« Who should make this decision come January 20? | Main | Another modest proposal »
What really caused the financial crisis?
Kashyap, Rajan, and Stein have lots of explanation but here is the initial bottom line:
The proximate cause of the credit crisis (as distinct from the housing crisis) was the interplay between two choices made by banks. First, substantial amounts of mortgage-backed securities with exposure to subprime risk were kept
on bank balance sheets even though the “originate and distribute” model of securitization that many banks ostensibly followed was supposed to transfer risk to those institutions better able to bear it, such as unleveraged pension funds. Second, across the board, banks financed these and other risky assets with short-term market borrowing.
In other words, one problem was not enough (!) securitization. They also call for counter-cyclical capital requirements. They like mandatory capital insurance -- with payments triggered by capital disasters -- even better. My main worry, of course, is how we should regulate (or not) the entities which offer this insurance. Will they too engage in liquidity transformation and if so who ensures them?
And, going back to banks, part of the governance problem was this:
...it is very hard, especially in the case of new products, to tell whether a financial manager is generating true excess returns adjusting for risk, or whether the current returns are simply compensation for a risk that has not yet shown itself but that will eventually materialize.
Their phrase "recapitalization as a public good" should not soon be forgotten. And it is leverage which is dangerous, a lesson becoming clearer every day.
This paper is essential reading for anyone following the crisis and it makes more sense than just about anything else I've read on the topic. I thank David L., a loyal MR reader, for the pointer.
Posted by Tyler Cowen on September 26, 2008 at 06:05 AM in Economics | Permalink
Comments
"And it is leverage which is dangerous, a lesson becoming clearer every day."
Clearer to whom? It was pelucidly clear to me in my youth, when I read Galbraith on The Great Crash. Lousy economist, perhaps, but a fine journalist and comic writer.
Posted by: dearieme at Sep 26, 2008 6:56:29 AM
The residue on bank balance sheets were not a symptom of not enough securitisation but a failure to pay attention to market signals saying "enough already" and has shown up in all sorts of other areas. The stuff was on the balance sheets purely because they couldn't sell it at an economic price but managed to rationalise it on the grounds of a dubious rating.
Similar things happened with the split-capital trust scandal but on a much smaller scale. Also similar is the (possibly apocryphal) story of the second cat bond. The first Florida hurricane risk bond is snapped up. The copycat bank is then surprised the same investors are not delighted by their almost identical issue.
The counter cyclical risk is not incidental or accidental, it is the risk that was supposed to be sold on and is intrinsic to mortgage securities.
The more you look at it the more it looks the same as all the other crises. The main difference is the sheer size of the market it applied to making diversification almost impossible.
Posted by: bunbury at Sep 26, 2008 7:04:07 AM
especially in the case of new product
How many of those new products are actually designed to optimize the risk portfolio, and how many are designed to obfuscate risk transaction and holding? If the latter >> the former, make all derivative contracts legally unenforceable except N types to be traded on m designated markets.
Posted by: ogmb at Sep 26, 2008 7:52:42 AM
ogmb, that's a question with no answer. For example with CDOs and CDSs they undoubtedly did provide risk optimisation at first and genuine trade benefits. They certainly helped banks recover from a previous crisis. Unfortunately, as the cat bond example was meant to suggest you can have too much of a good thing. Chopping up assets into slices to be sold to different investors is good until there is a slice you can't sell anymore. Unfortunately the banks didn't stop selling, they just kept the hard to shift slices on their own books. These slices were either equity containing most of the risk or supersenior which had to be sold at such low interest rates to make the deals work that external investors weren't interested and banks couldn't justify it unless their holdings were so extremely geared up that the holdings were dangerous again.
So banks ended up with large amounts of rubbish on their balance sheets. Even then proper banks have not fared too badly but investment banks which have looser capital requirements and which have been traditionally transactionally focussed have grown huge balance sheets.
Most of this is business as usual. In my lifetime, or nearly within it, this is the third blow up of the US mortgage market (and third bailout) and investment banks collapse regularly. Even Goldman Sachs needed bailing out in 1986 and has an eponymous scandal of its own which is in some ways quite similar. It's not long since rates were set to bail out Citibank. What makes this one is really a monoculture style problem. The asset class with problems is so large and the financial system is so much more centralised that noone involved can escape and the collateral damage is worse than before.
Derivatives are really not the distinguishing feature of the disaster. It has more in common with the Latin American debt disaster.
Even if it were possible, the gross intrusion on freedom of contract involved in tearing up trillions of dollars derivatives contracts would come a bit late.
I think the real question is how banks know when enough is enough. If they wait until they stop making money it may be too late.
The most constructive thing I can suggest is that more regulation and policy making should be pursued from the point of view of the asset class rather than the institutional level. So the more attention should be paid to what's coming instead of just looking at what it will find when it gets there.
Posted by: bunbury at Sep 26, 2008 8:37:45 AM
As this link says and I believe, securitization itself was not the problem.
The problem was that the banks really did not securitize fully. They decided that they needed to hold large numbers of these securities on their books instead of selling them to investors as had been traditionally thier business model.
The fact that they had to keep the assets on their books to get a deal done means to me that they were holding those securities at above market prices on the day they were done or at least they were speculating that they were worth more than market.
This combined with high leverage led to the problems. Had the banks stuck to thier traditional model of securitization and selling off all of the tranches of the mortgage backed securities they would not have been in this situation. Sure lots of investors would have lost money but it would not have been a threat to bring down our financial system.
Posted by: eccdogg at Sep 26, 2008 8:52:20 AM
Thanks, Tyler. This is easily the best piece out there on all three issues: cause, effect and solution. And the fact that it is two months old makes it even better. Its about to become required reading for my colleagues.
Posted by: JFalk at Sep 26, 2008 9:09:40 AM
Something that was pointed out in the paper and that has been bandied about by commentators but doesn't seem to be considered in any of the "official" proposals is flexible capitol requirements. It's never made much sense to me to have capitol requirements and then not allow that capitol to be used in the event that it's needed. All it does is tie up a certain amount of capitol permanently and gets creditors a few more cents on the dollar in the event of a bankruptcy.
I assume that the argument for strict and inflexible capitol requirements is some sort of moral hazard thing. As it stands, banks really have to keep some percentage of liquid capitol above the standards to use to pay when a depositor withdraws, pay interest, etc. If they knew that in a big crisis they could dip into their reserves, then they would sensibly hold less above the capitol requirements. Couldn't this be protected against by putting strict conditions or penalties on dipping into the emergency capitol?
Posted by: Podunk at Sep 26, 2008 9:47:56 AM
The banks failed to learn the two cardinal rules from The Producers
http://www.imdb.com/title/tt0395251/quotes
Max: The two cardinal rules of producing. One: Never put your own money in the show.
Leo: And two?
Max: [yelling] Never put your own money in the show!
Posted by: jfalk at Sep 26, 2008 10:35:00 AM
How about an even simpler proposal. I propose a new kind of convertible debt, one that converts when the stock price drops. This would mimic the insurance proposal suggested above but without worrying about the solvency of the insurer. Effectively this would be a kind of subordinated debt and would obviously require a higher rate of return than conventional debt.
Posted by: samik at Sep 26, 2008 10:41:09 AM
Why can't banks borrow from The Fed to get over the hump? Is it because this would increase the denominator of the capital requirements? Does providing a starting bid for their toxic tradeables decrease the numerator? They are stuck.
As long as they don't want to tell us what their junk is worth, why are we buying them at sticker price rather than just changing the rule that makes them tell us.
All banks are insolvent when there is a run.
Posted by: Andrew at Sep 26, 2008 11:09:14 AM
In other words, one problem was not enough (!) securitization.
Yes, but why? My guess is poor judgment on the part of the banks as to which mortgages to keep. The rates on the subprimes may have been so attractive that the banks were unwilling to accept market prices, which reflected higher (and more accurate) assessments of risks than the banks' own assessments.
An odd sort of adverse selection problem, perhaps, where the seller knows less about the risk than the buyer. Or maybe just greed overcoming good sense.
Posted by: Bernard Yomtov at Sep 26, 2008 11:25:27 AM
If I read it right, the gov't *made* $1.9 bn on the WaMu sale. If in fact WaMu was solvent, if not on a current basis, and they were in the worst shape (ok, maybe Wachovia), why doesn't that answer the question of whether it more like 2 or 8?
Posted by: jfalk at Sep 26, 2008 11:58:29 AM
It seems to me a lot of people are making this way too complicated. At the heart of the crisis, like most financial crises was excessive leverage on the part of financial institutions, or what some call the development of a shadow banking system. Regulation to deal with excessive leverage, which I think is going to happen, seems to me the least intrusive, most free market form of regulation. Everything else is micromanaging the system. You can regulate mortgage brokers, regulate CDOs, regulate CDSs, maybe some of these are good ideas, but ten years from now someone will invent something else. Here's another way of looking at it: the Fed tries to prevent excessive money supply growth to prevent inflation, now we need to prevent excessive credit growth to prevent asset bubbles.
Posted by: Phil P at Sep 26, 2008 12:12:25 PM
This is all a load of crap. What caused this crisis was human nature. This was a great ponzi sheme fuelled by exotic financial instruments which in themselves were not dangerous, rather the overseers of these instruments had lost all sense of prudence in their application. This whole debacle illustrates, with very few exceptions, that economics is the province of halfwits and spivs. This is a cultural problem and not an economic problem.
With the cultural acceptance of keynesian theory, every economic crisis was solved Government bailout. Net result, the market felt that risk could be ignored since "big brother" would always be there to back stop the financial idiot. This was THE implicit CDS of the financial industry. The most "toxic" CDS of all.
But to be fair to the pigmen, the general public was complicit in this idiot scheme. Had anyone tried to stop the run up of the Western housing boom they would have been pilloried by the proles. There have been people who have been prediciting this situation for the past 15 years: Bill Bonner, Eric Janzen and others. During the property boom they were ridiculed by all. Had any politician voted to put the brakes on housing expansion during its early phases he would have been voted out. Why are the bankers who approved the NINJA loans more culpable than the "grown ups" who voluntarily took these loans on?
No, human nature, not financial instruments are to blame for these circumstances we find ourselves in. No one wants the punchbowl taken away when the party is starting to go wild.
Posted by: slumlord at Sep 26, 2008 6:26:17 PM
Slumlord had a point. The problem was one of human nature. However, I don't think it was that complicated.
The problem was more one of laziness, and not just of economists but everyone involved. Particularly the highly paid executives of the various financial institutions, both private and public.
The financial institutions found it was easier to manage HR with less people in their employment=> engage agents on commission working from their garages to source mortgages on their behalf.
The garage based salespersons were able to keep their commissions on all mortgages that they sourced if the mortgagor did not default within the first three months of the loan. AFter that, presumable, the mortgages were supposed to be offloaded to an unsuspecting investor and the financial institution was supposed to be safe.
This modus operandi was great news for that part of the organisation that was paid bonuses for delivering lending services. It was a lazy way to make money. Some call it predatory lending. But the borrower isn't that stupid. Let's just call it lazy lending and predatory on the end investors in the mortgage backed securities. There is no need for us to feel sorry for the borrowers who had no hope of repaying the loan without significant promotions in salary, or inflation...It was the risk they took with someone else's money, compliments of the financial institution's lending arm.
However, someone forgot to tell the firm's investment management arm that these CDOs/CDSs were toxic and not meant for investment. And the investment managers just took for granted whatever the rating agencies said was the credit risk. Pure laziness. Both on the part of the rating agencies who obviously failed to properly analyse the risk in the bonds (assuming incompetence was not the reason that the bonds were not given a junk rating, given the 3 month "seasoning" period of the mortgages), and the investment managers since they appear to have not done their own analysis but relied on the rating agencies' assessments.
Unless we can regulate against stupidity and laziness, then increased regulation is not likely to avoid a similar problem in the future.
But boards should adopt a strategy for making bonus payments that take into account future performance based on current decisions, rather than merely past performance. That could be an interesting exercise!!
It should also be noted that Paulson's plan is far from optimal but appears to have suffered from the same lazy analytical approach as did the CDOs. Give the money to the social welfare agencies to buy foreclosed properties at their current market values. This will support the housing market, which is the real fear of the banks, consumers and end investors alike, and will also provide cash into the CDOs/CDSs/MBSs, which will support the holders of the securities.
Posted by: Peter Fane at Sep 27, 2008 12:14:36 AM
I did not read the paper, put off by the slightly arrogant title of the paper. In order to claim to know the real reasons one must have a theory that can predict well on unseen data - which is the only real judgement of a theory. With this criterion, widely adopted in the real sciences, real knowledge and understanding occurs.
If the data in question only has occurred twice (1930's and now), I am highly skeptical that one can find real reasons and control for other possible explanations.
I further assume, perhaps incorrectly so, that the authors did not predict this event (being worried about meltdowns is not the same as predicting them.). So, I remain unconvinced that they, or any body else for that matter knows the real reasons. All other explanations are just plausible guesses, but whose credibility is tainted by the lack of predictive value.
Posted by: Intrigued at Sep 27, 2008 2:11:02 AM
The first line should read, " ... put off by the slightly arrogant title of the post" and not paper.
Posted by: Intrigued at Sep 27, 2008 9:01:37 AM
Two words: agency cost.
I am very close to these issues - I have been in private equity and private credit for the last 11 years. I can say after all these years that agency costs are true costs, which can have massive incidence.
For all the reasons Shiller cites: the asymmetry of returns afforded an agent, the incentives created by compensation timeframes (year-end bonuses), the discrediting of naysayers in the early phases of a bubble, and the various biases (law of small numbers, confirmation bias, hindsight bias, insufficient account of mean reversion being the chief ones) combine to create an environment where risk becomes slowly mispriced, to extreme extent at the top of the cycle. Further, leverage magnitude and pricing are also brought to extreme levels, for these very same reasons. A key insight: leverage and liquidity drive valuation in the short term, and vice versa. This virtuous feedback loop then becomes vicious when mean reversion inevitably occurs. There are institutional and behavioral constraints which prevent the market from being truly efficient in the sense of being immune to non-rational spikes in values and leverage.
Last but not least, the "Greenspan put" really pulled the last self-correcting mechanism from the market. The associated malinvestment and mispricing of assets followed the Austrian textbook.
Posted by: Sunset Shazz at Sep 27, 2008 6:33:51 PM
In other words, one problem was not enough (!) securitization.
The post and the linked paper completely neglected the elephant in the room: the problems of moral hazard and adverse selection caused by securitization. Banks that offload their risk greatly reduce their incentives to properly judge credit quality -- their incentives become to push high volumes of loans out the door with much less concern about whether they will eventually default.
Securitization doesn't just distribute risk, it creates additional risk. Indeed, any distribution of risk, whether by insurance or securities, also creates new risk in the forms of adverse selection and moral hazard. For the most common kinds of securities (e.g. stocks and bonds) over hundreds of years we've evolved highly elaborate systems to minimize that risk creation, such as the elaborate accounting, auditing, internal control and governance, and external regulatory apparatus required to issue trustworthy stocks and bonds.
Under a highly simplistic set of mathematical models that captured the risk distribution but not the risk creation, the mortgage securitization market ignored the adverse selection and moral hazard created when risk is distributed and short-circuited the need to evolve (through extensive periods of trial and error) the reporting and control mechanisms needed for a securities market to work sustainably: to provide more value in distributing risk than it destroys by creating additional risk. (n.b. one can also refer to the problem as an agency problem between the security issuer/obligor and the security holder/obligee: moral hazard is the more general term: this agency terminology is often used for equity, i.e. stocks, but it's basically the same adverse selection and moral hazard problem as for debt and other methods of distributing risk).
To the extent banks didn't securitize completely (and it would be nice to have some real numbers here -- I didn't see any in the paper) I would imagine at least some investors demanded that the banks eat some of their own cooking before they would partake -- a crude signalling and incentive-sharing substitute for the necessary controls to sufficiently reduce the adverse selection and moral hazard created by distributing the risk. Given the massive wave of foreclosures, this didn't provide nearly enough reduction in the adverse selection and moral hazard of securitization, especially in the form of preposterously lax lending standards.
Posted by: Nick at Sep 27, 2008 7:47:04 PM
slumlord and Peter Fane, when does persistent incompetence become willful iniquity? I think that is the essential question here. It is truly a moral issue.
[For the record, I want to thank Randy Cohen, The Ethicist, The New York Times Magazine, for the turn of phrase used above, which I encountered in his column on Sept. 12, 2008, on an entirely different topic. I am in his debt.]
Posted by: Mr Fnortner at Sep 28, 2008 12:27:40 PM
Slumlord has a point about the cause of this downward spiral..human nature....more precisely greed! The failure to perform due diligence(the cornerstone of all prudent banking practise),the cavalier attitudes of bankers fuelled by the shockingingly excessive renumeration packages,the silencing of crtics who advised against over exposure a trait common to all the major players,all combined to send everything south the moment the system became gridlocked.Get back to basics and check the fundamentals...this great nation has endured much but this was the cruellest cut of all!
Posted by: gatekeep at Sep 29, 2008 6:49:19 AM
Alright fellas,let me get this straight. This mess we're in is because basically, nobody insured the insurers? And the insurers grossly underestimated (or ignored) the risk and potential danger of what they were supposedly insuring. Subsequently, when it came time to do some insuring, the insurers couldn't, or wouldn't insure?
Aren't our markets constructed to allay risk? Aren't we all leveraging something? Do we pay the full price of our life, home or auto as a yearly premium? When I was a boy, that last layaway payment for that item you saved for all year was a joy in itself. Instant credit now makes that joy immediate.
We looted and sold our country, industry, and way of life, so we can have it now. I am of the opinion that we are going to have to insure ourselves. Realize that our future has risks, and we will have to pay the piper if we are not cognizant of such. Maybe a little layaway mentality would benefit us right now.
Posted by: whalestein at Sep 30, 2008 2:25:47 AM
LOOK AT THE DATE!
By STEVEN A. HOLMES
Published: September 30, 1999
In a move that could help increase home ownership rates among minorities and low-income consumers, the Fannie Mae Corporation is easing the credit requirements on loans that it will purchase from banks and other lenders.
The action, which will begin as a pilot program involving 24 banks in 15 markets -- including the New York metropolitan region -- will encourage those banks to extend home mortgages to individuals whose credit is generally not good enough to qualify for conventional loans. Fannie Mae officials say they hope to make it a nationwide program by next spring.
Fannie Mae, the nation's biggest underwriter of home mortgages, has been under increasing pressure from the Clinton Administration to expand mortgage loans among low and moderate income people and felt pressure from stock holders to maintain its phenomenal growth in profits.
In addition, banks, thrift institutions and mortgage companies have been pressing Fannie Mae to help them make more loans to so-called subprime borrowers. These borrowers whose incomes, credit ratings and savings are not good enough to qualify for conventional loans, can only get loans from finance companies that charge much higher interest rates -- anywhere from three to four percentage points higher than conventional loans.
''Fannie Mae has expanded home ownership for millions of families in the 1990's by reducing down payment requirements,'' said Franklin D. Raines, Fannie Mae's chairman and chief executive officer. ''Yet there remain too many borrowers whose credit is just a notch below what our underwriting has required who have been relegated to paying significantly higher mortgage rates in the so-called subprime market.''
Demographic information on these borrowers is sketchy. But at least one study indicates that 18 percent of the loans in the subprime market went to black borrowers, compared to 5 per cent of loans in the conventional loan market.
In moving, even tentatively, into this new area of lending, Fannie Mae is taking on significantly more risk, which may not pose any difficulties during flush economic times. But the government-subsidized corporation may run into trouble in an economic downturn, prompting a government rescue similar to that of the savings and loan industry in the 1980's.
''From the perspective of many people, including me, this is another thrift industry growing up around us,'' said Peter Wallison a resident fellow at the American Enterprise Institute. ''If they fail, the government will have to step up and bail them out the way it stepped up and bailed out the thrift industry.''
Under Fannie Mae's pilot program, consumers who qualify can secure a mortgage with an interest rate one percentage point above that of a conventional, 30-year fixed rate mortgage of less than $240,000 -- a rate that currently averages about 7.76 per cent. If the borrower makes his or her monthly payments on time for two years, the one percentage point premium is dropped.
Fannie Mae, the nation's biggest underwriter of home mortgages, does not lend money directly to consumers. Instead, it purchases loans that banks make on what is called the secondary market. By expanding the type of loans that it will buy, Fannie Mae is hoping to spur banks to make more loans to people with less-than-stellar credit ratings.
Fannie Mae officials stress that the new mortgages will be extended to all potential borrowers who can qualify for a mortgage. But they add that the move is intended in part to increase the number of minority and low income home owners who tend to have worse credit ratings than non-Hispanic whites.
Home ownership has, in fact, exploded among minorities during the economic boom of the 1990's. The number of mortgages extended to Hispanic applicants jumped by 87.2 per cent from 1993 to 1998, according to Harvard University's Joint Center for Housing Studies. During that same period the number of African Americans who got mortgages to buy a home increased by 71.9 per cent and the number of Asian Americans by 46.3 per cent.
In contrast, the number of non-Hispanic whites who received loans for homes increased by 31.2 per cent.
Despite these gains, home ownership rates for minorities continue to lag behind non-Hispanic whites, in part because blacks and Hispanics in particular tend to have on average worse credit ratings.
In July, the Department of Housing and Urban Development proposed that by the year 2001, 50 percent of Fannie Mae's and Freddie Mac's portfolio be made up of loans to low and moderate-income borrowers. Last year, 44 percent of the loans Fannie Mae purchased were from these groups.
The change in policy also comes at the same time that HUD is investigating allegations of racial discrimination in the automated underwriting systems used by Fannie Mae and Freddie Mac to determine the credit-worthiness of credit applicants.
Posted by: ANGELOS BACKUS at Oct 2, 2008 10:30:28 AM
Here is the original link to the NY Times article by Steven A. Holmes. Good find Angelos.
Posted by: Mike Smith at Oct 9, 2008 9:50:18 PM
Trying again: NY Times article
Posted by: Mike Smith at Oct 9, 2008 9:52:05 PM
