgages over the assets, and more importantly how much it would cost to establish legal title to the underlying asset and foreclose on the loans. If you have a CDO with 100 loans over 100 houses you can work out the price of the houses, and even project how much those houses might be worth in the future. But the problem is that these loans came from different Banks all over the place and they didn't keep their paperwork in order. In all that shifting of loans around the paperwork for each loan has not been kept in order and the result is that you have own the loan but don't have the paperwork establishing legal title to the underlying asset. The uncertanty is then how much it will cost to establish legal title (if you can) and then if a judge will actually let you foreclose the loan. The cost might $50,000 in legal bills per loan. If the average loan size is $100,000 then half your CDO is gone right there. If you can't find the mortgage papers then you've got nothing. You can't legally foreclose on the loan and the loanee can tell you to get lost. That's why some CDO's might actually be worthless.

The sub-prime crisis is primarily a legal one. People don't actually say this much because its almost impossible to believe, but the heart of the problem is that lots of banks bought trillions of dollars of loans that turned out not to include clear legal title over the underlying asset.

Posted by: swio on March 17, 2008 at 9:40 PM | PERMALINK


Come on, what reasonable capitalist could have foreseen the need to establish ownership of assets?

Posted by: alex on March 17, 2008 at 9:55 PM | PERMALINK

I apologize if someone already said this upthread, but I didn't read the whole thing.

Credit markets are frozen, simply, because all credit extension is predicated on financial models that predict several things: future interest rates, inflation, and the rate of default on loans, among others. What the subprime mess has made most clear is that all of the models, including the models that apply to good credit risks, are very sadly flawed. Until more accurate results can be obtained, no one is willing to lend anything, because there really is no knowing any of the variables anymore, or how they can be assimilated into a true financial model.

Until recently, I was in the finance business. (I got out while the going was good).

Posted by: jprichva on March 17, 2008 at 9:57 PM | PERMALINK

> The value of the underlying assets
> can be worked out.

I doubt that. Not in California at least. To value an asset you have to have some reasonable chance of selling it, but if you tried to sell any significant percentage of Ontario-area real estate you would trigger off an even worse cliff dive that we have seen already. And there has to be someone to buy it - but now that people can't sell their old houses of 87,000% profit they can't buy the new ones.


In the Midwest the situation is nowhere near as dire, although I don't know what is happening in Chicago.

Posted by: Cranky Observer on March 17, 2008 at 10:03 PM | PERMALINK


You make it sound like I might be able to get away with just stopping my mortgage payments.


Posted by: FearItself on March 17, 2008 at 10:07 PM | PERMALINK

swio, excellent post. I would like to add just one point. You state that the value of the underlying assets can be calculated. That is true, at least approximately, if you assume that when present real estate bubble finishes bursting, prices will be at their normal, proper levels.

The problem is that this figure is so terrifyingly below what has been loaned out that the financial companies don't want to make the figures public.

Posted by: bobo the chimp on March 17, 2008 at 11:01 PM | PERMALINK

Financial Meltdown 101
1) Repayments unpredictable in a downturn. 2) "Insurance" may not pay off. 3) Cash is much better than a firesale.
1) Ultimately, these "assets" are the right to receive some stream of revenue. Various homeowners send in their monthly mortgage check. You get paid. For everyone who does not pay, the stream of revenue decreases. With house prices falling, many people who borrowed to much to pay too high will just walk away. No way to know how many will do this or how much this will cost the recipient of the mortgage stream. Therefore, there is no way to accurately value the revenue stream.
2) Mortgage payments and other revenue streams (from corporations repaying their loans, credit card payments) were converted into extremely complex derivatives and the likes of Bear Stearns bought huge piles of them. They protected themselves from any problems with these complex derivatives by paying insurance. But what if the issuer of the insurance can not pay? It is as though they bet huge amounts of money on both sides of a boxing match, paying both bets with money borrowed against the pay-off on the other bet. But there is a rumor that one or both of the bookies has been shot and won't be paying off. But they still owe the loan they took out to make the bet. And Tony doesn't like to wait for his money. oops.
3) Finally, various players are being forced to sell this paper so the prices have fallen below what they would be otherwise. (Maybe the "what they would be otherwise" will eventually fall even lower but maybe not.) So no one wants to sell their paper at firesale prices. The way to be sure you are not forced into a firesale (which could wipe you out) is to keep as much cash as possible on hand (real cash, not promises of cash from someone who may be worse off than you are). One way to keep cash on hand is to not loan it to anyone. So all the credit markets freeze solid.

There is plenty more complexity if you just can't resist, but the above three factors are the basic market dynamics. The social dynamics of how this mess was created is another matter. My simple explanation for that was that there was no one to take responsibility for the thing as a whole. For a few decades, we have mostly elected those who said that the market would take care of that automatically. I don't expect that belief to have as many followers when the dust settles. Markets depend on a complex social technology of trust. But our financial class seems to have understood the social investment that goes into creating networks of trust about as well as a stone hippie understands the network of hard work that went into making his Volkswagen bus.

Posted by: Jessica on March 17, 2008 at 11:22 PM | PERMALINK

The behavior of serially dumping misrepresented crap was pretty costless for four years, from the beginning of 2003 through the end of 2006, so mortgage originators took a tremendous and undeserved windfall when they sold mortgages for value premia related to true aggregate discount value... The New York City Math Teacher

Kevin, NYCMT has got it nailed. My predictions: I think that house prices will probably bottom at 2002 levels nationwide on average next Spring. Sluggish GDP growth will resume in the 3rd quarter of 2009. Christmas this year will suck in the worst possible way with a major retailer going under, then the new president oversees a bail-out of GM in 2009. Employers will start to hire again in Spring 2010. That's my on-the-fly appeal to intuitive processes that lie outside of standardly accepted scientific methodology.

Posted by: Doc at the Radar Station on March 17, 2008 at 11:23 PM | PERMALINK

If you're one of the lucky ones you migth just be able to do that. There people in living in million dollar homes around the country who have already worked this out. They stopped paying their paying their mortgages for up to two years, and they still haven't been evicted. Unfortunately the only way you can really know if your bank is screwed is if you stop paying your mortgage and dare them to foreclose on you.

See this link on "Foreclosure-proof Homeowners" for an example of what I'm talking about.

Now imagine that millions of times over and you'll see why its so hard to value CDO's and why some CDO's will turn out to be worthless.

Posted by: swio on March 17, 2008 at 11:34 PM | PERMALINK

swio: See this link on "Foreclosure-proof Homeowners" for an example of what I'm talking about.

Same ol', same ol'. You can squat in a $3M mansion for years, but try it with a $500k shack and I bet you're out in 6 months, tops.

Posted by: alex on March 17, 2008 at 11:52 PM | PERMALINK

bobo the chimp,

You're right about that. You can't say exactly what prices will be, but you can get forecasts that will probably be accurate within, say, 20% for different regions in America. They might be wrong, but all you have to do is make a guess at how far back prices will stabilise. Will they stabilise at 2002 levels? or 1999?. You can assign probabilities to each of these timeframes. Then all you have to do is find out what each property in a CDO was worth at that time and bingo, you have a reasonable guess as to the value. You can say there's a 20% chance you're out by a year either way, and 15% chance you're out by 2 years and so on. From that you can make an estimate that is based on hard numbers. Its not precise, it but its a real number and you have an idea how likely it is to be right or wrong. Its enough to make a valuation. I have no doubt that most of the banks are furiously working out these numbers as we speak and top banking executives are getting private super secret briefings on the results. You're not hearing about it in the media because no one wants you to know how bad the results are, and they know these numbers don't mean squat because the legal risk can't even be guessed at.

Posted by: swio on March 17, 2008 at 11:54 PM | PERMALINK

Rick B So your proposal is to keep known bad loans on the books at historical value or some arbitrary value close to it? That's worse than not being able to apply an accurate value.

Have you ever read a bank's financial statements? Have you ever heard of NPAs, NPLs, loan loss reserve, Tier 1 or Tier 2 capital. Every bank carries bad loans on its balance sheet, sometimes more of them, sometimes less. The bank, which is not an investment bank, values those bad loans every quarter and at least annually the Feds review the process as well as particular loans.

Your response shows a lack of understanding of bank lending. There are bad loans on the books and several footnotes to the financial statements explaining the bad loans.

What is the point of mark to market pricing when there is no market? What JPM did was to value the CDOs etc on Bear's books at zero. The Fed told JPM that they would lend on some value other than zero because there is no market.

Posted by: TJM on March 17, 2008 at 11:59 PM | PERMALINK

"This comment thread is a remarkable mish-mash of knowledgeable and insightful analysis and bat sh*t crazy rantings. Never seen anything quite like it."

You must not get out much if you think there's any real ranting going on here. FWIW, I've spend a lot of time on Calculated Risk, The Big Picture, BondDad, and Mish, all of which are great places to learn about various aspects of our sometimes crazy financial system. However, the comments put up here are as helpful as anything I've read anywhere in explaining the nuances of the credit markets and how they work, and don't work. Thanks to all.

Posted by: bluestatedon on March 18, 2008 at 2:04 AM | PERMALINK

What would happen if the government were to decree a debt pardon for every person and entity in the country all at once? By law no one has any legal claim to money owed to them by any other party prior to today. You wouldn't have to pay your mortgage but on the other hand you wouldn't collect on your 401 or other investment. But keep social security and PERS going to keep people from starving. Everything would be instantly transparent and you could start loaning and borrowing again with tighter regulation. Kind of a massive do over because the whole mess is beyond comprehension or cure.

Or maybe some variation on this? Or am I just insane?

Posted by: nameless bob on March 18, 2008 at 2:06 AM | PERMALINK

What is the point of mark to market pricing when there is no market?

Please. Mark-to-market is shorthand for a-clue-what-the-hell-the-asset-you're-holding-is-worth.

Saying "there is no market"--especially for those who were the nominal forerunners--is simply another way of saying "We don't have a clue. We are a menace to sociey. Please put us out of our misery."

To imply that what they hold deserver to be called "assets", or that they have have any similarity to loans carried by banks is specious. Might as well put the kids used diapers on the balance sheet.

FYI, "Without providing details, Fed officials insisted that the $30 billion loan was covered by even the most conservative estimates of the Bear Stearns holdings." So how ya figure they came up with that?

Posted by: on March 18, 2008 at 2:32 AM | PERMALINK

Nameless Bob, could I opt out of that?

You're assuming we all bought over-priced houses with mortgages we can't pay.

Posted by: Tilli (Mojave Desert) on March 18, 2008 at 4:11 AM | PERMALINK

Mark-to-market is shorthand for a-clue-what-the-hell-the-asset-you're-holding-is-worth.

No, mark to market has a specific meaning, marking to a publicaly traded value.

Not that the asset has worth, but that there is a clearing, functioning public market which has reasonable liquidity and depth to establish pricing

Many, many assets which have value are not marked to market because there is not a liquid, clearing market. Examples include as TJM notes, non-securitized loans held on a bank's books. Individual loans are not tradeable instruments, and indeed historically valuation was done by non "public market" (to be clearer, market in these references are public markets, like Bourses, etc) was done using various assumptions (based on historical observations) about default rates, etc.

There is absolutely no logic to marking to market when the said public market has ceased to function. It is not achieving transparency at all, transparency is predicated on there being proper information to value the assets, on reasonable liquidity. Reifying, or better fetishizing (public) markets is simply blind bollocks - this logic would have you treat the "market value" of a thinly traded equity in an emerging market with few actors and thin liquidity in the same manner as London, Euronext or NY. One doesn't for this very reason; it can be quite clear that the publicly traded price is utterly wrong due to a number of reasons such not enough qualified buyers, lack of liquidity, manipulation (perhaps perfectly legal low-balling).

Many mortgage backed securities are going to be perfectly fine - baring of course an utter collapse of the financial system and 1930s style depression, which one can reasonably expect to be avoided. The Fed valuation on the collateral on the USD 30bn highlights that, these are indeed assets, with income attached.

In anycase, the Naomi Klein commentator is smoking bad crack to think that a credit freeze up was "engineered" by the banks for some mysterious purposes. Pure illiterate conspiracy theory that.

She would be better served reading swio supra than the borderline illiteracies of Klein.

Posted by: The Lounsbury on March 18, 2008 at 7:45 AM | PERMALINK

But finance? Economics/economists? Those are the self-described "smartest guys in the room" **who created this situation**. I doubt they have any meaningful explanation. - Cranky

"Economics exists to make astrology look respectable."
J. K. Galbraith

Posted by: MsNThrope on March 18, 2008 at 8:03 AM | PERMALINK

Why are the credit markets frozen? Nobody is buying mortgage backed securities! No buyers....no credit.

Posted by: plane on March 18, 2008 at 9:44 AM | PERMALINK

So how ya figure they came up with that?
Oh nameless one, they guessed that the CDOs weren't worth zero.
Your financial naivete is charming, but there are numerous situations, like buying an asset or a company or a house where the market is what you are willing to pay.Broad brush assumptions that the MBS market travails mean the entire market is frozen are just wrong.

Posted by: TJM on March 18, 2008 at 9:58 AM | PERMALINK

Rick B - I think you are saying that firms like B-S provide the liquidity needed to fund new projects. They take savers' cash and direct it to existing and new assets. If savers cannot distinguish between gains from profitable new projects and losses from existing projects, they will shift their cash away from these financial intermediaries. Managers of profitable new projects would not get funds.

According to this argument, channels for evaluating assets and directing liquidity are quite restricted. But this assumption may be unreasonable. If a manager's project is truly profitable, then some investors somewhere would wish to fund it irrespective of what is happening at B-S (implication of MM Theorem). Funding sources such as venture capital, commercial paper and internal cash flow all remain available. So why is lending frozen up?

It could be that new projects are hard to evaluate, and all the experts who evaluate such projects are tied up with problem firms such as B-S. In other words, there is an economy-wide shortage of "evaluation expertise." These experts could always quit and form profitable new investment firms, but this process might take time (this is essentially what happens in my JME paper). Does this make sense? I wonder whether such an expertise shortage would be so sudden and broad-based.

Another point - Inability to distinguish gains and losses from housing sector assets makes sense. But is makes less sense to imagine that investors could not distinguish the values of existing housing assets and new defense sector assets, or service sector assets. The confusion argument has trouble explaining the propagation of the credit crunch across sectors.

Altoid - The MM Theorem says that in a perfectly competitive capital market, no matter how you slice and dice the cash flows from assets, the total value of the pieces always adds up to the value of the cash flows you started with. A revaluation might trigger a chain of liquidations when the structure of claims is complex, but in the end the total value of claims equals the value of the assets. Importantly, the chain itself cannot affect the value.

When assets decline in value, claimants will of course scramble for chunks of value, and some of them must realize losses. But in the end the value of remaining claims equals the value of the assets.

Most of the above comments address the chain of liquidations problem. swio comments on the high legal costs of transferring claims to mortgages. This surely introduces a significant departure from the MM Theorem as far as the claims to existing mortgages. But it does not explain why credit is frozen up for profitable new projects.

Jessica mentions liquidity hoarding. This is a very common story that underlies many theoretical models. I do not subscribe to these theories, however. In essence, unless the hoarded liquidity is the personal wealth of the managers (e.g., a mom and pop bank), it is easy to write side contracts that transfer liquidity where it is needed. For example, once a project is identified that everyone agrees is profitable, hard-up investors could fund it and use the valuable equity to ease their liquidity shortages. Again, there is no reason for the credit crunch to spread to profitable projects. Truly, it is hard to understand the "freezing up" phenomenon!

P.S. I am not sure what to say about paradigm shifts. Everything I discuss assumes simply that managers and investors seek to maximize personal wealth. Moreover, there have not been important changes in methods for evaluating the value of future cash flows from new projects.

Posted by: Prof. G. Ramey on March 18, 2008 at 10:23 AM | PERMALINK

Bobo at 11:01 PM

I think you have it nailed. One point to consider is the "value" of real estate.

A lot of people above seem to think that there is some objective value that can be applied to real estate if you just look for it, but in a market economy that's not true. Real estate has no financially measurable value beyond what some buyer will pay for it.

If no one can afford a home, then the price must drop to where someone who desires it has enough credit to borrow enough money to pay for it. Desire can range from zero to infinity. It is the capacity to pay - that is, find lending - that is the limitation on and the competitive factor between different possible buyers.

In recent years bankers were encouraged to offer loans that were unlikely to get repaid, and a lot of them could offer the loans, collect a fee, and pass the risk of non-payment off to someone else, while those ultimate investors were not shown the real risk they were taking. But since everyone else was doing it, it must be a good idea, and the promised interest rate was higher than a government bond. Smart people with fancy and impenatrable models agreed with Alan Greenspan that such loans were a good idea. So purchase money long ago passed sustainable levels, and that's how prices for homes in California, Florida, and a few other hotspots has been established for years now.

well, Greenspan was a liar, he's gone now, and the models didn't cover the current situation. There's no one else to catch the hot potato. There are also no real hints as to how low the realistic prices for those homes might be, and no one who has bought in recently wants to be the first to try to find out how low that price really is.

The people holding the security portfolios are holding stuff they paid a lot for and will get back only a fraction of what they paid. The very mechanisms for getting back any portion of their speculation are being presently worked out in back rooms and have not yet seen the light of day. Sell now and they lose a whole lot, wait a bit and they may get a little more back - if they can survive. Bear Stearns just found that it couldn't survive.

Who's next? And what are prices going to stabilize at? No one knows, and no one has enough information to guess, so paralysis is the order of the day. A lot of apparent wealth has literally disappeared, and those who don't move right now don't have to admit that some of that disappeared wealth was theirs.

So what are real estate values going to stabilize at? The answer to that will probably have to wait until a new class of lenders who know how much wealth they have appears. Until then the only sales will occur at fire sale prices, prices that those who think they own the property don't want to accept.

Someone above suggested that will be at about 2002 prices for SoCal. That's probably as good a guess as any. Lenders don't lend based on guesses.

Posted by: Rick B on March 18, 2008 at 11:42 AM | PERMALINK

"in the end the total value of claims equals the value of the assets"

Okay, Prof. G. Ramey (at 10:23), but if I understand the idea here, the missing term is time. Time is essentially what I was pointing to. If I securitize a stream of payments and sell it to you, what I'm doing is collecting from you a discounted payment of the whole amount. I'm capitalizing the flow, which means I'm compressing time.

Nothing prevents you, too, from securitizing the same cash flow and capitalizing it yourself. Probably you will as long as there's a market for it. You, like me, may think it's better to get whatever capitalized amount you can rather than take the risk that something might happen to that flow of cash.

Supposedly that risk is priced into the capitalization by discounting, right? But if something actually *does* happen to slow down the flow, and if it isn't apparently confined to a definable group of payers, what does that do to the risk assessment? A hypothetical becomes an actual event, and its extent is unknown.

If you knew that was beginning when I tried to sell you the initial security, would you buy it? If it happened when you were trying to securitize what you bought from me, who would *you* expect to buy it? You'd be stuck with it, and the declining cash flow would reduce its value by an unknowable amount-- unknowable until we could actually see the extent of the reduction in cash flow. But we can't know that until it's already happened. That takes time and hindsight. Only then do we have a definable set of conditions that we can model.

Basically, what I think I'm trying to say in a very naive way is that when something happens that looks like a macro-level change, the ceteris isn't paribus anymore and the marginal calculations are obsolete. And I think the calculations used in our current financial markets have become so complex, so finely-tuned, and so short-term that almost *anything* can qualify as a macro-level change in conditions.

The risk that's priced in as a discount assumes a certain set of conditions which I think is probably very narrowly-defined. That makes the financial system as a whole much more unstable, because when those conditions are shown to be wrong by new events outside the model, the model has to change. That takes time. The system is most unstable during this time when models are being redone in the absence of definite information about the extent of the change. That's about where we are now.

The same facilities that make for finely-tuned calculations also allow for more securitizing and more capitalizing. It may not multiply the actual assets or even the actual claims. But successive generations of securitization, I would argue, compound the instabilities because they remove the payees farther and farther from the payers and the conditions that govern them. The payees have less actual fundamental knowledge, in other words, and increasingly have to rely on finely-drawn models based on narrowly-specified conditions. Each seeks to compress time in the same way, but with succeedingly less knowledge about the conditions that make it possible.

That's why I don't think this is a technical economic question so much as it is a problem of modeling the real world, which resists being modeled. It's a problem of representation. The financial system is a representation of what actually happens out there. If it gets too far away from events, from experience, it isn't quite so believable anymore and people lose confidence in that interpretation. That's why I harp on the distance created by generations of securitization.

Well, this is pretty far afield. Fundamentally, I think the financial system has become full of fluff. It has always had a tendency to go in that direction. The job of regulation, because we don't have a JP Morgan (the person) anymore, is to countervail this tendency. But for a generation there's been less and less regulation. The fluffball is enormous. That fluff is all speculation on the future; too much of the future has been amassed (securitized and capitalized) so it can be used today. A certain proportion of the fluff has to go and once that's done, the countervailing force needs to be reinstated. This is really not going to be nice to live through.

Posted by: Altoid on March 19, 2008 at 2:22 PM | PERMALINK

Altoid - I like your comments. I think your main point is that macro-level changes have made securities less attractive by raising risk. Of course, this need not make the market less liquid. The market price falls instead. Assets become less valuable, and some previously profitable projects become unprofitable. But profitable projects should still get funds. There is no "freezing up" in this story.

You mention that macro risk may be exceptionally high right now due to changes in risk assessment and pricing models. Demonstrated deficiencies in these models can indeed raise investors' perceptions of risk. This argument makes sense, but maybe these failures are confined to a few sectors. The failures must be propagated somehow to the macroeconomy. And in any case the result is a repricing of risk, not illiquidity.

You also talk about how generations of securitization reduce investors' knowledge about the underlying sources of cash flows and their relation to the economy. Because of the complexity of their investments, investors become more reliant on deficient models. Very complex and far-reaching securitization could indeed provide a propagation mechanism. No one can be sure whether their claims are tainted!

These points argue for a sharp correction, but not for illiquidity per se. Even in the worst case, there will still be profitable projects. A "credit crunch" means that some of these projects cannot get funds. Why not? That is the puzzle.

Posted by: Prof. G. Ramey on March 20, 2008 at 3:12 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