May 13, 2008
Credit crunch: how we got here and how to get out
Fed Chair Ben Bernanke on Tuesday offered his perspective on the appropriate response of the Fed to the ongoing turmoil in financial markets. I still think he's overlooking a key element of what's been happening.
Meeting creditors' demands for payment requires holding liquidity-- cash, essentially, or close equivalents. But neither individual institutions, nor the private sector as a whole, can maintain enough cash on hand to meet a demand for liquidation of all, or even a substantial fraction of, short-term liabilities. Doing so would be both unprofitable and socially undesirable. It would be unprofitable because cash pays a lower return than other investments. And it would be socially undesirable, because an excessive preference for liquid assets reduces society's ability to fund longer-term investments that carry a high return but cannot be liquidated quickly.
However, holding liquid assets that are only a fraction of short-term liabilities presents an obvious risk. If most or all creditors, for lack of confidence or some other reason, demand cash at the same time, a borrower that finances longer-term assets with liquid liabilities will not be able to meet the demand. It would be forced either to defer or suspend payments or to sell some of its less-liquid assets (presumably at steep discounts) to make the payments. Either option may lead to the failure of the borrower, so that the loss of confidence, even if not originally justified by fundamentals, will tend to be self-confirming. If the loss of confidence becomes more general, a broader crisis may ensue.
Bernanke concludes that it's the responsibility of the central bank to stop such self-fulfilling instability. But he neglects to discuss the key feature of a healthy financial system that is supposed to prevent such a problem from ever arising. Specifically, any institution that is in this position of borrowing short and lending long needs to ensure that a certain fraction of the funds it is lending came not from borrowers but instead from the owners of the institution itself, in the form of net equity. The goal is for the size of this net equity to be larger than the losses the institution would incur from selling its less-liquid assets at steep discounts. As long as it is, no creditors ever have reason to demand cash, and there would be no need for the central bank to step in to prevent a self-fulfilling breakdown.
And the core reason we are in the mess we are today is that these equity stakes were nowhere near sufficient for this purpose. Instead, financial institutions were allowed to take highly leveraged positions whose details are largely opaque to readers of publicly available financial statements. Exhibit A here might be Bear Stearns, whose 2007 10-K reported that Bear had outstanding derivative contracts whose notional value was $13.4 trillion. Much of these were credit-default swaps, in which the seller receives a fee in exchange for promising to pay any losses incurred by the buyer on some specified asset and time interval. If every such asset lost 100% of its value over the period, then maybe Bear is supposed to pay or receive $13.4 trillion. In practice, the actual price moves and net sum owed would be a small fraction of that notional total.
Now, there is nothing inherently wrong in making financial investments in the form of derivative contracts rather than outright loans. You're doing something similar whenever you buy or sell an option rather than the stock itself. But, if you were to sell an option through an organized exchange, the exchange would require you to satisfy a margin requirement, delivering for safekeeping good funds such that if the price of the underlying asset against which the derivative is written moves against you, you are able to make good on your commitment.
If anything like a reasonable margin requirement had been in effect, Bear Stearns could not possibly have gotten into contracts totaling $13.4 trillion notional. But these weren't traded on a regular exchange, so there was no margin requirement, and apparently no real limit on the size of the exposures that Bear Stearns could take on, or the size of what they could bring down with them if they fell.
And that raises the question, Why were counterparties willing to accept these trades with no margin to guarantee payment? To this I'm afraid the answer is, they figured Bear was too big for the Fed to allow it to fail. And on this, I'm afraid they proved to be exactly correct.
I would feel better if Bernanke were less focused on how to "provide liquidity" and more focused on how to get the system deleveraged and more transparent.
Posted by James Hamilton at May 13, 2008 08:16 PMdigg this | reddit
Listed below are links to weblogs that reference Credit crunch: how we got here and how to get out:
» Leverage, Bear Stearns & Econbrowser from PrefBlog
The usually reliable Prof. James Hamilton of Econbrowser disappointed me today with a rather alarmist post on leverage and Bear Stearns: And the core reason we are in the mess we are today is that these equity stakes were nowhere near sufficient for th... [Read More]
Tracked on May 14, 2008 12:48 PM
» May 15, 2008 from PrefBlog
I’m becoming more and more convinced that the Credit Crunch has evolved from its fundamental role as Reducer of Excesses to a new position as Political Football. My thoughts on this are influenced by many things. For instance, compare the sub-pr... [Read More]
Tracked on May 15, 2008 06:53 PM
Your argument is perfectly logical and, as you point out, Bear couldn't have possibly racked up that kind of exposure if it had been required to post a sufficient margin to cover...which is likely why it didn't happen.
Posted by: Gegner at May 13, 2008 10:51 PM
Jim -- You don't want to overlook this part of the speech:
"Although central banks should give careful consideration to their criteria for invoking extraordinary liquidity measures, the problem of moral hazard can perhaps be most effectively addressed by prudential supervision and regulation that ensures that financial institutions manage their liquidity risks effectively in advance of the crisis. Recall Bagehot's advice: "The time for economy and for accumulation is before. A good banker will have accumulated in ordinary times the reserve he is to make use of in extraordinary times" (p. 24). Indeed, under the international Basel II capital accord, supervisors are expected to require that institutions have adequate processes in place to measure and manage risk, importantly including liquidity risk. In light of the recent experience, and following the recommendations of the President's Working Group on Financial Markets (2008), the Federal Reserve and other supervisors are reviewing their policies and guidance regarding liquidity risk management to determine what improvements can be made."
The fire prevention efforts sometimes get less devotion in the middle of the blaze, but that doesn't mean they've been forgotten.
Posted by: Dave Altig at May 14, 2008 04:30 AM
Lending standards are simply stupid, stupid, stupid, stupid if you actually look at the properties of debt at different rates.
I did a post with my suggestions for sensible lending standards, http://makingsenseofmyworld.blogspot.com/2008/04/sensible-lending-standards.html.
This was the third in a series, the first was looking at Canadian housing bubbles that I've seen, http://makingsenseofmyworld.blogspot.com/2008/04/what-can-housing-bubble-teach.html.
The second was my thesis as to why low interest rates are far more damaging to the economy from a household perspective.
Posted by: Deborah at May 14, 2008 06:02 AM
A problem arises with you proposal, considering that allegedly there is 700 trillion in derivatives outstanding(15 times global gdp), and amazing and scary number. Losses but not all are hidden in level 3, can you figure out the cost of that setlement and who will pay for it? Besides the delevaring of that would take years to normalize the financial system impacting in the real economy.
Posted by: carmelo cortes at May 14, 2008 08:05 AM
Carmelo Cortes, what do you think of this idea? Declare that effective (say) January 2009, any new derivative contracts entered into after Jan 2009 that are outside of organized exchanges will not be enforceable through U.S. courts.
Posted by: JDH at May 14, 2008 08:23 AM
Makes perfect sense, Professor. The leverage/exposure ratios at the banks, FNMA, and insurers (e.g., MBIA) are incredible, and their equity cushions will evaporate quickly in time of trouble (coming soon).
When do you get a voting seat on the Board of Governors, to bring some clear, persuasive thinking to the group?
Posted by: jg at May 14, 2008 08:39 AM
And that raises the question, Why were counterparties willing to accept these trades with no margin to guarantee payment?
Counterparties to credit derivatives contracts, for the most part, have been investment and money-center banks. Their willingness to make private derivative contracts with each other with little margin coverage are manifold:
- The fees generated by coining the derivative offer instant profit to the issuer.
- Many derivative contracts (CMOs CDOs etc.), artificially enhanced the surety of the underlying security (increased the security rating). This allowed the issuer to arbitrage the point spread between the underlying security as issued, and the new derivative product.
- Banks structured many of the most toxic of these derivatives in off- balance sheet entities (SIVs). By so doing, the banks thought that they could isolate their liability to fraud litigation, in the event that the derivative securities sold to third parties lost value.
- Derivative contracts offered a method of increasing bank investment leverage. Because OTC derivative contracts are not regulated, reserve requirements for their issuance are absent.
- Derivative contracts without reserves offer a means for issuing banks to increase money supply (credit), independent of central bank regulation and control.
It will take considerably more than the Federal Reserve to get the Derivatives Mess under control. Notional value of all derivatives in force is now greater than $600 Trillion. It will take a conscientious effort by central bankers and governments world-wide more than a decade to agree on a regulatory framework as well as to create exchanges to force derivative transactions into the sunlight.
I will not hold my breath waiting for the Federal Government to force the Banks to
cough-up the profits made in spawning fraudulent securities, or government forcing the fast unwinding of much of the existing credit derivative market...Congress is far too ignorant and too ideological to write and pass enabling legislation. I plan on a long drawn-out ordeal with lots of political jockeying, obfuscation, and ultimately NO MEANINGFUL ACTION until the financial system melts-down in crisis… The money-center banks will not easily relinquish the power and freedom they accrued in the repeal of Glass-Stiegel or the fast profits they made in its destruction.
Posted by: MarkS at May 14, 2008 10:40 AM
Seems like the analog is insurance.
We have state regulation of insurance issuers to require adequate liquidity. One would hope that any sane insurance commissioner would never allow this lack of coverage.
Insurable events tend to be random and not corrolated so that one can develop actuarial predictions of cash outlays. Mortgage failure rates have tended to have some underlying corrolation in the business/employment cycles but even those have a long history and data set.
What's new is a common mode failure that's not random and not business cycle related. Seems to me that the actual failure rates are not so high to cause the upset. Rather, it is the lack of understanding and predictability of the common failure mode.
In other words, a lack of transparency is making matters worst than they need to be.
Of course, the Professor has said as much here before. It's just soaking in to my brain.
Posted by: Joseph Somsel at May 14, 2008 10:49 AM
Declare that effective (say) January 2009, any new derivative contracts entered into after Jan 2009 that are outside of organized exchanges will not be enforceable through U.S. courts.
If somebody wishes to guarantee his son's car loan, will an exchange listing for this derivative be required?
Posted by: James I. Hymas at May 14, 2008 11:08 AM
James Hymas, what if we specified contracts with notional value in excess of $1 million, say?
Posted by: JDH at May 14, 2008 11:51 AM
James_Hamilton: I get what you're trying to accomplish with the lifting of US Court enforcment change, but you have to plug a lot more holes than that. What about co-signing on a loan for my business partners to buy some new office equipment etc?
Then again, I didn't know a co-sign on a car loan counted as a derivative? Does a bank have to list the estimated value of "Daddy's promise to pay whatever Junior can't"?
Posted by: Silas Barta (formerly Person) at May 14, 2008 12:18 PM
what if we specified contracts with notional value in excess of $1 million, say?
That would work, but it seems to be a very complex way of moving the problem from SEC/Dealer to SEC/Exchange/Dealer.
I didn't know a co-sign on a car loan counted as a derivative
I'll admit, I was being mischievous. But why not? The bank has made a loan to Junior and bought credit protection from Dad (presumably at a rate of 0bp). The guarantee is a Credit Default Swap in all but name, the bank's position as a whole is a basis package.
Posted by: James I. Hymas at May 14, 2008 01:35 PM
Are you suggesting that the government set interest rates based on the length of the mortgage? I assume you mean at a fixed rate? What do you do with a variable rate mortgage?
Posted by: DickF at May 14, 2008 02:48 PM
Part of the problem may be asymmetrical executive compensation that encourages too much risk - stock options give executives huge rewards on the upside, but the punishment for losing is much, much smaller.
Posted by: don at May 14, 2008 03:54 PM
I am suggesting that the maximum mortgage term allowable for mortgages adjust with rates. As rates decline, the maximum length of the term declines.
I worked in banking. People used to be motivated to pay back debt faster.
Posted by: Deborah at May 14, 2008 04:11 PM
A rather artful comment by Dave Altig of the Federal Reserve. One might be more impressed if the Fed had been doing its fire prevention job during the last 5 years, insstead of providing accelerant.
Posted by: Paul at May 14, 2008 04:55 PM
" he neglects to discuss the key feature of a healthy financial system that is supposed to prevent such a problem from ever arising "
JH.. Of course he neglected to do this. He's a part of the same game that Greenspan was playing. "It's not my fault, we couldn't do anything about this.. so reward us for sleeping on the job for 6 years by giving us more power!"
If Bernanke actually wanted to restore the banking system, he'd enact these very same procedures that you describe. But he doesn't.
Eventually, actions define the man.
Posted by: Unsympathetic at May 15, 2008 02:16 PM
So, re Ben's latest remarks. do you suppose he visited your website? It is disappointing, though, that he gives mere lip service to the problem.
Posted by: don at May 15, 2008 08:28 PM
James_I._Hymas: I'll admit, I was being mischievous. But why not? The bank has made a loan to Junior and bought credit protection from Dad (presumably at a rate of 0bp). The guarantee is a Credit Default Swap in all but name, the bank's position as a whole is a basis package.
Just to clarify, I do understand how, seen from the bank's perspective, Daddy's promise is a derivative, similar to those really traded. I just didn't know if banks' balance sheets had to count it that way. That is, do they just value the loan as a whole (with the co-signing), or do they have to say, we have this junk loan to some kid, PLUS a derivative that happens to be valuable enough that value(Daddy derivative) + value(junior loan) just happens to have the value of a prime loan?
Posted by: Silas Barta (formerly Person) at May 16, 2008 11:08 AM