The Bubble and the Crisis: A Picture Book
Pictures Of The Crisis: Timing Is Everything
In the spirit of show, don't tell, this week is about visual enhancement, using charts to place familiar concepts into a broader context. For many this is old and familiar news; for others the info might be fresh.
Right away the charts tell you what every real estate lender knows: (a) Location, location location; and (b) Timing is everything. The point in the housing cycle when you booked the loan drives your recovery. Or more specifically, home price appreciation, positive or negative, post-closing drives loan recovery more than any other factor. This has been especially true for subprime loans.
The Bubble in Four Key Markets
As the Financial Crisis Inquiry Commission illustrated how the housing bubble and loan losses were concentrated in the four sand states: California, Florida, Arizona, Nevada.
"Which Housing Markets Have the Most to Worry About," by Nick Timiraos for The Wall Street Journal, is an excellent resource for evaluating most major real estate markets today. Below are four of the charts that Nick assembled. The good news: None of the four markets below (many more are highlighted in the edifying article) qualify as "trouble spots." But these four markets were at the center of the housing bubble the preceded the crash.
Remember, mortgage lenders care about nominal prices, which determine loan-to-value. Whereas real prices tell us more about macroeconomic issues such as housing affordability.
Right away, we can see that housing cycles are very, very long. Certainly in real terms, home prices in these markets were essentially flat during most of the 1990s and into the early 2000s. In nominal terms, home price appreciation was modest. The nominal price of a home in Los Angeles at the beginning of 2000 was exactly what it was at the beginning of 1990.
These and other charts in the Journal eviscerate the oft-repeated claim that the bubble started under Bill Clinton. In each instance, the slope of the curve changes dramatically after 2000, when Greenspan slashed rates and kept them low through most of 2004. And when Greenspan and the Bush Administration chose to turn a blind eye to an epidemic of mortgage fraud.
GSEs and the bubble: Here's another graphic, courtesy of the good people at Trust Company of the West, which illustrates how home price appreciation impacted the GSEs in comparison to the standard Case-Shiller benchmark composite. FHFA data shows that home prices in Los Angeles rose just as dramatically as they did in the Case-Shiller composite, but the GSEs' geographic exposure was much more diversified. The GSEs' mandate was to provide financing to underserved markets.
What caused the spike in homeownership? Then there's that other familiar trope, popular among the blame-everything-on-Fannie-and-Freddie crowd, which claims that the Clinton Adminsistration artificially inflated home ownership with affordable housing goals. As it turns out, the Census Bureau shows us that homeownership did spike, among Americans who live alone. That trend began in the 1980s. Whereas by 2000, the homeownership rate among households of two or more was close to what it was in 1980. America was in the middle of a housing slump in 1990. Live-alone individuals were not targeted by affordable housing goals.
Keep all that in mind whenever you read nonsense written by those who traffic in stereotypes that pander to class bigotry. Exhibit A: "Bubba & the Housing Bubble" by Charlie Gasparino.
The Housing Department was Fannie and Freddie’s top regulator — and under [Andrew] Cuomo the mortgage giants were forced to start ramping up programs to issue more subprime loans to the riskiest of borrowers.We know how that turned out: Fannie and Freddie help stoke a housing bubble that actually made homeownership less affordable unless borrowers took out ever-more-risky loans.
And then for those who respect truth, here's what really happened. The Clinton Administration--both Cuomo at HUD and Larry Summers at Treasury, studied predatory (i.e. fraudulent) lending in the subprime segment. They found rampant abuses, and sought to curtail those abuses through legislation. One study showed that half of borrowers with subprime mortgages would have qualified for a GSE mortgage, and Cuomo wanted Fannie and Freddie to market to them. (See how Gasparino twists everything around.)
Mortgage Lending and the Bubble and the GSEs
The National Deliqnuency Survey, published quarterly by the Mortgage Bankers Association, is generally used as the standard benchmark for loan performance. And the most commonly used metric is the Serious Deliquency Rate, which applies to loans that are 90+ days delinquent or worse, because the date of foreclosure, or loan liquidation, relies on the discretion of loan servicers.
What does "subprime"mean? Subprime, Alt-A and Prime are all squishy words that have different meanings in different contexts, and those meanings have evolved over time. In generic terms subprime meant the borrower has bad credit, Alt-A meant the borrower has good credit but there's some slight flaw in the documentation, and prime meant the loan is well underwritten and the borrower has good credit. But there are subprime borrowers and subprime loan products. There are many instances when otherwise prime borrowers have been steered toward subprime products by mortgage brokers. Balance sheet lenders use one, very loosy goosy definition provided by regulators. For private label securitizations and for the rating agencies, everything is driven by FICO scores. A securitization is either a subprime, an Alt-A or a jumbo (i.e. prime) based on the weighted average FICO score in the pool.
In other words, a subprime securitization deal may have prime or Alt-A loans included in the portfolio and vice versa. Laurie Goodman, et al. illustrated the private label overlap in their book, Subprime Mortgage Credit Derivatives, which, I can assure you, will not make a comeback any time soon.
GSEs and the market: How do GSE loans fit into the MBA's nationwide averages? The MBA doesn't say, though it's a very safe bet that GSE loans represent a big percentage of the total and skew the nationwide averages accordingly. The chart below, prepared by me, shows that Fannie Mae serious delinquency rates closely track those of Prime fixed rate mortgages. The MBA uses a squishy definition of subprime, but, as one can see from reviewing the data, that subprime represents a small percentage of the overall total, and subprime loans have always performed disproportionately worse than prime. (Remember, most subprime loans are ARMs and most prime loans are fixed rate.)
Charts that illustrate the falsity of The Big Lie.
GSEs and the banks: Below is the only chart I prepared myself; everything else is from public documents. I used Fannie Mae as a proxy for both GSEs. Apples and oranges alert: These percentages reflect realized losses, from loans that were permanently removed from the books of the banks and the GSEs.
The data tells us several things. First, it reveals, in the starkest terms possible, the impact of the Federal Housing Enterprises Financial Safety and Soundness Act of 1992, which imposed affordable housing goals on Fannie and Freddie. From 1993 through the end of 2007 Fannie's average annual loan losses were about 2.7 points.
Of course, during the last half of that 15-year stretch Fannie benefitted from a housing boom. Annual credit losses, which ranged between four to six basis points during most of the 1990s, plummeted to two basis points in 1999, before a six-year stretch, from 2000 through 2005, when Fannie's average annual credit losse was one basis point. Who knows? Without a boom, Fannie's credit losses could have doubled.
Losses from 2008 onward were clearly disproporationate to anything Fannie experienced before. But, as Barney Frank likes to say when he's quoting Henny Youngman, compared to what? Compared to commercial banks and private label securities, the GSEs still did great. And they added liquidity to the mortgage markets when no other private players would.
GSEs and the Market, Per The Mortgage Bankers Association
by James Lockhart/ Public Domain
The chart above was prepared by James Lockhart as part of a presentation he gave at the American Enterprise Institute on January 25, 2011, two days before release of the FCIC Report. As you can see, AEI scrubbed the presentation from its website.
Why it's called a lie: Now, I feel compelled to articulate a few things that I had presumed were so obvious that they went without saying. In fact, the first time I wrote that the oft repeated claim--that most of Fannie's and Freddie's loans were "high risk"-- I got fierce blowback.
I come from capitalismland, which is dictated by certain rules. Lending is all about getting your money back with interest and fees. Risk never exists on a vacuum but on a spectrum, from lowest to highest. Risk analysis begins with a review of historical performance. The riskiness of a loan is based on a combination of many factors. Success or failure is measured by overall portfolio performance. Where I came from, anyone who disregarded or repudiated those rules would be considered an imbecile and/or a huckster. So I was shocked to learn that professors at elite business schools and other financial "experts" take an oppositional view.
And because lending is about getting your money back, and the best performing loans anywhere were underwritten by Fannie and Freddie with a stretch of outstanding performance extended for decades, the rules dictated that there was simply no way anyone could plausibly say that affordable housing goals contributed to the crash, or that the GSEs led the race to the bottom in credit standards, or that the GSEs were destined to fail, or that they were a major contributor to the financial crisis. If risk exists on a spectrum, then the best performing loans cannot be called high risk loans; the idea is oxymoronic. In fact, it's a big fat lie. It is not a "competing narrative" any more than an argument against the dangers of cigarette smoke is a "competing narrative."
And yet, a large army of "experts" and media types and "economists" and professors at elite business schools and politicians and high ranking government officials all keep repeating the calumnies listed in the foregoing paragraph, as if they were truths chiseled in stone. Mention the words "Fannie and Freddie" and big percetage of Americans start salivating like Pavlov's dogs, anxious and angry at companies they have been told are evil incarnate.
Make no mistake, the principles of free market capitalism--reliance on empirical data, transparency, personal accountability, the rule of law--are under attack in this country. They are under attack by "experts" paid by conservative think tanks who invoke class bigotry and race baiting to demonize the GSEs to distract away from an epidemic of corporate fraud.
GSE loans by vintage: Four economists presented the following two charts in the back of their paper for the New York Fed, "The Rescue of Fannie Mae And Freddie Mac." The authors all subcribe to the GSEs-are-fatally-flawed-and-must-be-abolished school of economic thought. As a result, they illustrate the deterioration of the GSEs' loan performance, while choosing to ignore any comparisons with the rest of the market.
Apples and oranges alert: These percentages reflect defaults, not losses. Losses are a fraction of defaults, reflecting the loss severity at the time of liquidation. So if you have a $100 loan and recover $60,your loss severity is 40%. Also,these charts show defaults by vintage, or year of origination. Once again, timing is verything. The chart I prepaerd showed loan losses from all vintages recognized in a calendar year. Banks and other balance sheet lenders have constantly changing loan portfolios.
These are defaults by year of origination. As you can see in the charts that follow, loans booked just before and just after the peak of the bubble, years 2006-2007, created most of the problems. Those homes suffered the most rapid declines in value, when homeowners had no benefit from the home price appreciation that preceded 2006.
The charts below were all lifted from Moody's Global Structured Finance Collateral Performance Review as of October 2015. They measure weighted average RMBS losses by year of origination. The line charts show actual realized losses from loans that have been liquidated, and then, below, they show Moody's estimates of what total losses will be when the entire securitization is no more. So, for instance, Moody's shows that 2007-vintage subprime securitization have realized a 36% loss rate, but that, eventually, the loss rate is expected to increase to 48%.
No matter how you look at them, private label losses were of an entirely different magnitude than those incurred by the GSEs.
Again, Moody's numbers are averages. Many securitizations performed far worse than the average, and many performed much better. But, given how different categories of RMBS deals were all defined by the same cookie-cutter capital structure, we can draw some inferences. If a single subprime deal has a loss rate higher than 20%, all of the sub-triple-A tranches got wiped out. If a single Alt-A deal had a loss rate higher than 10%, all of the sub-triple-A tranches got wiped out. If a single jumbo deal had a loss rate higher than 6%, all of the sub-triple-A tranches got wiped out.
You know who bought almost all of those sub-triple-A tranches? CDOs.
1 comment - The Bubble and the Crisis: A Picture Book
Leave a reply
RNC Suggests GSE Recap & Release is a Viable Option
15 Sep, 2017
Are credit risk transfers fatally flawed?
6 Aug, 2017
BlackRock document refutes Hank Paulson's story about GSE capital
3 Aug, 2017
More nonsense embedded within Perry v. Mnuchin, the "new capital paradigm"
19 Mar, 2017
More nonsense embedded within Perry v. Mnuchin, the DTAs
16 Mar, 2017
Perry v. Mnuchin: A case study in disingenuity
10 Mar, 2017
Kroll Rating Agency Frames GSE Reform Around Urban Myths
27 Jan, 2017