In the post from Thursday, I argued that the current recession is the result of classic over investment, in this case in housing. This brings up the very important question of why we had an over investment in housing. (I'm posting a new poll on this subject, but first, you read.)
The potential suspects I'm going to consider are: the GSEs, the Fed's low interest rate policy in 2002-2003, and the rise of structured finance, especially CDOs.
First, one suspect I'm immediately tossing out. Lack of government regulation on lending. There is a perfectly legitimate argument that regulation was too lax. But I'd rather investigate why banks were so willing to underwrite so many sketchy mortgages. Because if there was some perverse incentive to lend money recklessly, then no regulatory scheme would have prevented it. Had mortgage regulations been more stringent, the lending would have flowed someplace else anyway.
GSEs
Blaming the GSEs is complicated, because the GSEs have been around a very long time. It is undeniably true that without the GSEs there wouldn't have been as much supply of available funds for loans. But in my opinion, the GSEs incremental impact on loanable funds didn't wildly change from 2001-2007. In fact, the evidence is that the GSE's market share declined during this period, as other types of securitization gained in popularity. More on that later.
1% Fed Funds in 2003
This is a popular argument, that ultra low rates lead to ultra easy liquidity. Plus ultra low government bond yields lead investors to aggressively seek out higher yielding investments.
There is something to this, but to me it doesn't explain why availability of mortgage capital actually accelerated in 2004-2006 as interest rates were rising.
Structured Finance
I argued, over a year ago, that CDOs were the primary culprit in causing the sub-prime problem. Consider: why were mortgage brokers willing to underwrite every loan they saw? Because they knew they could sell the loan. Who was the buyer? Structured finance.
On the lower end of the credit spectrum, structured finance created the leverage. On the top end of the credit spectrum, banks levered their positions through SIVs. All that liquidity flowed right into mortgages, and thus into houses.
Now, you could argue that 1% Fed Funds helped to create demand for structured finance. Sure. There is plenty of inter-related issues here. But in my opinion, without CDOs, the Fed's interest rate policies wouldn't have caused the housing bubble we're seeing now. I'd also argue that low volatility, not low interest rates, were the primary reason why structured finance flourished. The Fed can't really be blamed for creating a low volatility environment, after all, that's their job.
If readers have other potential suspects, go ahead and post a comment.
From your blog to the Banking/Finance subcommittee ears.
ReplyDeleteYou've got the right question for the topic of a book rather than a blog.
Very insightful post. Thank you for blogging, it's almost impossible to get honest analysis in the mainstream press. Keep up the great work.
The reason it "accelerated" is the same reason a snow ball accelerates when it begins barreling down a hill. The important thing is who created the snow ball in the first place.
ReplyDeleteC'mon now vijay...there are innumerable snowballs created every day by people who hope they will gather snow as it rolls downhill. The important thing is why did this own get so big.
ReplyDeleteYou can't blame CDOs without also implicating the rating agencies. Their stamp of approval gave investors comfort that even with the extra yield, they were still getting a tranche of a certain rating.
ReplyDeleteBut let's go further.
Vijay has a good point, where did this start? Underwriting standards were terrible. Years of price appreciation in real estate kept defaults artificially low. Everyone was flipping houses and banks weren't that concerned about credit because if a borrower got in trouble, he could sell for a profit.
Builders opened mortgage banks that gave loans on the houses they couldn't sell. Then the mortgage banks, like sharks they can't keep moving unless they keep making loans. And when the loans ran out, or slowed down or got returned the mortgage banks went under.
Strike that, first they got bought by the I-banks that thought it would be a softer landing than letting them fail (this is in 2006).
But go ahead and blame the CDOs. The ABS backed CDO is dead, no new issuance and deals unwinding every week. They were buying everything under the sun and creating a demand for even sub prime assets.
If I buy bonds, I will be paid more interest for a higher risk that I will lose the principal. There is risk associated with the loss of this particular investment.
ReplyDeleteThe money I use for my bonds is part of the money or assets I own. When I invest in a particular bond, however risky, I must consider the risk of that loss to my total holdings. This risk is different than the risk of the individual investment.
CDO's are no different in your excellent description than any other investment. To the extent that any investment is unclear or the risk uncertain, it is a poor investment. The CDO's have a ladder of risk and rewards. Period.
In other words, one can question the sense of investing in CDO's. One can also consider the risk to your total holdings, i.e., how much you invest in CDO's of your total wealth.
GSE's main problem was the system of implicit and explicit government guarantees to intervene.
While it makes sense to look for investments that pay more than the safest investments, there is no rule that you need to invest foolishly. That doesn't wash. At most, it's a necessary condition.
The CDO's are like every investment. The sound rules of investment apply also here.
"Because they knew they could sell the loan. Who was the buyer? Structured finance."
I cannot conceive that the amounts risked would have been undertaken without the understanding that the government and Fed would intervene to protect those investments. Otherwise, it simply comes down to poor investing, fraud, negligence, wishful thinking, fiduciary irresponsibility, etc. If there were riskier investments called for, it was because individuals were willing to take more risks. That's what needs to be explained. Not the instruments created to serve the need. If it wouldn't have been CDO's, it would have been foreign investments, riskier bonds, whatever.
To the extent that a person can be fooled by risk, there is no solution on earth to that.
This is a human agency as opposed to a mechanistic explanation as I term them.
I'm on my own in thinking of these things this way, but that's okay.
However, I want to complement this blog and the comments on the last two days especially. They were excellent. I'll keep reading, because I'm learning so much. Take care.
Don the libertarian Democrat
This is not the first time excess credit expandsion has created inflated assets nor will it be the last. The search for the answer to your questions will fill business text books for generations and will consume the intellectual life of many a professor to come.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteOK, everyone mentions the 1 year at 1%, but ignores the 3 years at 2% or under. That's massive stimulus that can't be ignored.
ReplyDeleteLow rates in the absense of ultra-low inflation always (virtually always) creates these kind of loose lending conditions.
Of course, the separation of the lender and borrower via CDO's etc. didn't help, but the prime number in this was 2 (as in 2% funds for three years).
Jay
The Confused Capitalist
(Yes, I know that 2 isn't a prime number, OK).
I think of the bubble as more of a cause than an effect. If prices are steadily increasing every year, people are going to "innovate" and create more and more leveraged products or behavior to take advantage of the easy money. At some point this blows up.
ReplyDeleteFor example, in Spain, despite banks being extremely regulated and lacking CDOs (due to a previous blowup), property prices still went through the roof. Now they are crashing and will probably take a large chunk of the banking sector with it.
Let's not forget that the home builders, mortgage brokers, banks, investment banks, and ratings agencies are all made up of people. They were people that were paid for short term results on products that could not be valued except on a long term basis. They made money while the bonus money was good.
ReplyDeleteThe only folks who lose are the home buyers (who didn't sell), the stock holders (who didn't sell), the CDO holders (who didn't sell), the CDS writers (who did sell), and the taxpayers (who pay their taxes).
This gets to one of the fallacies of corporate governance, that anyone really cares about the stockholder over more than ~1year time span. The CEOs don't care, the middle management doesn't care, and the "little guy" is just hoping to keep his job while the guy next to him sells more toxic crap by ignoring the future. There was plenty of follow the leader going on... and how could you lose except by not playing the game?
Lot's of folks are to blame, but the folks who really got taken were the ones with the least knowledge: the stock holders of the companies, the pensioners, those who depend on insurance, and the backers of US Treasuries.
The employees made out like bandits. They made profits far in excess of the actual value of their work for a few years. Although they got used to the income, and perhaps borrowed against it, they should have been able to save enough for the years it will take to get a real job.
Perhaps the stockholders and tax payers would have wised up a little earlier if someone at the FED who wrote a dissertation on housing bubbles had called a spade a spade.
http://online.barrons.com/public/article/SB120917419049046805.html?mod=mktw
There wasn't even an "irrational exuberance" quote. And it's not the fault of the markets, it's the fault of those who allowed human beings who risk only OtherPeoplesMoney to make huge profits by misinterpreting risk... a Ponzi scheme is never far away.
Hedge funds are the next good example.
The housing bubble wasn't "created" by the regulators, the rating agencies, loose credit, etc. They are symptoms of the problem. Over the past 10 years it became a widely held theory that the housing market would appreciate at 5% per year forever. That was based on the previous 50 years of housing prices. It was a classic bubble mentality.
ReplyDeleteThe thinking at the rating agencies was no different than the consensus of the general public. Their models were based on the housing market appreciating 5% per year. If the asset is to appreciate at that rate the borrower is not important. If the borrower cannot pay the mortgage, the asset will appreciate and the lender will be protected. Unfortunately markets don't go in one direction and now that the theory has failed we are realizing the absurdity of the theory.
What is compounding the problem is the mispricing of the mortgage market. The ability of the borrower to cash out as the asset appreciates or to walk away when the asset depreciates will eventually be proven to be the undoing of the capital market system.
You are arguing over which pebble is to blame for the avalanche.
ReplyDeleteThis is a social disaster with numerous causes, many of which are not subject to easy regulation: entitlement indoctrination of children, the peace dividend from the end of the Cold War devaluing labor, the increased velocity of money due to software and networks, and many others. Comprehensive repair of any one problem would have just shifted the ultimate failure to somewhere else.
Basel II.
ReplyDeleteWe Should Move Away From Foreclosure System
ReplyDeleteI want to propose a thesis. The cause of this crisis was the foreclosure system itself. The crisis could have been avoided had a system been set in place that renegotiated mortgages immediately upon delinquent payments.
The cause of the crisis in the housing sector was a misallocation of resources by individual homeowners. The borrowers took mortgages that were eventually too high for them to afford. The loans/mortgages were based upon terms that were much more likely to lead to default/foreclosure.
But what if foreclosure had not been the option? What if the first option had been to renegotiate the mortgage/loan to a payment affordable to the borrower? The level of payment that the borrower could afford would then trigger a set of possible options:
1) Increase the price of the house
2) Increase the length of the mortgage
3) Base payments on percentage of income, so that they will rise as borrower's income rises.
Now, however it would be done, wouldn't this have been a better situation than the one we currently face? The reason to try a mortgage renegotiation plan now is to try and see if we can move away from the foreclosure system and devise one that keeps borrowers in their houses. Isn't that a sensible proposal?
Don the libertarian Democrat
Changing of lending from Originate and Hold to Originate and Distribute, excess demand for mortgage securities because interest rates were so low, etc
ReplyDeleteAs to the origins of the current problem, there are probably social/political one that go deeper in history, but the single origin I think is the colossal credit explosion that started in 1995. This has created a wall of money that has bubbled up asset class after asset class, until housing was bubbled up most destructively (because at least most other asset classes finance production in some part, housing is just a cost).
ReplyDeleteJust about every finance-related graph shows a sharp upturn in 1995, where the trend line rises far more steeply, and in particular faster than GDP. For example stocks bought on margin, and M3/MZM:
http://bigpicture.typepad.com/2007/10/margin-debt-gro.html
http://www.nowandfutures.com/key_stats.html
But then also USA debt:
http://blogs.cfr.org/setser/2008/09/08/the-stealth-bailout-illustrated-in-close-to-real-time/
and blue-chip stock prices, financial and non financial:
http://finance.yahoo.com/q/bc?t=my&s=MER&l=off&z=l&q=l
http://finance.yahoo.com/q/bc?t=my&s=FNM&l=off&z=l&q=l
http://finance.yahoo.com/q/bc?t=my&s=IBM&l=off&z=l&q=l
http://finance.yahoo.com/q/bc?t=my&s=GE&l=off&z=l&q=l
So something important happened in 1995. From the graphs it looks like that it was a gigantic expansion in the availability of "money", and in particular of short term credit.
The only plausible explanation that I found is amazingly from a gold bug (who was inspired by Luskin...):
http://www.signallake.com/innovation/FedReserve1995.pdf
«The key event that happened around 1995 is that the fractional reserve ratio was not only lowered, it was effectively eliminated entirely. You read that right. The net result of changes during that period is that banks are not required to back assets which largely correspond to M3 or "broad money'' with cash reserves. As a consequence, banks can effectively create money without limitation. I know that sounds hard to believe, but let's look at the facts.»
I agree with several comments here that suggest there is a deeper reason for our problem. I'd say it gets down to a principal/agent problem. From a separation of mortgage origination from mortage investing to corporate managers not being aligned with long-term shareholder value. Its all principal/agent.
ReplyDeleteAnd clearly the long-term evidence for positive home price appreciation lulled people to thinking these securities were safer than they turned out to be. But what happened to cause this long-term trend to end?
So in my post I'm thinking more specifically to what triggered this housing bubble. I'd say it was creativity in creating mortgage structures.
We know with hindsight that once CDOs started re-packaging mezz ABS products, the game changed. This created a ready market for the difficult-to-sell portion of ABS deals: the mezz. And because the collateral was already investment-grade, it was relatively easy for ABS CDOs to get AAA ratings on their senior pieces.
And thus demand for ABS (and the thus loans) expanded greatly. That's the proximate cause for the bubble.
As kristof pointed out, other countries like Spain haven´t had the Agent-Principal conflict and still there was a bubble.
ReplyDeleteIn Spain, the model is hold-to-maturity so banks have no incentives to give bad mortgages. And customers don´t have the incentive to take them, since the debt is full recourse. Yet the bubble happened. I would like to point out, however that its bursting is unlikely to be as bad as in the US. Currently, with 11%+ unemployement the rate of default is below 2%.
So I am blaming low interest rates.
Good post again. I think that several things happened at once that created the perfect storm. Imho, there are several things to remember:
ReplyDelete1. The financial industry is already very highly regulated. The banks with the biggest problems like WFC are the most highly regulated of all of the institutions. People forget that WFC has +100b of HELOC's on their balance sheet that they carry as bank loans at 95 cents right now. The true market price is probably closer to 20 cents.
2. The Fed really only manipulated the front end, so we still have to understand why the rest of the curve became so cheap. I think that is just b/c foreign central banks were big buys across the curve.
3. The switch to basel 2 made ratings more important in computing capital requirements. Most IT systems download the banks books, consolidate, and then pair off positions against each-other based on risk characteristics. You can imagine the logic in each of these programs looked at an AAA treasury in the same way as an AAA CDO tranche.
4. Changing the tax rules to allow 500k gains on a primary residence added fuel to this.
5. Until about 15 years ago there were strict limits ( I beleive to the MSA level) for GSE lending. I think people used to call this red-lining. Congress then reversed itself and started forcing expanded lending to "under severed" areas.
This includes the acceptance of down payment gift programs. That's got to be one of the craziest things and it was encouraged by congress and the administrations (Clinton-Bush).
6. IMHO, the emerging market players were smart in a way. The BRIC countries didn't want to leave their countries vulnerable to currency shocks, so they accumulated a buffer in treasuries and GSE debt, encouraging issuance.
7. This inlow of foreign central-bank money created the search for yield, which drove the CDO business.
8. The mortgage banking business blossomed. I can't count how many friends that I have who became mortgage brokers. They figured out that they could issue a mortgage to anyone (often with fraudulent W2's and paid-off appraisers) and then just pass it off to the GSE's. This flood of GSE money helped raised to floor radically on the entire housing market.
And at the end of the day, asset bubbles happen and will happen again. Maybe there's not too much that we can do...
Housing had a positive "carry" for 3+ years.
ReplyDeleteIt is like decades of infomercials suddenly came true, and you didn't have to send in money for a special real estate course or go to a seminar.
The man on the street isn't into deep financial thinking -- he is just thinking monthly payments and personal cash flow.
He also knew a deal when he saw it. Pay 4% and collect 10% appreciation. He didn't even have to wait until he sold -- just take it out via home equity loans.
So you had the so called "smartest people in the room" or at least the nerdiest creating byzantine structured finance securities and the man on the street knowing on some visceral level that assets were being mispriced.
You had highly paid finance ppl creating asset valuation models based on aggregate statistics and other side of the deal were people that could only think in monthly payments.
Nothing good could come out of this.
I agree with most of the other comments as well.
One additional element. In the 90's, most mortgages were close to 8% and in the 00's they were closer to 6%. From the monthly payment perspective, housing could appreciate 25% without payments going up. This might be the tinder that started the bubble.
I do think that Basel II had a lot to do with the rise of CDOs. Suddenly AAA wasn't just an opinion, it was the law. I say its a classic example of regulation creating rules just strict enough to be gamed.
ReplyDeleteTalk about moral hazard...
Spain is an interesting case. I don't really know enough about it to comment.
Related to the rating agency issue and many other problems that have been exposed lately is the 'problem of data'. There are many aspects to this, but here are a couple
ReplyDeletePeople like to have a lot of data when they build models. If there are more numbers in my spreadsheet, it must be more accurate, right??? If I have daily data, it's better than monthly, etc. The pinnacle is tick data... There is a great bid for spurious accuracy.
The amount of data available has grown exponentially, particularly over the last 25 years or so. If you go back 100 years, data is incomplete, infrequent, and aggregated. This is not suitable for complex correlation models, so it's easier to just use the last 20 years instead.
As a result, very little analysis actually looks at the "long term". Oft cited factoids about long term returns (i.e. stocks outperform bonds over 20 year periods) neglect to mention a simple fact; they are referring to a sample in which there are only 5 or 6 independent observations!
No matter how good or bad the theoretical framework was for all of these models, there was no excuse for putting so much faith in something that is at best a good guess.
An extension of your CDO theory:
ReplyDeleteI think hybrid CDOs with 75% synthetic exposure were the worst culprits. It's pretty obvious that they increased the number of CDOs produced with any given number of actual mortgages -- this increased the fee income that a single mortgage could generate and exacerbated the agency problems.
Furthermore, I think there's something perverse about selling a synthetic asset that guaranteed to cause problems. Suddenly we've left the financial world where we believe that our actions are making the economic pie bigger and instead have entered into a zero-sum game. In a zero sum game that's marketed as an investment, somebody is the stooge.
(1) The commercial banks create NEW money when making loans to, or buying securities from, the non-bank public; whereas lending by financial intermediaries activates EXISTING money.
ReplyDelete(2) Bank lending expands the VOLUME of money and directly affects the VELOCITY of money, while intermediary lending directly affects only the VELOCITY.
(3) Monetary savings are never transferred FROM the commercial banks TO the intermediaries; rather are monetary savings always transferred THROUGH the intermediaries. The funds DON’T LEAVE the system.
(4) Inflation results from a long-term excessive flow of money relative to the volume of real output of goods and series offered in the markets.
From 2003-2008 Bank credit of all commercial banks expanded at excessively high rates-of-change - c. 10% annualized. During part of the housing boom, the “money multiplier” doubled in a six year period. The technicians at the FED announced that reserves were no longer “binding”.
The sales volume of house-hold home-mortgages grew by 246% from 1998-2007 to $10,542b. GSE’s grew by 243% to $887b (Z.1) – Accrued Interest is right about GSE’s contribution.
But the growth of bank credit doesn’t explain the inflation that took place in housing. Structured finance made long-term commitments more attractive. This segmented distribution of available credit was fueled by an increase in the supply of loan-funds.
I.e., the addition to the supply of loanable funds kept interest rate low in the long-term markets, which stimulated spending. A higher money velocity is associated with shifts FROM money creating depository institutions TO financial intermediaries or SIVs. That flood of hot money vastly accelerated the velocity of money.
Even more fundamentally, appraisal at current market value all by itself can lead to a bubble. If prices go up 20% a year for a sustained period of time, a "conservative" 80% exposure for the banks isn't really removing very much risk.
ReplyDeleteI'm not sure how much structured finance contributed to inflating the bubble, but in terms of collateral damage after the fact, that's going to be huge.
Imagine the banks as drunk drivers... not wearing a seatbelt didn't cause the accident, but may have fatal consequences.
Investigation is the process of inquiring into a matter through research, follow-up, study of discovery.
ReplyDeletefutures, trading, value, investing, forex, stock