Princeton Professor Alan Blinder offers his thoughts in the New York Times on who’s to blame for the mortgage mess, getting the attention of Mark Thoma, Dave Iverson, Brad DeLong, and Greg Mankiw. Here are my two cents.
Blinder explains very nicely why we all should be giving this some thought:
Finger-pointing is often decried both as mean-spirited and as a distraction from the more important task of finding remedies. I beg to differ. Until we diagnose what went wrong with subprime, we cannot even begin to devise policy changes that might protect us from a repeat performance.
Amen to that. So then Blinder begins to tick off his list of folks at whom to point the finger of blame: (1) home buyers who took on mortgages they couldn’t repay; (2) mortgage originators, for issuing same; (3) bank regulators, who didn’t have the inclination or authority to monitor this closely enough; (4) the investors who ultimately provided the funds for the mortgages, and (5) securitization, which led to assets that are “too complex for anyone’s good”. Running out of fingers on one hand, Blinder brings out the dreaded economist’s other hand for (6) ratings agencies which underestimated the risk.
Now, I’m all for two-handed, six-fingered economists, but I really think the big fat finger we should all be focusing on is #4, without which none of the others would have mattered. A dumb borrower requires an even dumber lender. And if there is always an investor to dump the product off on, then of course there is an incentive for someone to originate and then sell off the loan. Evidently securitized tranches is the form in which investors wanted to hold these assets. As for (6), the ratings agencies are only offering investment advice– I see the person who puts up the money as the one who unambiguously must take responsibility for the decision. And how exactly are regulators in a better position to tell an investor that he or she is making a stupid investment, if that is indeed the core problem?
For these reasons, the question I would most like to see answered is, Why did investors buy up this junk? Here is Blinder’s answer:
By now, it is abundantly clear that many investors, swept up in the euphoria of the moment, failed to pay close attention to what they were buying. Why did they behave so foolishly? Part of the answer is that the securities, especially the now-notorious C.D.O.’s, for collateralized debt obligations, were probably too complex for anyone’s good– which points a fifth finger…
In his rush to get on to the next finger, Blinder doesn’t seem to give us a lot to go on with understanding Finger #4, beyond the notion that these instruments were new and complicated and investors were stupid. Stupid, I might add, to the tune of hundreds of billions of dollars.
Perhaps that’s all there will ever prove to be to this story. But I can’t help looking for more, thinking there is likely to be something special that caused the usual incentive structure to break down here, something we might be able to understand with more orthodox economic methodology. In my remarks at Jackson Hole, I suggested an interaction between monetary policy and implicit government guarantees as providing one possible basis for a rational calculation on the part of investors. Jin Cao and Gerhard Illing of the University of Munich have an interesting new research paper spelling out the details of exactly how such an equilibrium might play out. Professor Illing lays out the implications for practical policy-making here.
As I also said in Jackson Hole, I am not sure why investors perceived it to be in their best interests to buy these assets. But I am sure that this is the right question, and would encourage young economic researchers seeking to make a name for themselves to take a swing at it.
Because I basically agree with Blinder– until we know the answer, it’s not clear exactly how to fix the problem.
Technorati Tags: macroeconomics,
subprime,
Federal Reserve,
economics
As for (6), the ratings agencies are only offering investment advice– I see the person who puts up the money as the one who unambiguously must take responsibility for the decision.
Bravo!
(5) securitization, which led to assets that are “too complex for anyone’s good”.
I’m not yet convinced that the problem is complexity. A futures contract on a thirty-year treasury is fiendishly complex, but nobody seems to mind them too much. In the FX world you run into things like double-knockout options that aren’t exactly the easiest things in the world to price.
CDO’s are volatile and the stability of their credit ratings leaves much to be desired. However, the IMF notes that: While many structured credit products were bought under the assumption that they would be held to maturity, those market participants who mark their securities to market have been (and will continue to be) forced to recognize much higher losses than those who do not mark their portfolios to market. So far, actual cash flow losses have been relatively small, suggesting that many highly rated structured credit products may have limited losses if held to maturity
I suggest that part of the problem – if not most of the problem – is that CDOs are specialized instruments, many of which have been designed to suit a particular investor. Each one has a unique pattern of expected cash flows and risks related to that cash flow, which makes it a much less generic investment than ‘normal’ bonds.
I like the homely example of a suit. If I buy, say, a $500 suit off the rack, it’ll fit me OK – and it will probably fit a lot of other people OK, so if I have to sell it, maybe I can get a good price.
I could also buy a $4,000 hand-tailored suit. That will fit me absolutely perfectly, but in so doing will destroy the value to anybody else.
And so it is, possibly, with CDOs. The market for each individual tranche is simply too shallow. This happens all the time in the investment market – bank ‘perps’ (100-year floating rate bonds) were hot in the eighties; it’s very hard to get a decent bid on them now. CARS & PARS (synthetic bonds created by the partial stripping of a high coupon bond to create an instrument selling at around par) have been popular … just don’t try to sell them once you’ve bought them. Monthly Auction Preferred Shares … see above.
I regret that I’m not a specialist in sub-prime CDOs, but I suggest this hypothesis – that it’s a liquidity/mark-to-non-existent-market problem, not an actual “value” problem – is as worthy of investigation as the “stupid investor” theory.
Accrued Interest has a good discussion of what went wrong. Here’s my grossly simplified retelling of his story:
1. The concept of securitization is that by combining risky assets into a pool and sharing the costs of default, you will find that the risk of a share in the pool is lower than than the risk of the individual assets. Conceptually as long as you have a good idea of the probabilities of default and correlations between these probabilities for the different securities in your pool, you’ll have a pretty good idea of the return on your pool and thus the value of your pooled asset.
2. CDO issuers crunched historic data and found that they could predict very nice returns on their pooled asset by including in their mix high-risk (and therefore high rate) loans with low documentation, high loan to value, or other subprime characteristics. Of course, if they left the high rate loans out of the pool completely, they wouldn’t have anything to generate very nice returns. This generated a demand for high-risk loans. The efficiency of our mortgage markets meant that mortgage brokers were paid extra for producing things like high-rate, no documentation loans. It’s hardly surprising that the brokers found ways to meet this demand.
3. Because the demand for high-risk loans was a new thing, it changed the nature of the market for sub-prime loans. Thus the loans that were going into the CDOs were not comparable to the loans on which the CDO issuers (and rating agencies) had data. So the models being used by the CDO issuers to design their CDOs were flawed because “they seemed to forget Statistics 101” (i.e. if your data is about apples, you can’t use it to make inferences about oranges).
Note that the point of Accrued Interest’s blog is that there is nothing inherently wrong with CDOs. Accrued Interest has recommendations for how to make the industry more successful in the future.
I guess I’m still having trouble understanding why this is a problem that needs fixing? Those on the wrong side of the trades lose money and vise-versa. Sure there will be a correction in home prices but long term I can’t see the problem. Those that do get burned will be less likely to sell the instruments without a more appropriate risk premium and under-writing. If they do, shame on them twice.
We don’t call for an overhaul every time the S&P dives 10%, why then when housing is involved? Speculators lose, thats capitalism!
Why did investors buy up this junk?
SotT brought up the concept “that by combining risky assets into a pool and sharing the costs of default, you will find that the risk of a share in the pool is lower than than the risk of the individual assets.” The quick way to express this is that you can make mortgage securities slightly less risky than the market in general does, so whatever the market charges you can underbid by a little bit. But if the rest of the market gets hold of those tools, you end up in a foolish bidding war.
Which might be a long-winded way of saying “they failed to realize the market would adapt to them.”
SotT’s point (1) is true up to a point, but is not the driving force behind sub-prime securitization.
Combining assets into a pool and selling a stake in the pool as a whole is how traditional mortgage-backed securities work. Mutual funds also work this way.
Sub-prime mortgage securitization does not sell undifferentiated slices of the pooled pie, though. It tranches the securitization, with the expected losses on the entire pool being applied to the riskiest tranche first and to the least risky tranche last. It’s as if investors in the least risky tranche get to choose the individual mortgages in their pool AFTER they know which ones have defaulted.
See Page 20 of S&P’s Senate testimony for an explanation of Credit Enhancement : How Securities Backed By Subprime Mortgages Can Receive, and Merit, Investment Grade Ratings.
the rating agencies share at least some blame if they had a vested interest in the marketing of the securities as did investment banks that wouldn’t allow sell ratings on garbage internet stocks.
Thank you Professor. All I hear is blame for the borrower and lender, but neither would have taken action if they had not felt that they would have gained in the transaction. This is especially true of the lender. Most are professionals and so something changed in their calculations.
Only the government or its proxies can change the rules so quickly. Was it congress? I do not remember any congressional action that made a major change. Was it the administration or judiciary? Nope. The most obvious perpetrator is the FED. BB has created problems with high interest rates while creating liquidity.
And right now he is creating a situation that could cause serious problems in the not to distant future. He has cut rates but he has not reduced liquidity. His rate cuts could increase business demand for money that would prevent inflation, but right now the FED is introducing so much liquidity into the system that it is doubtful that monetary demand can offset inflation.
James_Hamilton: A while ago you wrote about the San Diego public employees’ pension fund investing in Amaranth. Would the same reason for this questionable investment choice explain the reasoning for the overbidding on CDOs?
JS, I agree that if your decision causes you to lose your money, it’s your problem. But if your decision causes the economy to go into a recession, it’s a problem for all of us.
Yes, Person, that’s another possibility, as is the interaction of deposit insurance or guarantees with these decisions, either at U.S. or foreign banks. We’ll have a clearer idea when we see who’s actually holding the bag as the losses come down.
Sub-Prime! Blinder Wants Changes … of Some Kind
I have great respect for Alan Blinder, a former Vice Chairman of the Board of Governors of the Federal Reserve System, but his recent op-ed in the New York Times is not the type of thing of which reputations are made.
As enumerated by James Hamilton of…
Not to let the dumb investors entirely off the hook, but most of the commentary here is missing the point that the MBS and CDO many-tranch issue was CONSTRUCTED to be complicated and opaque. The sum of the whole was to earn an average rating higher than the sum of the parts including the top tranches which had a back and forth with the rating agencies until they earned the misleading AAA rating.
If you doubt the complexity part, consider how the inability and timelag of necessary workouts will play out now between servicers and investors in a mortgaged bundled tranch world with with early payoffs and higher than expected foreclosure experiences now putting different tranches into early and confused conflict. When are workouts too good for the beleagured borrower with a stepped up ARM and will the market keep sinking so settling low for a workout now is better than foreclosure into a sinking market?
And the investment banks who packaged and sold this complexity and presumbably understood it well made sure they held as little risk as possible. And we can be expected to see interesting lawsuits about whether mortgage originators and MBS and CDO packagers represented documentation of income and other potential fraud and loan to value issues accurately.
Complexity and opacity has a cost and it will soon be evident as the liability lawyers full employment act of 2008-2010.
How about this: US/EU/Japan central bankers cutting interest rates to record lows coupled with EM’s central banks buying reserve assets drove private investors towards high risk assets to seek yield. This coupled with the agency problem of the asset managers (built-in incentives to take risks due to the compensation structure) and the perverse nature of the housing boom (high yield with record low default rate) led to the inevitable subprime debt binge. CDO structure is probably only a minor issue though it did make the flame-out a lot faster and more dramatic.
Risk management failed everywhere. Risk management does not exist in one body or group of persons, it is inherent to all investment decision making processes whether that be the person buying a home, a corporate risk management department in a financial institutions, a pension plan’s corporate governance framework, a rating agency, a bank regulator, the Fed, risk management professionals in hedge funds/pension funds/ etc.
The risk management failure was placing too low a probability on the adjustable rate mortgage reset risk; that being the risk of a combination of higher interest rates, a flatter yield curve and reductions in home price appreciation.
There are lots of excuses for this bad risk management like – we assumed the mortgage originator properly tested the borrowers ability to pay not only at origination but also various renewal scenarios; another common excuse might be that the underlying risk was too complicated to model and stress test due to all the trancheing in the CDO structure with leverage and all so we relied on the rating agency.
A more fundamental cause of the risk management failure might be the lack of independence and power (or whatever you want to call it; backbone, approval authority) in the hands of the various risk management persons/functions. Risk managers tend to be nerdy and are often ineffective at saying no to flashy aggressive extroverted asset managers/investment banks/CEOs etc. Further, these sames asset managers/investment bankers/CEOs are motivated by an inherent moral hazard that persists in the capital markets (for good reason) whereby they typically have a compensation arrangement with little downside (no bonus if the risk managers worst case scenario comes true) but big upside in most normal scenarios. Risk management compensation on the otherhand is fairly stable with salary and bonus that is often a function of overall firm profitabilty or some qualitative objectives.
It is interesting, given the significant advances in modeling and risk management frameworks in the last 20 years, there has obviously been very little progress in developing an effective mechansim to check and balance this moral hazard in the investment business and capital markets.
Prof Hamilton,
either the rating agencies shouldn’t have provided the favourable ratings for the CDO’s in the first place – and there’s an obvious problem here as they were paid by the issuer after close scrutiny of the product – or they provide rubbish investment advice: I know that a US gov’t bond is AAA; it’s for complex products that I need them.
So what are rating agencies good for in the end?
JDH –
At the top level of investors didn’t we have top quality outfits like UBS buying stuff created and marketed by other top quality outfits like Goldman Sachs?
Why does an investor buy stocks like GE or Berkshire Hathaway (Warren Buffett) except they have done well in the past and seem poised to continue? Perhaps we all need to be doing more work before we invest.
Blinder did a good job but left off the finger of the FASB and the Public Accounting profession.
In the good old days a bank made a mortgage and recognized revenue and profits over the course of the loan. financially conservative.
Now the subprime mortgage comes along and is bundled with other subprime mortgages and sold to investment bankers. The profit at the mortgage maker is recognized immediately EVEN THOUGH there is this little tiny buy-back clause if the loan goes bad.
However, like Moody’s et al and real estate appraisers the Public Accounting business runs on rules AND Transaction History. the history of subprime defaults looked good UNTIL you saw:
a. refinancing to keep the loan alive and this refinancing only possible because of high inflation rate of home prices.
b. Terms of mortgages that made payments OPTIONAL.
[Among other things a speculator could buy a house and resell in a year making few or even NO payments because for the first few missed payments were allowed by the terms of the loan so the borrower was not in default. While the payments missed later would provoke a process that would result in foreclosure… it would NOT if you sold it at a profit at the right moment.]
At a first level of examination the audit looks for Non Performing Loans. Not seeing any the then look at the reserves for bad debts and see if they are consistent with loan performance. voila! subprime loans do not default often and thus only small reserves are needed and profits can be recognized.
EXCEPT that auditors should be examining the terms of the standard loan agreements and sampling some for a full history of to whom credit was extended and why did it appear that these borrowers were doing so well. I sniffed this rot out in a flash. any competent auditor should have as well.
PS I was much amused at the irony of failing to pay on a pay-option loan causes the value of the loan asset to increase because the unpaid amount is added to the principal owed. neat trick.
I agree with akpundit’s suggestion that informed parties likely are collecting fees and shedding risk, though I thought that, in principle, the investment banks that package loans and issue MBS/ABS retained exposure through the equity tranche. Have they been able to lay off that risk? Even if the investment banks did retain the equity tranche, those tranches in some cases must be nearly worthless–does anyone know whether that will have any economically significant incentive effects over the remainder of the vehicle’s life?
Regarding James Hymas’ earlier post, I think that there is a difference between the complexity of the T-bond futures contract or knock-out options on the one hand and securitization on the other. In the former case there is a deep, liquid market for the underlying, making it possible to price the derivative. All of the complexity relates to the pricing model for the contingent claim. In the case of sub-prime ABS and CDOs of ABS, I think that investors are badly uninformed about both the underlying credit and how the vehicles work and will find that they come out on the short end of the stick on both counts.
The politicized fed knew what it was doing. They needed/used these products to stimulate beyond the tax cuts, and war mongering to get a 2nd term for Bush. Easy money (Greenspan monetary policy) was used to manipulate the mass psychology of housing desire to get beyond the close election experienced in 2000. It is all coming back in the inverse now.
Since someone already pointed to my blog post (thanks SotT), let me chime in with a couple points from the bond trading world.
1) The end investor drives everything. The Professor is right that with no investor, nothing else matters. Now the ratings agencies, borrowers, and lenders aren’t blameless. But the lenders created perverse incentives for borrowers and investors created perverse incentives for lenders.
2) The conflict of interest issue with S&P and Moody’s gets more play than it should. I’m not saying there is no issue of conflict, but to focus too much on the conflict issue misses a bigger problem with CDO ratings. That is the proper estimation of default level, timing and correlation. See here.
Any good reason for not using a finger to count “Real Estate Appraisers” into the blame pool?
TDDG,
Thanks for the comment. A question, shouldn’t you push back farther in the chain than lenders and investors since to create incentives they need FED complicity?
Not that I hold the Fed blameless either, but had underwriters forced borrowers to substantiate their income as well as qualify under a fully indexed ARM rate, we wouldn’t be in this mess. It was investors, primarily CDOs, that eagerly bought up these pools. Because there was an investor (greater fool?) to buy the mortgage loan, originators were happy to underwrite it.
Direlink misstates governmental policy over the last 7 years.
Bush has pursued a formally bipartisan economic policy to deal with the recession and 9/11 since the beginning of his first term.
For the liberals, he adopted (or acquiesced) to Keynesian remedies – budget deficits. For conservatives, tax cuts, easy money and low interest rates.
As to houses, one either pays the banker or pays the owner. The banker was too easy so the owner got the appreciation. Now bankers no longer want to play (tougher loan origination rules) so owners can’t play either. Equilibrium will be regained since no market for homes is politically untenable.
Joseph,
I think I understand your point. I might replace “budget deficits” with “massive stimulative spending.” The budget deficits have actually been in decline due to the tax cuts.
While we are pointing fingers: I think the relentless pushing of the “American dream”, “ownership society” or whatever various administrations called it deserves some blame. It was poor policy to try to get everybody to own their home. Financially it just doesn’t make sense for a lot of people (whether they can afford to buy a home or not).
The Golden Grail of financial investing is an above market rate of return. That requires shouldering more risk by investing in plays that promise quasi-rents. Nature usually provides the quasi-rent plays, but saavy institutions can create quasi-rent plays by erecting their own barriers to entry or in this case barriers to information.
Trying to regulate the latter behaviour in a world of unlimited technological possibilites is going to be inevitably tough.
Overcoming the time inconsistency problem by formally and credibly adopting inflation targeting or a similar price level averaging policy will deliver a better return for the policy shock buck.
I would point one finger at US federal reserve real output mandate and another at the suffocating wave of hyper-vigilant mass hysteria that overwhelmed the USA in the wake of the September 11 terrorist attacks.