Will stated-income (liar) loans make a comeback?
Some form of stated-income loan will likely be offered by portfolio lenders in the future, but it will be limited to borrowers with large down payments.
When lenders underwrite new loans, one of the fundamental tasks they perform is determining whether or not the borrower can repay; therefore, allowing borrowers to simply state their income with no verification is an abdication of an underwriter’s responsibility. Stated-income loans (aka liar loans) were the worst financial innovation of the housing bubble because these loans undermined one of the pillars of lending: borrower capacity.
Liar loans were the worst financial innovation of the housing bubble because these loans caused investors in mortgage-backed security pools to question the financial representations of all borrowers in all loan pools. This doubt about the veracity of loan qualifications spread from the pools that specifically allowed liar loans to all MBS pools, causing investors to abruptly stop buying mortgage-backed securities — a problem we still have today.
The abrupt halt of the flow of investor money resulted in the mortgage credit crunch of August 2007 that triggered the financial collapse of 2008. Investors have good reason to be very cautious about buying these pools, and given the risk these loans posed to the financial system, regulators have good reason to be cautious about allowing them at all.
The financial crisis was caused in part by widespread fraud, which may seem like an obvious point. But it remains surprisingly controversial. …
New academic research therefore deserves attention for providing evidence that the lending industry’s conduct during the housing boom often broke the law. The paper by the economists Atif Mian of Princeton University and Amir Sufi of the University of Chicago focuses on a particular kind of fraud: the practice of overstating a borrower’s income in order to obtain a larger loan.
In his zeal to write an anti-bank screed, the reporter is missing something very basic. Notice above he talks about the lending industry’s conduct when he says a borrower’s income was overstated. Well, didn’t the borrower overstate their income, not the lender?
Are we really expected to believe every borrower, or even most borrowers, were unknowingly duped into filling out fraudulent loan applications? When a borrower making $50,000 per year signs an application stating they make $130,000 per year, do they really bear no responsibility for that?
Many on the political left believe they don’t.
They found that incomes reported on mortgage applications in ZIP codes with high rates of subprime lending increased much more quickly than incomes reported on tax returns in those same ZIP codes between 2002 and 2005.
“Englewood and Garfield Park are two of the poorest neighborhoods in Chicago,” they wrote. “Englewood and Garfield Park were very poor in 2000, saw incomes decline from 2002 to 2005, and they remain very poor neighborhoods today.” Yet between 2002 and 2005, the annualized increase in income reported on home purchase mortgage applications in those areas was 7.7 percent, strongly suggesting borrowers’ incomes were overstated.
Liar loans were more prevalent in poor areas. So what? They were offered a choice: lie and get a house, or tell the truth and not get a house. Is it surprising that many of them lied?
Wealthier neighborhoods also had many liars, but for them the circumstances were slightly different. If they lied, they got a better house, but they could probably afford a house either way. Since more affluent borrowers could own either way, fewer of them lied.
The study is particularly noteworthy because in a study published this year, three economists argued the pattern was a result of gentrification rather than fraud. “Home buyers had increasingly higher income than the average residents in an area,” wrote Manuel Adelino of Duke University, Antoinette Schoar of M.I.T. and Felipe Severino of Dartmouth.
Academic nonsense. Anyone who knew what was going on during the housing bubble knew liar loans were extremely common.
The three economists also argued that lending in lower-income areas played only a small role in the crisis. Most defaults were in wealthier neighborhoods, where income overstatement was less common.
“The mistake that the banks made was not that they over-levered crazily the poor in a systemic fashion,” Ms. Schoar said. “The banks were not understanding or not wanting to understand that they were increasing the leverage of the country as a whole. They were forgetting or ignoring that house prices can drop.”
The new paper by Mr. Mian and Mr. Sufi is a rebuttal. Their basic point is that the incomes reported on applications should not be taken seriously. They note that income reported to the I.R.S. in these ZIP codes fell in subsequent years, a pattern inconsistent with gentrification. Moreover, the borrowers defaulted at very high rates, behaving like people who borrowed more than they could afford. And the pattern is particular to areas of concentrated subprime lending. There is no income gap in ZIP codes where people mostly took conventional loans.
“Buyer income overstatement was higher in low-credit score ZIP codes because of fraudulent misreporting of buyers’ true income,” Mr. Mian and Mr. Sufi wrote. …
They are also right.
Both groups of academics are correct in their conclusions, and their conclusions are not contradictory. Liar loans were very common, and just because they were more common among poor people doesn’t mean those loans caused the housing bubble — the poor didn’t cause the housing bubble.
The main culprit in the housing bubble was the negative amortization loan. This unstable loan product is what allowed borrowers to leverage unsafe multiples of their income. Liar loans were another mechanism used to overborrow, but it was the Option ARM that caused house prices to get severely inflated, not liar loans.
The prevalence of income overstatement is sometimes presented as evidence that borrowers cheated lenders. No doubt that happened in some cases. But it is not a likely explanation for the broad pattern. It is far-fetched to think that most borrowers would have known what lies to tell, or how, without inside help.
Yes, borrowers had help, but it was still the borrower who knowingly told the lie.
We do not have a comprehensive accounting of the responsibility for each instance of fraud — how many by brokers, by borrowers, by both together.
Some fraud was clearly collaborative: Brokers and borrowers worked together to game the system. “I am confident at times borrowers were coached to fill out applications with overstated incomes or net worth to meet the minimum underwriting requirements,” James Vanasek, the chief risk officer at Washington Mutual from 1999 to 2005, told Senate investigators in 2011.
Most fraud was collaborative.
Think about the typical conversation back in 2005: the borrower would say they wanted to buy a $500,000 home on their $60,000 a year income. The mortgage broker would say they couldn’t buy a $500,000 home with a $60,000 income, but they could buy the home with a $100,000 income, and if they were willing to sign papers stating they made $100,000 a year, they could have the house. What would you expect most borrowers to do?
For as much as I would like to keep an open mind about “innovative” loan products, I can’t see any reason stated-income loans (liar loans) should be allowed to exist. These loans undermine the foundation of mortgage lending, and their only future is as a niche product offered by portfolio lenders who understand the risk and are willing to assume it.