July 23, 2009
Looking for an exit: Part 2
In my previous post I commented on Ben Bernanke's recent communication of the Fed's exit strategy for getting its balance sheet and daily operations back to historical norms. I suggested that one necessary ingredient to convince the public that we will see a return to a stable monetary regime would be a credible explanation of how the United States government will be able to meet its enormous current and implicit future fiscal obligations. Today I'd like to discuss a second element that I feel is missing from the exit strategy articulated by Bernanke, and this is a compelling vision of what a healthy financial market not propped up by the Treasury and the Fed would look like.
Let me frame this issue using as an example one of the ongoing Fed-Treasury initiatives that troubles me the most, which is the Term Asset-Backed Securities Loan Facility, or TALF as it is affectionately known. According to the Fed,
The TALF is intended to make credit available to consumers and businesses on more favorable terms by facilitating the issuance of asset-backed securities (ABS) and improving the market conditions for ABS more generally.
Let's see if we can start by agreeing on some basic facts. First, the process of securitization was unquestionably enormously successful in the early part of this decade in attracting huge sums of capital from all around the world to fund loans to U.S. households and firms. Without securitization, it is inconceivable that we would have seen anything like the $4.3 trillion in new non-agency household mortgage loans issued between 2004 and 2006.
Second, surely we can agree today (though some may have still thought otherwise as recently as May of 2007) that this success was absolutely not
spurred in large part by innovations that reduced the costs for lenders of assessing and pricing risks. In particular, technological advances facilitated credit scoring by making it easier for lenders to collect and disseminate information on the creditworthiness of prospective borrowers. In addition, lenders developed new techniques for using this information to determine underwriting standards, set interest rates, and manage their risks.
I hope that instead we would agree today that the success of securitization in 2004-2006 was due to (1) misperception on the part of the buyers and raters of the ABS of the degree to which aggregate risks could be diversified by pooling disparate loans, and (2) a misplaced confidence in the implicit or explicit guarantees provided by the securitizer, by derivatives used to hedge ABS holdings, or by the U.S. government, the latter in particular introducing a moral hazard component that rendered the system seriously unstable. And I hope we'd further agree that these factors resulted in a profoundly malfunctioning credit market that caused an unsustainable run-up in U.S. real estate prices and household debt that are the primary reason we're in such trouble today.
A successful exit strategy requires a vision of where you want to go, and how it's going to be different from the place from which you just came.
So the natural questions are, do we believe that problems (1) and (2) above have been resolved, and what persuades us that securitization could succeed on anything remotely like its previous scale in the absence of features (1) and (2) above?
I'm worried that Bernanke's answer for why these problems won't recur may be that underwriters, investors and rating agencies have learned the lesson from previous mistakes.
I guess I'm looking for evidence that the Treasury and the Fed have learned the lesson from previous mistakes.
Posted by James Hamilton at July 23, 2009 08:37 PMdigg this | reddit
Cannot agree more,the structural problems of the Western economies (that is the rebalancing of the major components, CA BOT) are far from addressed.
The old mantra is prevalent and the credit supply side the sanctified Graal.
Reflation in the casino style is maintained.
There are no changes on substance as they are limited by the individuals,corporations,states indebtness.
As regards the CBs ability to prevent inflation the how is available,missing the when (not subjective)the how much ?
The only certainty is a vanishing global village concept when it comes to the the financial industry.
Posted by: ppcm at July 23, 2009 09:41 PM
Your "Looking for an exit" entries are excellent.
Why are policymakers failing to act on this and other sound analyses of the crisis? Are the Treasury and Fed controlled by the big banks? Where is our regulatory reform?
Posted by: Leland Smith at July 24, 2009 05:53 AM
Your reasons (1) and (2) are too simplistic, and you are focusing on too narrow a timeframe; after all, securitization started in the early '80s, and became a critical path for virtually all debt origination throughout the '90s. For those of us on the buy side who put in the effort (and enormous capital investment) to analyze the risk, we saw that what distinguished the '04-onward timeframe was the deterioration of underwriting credit quality while origination volume continued to skyrocket. Unfortunately, there are a lot of reasons why my firm was (and is!) in the strict minority in terms of ability and motivation to actually do the analysis work rather than just 'buy on the rating'.
Securitization itself is not, per se, the problem -- after all, an individual RMBS securitization is not all that different than an S&L/thrift (the RMBS is more geo diverse, the active management of the S&L portfolio may work for you or against you as a bond holder, and the RMBS portfolio is vastly more transparent than the S&L portfolio). Indeed, with properly motivated and knowledgable originators, sellers, and buyers, securitization has an excellent track record, even in very difficult economic situations.
Thus to me, the problem is not securitization, but rather the fact that the buy side never properly priced the risk: as the underwriting standards slid down the slipperly slope, the buy side bid persisted, with all deals being oversubscribed regardless of how poorly the credits were underwritten even as the pools became increasingly homogenous.
Consequently, if we 'solve' this problem by replacing one indiscriminant bid by another, we will in short course be repeating history! Of course, if you measure things by credit volume, or credit availability, this seems like a correct near-term path, regardless of the long term consequences: kick the can down the road, as the pols are fond of saying.
Instead, the Fed and others should be looking at the price of credit across a spectrum of benchmark credit quality: credit should be available to even low quality risks, but the price of said credit should be dear. Had the regulators and others been watching (as we have been) this simple metric, seeing that all risks were priced to the same low cost, they would have seen this crisis coming years in advance. Liquidity is a good thing iff risks are properly priced!
And of course, the world needs to come to terms with the fact that ratings on structured securities are not, and never will be, comparable to ratings on corporates - and the RA's should fix their rating categories to educate the buy-side to this fact. And we (all of Wall St) need to look carefully and figure out exactly why we seem to have had a broad-based failure of due dilligence over the past half-decade plus: Madoff was not so much an aberation as an extreme.
So from my POV (structured finance quant on the buy side), the problems haven't been resolved and while the bid has been pulled back, even that is not from lessons learned but rather simply because the recent experience was painful. TALF may help keep the plane from flying into the dirt, but we do need to address the real problems if we are to avert future crises. And yes, securitization is both better than the alternatives (back to the S&L's? Yuk.) and essential to the modern economy.
Posted by: Dr. D at July 24, 2009 07:06 AM
"do we believe that problems (1) and (2) above have been resolved"
I think the financial market has learned its lesson that securitization does not lessen risk, only spread it around. I think the financial market has learned a dangerous lesson that the US government will bail it out if it gets into systemic trouble.
"what persuades us that securitization could succeed on anything remotely like its previous scale in the absence of features (1) and (2) above?"
Absolutely nothing, and we wouldn't want it to. Large-scale foreign investment in real estate will create unhealthy bubbles in any country, regardless of whether it is done through securitization or some other way (eg European investment into Central/East European property).
I do think securitization will gradually revive to some extent, but I don't think it will ever again in my lifetime become a major factor in drawing foreign investment into the US.
The real substitute that has emerged for the securitization industry is Treasuries. This is also unsound and unsustainable.
The sound way forward would be to focus on efficient production that would justify foreign investment. But who wants to work when you can just issue IOUs?
Posted by: Tom at July 24, 2009 07:20 AM
Dr. D, you seem to ignore the possibility that this crisis began in the 1980s, as part of a super-credit bubble.
To me this touches on the larger question that few economists seem to think is worth exploring. After being reasonably stable for 40 years, household debt, as measured against either income or GDP, exploded for 25 years beginning about 1984.
Clearly, that is a large part of the problem. But what is a viable level of household debt, and how do we reach that level? Are any economists curious about this?
Through defaults, a decrease in loan availability, and increases in income and transfer payments/tax cuts, household debt to disposable personal income has gone from 141% in 2007Q4 to 131% in 2009Q1. As recently as 1999 it was 101%, in 1984 it was 69%.
Just to get back to 1999's elevated levels, and assuming DPI increases 10% during the adjustment period, we need to eliminate $2T in household debt. To reach 1984 levels, we would need to eliminate $6T in household debt.
My question is, what is a viable level of household debt and how do we reach that level?
Posted by: Bob_in_MA at July 24, 2009 08:56 AM
Dr. D says
"...and the RMBS portfolio is vastly more transparent than the S&L portfolio)..."
A nice post which seems to focus on the failure to discriminate between different credit qualities.
But how can a geographically diverse aggregation of mortgages be more transparent than the individual mortgages contained within? To me this appears impossible. And from my viewpoint represents a critical misperception about derivatives themselves.
Posted by: beezer at July 24, 2009 12:24 PM
Beezer, RMBS mortgage pools are completely transparent: as an investor, we can get full details on each and every loan in the pool (thanks in part to Reg FD). Analyzing these pools at this depth -- effectively re-underwriting them -- is both time consuming and phenomenally expensive. And yet, we believe it essential for investors to do this; no 'rating' can substitute for full knowledge of the risks embedded in the pool of mortgages, regardless of whether investment is intended at the AAA-level, the mezz level, etc. Alas, my firm is one of the very few firms equipped to do this....
Contrast this to the portfolio of mortgages held by an S&L/thrift: as a bond or equity investor in the entity, you have zero ability to extract any information about their mortgage portfolio beyond high-level statistics (e.g.: x% construction loans coming due in 18 months). So, no ability to re-underwite the risks, no ability to determine the quality of the origination underwriting. Thus, less transparent.
It's not the geo diversity that makes the aggregation transparent; rather, geo diversity is an additional feature of securitization (in addition to transparency). I highlight geo diversity as it is well known to be one of the very large advantages of securitization versus local banking: compare the performance of securitizations containing loans made in oil-patch areas vs investments in local banks....
Posted by: Dr. D at July 24, 2009 12:58 PM
First, the process of securitization was unquestionably enormously successful in the early part of this decade in attracting huge sums of capital from all around the world to fund loans to U.S. households and firms. Without securitization, it is inconceivable that we would have seen anything like the $4.3 trillion in new non-agency household mortgage loans issued between 2004 and 2006.
Very good analysis professor, as far as it goes, but just what facilitated the secruitization? Such securitization required dollars whether it came from foreign sources or from domestic sources. The securities were dollar denominated securities and so required dollars for purchase. Where did those dollars come from? Hint: who creates money?
Secondly, I am not sure I would use the phrase "unquestionably enormously successful" when the securitization resulted in a huge malinvestment as described in your items 1) and 2).
Interesting, excessive money creation leading to an unsustainable boom accompanied by malinvestment terminating in an economic bust. Now where have I heard that before? Hint: ABCT
Posted by: DickF at July 24, 2009 01:21 PM
The basic problem is that domestic savings do not equal domestic demand for credit. Foreign savings are notoriously volatile. Many nations have learned this lesson over the years, as large quantities of foreign savings came in, then suddenly stopped. Domestic savings tend to be far more stable.
The Fed's de facto policy of trying to force US citizens to save nothing at all to compensate for excessive overseas savings is not a strategy that is sustainable in the long run. There is no long run model of securitization that will work, because there is no long run model of endless trade deficits that will work.
In the long run, every nation must have a 10 to 15% national savings rate for their economy to be sustainable in the long run. One nation cannot forever compensate for excessive savings in other nations. Eventually, debt to GDP ratios become too large to service.
Posted by: Balance at July 24, 2009 02:25 PM
If one looks narrowly at the securitization process, it is a neutral numbers exercise. If one's view of securitization expands to view the supporting infrastructure, many problems are visible. Rating agencies are an example. If I could move my school tax dollars around to the teacher who gives my children good grades, it would be considered unethical, maybe illegal. But let mortgage companies move their business around to rating agencies that give their securitized mortgages "good grades" and it is currently considered standard operating practice on Wall Street. Some other business model for ratings must be found in a healthy financial market. Another example is undocumented loans - sometimes called "liars loans". Where was Federal Reserve regulation when it was obvious to casual observers that it was needed? And what reason do we have to be confident that the Fed can regulate "more and better" in an envisioned healthy financial market? Another example is Credit Default Swaps - CDSs. Say I'm an insider and I know from trading and other sources that some firm is full of toxic mortgage assets. Maybe my name is Goldy Q. Sacks. I purchase fire insurance on my neighbors house - otherwise known as CDSs - and the entire support system, including regulators, acts to keep this information secret even after the purchases. Are incentives for good behavior in the right places in this securitized market?
I agree that a vision of a healthy financial market is needed, and that the Fed has not even started to provide that vision. Since the support system is broken in so many places, I am skeptical that the Fed is able, or even appropriate, to provide that vision. But it is sorely needed, no doubt.
Posted by: Mike Laird at July 24, 2009 02:59 PM
Dr. D: In light of the behavior RMBS prices (and the prices of lots of other securitized debt) over the past couple years, how has your firm performed? I'm not skeptical; just curious.
Posted by: Phil Rothman at July 25, 2009 12:51 AM
There are multiple risks needed to be cleaned before the rebirth of securitization:
1 The securitization were not pari passu that means the agent(s)and book runners were merely downloading risks and funding on the end users.(the end story,during the debacle there was no legal entity in charge of remedies or work out)
Ramifications of this status can be expanded at will.
2 Banks have lost the needed trust for undertaking this trivial function, few of them like Goldman Sacks were not only packaging the risks but shorting its abstracts in the CDS market.
# If banks were to be reintroduced in these undertaking one has to trust their perenity in the business(that is their solvency, their conflict of interests)
# The CDS market should be regulated if not totally eleminated (banks and financial agents are too prompt to manipulate these markets see the latest May June July prices in the CDS)
Should Securitization be a vital components of funding when savings are scarce.
The only prompt resources would be through third party supervision bodies (mostly government own entities)
The CDS will always be a Damocles sword, and never forget Investments banks are weak long term holders of assets loans and poor assessors of long term risks.
Posted by: ppcm at July 25, 2009 09:16 AM
Dr. D: Here's an addendum to my earlier question. In a 7/21/09 WSJ op-ed piece titled 'Why Toxic Assets Are So Hard to Clean Up,' Kenneth Scott and John Taylor write, 'While the original MBS pools were often Securities and Exchange Commission (SEC) registered public offerings with considerable detail, CDOs were sold in private placements with confidentiality agreements.' In light of those 'confidentiality agreements,' how does your firm go about valuing the CDOs?
Posted by: Phil Rothman at July 25, 2009 10:05 AM
Bob in MA: I certainly agree that we have seen a dramatic increase in household (and other) debt in the last 25 years, in great contrast to the earlier part of the 20th century. But as to whether this is a "super credit bubble", well, I leave such labeling to others more qualified. I will note that at the early end of this timeframe there were large changes going on in both the US and the world community that resulted in the US being a very attractive investment for a large and growing amount of global liquidity....
And I like you also wish that economists would spend more time/effort on discussing and answering this most important question! But it is a complex question: after all, consider that a significant volume of the internet boom was in fact financed with home equity extraction, actively invested into small home-based businesses, businesses that resulted in significant positive returns for their owners as well as society at large.
But sidestepping your fundamental question, I would sum up my position as structured finance is neither inherently evil nor inherently good; it's all in how it's used and by whom for what purpose. (Perhaps a variation on the theme of "guns don't kill people, rather people kill people.") Structured finance has both advantages and disadvantages when compared with, say, banks/thrift structures.
Consider that when LTCM imploded and nearly took the financial system with it(ah, to be back in those days of relative calm!), few if any called for the elimination of financial futures & options - we should be thinking likewise now. Ditto for that infamous and overly broad subject, derivatives. Note that Buffet's famous quip was actually in relation to interest rate swaps, not CDS...! Similar yes, but the experience has been vastly different - and interest rate swaps have become essential to modern financial management.
Thus to focus solely on structured finance is to overlook the far larger - and more interesting - problems lurking behind the obvious target. And indeed, if we don't understand those problems and address them directly, we shall be doomed to repeat our recent experience: different products with new acronyms perhaps, but the same fundamental drivers.
What caused such insatiable investment demand that the buy side lost all control over pricing credit risk? What motivated Wall St. to fill these orders while virtually completely overlooking all fiduciary duty to perform adequate and thorough due dilligence? Why were the investors so insatiable that they ignored the very real (and well-known) principal agent problem they were creating and reinforcing? Why didn't the central bankers and the market as a whole recognize that funded venture opportunites as well as all forms of lending were both reaching new quality lows? (A strong indicator if ever there was one of too much money sloshing around in the system.) Why was the inverted yield curve marveled at but not taken seriously as an indicator of the mispricing of risk? Eliminating or scaling back structured finance does nothing to fix the problems underlying these issues; indeed, it only serves to stoke the money creation fires back to life!
Posted by: Dr. D at July 25, 2009 01:46 PM
Don't forget to celebrate our annual tribute to unemployment. The minimum wage just went up another 13% or so.
Now why didn't I think of increasing the minimum wage at a time when our unemployment is almost as high as during the Great Depression. That progressive congress is so brilliant!!
Posted by: DickF at July 25, 2009 02:50 PM
It isn't clear cut to what extent a raise in the minimum wage has on employment. Rather than go into great detail about the empirical data and arguments, may I refer you to this paper:
All in all, it may not have anywhere as near a negative effect as the opponents of the minimum wage argue, but not as minor as the proponents believe.
As for our current unemployment compared to the Great Depression: It's difficult to assess since how we measure employment has changed, but the best estimate is the the unemployment rate was twice as high as it is now (using the U-6 measurement, the broadest and arguably the best indicator of employment "health).
This recession has characteristics that are far more similar to those of the pre-WW2 era, but not hardly as severe as the 30's contraction.
Posted by: SecondLook at July 25, 2009 09:39 PM
Dr D and others,
I agree with most of the comments made by Dr. D; IMO, it was the relaxation of credit underwriting standards that was perhaps the main reason for the subsequent collapse (particularly in subprime and Alt-A markets).
Now people can argue that securitisation - by divorcing origination from ownership (the "originate and distribute" model) - facilitated this reduction in credit standards. Even with this argument, one rebuttal is that this was only possible because (most) investors outsourced credit risk analysis to the ratings agencies. This wasn't a problem right??
The fact that securitisation has become a four-letter word reflects, IMO, people's lack of skepticism (perhaps even cynicism) in the ratings bestowed by the agencies during the boom years.
One example, to me, clearly illustrates people's outsourcing of credit risk: could people really believe that a CDO of BBB-rated tranches of subprime loans (a 'subprime-mezzanine CDO') could really have a senior tranche rated AAA?
I think the crash illustrated the problem that investors relied exclusively on 3rd party opinions of risk - rather than doing their own due diligence - as much as the crash illustrated the inherent 'originate and distribute' problem with securitisation
Posted by: Alan at July 26, 2009 05:57 PM
Bob in MA: I think that the viable level of household debt is a function of the level that households can afford at any point in time given (hopefully) some imperfect foresight about the cost of future financing. This is when you are not in a bubble scenario and a well informed borrower.
If you construct a series of the average annual debt repayment (absolute debt times your choice of annual mortgage rate) as % of income or GDP you will find it a lot more stable than absolute amounts of debt. I see this as an equilibrium series representing the clearing point through time of an impotetical aggregate supply-demand for credit with price (i.e. rates) and volumes on the axis. This puts together your point and Dr D's: the cost of credit was too cheap through the 1990s and 00s leading to too much debt accumulation. I think both the cost of credit and its absolute levels need to be monitored by regulators/Fed. I agree with you that the key question remains what is excessive.
With hindsight it is a simple principal-agent problem, where the government failed to put in place the right incentive schemes for financial institutions to rein in the absolute levels of lending. As a result you get an unsustainable outcome where risk is consistently mis-priced by financial institutions (the agents) because they are short-sighted (risk-unadjusted!) profit maximisers. From what I could see from the inside from 2005 to 2007-8, lower and lower rates were a simple reflection of excess credit supply as financial institutions were undercutting eachother to mantain/gain market share regardless of risk. Again, this could have only happened with a complete mis pricing of risk. Maybe it is an incomplete market situation: as banks believed that they were diversifying risk thanks to more and more securitisation there was no actual market with a clearing price for the actual level of aggregate risk and banks assumed it constant until they realised (too late) that it wasnt. Maybe this is the answer: we need a market for aggregate risk and need to get banks to profit maximise over it! Maybe this is ?all? you need and do no even have to answer what is excessive?
Posted by: AndreaZ at July 27, 2009 02:44 AM
Barry Eichengreen has made the claim that those countries that held to the gold standard longer during the Great Depression actually had longer recoveries than those who abandoned gold early. Eichengreen has many problems with this hypothesis. One is that countries that attempted to keep their currency tied to gold actually had much more extensive public works programs, France and the US are prime examples.
Now Eichengreen, in attempting to bolster his argument, has actually introduced another variable that he must deal with. Those countries that were slower in coming out of the Great Depression had stronger trade restrictions than those that came out early. It appears that Eichengreenís argument gets weaker and weaker as he does more research.
What the author of the article fails to grasp is that under a real gold standard there cannot be trade barriers. The presence of trade barriers proves that the government is attempting to circumvent the operation of the gold standard and the author's analysis rings of Keynesian stimulus analysis rather than classical gold standard analysis. The claim that the gold standard contributed to the Great Depression does not hold up to reasonable analysis.
From the Economist:
A new paper by Barry Eichengreen of the University of California, Berkeley, and Douglas Irwin of Dartmouth College adds a fresh dimension to this analysis. It examines how the decision to quit gold or to cleave to it affected trade policies. The picture painted by most accounts is of a rush to erect new barriers to trade, as each country tried to shield its businesses and workers from the deepening slump and responded in kind to the protectionism of others. In fact, say the authors, trade barriers were not imposed uniformly. Countries that stuck with gold were more protectionist than those that abandoned it. Tariffs were a poor substitute for policy stimulus, but in an era of balanced budgets they were all that gold-standard countries had.
Posted by: DickF at July 27, 2009 01:56 PM
Susan Hickok of the OCC gave a preview of a paper you all might find rather interesting: "The US Financial System in 2011: How Will Sufficient Credit Be Provided?" Should be available from the OCC in the next few weeks....
Posted by: Dr. D at July 29, 2009 06:41 AM
According to this paper --
the Fed deliberately drove up excess reserves by paying interest on reserves.
IOW, the large expansion in the monetary base was not caused by the collapse in the money multiplier, but rather the Fed deliberately drove down the money multiplier to offset the massive expansion in the monetary base. I am outraged.
Posted by: Richard A. at July 29, 2009 11:40 AM
President Obama wants us to believe that increasing taxes on the rich allows him to give more to the poor, ala Robin Hood. However, there is no historical relationship between federal spending and federal taxing.
Yes, raising the tax rate on the wealthiest Americans may address a psychological need to punish the rich for being rich. But, if this tax increase succeeds in collecting more taxes from any group, rich or poor, it will hurt poor people. All taxes remove money from the economy, which slows or reverses economic growth. Even now, with so many unemployed, the poor remain slow to understand that their jobs depend on the financial health of the rich.
The economy runs on money. One of the government's most important jobs is to create money and to spend it. In 1971 we went off the gold standard specifically to give the government the unlimited ability to create and spend money. When the government fails to do that job, the economy falters. Every depression in U.S. history and every recession in the past 40 years followed periods of reduced federal deficit growth, and every recovery as been associated with increased federal deficits.
The media and the politicians suffer from federal debt paranoia, which is responsible for more misery in America than all the crime and illness combined. Federal debt paranoia keeps us from providing universal health care, fully funded Social Security and improved education, military, infrastructure, energy and policing.
I call it "paranoia," because, history indicates large federal deficits do not cause inflation, recession or any other negative economic effect. On the contrary, large and growing deficits stimulate the economy. We keep falling back into recessions because of federal debt paranoia.
Although President Obama wants to redistribute wealth, increasing any taxes will reduce overall economic growth, taking from the rich and from the poor.
Rodger Malcolm Mitchell
Posted by: Rodger Malcolm Mitchell at August 1, 2009 08:34 AM
Susan Hickok of the OCC gave a preview of a paper you all might find rather interesting: "The US Financial System in 2011: How Will Sufficient Credit Be Provided?" Should be available from the OCC in the next few weeks....
Posted by: Dr. D at August 11, 2009 08:52 AM