January 31, 2010
John Cochrane on the credit crisis
The signature event of this financial crisis was the "run," "panic," "flight to quality," or whatever you choose to call it, that started in late September of 2008 and receded over the winter. Short-term credit dried up, including the normally straightforward repurchase agreement, inter-bank lending, and commercial paper markets. If that panic had not occurred, it is likely that any economic contraction following the housing bust would have been no worse than the mild 2001 recession that followed the dot-com bust....
Why was there a financial panic? There were two obvious precipitating events: the failure of Lehman Brothers investment bank in the context of the Bear Stearns, Fannie Mae, Freddie Mac and AIG bailouts; and the chaotic days in Washington surrounding the passage of legislation establishing the Troubled Asset Relief Program (TARP).
Why would Lehman's failure cause a panic? Why, after seeing Lehman go to bankruptcy court, would people stop lending to, say, Citigroup, and demand much higher prices for its credit default swaps (insurance against Citi failure)? Nothing technical in the Lehman bankruptcy caused a panic. The usual "systemic" bankruptcy stories did not happen: We did not see a secondary wave of creditors forced into bankruptcy by Lehman losses. Most of Lehman's operations were up and running in days under new owners. Lehman credit default swaps (CDSs) paid off. Sure, there was some mess-- repos in the United Kingdom got stuck in bankruptcy court, some money market funds "broke the buck" and had to borrow from the Fed-- but those issues are easy to fix and they do not explain why Lehman's failure would cause a widespread panic. What is more, Lehman's failure did not carry any news about asset values; it was obvious already that those assets were not worth much and illiquid anyway.
We are left with only one plausible explanation for why Lehman's failure could have had such wide-ranging effect: After the Bear Stearns bailout earlier in the year, markets came to the conclusion that investment banks and bank holding companies were "too big to fail" and would be bailed out. But when the government did not bail out Lehman, and in fact said it lacked the legal authority to do so, everyone reassessed that expectation. "Maybe the government will not, or cannot, bail out Citigroup?" Suddenly, it made perfect sense to run like mad....
Let us go back one step further. Why did Lehman fail-- along with Fannie Mae, Freddie Mac, AIG, Wamu, and very nearly Citigroup and Bank of America? Here is where I part company on the usual worries about bubbles, imbalances, silly mortgages, and so on.
The underlying decline in wealth from the housing bust was not that large.... Most estimates put subprime losses around $400 billion. The stock market absorbs losses like that in days. But it turned out that housing risks are spread very differently from stock market risks.
The difference is that mortgages were held in very fragile financial structures. An extreme example: many mortgages were pooled into securities, and the securities were held in special purpose vehicles (SPVs), funded by rolling over short term commercial paper with an off-the-books credit guarantee from a large bank. Less extreme: when Bear Stearns failed, it was holding a large portfolio of mortgage-backed securities (MBSs) funded at 30-to-1 leverage by overnight debt. In both cases, when the mortgages lose value, the debt-holders refuse to renew their loans and the whole thing blows up. In contrast, when your (and my) pension account loses value, we cannot run for the exits and try to make someone else hold the losses.
John raises a number of interesting points. But I must say that I put more weight than he does on "the usual worries about bubbles, imbalances, silly mortgages, and so on." The problem is not just the fragility of the system in 2008, but the huge flows of capital into these devices in the years leading up to the crisis. U.S. home mortgage debt grew much more quickly than income, with almost $8 trillion in gross new U.S. household mortgage debt issued between 2004 and 2006 alone.
A graph of real home prices is remarkable not just for the dramatic descent since 2006 (which began two years before John picks up the story) but moreover for the meteoric rise between 2000 and 2005.
My view is that there were very serious problems with the U.S. financial system prior to 2008. But I think John has correctly described part of the reason why the failure of Lehman was such an important event.
Posted by James Hamilton at January 31, 2010 06:34 AMdigg this | reddit
Nicely put. If one accepts the view that new information arrived with the Lehman failure that caused a recalculation of asset prices that sort of begs the question of what was happening to information sets and prices in the run up before that.
Posted by: Robert Bell at January 31, 2010 07:41 AM
Can a case be made that the '01 recession never
ended ? If yes (even with qualifications), then
one can understand the meteoric rise in terms of
the political necessity to convince voters that
it did end ("the tax cuts are working" and other
Posted by: Peter at January 31, 2010 07:54 AM
On this subject, I believe everyone is right. And trying to identify just one layer in the Ponzi scheme is inadequate.
I would call Lehman the "catalyst" that got everyone to question TBTF, tho I don't really see much difference in the actual process with Lehman and the rest that went down. The USG told Bear "you are now being taken over and your stock is $2." In the case of Lehman, they said "we can't bail you out", then Barclays, after first declining to be the official receiver of Lehman, came in and picked up the pieces they wanted.
I think the author is a little light on this figure. " Most estimates put subprime losses around $400 billion.". At the start of 2009, both Roubini and the IMF put US(only) banking losses at $1.1T. It probably wasn't all official subprime, but lax underwriting and basically fraudulent securitization made what was once subprime standards lending lending into AAA paper. Then there is the uncertainty about when it finally ends. Not yet, imo.
He's right about equities being less intertwined in the workings of the financial sector and the economy. Simple reason is stockholders are easy to kill(and no one cares unless they are one), but AAA debt is perfectly fine for banks to hold as "capital". Except when it's there because IT IS masking "bubbles, imbalances, silly mortgages, and so on".
And funding 30:1 leverage and SIVs containing this stuff with overnight money , at rates and oversight(none from official sources, and it sounds like the CEOs were out golfing) extended to the highly rated bank holding company, is pure insanity.
Posted by: Cedric Regula at January 31, 2010 08:05 AM
who believes this stuff ?
little factor, goes by the name "RISK" - ever hear of it ? means you can't make unlimited money forever at any percentage rate.
Posted by: biker at January 31, 2010 08:21 AM
Rien ne se perd, rien ne se cree, tout se transforme :
The metrix of this financial crisis should be much larger,in components,in participants, in instruments.
GDP breakdown C+I+ (Gv expenditures - taxes)
Evolution on a 10/15 years life span
Current account,balance of trade
Evolution on a 10/15 years life span
Liquidity,money supply,central banks.
Evolution on a 10/15 years life span when compared to GDP
see FRED and
Evolution The Evolution of the US Financial Industry from 1860 to
2007: Theory and Evidence.
New York University, NBER, CEPR
see above and FRED http://research.stlouisfed.org/fred2/series/LLRNPT5
see oocc derivatives report
securitization see litigations
For the spirit of this crisis see: Fantasia Walt Disney
Posted by: ppcm at January 31, 2010 08:22 AM
Forgot one key element. Then someone comes along and says "We'll insure it all!".
Posted by: Cedric Regula at January 31, 2010 08:23 AM
it's interesting (or not) how little discussion there has been regarding lehman and _why_ lehman was so important. people take for granted that it was a disaster, but there has been very little nuanced discussion.
i don't have a very complete picture, but i think this had something to do with it: http://brontecapital.blogspot.com/2008/10/why-lehman-mattered.html
Posted by: q at January 31, 2010 09:18 AM
Suprised not to see mentioned the termination of the uptick rule for shorting markets in 2007 and the imposition of marking to market certain assets way before Lehman bk. Some groups understood the extreme overpricing and extreme weakness of the system. There were bear raids all over the place, leading to collapse in stock prices, leading to worsening balance sheets, leading to downgrades of debt, leading to a quest for more collateral that was not available creating a cascade of forced selling, leading to more naked short selling, leading to triggering credit default swaps, all in a vicious cycle... exposing complete fragility across the board. Lehman was just one six letter word in the midst of this mayhem. And everyone thought after it there was really no safe harbor. Proof: the corrective measures taken later (banning shortselling, terminating mark-to-market accounting, etc.).
It all looked like an experiment gone wrong.
Posted by: Simpleton at January 31, 2010 09:58 AM
Those are key issues too, and I think the answer is we need to take a look at what our core definitions of "capital", and also maybe even more importantly what "insurance", are.
Mark-to-market was a logical move considering that securitization gave us instruments that banks could hold as Tier I and 2 capital. But they fluctuate in value, and banks still need a liquidity buffer, so mark-to-market supposedly valued securitized paper at what it's liquid (sell at any point in time) value is. Makes sense, except the paper turned out to be crap, and the problem we face in future is interest rates can only go up. But you can still insure it!
There there is CDS. Critics say the industry calls them "swaps" instead of "insurance" because then they don't get regulated as insurance, even tho they function as insurance. True, but you
could say the same about collars in the commodity futures markets, or the re-insurance business diversifying risk of conventional insurance.
So some re-definition is in order here, tho I don't know what. But CDS is a large part of the derivospere, now estimated to be 10X global GDP, so what may be lurking out there is scary.
I too have been wondering what was so important about the Lehman tragedy, relative to everything else. Maybe I'll find out when I get to reading Q's link next.
Posted by: Cedric Regula at January 31, 2010 11:37 AM
I have the Baltic Dry index in free fall by late July 2008, down 40% by Sept 2008. So, using cause and effect, there must have been something Lehman was supposed to do in May or June of 2008 in expectations of a trade collapse in July.
Posted by: Anonymous at January 31, 2010 12:20 PM
It's interesting that mainstream economists are finally emphasizing household debt. I knew you all would come around sooner or later.
But the problem is really much wider than household debt. For instance, Lehman's exposure was mostly to CRE. And don't forget the leveraged buyouts and the huge junk bond market, currently supported by ocean of liquidity. Now we have unsustainable sovereign debt. And China is coming along nicely as the next leg of the debt crisis. Just another $T of new loans this year should do the trick.
Posted by: Bob_in_MA at January 31, 2010 01:18 PM
There seems to be a contradiction in Cochrane's argument. He claims that, after Lehman, everyone's expectation of bailouts changed and which resulted in panic. But the government immediately began to bail out everybody -- Goldman Sachs, JP Morgan, AIG, GM, Chrysler, Fannie and Freddie. If diminished expectations of a bailout caused the panic, why didn't all of these extraordinary bailouts end the panic?
Further, Cochrane goes on to say that we must put into place legal guarantees that we will never bailout again so the banks will believe it.
He can't have it both ways -- saying that the banks didn't believe that there would be a bailout, which caused a panic, but now the banks have to be convinced that there won't be a bailout.
Posted by: Joseph at January 31, 2010 04:22 PM
I think it's interesting that everyone now talks as if the financial crises is over. That somehow the worst is behind us. I my opinion who really cares what the "signature event" was. Lehman was a casualty, not a significant event. Theoretically, every wall street investment bank should be dust right now. The US and the Fed stepped in and gave bank status to former investment "casino" banks in order to prevent the inevitable. The consequences from such actions won't even begin to be realized down the road, when they're is an even bigger crises from all the moral hazard that has taken place. Lehman Brothers is gone, does that affect anybody's daily life. Not really, most people haven't even had a dent. So, why is it we think that if none of the investment banks existed it would be the end of the world? Jim Rogers says the same things over and over again every time he goes on Bloomberg or CNBC but no one really listens to him. I agree with him. Investment banks have failed in the past and they should be allowed to fail in the future.
Posted by: Niki Arinze at February 1, 2010 01:53 AM
Anonymous: I don't get the connection to the BDI. That index usually follows commodity prices. Oil and copper spot prices started their big plunge in July: hence the BDI fall. Is there something about Lehman's involvement in trade finance or selling oil hedges that you're alluding to?
Posted by: sjp at February 1, 2010 08:47 AM
Basically agree with pretty much all you say here, Jim. Let me simply note that housing prices began to rise noticeably (and the price to rent ratio as well) in 1998, not in 2000, although nobody noticed it at the time as they were so fixated on the much more dramatic dot.com bubble that was nearing its climax.
Posted by: Barkley Rosser at February 1, 2010 10:01 AM
While the Dow Jones Industrial Average is not a perfect indicator of events it is probably the best indicator we have of a day by day hour by hour assessment of the impact of economic events. If the DJIA is considered in this light most of the popular mythology concerning the economic crash in September – October 2008 falls away.
Let’s look at the major events for the period in question for September – early October, 2008.
Friday, September 5, 2008
Rumors are swirling the last business day before Fannie and Freddie go into conservatorship. Sunday, September 7, 2008 they are taken over by the federal government.
DJIA Open 11,185.63
DJIA Close 11,220.96
Daily Change +35.33
Monday, September 8, 2008
The market reacts to the demise of Fannie and Freddie. DJIA opens up and climbs throughout the day.
DJIA Open 11,224.87
DJIA Close 11,510.74
Daily Change +285.87
Monday, September 15, 2008
Lehman Brothers announces bankruptcy
DJIA Open 11,416.37
DJIA Close 10,917.51
Daily Change -498.86
Tuesday, September 16, 2008
Barclay’s Bank announces it will buy Lehman. DJIA begins recovery.
DJIA Open 10,905.62
DJIA Close 11,059.02
Daily Change +153.40
Friday, September 19, 2008
By Friday immediately before the Lehman purchase is approved on Saturday the DJIA regains strength to slightly below the level before Lehman’s announcement. The shock of the Lehman bankruptcy is virtually over.
DJIA Open 11,027.51
DJIA Close 11,388.44
Daily Change +360.93
Monday, September 29, 2008
The House rejects the TARP bill in afternoon 2:00 PM. DJIA falls all day and is down significantly after vote to close. News media blames House vote. A CNNMoney.com poll asks if the house should reconsider its TARP vote and 58% respond no.
DJIA Open 11,019.04
DJIA Close 10,365.45
Daily Change -653.59
Tuesday, September 30, 2008
DJIA opens higher than Monday close and rebounds almost totally recovering Monday’s loss after investors digest House vote.
DJIA Open 10,371.58
DJIA Close 10,850.66
Daily Change +479.08
Wednesday, October 1, 2008
McCain forces Obama back to Washington DC for TARP vote. Both vote yea against the wishes of most Americans. McCain proves himself to be no different from Obama essentially guaranteeing his loss in the presidential race.
DJIA Open 10,847.40
DJIA Close 10,831.01
Daily Change -16.33
Friday, October 3, 2008
House reverses itself voting for Senate amendment creating TARP at 2:22 PM. The DJIA began to slide immediately after House vote. It had been up about 200 points but declined 350 points before the close. Bush signed the bill late in the afternoon.
DJIA Open 10,483.96
DJIA Close 10,325.38
Daily Change -158.58
Monday, October 6, 2008
The market spent the weekend digesting the passage of the largest bailout in US history. The DJIA opened up down slightly but closed the day down sharply closing below 10,000 for the first time in years.
DJIA Open 10,322.52
DJIA Close 9,955.50
Daily Change -367.02
Tuesday, October 7, 2008
The market resumed its decline that continued through November falling below 8,000.
DJIA Open 10,955.42
DJIA Close 10,447.11
Daily Change -508.31
Based on the DJIA as the indicator there can be no doubt that the economic crash of 2008 was caused directly by the passage of the TARP bill and not as popular myth has come to promote, the bankruptcy of Lehman Brothers.
Posted by: RicardoZ at February 1, 2010 10:30 AM
In case anonymous has disappeared into the ether, I took his comment as sarcasm. In other words, weakness in the economy and fragility in the financial system began to come to a head prior to the failure of Lehman, and Lehman was one of the weak links in the chain that broke first.
Or the way Warren buffet puts it, when the tide goes out, we see who has been swimming naked.
Very true, and there are some reasons for it too. One was, starting around 1995, the Admin and some members of Congress started arm twisting the GSEs to get into subprime lending. The GSEs began ramping it up in earnest thru the later 90s. Then there where some tax law changes in '97, I believe, intended to stimulate the builders.
After the mini boom-bust of 89-90 (along with the S&Ls, RTC, etc...) the housing market was weak in the early 90s and the USG thought it needed help.
Posted by: Cedric Regula at February 1, 2010 11:00 AM
What anonymous is getting at is that the BDI was signaling economic trouble long before Lehman failed. As was the rise in the dollar and the fall in commodity prices. Lehman was a victim of the economic slowdown that was already evident in July. I agree with Ricardo Z that the cause of the crisis phase was TARP and the uncertainty surrounding the election of Obama. In fact it seems the passage of TARP is what set off alarm bells. I think there was an expectation that something concrete would happen after it passed Congress and it quickly became clear that Paulson & Co. had no idea how to implement the thing. Obama's election just added uncertainty since no one really knew what his policies would be.
As for whether anything could have been done about the drop in NGDP expectations well visit Scott Sumner's blog for a lively debate. He thinks the Fed could have prevented the crisis with monetary policy; I'm not so sure.
Posted by: Joe Calhoun at February 1, 2010 01:49 PM
Here's an article about how the SEC, in 2004, raised allowable leverage for the 5 big IBs from 12:1 to 40:1. This is a meeting that Paulson hasn't talked about lately. Maybe it makes him swoon to think about it.
Of course running leverage is nowhere near as much fun with short rates at, say, 5% rather than a tad over 1%.
Posted by: Cedric Regula at February 1, 2010 03:29 PM
The focus on system fragility is good, but it isn't the linchpin. And identifying the linchpin, or root cause, is important to shape policy.
All these issues can't be the root cause: institutions and some markets (SIVs, counterparty credit arrangements, etc.) were overleveraged +/or poorly structured, so they blew up when asset values plunged.
Do you think there would have been a much lesser impact on aggregate demand and wealth if Street firms had been leveraged 20:1 rather than 45:1? Would Fannie & Freddie be ok today had they been leveraged 30:1 rather than @70:1? Would the bond insurers had survived had they been leveraged 50:1 rather than 250:1?
Equity on the US private household balance sheet was sliced in half from end of '06, predominantly as a function of res. real estate and equity price changes. Why?
The root cause was the collision of (i) a resource constraint with (ii) mis-allocation of credit, most importantly in US residential real estate.
Resource pricing, primarily oil, provided new information over the course of late '06 to late '07 that global growth & income expectations were wrong -- a global top line 4.5% - 5.5%, and importantly, what that implied for corporate profits & household incomes, was incorrect and the new number would be lower, @1.75% - 2% or so.
In other words, the parabolic change in resource pricing wasn't coincidental but causal. Anecdotally, it was most evident in small business margins when US gas prices shot up.
This shift in growth/income expectations caused the wealth effect from repricing of assets globally, not only Western real estate (incl Spain, Ireland, etc.) but most other items which had greatly risen in price...seafood, art, wine, gambling activity, prestige automobiles, etc.
And when the music stops playing, the guy in the room with the highest leverage and the highest fixed expenses is screwed.
That's the US, as a function of the long period of very accomodative real interest rates, years of extraordinary bipartisan meddling in housing credit, and perhaps cultural factors.
Posted by: Tim VH at February 1, 2010 09:19 PM
"the usual worries about bubbles, imbalances"
Professor, you forgot to include an analysis of one key part of Cochrane's (flawed) reasoning. He's a believer in EMH.. so of course any analysis of a bubble or imbalance is immediately thrown out, because all information is correct and all players have access to all information!
Of course, any facts that run contrary to the EMH must be fantasy.. because the EMH is an immutable law of nature!
Until it isn't.
Please remember that a belief in EMH is part of the requirement for tenure in the UC econ dept.
Posted by: Unsympathetic at February 2, 2010 09:51 AM
Little noticed in the above discussions was the the sword of Damocles hanging over the markets: trillions in naked CDS contracts that could become triggered by cascaded bankruptcies. The bankruptcies of Ibanks themselves would just involve some paper shuffling and high fees for some lawyers. These naked CDSs, however, involved trillions in unhedgeable side bets(industry claims notwithstanding: markets are not complete and filtrations are anything but Brownian). Even a cascade of well-deserved Ibank bankruptcies and the cascade of those onto other sectors would not per se create a true, lasting crisis. But calling the collateral on the naked CDS contracts involved in such cascades would precipitate a very deep, Depression-level crisis that would make the Panic of 1907, which led to the banning of similar products, look like child's play. This specter was ever in the background of discussions surrounding AIG & TARP; and it was that specter, not TARP and not events leading up to it, that was spooking the markets, esp. from Oct. '08 through Feb '09. CDS spreads exploded despite Fed, BOfE, etc. interventions. As they did, corporate grade bonds tanked vs. Treasuries and implicit asset levels were driven down pari pasu, driving down stock values along with them (recall that a stock is, per the Merton model, is just a call on the assets with strike equal to the bonds' par).
Posted by: d4winds at February 2, 2010 02:29 PM
Agree with you, and with Unsympathetic and Ricardo Z. It seems to me that explanations for our current malaise that ignore the run up in asset values and debt (and the attendant imbalances in balance sheets of individuals), and instead attribute them to "financial panic," or to over-cautious lending, are displays of willful ignorance.
Isn't Cochrane one of those illustrious fellows who rediscovered the old depression era "Treasury View?"
Posted by: don at February 2, 2010 02:32 PM
I recall much more panic over "counterparty risk", than TARP, or subprime, or falling house prices, or weakening economy, or Obama, or Bush running for a third term, or anything else.
This was a reference to CDS derivatives, and concerns about whether these "net out" smoothly, especially if a counterparty doesn't really have the money. And I'll bet Wall Street panicked over it because they knew what a house of cards it was.
And lots of them are still in the derivosphere, and no changes to the system to deal with it yet.
I'll also give the EMH and Chicago people something to ponder. If your insure a asset you increase the value of it. If you find out the insurance is no good, the value falls. Works that way with muni bond insurance too.
Maybe with a little contortion they can make that work in the model.
Posted by: Cedric Regula at February 2, 2010 04:09 PM
"We are left with only one plausible explanation for why Lehman's failure could have had such wide-ranging effect: After the Bear Stearns bailout earlier in the year, markets came to the conclusion that investment banks and bank holding companies were "too big to fail" and would be bailed out. But when the government did not bail out Lehman, and in fact said it lacked the legal authority to do so, everyone reassessed that expectation. "Maybe the government will not, or cannot, bail out Citigroup?" Suddenly, it made perfect sense to run like mad..."
Actually it is equivalent to say that credit risk of large financial institutions was mispriced. After Lehman default there was a "run" to reprice such risk. Usually large market movements do not occur in a linear and painless way. They are a process of "reciprocal price discovery" that require additional adjustments of risk appetites
Posted by: amadeus at February 3, 2010 05:43 AM
Interesting thread. It does seem plausible that the linchpins were 1) the misallocation of credit to US households that was exposed by the oil price spike of 2007-8, and 2) the subsequent murderous games with CDS positions that left AIG in the arms of NY Fed.
The key reforms would then appear to be those that are related to CDS pricing and risk. Central clearing and collateral management is obviously one line of reform, as are "living wills" for financial entities.
Regulators also need better strategies for intervening in credit allocation processes. The solution offered by CDS is unstable, since the natural long-CDS hedge is a short equity position. Games played with CDS cannot end well except for the lucky few left standing.
Posted by: Alan King at February 3, 2010 08:18 AM
"Regulators also need better strategies for intervening in credit allocation processes.The solution offered by CDS is unstable, since the natural long-CDS hedge is a short equity position. Games played with CDS cannot end well except for the lucky few left standing.
Let's take that one step further. We have to limit what can be "insured", or hedged. Right now we are hedging interest rate moves, corporate solvency and things we find move synchronously on a global scale. Like insurance against a global car wreak. Then someone has one and we get a global car wreak, because we find out that there wasn't enough money to put that right again.
This is like trying to say you have a market economy, but you can insure it to be a command economy with price of credit and risk that doesn't change. Not surprising that doesn't work.
Posted by: Cedric Regula at February 3, 2010 11:27 AM
Looks like we may get a chance to watch a replay, except in slow motion, with perhaps slightly smaller excesses(so much has been absorbed by governments)....and shock value...No Bear-Lehman this time.
"The imminent round 2 of the foreign bank dollar funding crisis, or the eurodollar squeeze redux"
Posted by: Cedric Regula at February 4, 2010 10:08 AM