Tuesday, 3 March 2009

Myron Scholes - crisis to end in March

Last year Myron Scholes predicted the credit crisis would be over on 7th March 2009.

That's coming up this Saturday, so how's the prediction looking?

Well, AIG has lost $60bn and received a further capital injection; the US government has just converted a bunch of debt into 30-odd percent of Citigroup; Martin Wolf has written another article with some nasty-sounding predictions. We're running short of time to fix the system if Myron is to be proved right.

To be clear, this is not about the recession. I don't think we expect economic growth to start in the next four days. It's about fixing the problems in the financial system so that money can start flowing and the recovery can properly begin.

The debate in the US over nationalisation is growing louder. Unusually, the argument is getting simpler rather than more complicated: the losses are there, shareholders have been (as close as makes no difference) wiped out, so do the remaining losses go to creditors or taxpayers?

If no government action is taken, and banks go under, creditors will lose substantially (in the short term). The argument against this (made by Justin Fox at Time) is that if creditors take a loss, they will immediately stop lending to all financial institutions. This would be disastrous for economic efficiency and make the recession much worse. Another argument which he doesn't mention is that even the possibility of this happening will cause a series of bank runs as creditors try to take their money out. However, sufficient government guarantees have been given already to probably prevent this from happening in the short term.

On this argument, taxpayers need to take the hit. But there is at least one counterargument.

Lending is, in theory, based on the returns people expect in future and not those they have received in the past. Therefore, if bank creditors take a haircut in a way that is credibly not going to happen again in the future, their lending behaviour should - in theory - be unaffected.

For example, if this policy were instituted in the US, simultaneously with a constitutional amendment saying that it could never be done again, owners of wealth might have the confidence to keep lending. In theory. Note that the amendment would probably need to be backed up with substantial regulatory legislation to protect taxpayers from the unlimited future losses that might result from the moral hazard created.

Or if the system were credibly recapitalised and regulated in such a way that creditors felt a financial bubble could never happen again, it might work. In theory.

But in reality, creditors who have just lost 10% - or 40% - of their money are unlikely to trust in such promises. Even in the unlikely event that they behave with perfect rationality, it's probably rational to wait a while before reinvesting to protect yourself against legislative instability. And they're not likely to be rational; they're likely to, frankly, fly into a tantrum and start pulling their money out of all sorts of places.

If only there were a way for governments to invisibly reduce creditors' claims without openly confiscating their assets - and without destroying future confidence in property rights. A way for creditors and taxpayers to wink at each other; colluding in sharing the pain, without losing the faith in credit that has built our economic system. Game theory suggests that rational players can lose a round or two if they believe the game will be played enough times in the future to make it worthwhile to keep playing.

There is, of course, such a way - inflation. 10-15% inflation over the next three years would probably haircut creditors sufficiently to eliminate the banking system's deficits. A number of respectable people (Scott Sumner for example) are arguing for an official, pre-announced inflation target (or alternatively a 3 or 5-year target for nominal prices) which would presumably give the banks enough breathing room to sort themselves out within that timeframe.

In rational terms, this is no different than chopping off 3% off creditors' assets each year; but it's within the rules of the 'con game' that is the world monetary system and economy (and I use that term with affection). And thus, it won't have the same destructive effect on confidence.

Of course it is a bit unfair; it forces creditors everywhere to subsidise the creditors who lent to insolvent banks; and it will take some time to work. It creates its own moral hazard problems too. But maybe we have to be utilitarian, let the kids threaten their temper tantrum and give in because it's just easier.

If President Obama and the IMF could just announce it all by Friday, Myron Scholes will have been right and I'll buy him a bottle of champagne. Let me know if you want to claim it, Professor Scholes.

4 comments:

jm said...

But how would you create the 10-15% inflation? You would need to expand the money supply. But because it is so dependent on the multiplicative effect of fractional reserve banking, that means you have to get people to borrow, and banks to lend.

Even supposing the banks were willing to lend, what compelling reasons would there be for people to borrow and spend enough to generate 10-15% inflation?

There is a glut of housing, much of it standing vacant. People can't live in more than one house at a time. Why should they buy another? To move up? That requires there be someone to buy the house they're moving out of. But the supply of young households not already owning homes was severely depleted by the bubble -- many who otherwise would have bought in coming years have already bought, and are now upside down and/or being foreclosed upon as they lose their jobs. The glut of supply and dearth of demand are too obvious for the housing bubble to be reinflated.

Indeed that was so as long ago as 2005, when the bubble peaked -- the reason the worst mortgage loans were made after that was that the pool of rational, qualified buyers had been heavily depleted, so that a larger and larger fraction of the sales and mortgages had to be made to the irrational and unqualified.

There is also a glut of cars and trucks. The fraction of the population that really needs a new car is smaller than ever.

Big-screen TVs? They're crashing in price because the Asians have created massive overcapacity to produce them -- and most everything else electronic, too. We're not going to get inflation in electronics.

And in general, Asia stands ready as a vast deflationary sink ready to absorb as many dollars as we can print.

It's hard to imagine how much money the government would have to print to get inflation started. And how it would get the money out into the pockets of people willing to buy.

Don said...

A Fable:

There was an even called the S & L Crisis:

http://www.uwsa.com/issues/cfr/quicksandl.html

"The S&L crisis was caused by economic factors but was greatly exacerbated by five specific policy decisions:

* Lending requirements were loosened as described above.
* Deposit Insurance was raised from $40,000 to $100,000.
* Enforcement of the law by banking regulators was decreased.
* Politicians actively interfered in investigations of failing thrifts on behalf of specific S&L owners .
* Implementation of the solution was delayed until the industry itself couldn't possibly pay for it and taxpayers were stuck with the bill."

And then:

http://www.washingtonpost.com/wp-dyn/content/article/2008/11/24/AR2008112402593_pf.html

"Looser enforcement of banking laws would eventually allow Citi to expand into other forms of banking. And when regulations got in the way, Wriston, like Mitchell before him, found ways around them.

At the time, national banking regulations kept Citi out of the market for commercial paper -- a form of debt sold by companies to help meet short-term credit needs. Citi figured out that those restrictions applied only to multi-bank holding companies and set up Citicorp as a single-bank holding company, with Citibank as its subsidiary.

In the 1980s, Citi and other banks ran into trouble over loans made to developing countries in Latin America on the assumption that sovereign nations don't go bankrupt. Citi and other banks faced billions of dollars in potential losses when Mexico became insolvent in 1982. The government stepped in with the Brady Plan to help sell Latin American nations' debt and have the banks write off some of it.

Citi needed help again during the recession of the early 1990s. It accumulated $10 billion in bad loans, many related to leveraged buyouts or commercial real estate. Its stock price fell, and rumors of bankruptcy began to circulate. The Federal Reserve came to the rescue with several interest rate cuts.

To weather future crises better, Citi executives believed the company needed to become even larger and more diversified, a true one-stop financial services firm where consumers could go for checking, brokerage and insurance services."

And then:

http://www.gao.gov/archive/2000/gg00067r.pdf

11. Did the Federal Reserve’s intervention create new incentives
for other large financial institutions to take huge financial market
risks in the future?
Any type of intervention creates the potential for increased moral hazard;
however, the long-term implications of FRBNY’s involvement in the
recapitalization are unknown. Although the FRBNY stressed that its
actions were dictated by the state of worldwide financial markets at that
time, its actions raised concerns among some industry officials about
moral hazard. Some industry officials said that FRBNY’s involvement in
the rescue, however benign, would encourage large financial institutions to
assume more risk, in the belief that the Federal Reserve would intervene
on their behalf. According to FRBNY officials, it is unlikely LTCM’s
creditors would have been able to work together to avoid the rapid
liquidation of the Fund if FRBNY officials had not intervened. Thus,
FRBNY’s intervention probably affected the outcome in this case and, over
time, such actions could increase moral hazard and potentially undermine
the effectiveness of market discipline.
12. Did the Federal Reserve’s intervention in the rescue of LTCM
create unacceptable risks to the federal deposit insurance system
or expose American taxpayers to a threat of future hedge fund
bailouts?
Enclosure
Questions Concerning LTCM and Our Responses
Page 15 GAO/GGD-00-67R Questions Concerning LTCM and Our Responses
Although no federal dollars were involved in the recapitalization of LTCM,
the Federal Reserve’s involvement has raised concerns among some that
the “too big to fail” doctrine has been expanded to include hedge funds.
Federal Reserve officials have testified that its facilitation of the
recapitalization of LTCM was not an expansion of “too big to fail”13 and had
the private-sector recapitalization not come together, LTCM would have
been allowed to fail. However, if companies believe that the federal safety
net has been expanded, it may encourage more risky business practices.
Based on the LTCM experience, if problems surface during periods of
market turmoil, regulators may decide that some form of federal
intervention, albeit nonfinancial, may once again be necessary."

And then, in 2005, this was not posted on the internet:

http://www.msfinance.ch/pdfs/AnnelisLuescher.pdf

"It should by now have become clear that the Gauss one factor model is not the
right model to price CDOs and that we are in desperate need for a convincing
alternative. In this thesis, three promising models for pricing synthetic CDOs
were for the first time compared to each other and to the Gauss one factor
model. Further tests using current market data are necessary before we can
conclude on a superiority of any of the three models."

These are things which couldn't have happened, since they show that moral hazard matters, and that there simply wasn't enough data, under the best of circumstances, to rely on CDO risk models. These are such obvious, prudent points, that, in the real world, where people aren't idiots, the current crisis couldn't have occurred. And you surely couldn't have had a housing bubble, following on a tech bubble. That would have shown that bubbles can still occur. And people couldn't have loaned money to the riskiest borrowers at the top of a market, with loans that didn't cement in low rates.

This is all a fable. A counterfactual. Right? But I'm a Modal Realist, so, it did happen in some world, just not ours.

Remember, I'm a novelist. In fact, I write horror and satire. I'll let the reader decide which this fable qualifies as.

Don the libertarian Democrat

Anonymous said...

I think the Sumner suggestion is brilliant. As to how to create the inflation - read his blog posts. It is really a matter of Fed communication strategy - if they promise inflation we will expect it and then we will get it - because our expectations will lead us to borrow and spend when we might not have done so otherwise. This is the monetary theory of Michael Woodford - expectations of future prices matter more for future prices than present prices do.

It is also a matter of credible signaling theory. A credible signal is one that a person believes exactly because they know that the giver of the signal can make the signal come true no matter what you or any one else does. And if the Fed purchases assets on a large enough scale we sure do know that they can create inflation all by themselves. Or they could finance the whole Federal deficit with fresh cash. We do know for sure that the government will spend that.

If the Fed gives a press statement promising that the U.S. price level will be x percent higher y years from now - and they promise to buy assets on whatever scale necessary to do that - and if they stop paying interest on excess reserves and rather start charging interest on them - expectations will change - behavior will change - and they likely wont have to print any new money at all.

And as he also says - Roosevelt did this by going off gold early in his first term - when the banking system was worse than it is now. Inflation first - then the recap.

Doug said...

Well, it's Sunday now. Glad that whole mess is over with.