« A little good news | Main | Hypotheses which are too simple to be true as stated »
Humpty-Dumpty theories
One version of pessimism is to wonder how easily a banking system can be put back together again. Just as a physical bubble is an asymmetric process (it is easier to pop one than to rebuild it), maybe a banking system is similar. It's built up over time by lots of lattice work and investment in complementary processes and assets. Once it pops away it can't be easily reconstructed, even if the reconstruction plan targets the initial cause of the problem (low capitalization). One implication of this view is that initial recapitalization of banks -- say about a year ago --was a much more urgent matter than we realized at the time.
A more optimistic take is the "bounce" view: things have to fall far enough so that they hit the floor and get a bounce, pushing them back up again.
In the last two days I have started to entertain the possibility that the Humpty-Dumpty view is in fact correct.
Posted by Tyler Cowen on October 10, 2008 at 09:31 AM in Economics | Permalink
Comments
Humpty Dumpty sat on a wall (street).
Humpty Dumpty had a great fall (40% YTD).
All the king's horses and all the king's men
Couldn't put Humpty together again.
Posted by: mother goose at Oct 10, 2008 9:50:37 AM
With hundreds of trillions in derivatives out there, a 1% optimism in valuing those derivatives will result in trillions of dollars of fake assets. Credit default swaps were being used to speculate instead of move risk, so we've got a huge set of mousetraps and ping-pong balls lined up. International trade was built on a premise of cheap fuel that just got yanked away. People using bad machine learning techniques made fake assumptions of conditional independence and concentrated their risk into giant Black Swan conditions that are now starting to materialize.
I'm not an economist. I'm a programmer, and when something gets this awful in computer programming, what we want to do is throw away the whole thing and start over. Now, I can understand that we might not want to do that here - that we might prefer some kind of orderly unwinding.
But the idea that the current fabric of the financial system is something that has been carefully built up and has to be *preserved* strikes me as an awful idea. It's been carefully built up *the wrong way* and now has to be *unwound*. Was Japan's mistake trying to unwind too slowly, or trying not to unwind at all? Was the Great Depression trying to unwind too quickly? But the question we have to ask is how to rebuild the financial system, not how to keep it intact. It's already broken in a very real sense, held together by inertia and hidden losses.
I'm not an economist or a trader, but that's the sense I get.
Posted by: Eliezer Yudkowsky at Oct 10, 2008 9:53:04 AM
*sigh*
We really need to put three or four people who know how to value mortgage securities and credit-default swaps in a room with "economists" and non market people, so they can talk.
If we put a floor under the value of mortgage securities, we can reflate the entire system overnight, including mortgage-related CDS, CDOs, CDO-sqr.
If this problem is as big as we might imagine, then "capital injections" is probably not the way to go. It may work, but it may not. Buying up securities in a slush fund is ... okay, but will not reflate the system, so it's a second-best, really.
If the banks have problems with other assets, like ABS (auto-loans), credit cards, commercial real-estate, etc., then those problems ought to be addressed the old-fashioned way, with a little more capital.
Finally, we ought to consider regulations that put some firewalls, so we don't have to go through this again.
Posted by: Amicus at Oct 10, 2008 9:59:12 AM
What do you think of Casey Mulligan's opinion piece in the NY Times today? His central argument is that the "real economy" is strong despite the financial crisis. His proof is that the marginal product of capital is above historical averages.
Posted by: Josh S at Oct 10, 2008 10:10:00 AM
I doubt whether "experts" are going to solve this problem. Bernanke is as much an expert as anyone around and so far he has failed miserably.
He failed to anticipate the problem
He failed to realize the depth of the problem when it appeared
He has failed to elucidate a plan to fix the problem
I suspect "his" problem is one of bias. He, like most, really believes that markets do solve problems and not make them.
Posted by: RobbL at Oct 10, 2008 10:11:26 AM
Why would we want a floor on CDO prices? I don't understand that. What about all of the CDOs worth zero now, and which may be worth zero in the future? Put a floor on them?
Many CDOs (look at the Lone Star transaction, for ex), are already trading at or near zero (as little as 6 cents on the dollar for "super senior" CDO in Lone Star sale). Would you set a different floor for each CDO? I just don't get the concept. Seems a bit like setting a floor on stocks. Think of how much money you'd have lost by setting a floor for Lehman, or WaMu.
Posted by: Mercutio.Mont at Oct 10, 2008 10:18:57 AM
I doubt whether "experts" are going to solve this problem.
=========
People are talking past each other.
It is a question of getting the right expertise in the room and then making hard choices.
If it is just a matter of reflating mortgage-related securities (including derivatives), it can be done.
It may have direct costs to the taxpayer and it may *greatly* offend the political sensibilities of just about everyone (liberal and conservative alike), but ... it can be done. Quickly.
Posted by: Amicus at Oct 10, 2008 10:19:18 AM
Why would we want a floor on CDO prices?
========
Because we want to "reflate" the banking system, to restore confidence in the lion's share of unknowns about asset quality, going forward.
We will need no "bailouts", government participation in company boards, and massive capital injections, if we restore investor confidence that banks have a handle on the value of the assets them own. You can get by with $1 in capital, if everyone believes that what you are in invested in will not fall greatly in value, right?
Posted by: Amicus at Oct 10, 2008 10:22:34 AM
if a nation's physical capital stock is wiped out quickly, following war or disaster or something, that country often can quickly get back to where it was before--like w. europe after ww2, for example.
so i'm just going to cross my fingers and substitute in ''banking sector'' or even ''economy'' for ''physical capital stock.''
Posted by: pants at Oct 10, 2008 10:26:11 AM
Would you set a different floor for each CDO? I just don't get the concept.
===========
missed this, first pass.
Let me do this Socratically, at the risk of making this thread too long, but with the added benefit of not being another 2-cents.
What are the variables, the key inputs, to valuing a mortgage-related CDO or a sub-prime or alt-a mortgage pool?
Once you have those, think about how you would move to making those variables unkown, still, but fairly tightly bounded, using tools at the disposal of the government.
Posted by: Amicus at Oct 10, 2008 10:33:17 AM
Let the Fed write credit default swaps and it would take off again, and make a little money on the side.
Posted by: rhhardin at Oct 10, 2008 10:41:51 AM
encouraging to me to see wells fargo and citi fighting like cats and dogs over wachovia while everyone else is crying lookout below.
Posted by: at Oct 10, 2008 10:46:16 AM
What are the variables, the key inputs, to valuing a mortgage-related CDO or a sub-prime or alt-a mortgage pool?
=========
I like your format, so I'm copying it :]
Suspect that major variables in CDO valuation are:
1) Seniority of CDO
2) Amount of money coming in right now.
3) Expectation of how much money will come in in the future:
a) Future default rate
b) Future property value
Those might be the big ones.* So once we get the future default rate and future property values figured out, we'll be good to go.
Because there is pessimism about future default rates and future prop values, the CDOs aren't worth much. And rightly so, in my opinion.
*I'm just shooting from the hip here, not a professional in this field.
Posted by: Mercutio.Mont at Oct 10, 2008 10:56:03 AM
Let the Fed write credit default swaps
=========
Here's a separate question, one about unwinding derivatives risk.
Say, the "core" underlying problem is related to $700B-$1,000B in sub-prime securities, for the sake of argument.
Now, derivatives multiply that by several orders of magnitude, most likely.
If you had to come up with an unwind scheme to limit the financial loss to just those of the underlying (which we might guess is about 20-50% of the face, or $350B-$500B), what would you do?
Could it be done?
Posted by: Amicus at Oct 10, 2008 10:57:01 AM
1) Seniority of CDO
2) Amount of money coming in right now.
3) Expectation of how much money will come in in the future:
a) Future default rate
b) Future property value
========
Cool.
Here we go.
The government puts a floor under the default rate, by buying up defaults, restructuring the loans, and making payments at the restructured terms to the lenders.
Now, we can "argue" about the terms of restructuring and the logistics, which I think reasonable people can figure out, even though we'll start fighting along partisan lines, quickly, I suspect.
Next up, house prices ... that's really tough. Most people think that there might be another 15-20% to go, on some indexes, right? They are worried, still, that it might be 30-40%, if the real-estate market gets really depressed.
What if we come up with a restructuring scheme in which the government "sucks up" half of the projected price decline, for a targeted set of securities only (right now for sub-prime and alt-a)? If the property you lent against is worth just 60% now, the government will share half of that with you, as part of the work-out terms. (They government may want an equity stake from the homeowner, to help recoup it's near-term underwriting, but we can argue about that, too, along partisan lines...).
So now, your "worst case" valuation just got cut in half. To get to the disaster valuations, the "fire-sale" prices, you suddenly have to assume something like a 70% price decline. That seems ridiculous. In other words, a little bit of 'price insurance' stretches a long way, perhaps longer than is intuitive, without completely assuming all the risk of price declines, either.
Last, you are now a credit rating agency.
The government has passed out some additional backstop / security.
What do you do?
Posted by: Amicus at Oct 10, 2008 11:10:34 AM
uh...correction: "..floor under the default rate, by buying up defaults...", should be "...floor under the default rate, by buying up *some* defaults..."
Posted by: Amicus at Oct 10, 2008 11:16:28 AM
That would cost a tremendous amount of money.
Posted by: Mercutio.Mont at Oct 10, 2008 11:21:26 AM
Might such a theory have wider application -- r.e., the difficult underdeveloped countries have in taking off?
Posted by: John B. Chilton at Oct 10, 2008 11:25:49 AM
I'm trying to think creatively and would be interested in comments:
Instead of raising taxes on the wealthy, the government could give anyone who can afford to do so a substantial tax break for DESTROYING their major assets (houses, cars, computers, yachts, anything really...) as long as they immediately replaced them with newly-made equivalents of equal or greater value.
Let's use an analogy. Over the first two decades of the last century, wheat production grew faster than the population. Wheat farmers could earn a small fortune by growing more and more of it. City dwellers (known then as "suitcase farmers") fled for countryside farms tracts looking to cash in on the riches.
By the 1920's, glut crashed the wheat market. The irony lies in the fact that, by the 1930's, grain elevators were stuffed full, crops were rotting... while Americans went hungry.
Instead of thinking about the problem as an oversupply of wheat, suppose we recast it as an oversupply of labor. This would explain why farmers tried to grow even MORE wheat as the glut grew more pervasive - because they hoped to sell enough at rock-bottom prices to merely survive.
At the risk of oversimplification, the oversupply arguably persisted until the government paid (subsidized) farmers just to plow their crops under.
Could some form of this theory work on a similar scale today with homes, cars, or anything else in which "forced creative destruction" together with "subsidized replacement" created new jobs and opportunities and in the process reinvograted the economy?
Posted by: Michael Zimmerman at Oct 10, 2008 11:28:46 AM
Also, by renegotiating the loans you are decreasing the expected future amount of money coming in in the future. This may not depress CDO prices as much as a higher default rate would have. So if it does inflate CDO prices, it wouldn't be as dramatic.
Posted by: Mercutio.Mont at Oct 10, 2008 11:31:35 AM
That would cost a tremendous amount of money
======
Sensible concern. But ... Prove it.
On the restructured loans, the government borrows at less than 4%, today, and I think we can assume that it may be able to collect 150-200bps (maybe even 250) more than that, for most/many sub-prime borrowers.
Price protection ... yes, that's costly. But, if the total, *recognized* (not realized) price decline is another 30% from here, and the government takes half, or 15%, we have 15% of $700B-1,000B (expressed in loan value). That's $105-150 billion. If some loans borrowers are let to default (say 4%), the figure is that much less. Going forward, there will be "defaults" on the restructured loans, but they will be a LOT less than if we have a severe recession for having done nothing. Add another $100 billion, still, just to make a fair calculation. So, maybe $250 billion, altogether.
Last, the government may be able to take a passive equity stake. If you believe that homeowner's prices will recover, the government may mitigate it's losses, just as it did with the program during the Great Depression.
On the other hand, we have $85 billion at risk to AIG, who wrote checks they couldn't cash. We have an untold amount going into "capital injections", for firms that may go to zero, yet (maybe not, we cannot tell). We have proposals on the table for economic "stimulus" for 1-1.5% of GDP.
Cost is relative.
Posted by: Amicus at Oct 10, 2008 11:36:44 AM
Might such a theory have wider application -- r.e., the difficult underdeveloped countries have in taking off?
=========
You mean, like the Brady plan?
Yes, it's been tried and ... it worked!
Posted by: Amicus at Oct 10, 2008 11:38:05 AM
oh, add to the costs the $15 trillion or so of equity market wealth (about half of the $30 trillion total) that was just wiped out, while the markets decided that the 'credit seizure' was severe enough that it might just provoke a deep, global recession.
In my mind, that was an avoidable cost.
$15T, is of course, too high. Use your macro book to figure out what the wealth-impact to consumption GDP will be, though. What did you come up with for an elasticity? Is it 1-3 cents on the dollar?
Posted by: Amicus at Oct 10, 2008 11:47:24 AM
When the final history is written what regulators will regret most was letting Lehman enter into bankruptcy rather than forcing a merger with some other partner. The cost to the entire economy would have been much less. If there is one criticism I have of the Bernanke Fed is the over reliance on their slow academic processes and their lack of understanding of the shadow banking system and the way financial firms are very tightly woven to other financial firms.
Posted by: bill at Oct 10, 2008 11:53:01 AM
I keep hearing that the financial sector is a main street problem (as opposed to the auto, airlines, aerospace industry, steel, and other sectors), but I haven't heard any evidence that makes it seem true. What if the financial sector was just another sector suffering a severe decline in demand and shedding jobs? What if the 9,000+ banks not directly involved with CDS or other swaps, not holding MBS and not requiring trillion dollar interbank loans can continue the traditional banking business in this country with nothing more than normal economic recession retrenchment? Has anyone smarter than me investigated this line of thinking?
If I’m right, these interventions are just increasing panic and stalling the inevitable.
Anecdote is poor evidence, but I have acquired 2 mortgages on rental property, financed a new car, and received the ordinary amount of credit card balance transfer offers over the past month.
Posted by: guy in the veal calf office at Oct 10, 2008 12:05:29 PM
