For the past three years, the Justice Department has filed a series of bruising lawsuits against U.S. lenders for race discrimination while hiding behind opaque investigative methods and hoping banks don't challenge the cases in court. But while most banks have quietly settled the record number of lending-bias claims, some are fighting back against what they call a never-ending "witch hunt.
In interviews with IBD, bankers and their lawyers complain that Attorney General Eric Holder and his army of civil-rights prosecutors have built their cases around dubious statistical models instead of complaints from actual victims of discrimination.
The DOJ filed at least 55 race-bias claims vs. lenders in 2009-11, up from just 30 in the prior eight years.
Bank executives say the government's econometric models, which crunch reams of public loan data, don't dig down into race-neutral factors. If prosecutors did, bankers say, they would see that differences in credit risk explain racial differences in lending and pricing, and that any "disparities" reflect prudent business decisions.
They point out, moreover, that most of Holder's racism charges are based on "disparate impact.
The legal theory reduces the standard of proof to mere speculation of intent to discriminate. Basically, all prosecutors have to show is that lending decisions adversely impact minorities in a "statistically significant" way.
The Supreme Court could reject or restrict "disparate impact" in lending if justices ever hear such a case. Holder apparently agrees.
Earlier this year, Justice quietly killed a petition to the high court that threatened to take away one of his biggest weapons vs. banks.
Through July, nearly 20 banks — including the nation's largest mortgage lender and largest commercial bank — have settled DOJ charges of lending bias while denying any wrongdoing.
Holder's increasingly aggressive prosecutors have managed to wring more than $550 million in payouts from bank defendants, not including the costs, in some cases, of having to open branches in depressed urban areas.
Lawyers say regulators have pressured their clients to settle to avoid having acquisitions and other expansion plans blocked during litigation.
Banks also cite mounting legal costs and bad press as reasons for cutting deals.
"Banks are very unpopular right now," said Washington lawyer Andrew Sandler of BuckleySandler LLP. "It's only going to end," he said of the federal lawsuit frenzy, "when we get some of these cases in court and beat them.
With another 30 lenders under active federal investigation and the size of payouts growing, some are fighting back.
New York-based GFI Mortgage Bankers is contesting charges it overcharged black and Latino borrowers based on their race. Sandler has filed a motion to dismiss the case.
Los Angeles-based Nara Bank is fighting a Justice lawsuit claiming it overcharged blacks and Latinos for car loans.
A district judge dismissed the suit after agreeing that DOJ's statistical analysis of loan terms was faulty because it didn't consider other important factors. The case is on appeal.
After Justice informed McLean, Va.-based Cardinal Bank that it was under investigation for failing to serve minority communities equitably, Cardinal fought back in the media. It argued it hadn't broken the law and "would never settle." The bank pointed to the satisfactory 2010 rating that regulators gave it for minority lending under the Community Reinvestment Act. In May, Justice dropped its case.
Stop Suing Us
Large bank defendants so far have not fought back, although Wells Fargo beat a related lawsuit filed by the city of Baltimore.
But the American Bankers Association last month fired off a letter to Holder calling on him to stop suing banks under disparate impact, saying it "has no valid statutory foundation" and is causing undue alarm among banks.
The Independent Community Bankers of America went further. In an earlier letter to Holder, the group called his tactics "an abuse of authority.
Justice denies ignoring creditworthiness and other legitimate business reasons for loan disparities. It argues that in all cases, it compared black and Latino applicants with "similarly situated" whites before drawing conclusions about bias.
Asked to define "similarly situated," Justice referred IBD to court-filed complaints. But none of them fully defined the term.
Prosecutors have devised various screens to trawl mortgage data for racial disparities. But the Treasury Department databases don't include detailed personal credit information that factor prominently in lending decisions.
Missing are borrowers' credit scores, household debt-to-income ratios, loan-to-property-value ratios, down payments, employment history, documented income, assets and other key variables.
"They're taking a shortcut and not getting a true picture" of individual creditworthiness, said Sandler, who has represented several bank defendants.
Last month, Wells Fargo asserted in court documents that its data prove that its subprime borrowers had "significantly weaker credit characteristics" than its prime borrowers. It also noted that "an appropriate analysis" shows "no disparate impact in product placement against African-American or Hispanic borrowers.
A Virginia banker sued last year by Justice for predatory lending complaints that prosecutors didn't control for down-payment amounts, household debt and other key variables. Such factors explained minor markups in the 200 loans analyzed, says C&F Mortgage President Larry Dillon. Yet prosecutors refused to consider them, he said.
"They would not let us use loan-to-value ratios or debt-to-income ratios," Dillon said. "If they had, there would be little, if any, disparity.
He says this is why Justice has insisted bankers sign "nondisclosure agreements" barring them from talking about the methods that prosecutors use to formulate charges.
Dillon called the multiyear investigation of his relatively small bank a "witch hunt.
"That our federal government can behave in this manner is very scary," the CEO told IBD. "I don't know how they sleep at night.
Sandler says the Justice Department is guilty of "bucket analysis." "They take different groups of borrowers and throw their loans into different buckets," he said. "And they come up with an average price without meaningfully seeking to hold constant any differences in the risk profiles of the folks in each of the buckets.
Sandler added, "Using those kinds of superficial statistics, they can make a case against virtually any lender in the country.
Dillon ultimately settled with Justice in a deal that requires C&F to pay $140,000 to alleged victims and, among other things, hang posters in all its branches stating: "It is illegal to discriminate in any credit transaction because of race (or) because income is from public assistance.
Bank officials complain the department is more interested in engineering social outcomes than enforcing the law.
They note that civil-rights division chief Thomas Perez, who has likened bank defendants to cross-burning Klansmen, has said in recent speeches he's using banks to "repair" and "rebuild" entire "min ority communities" devastated by the recession.
Perez's top prosecutor, a former inner-city activist, shares his agenda.
As special counsel for fair lending, a new position created by President Obama, Eric Halperin directs the econometric research fueling race-bias suits. He previously headed the Washington office of the liberal Center for Responsible Lending, which produced similar research.
Before the housing crisis, his center harassed Fannie Mae and primary lenders to ease underwriting rules for minority borrowers. Halperin's old boss, Martin Eakes, who co-founded the center, has called minority homeownership key to "economic justice.
"For us, becoming involved in homeownership was the single way of trying to undo the legacy of slavery," Eakes told PBS before the crisis.
It turns out Justice has modeled its statistical analyses after those developed by the center.
In defending his models, Perez recently cited a years-old study published by the center that compared, in the center's words, "similarly situated" homebuyers by race. He noted it found the same kind of disparities in loan pricing.
However, the 2006 study was widely criticized for failing to adjust data to account for debt-to-income ratios and other reliable risk factors lenders used to price loans.
A more comprehensive Federal Reserve study at the time found no racial bias in loan pricing after controlling for all risk factors.
Still, the flurry of stat-based lawsuits has sent a chill throughout the industry. "I talk to lenders everyday who are so fearful of being accused of discrimination based on statistics that it results in them being much more limited in the kinds of lending they're prepared to do," Sandler said.
There is also concern that anti-redlining suits will pressure banks to further degrade underwriting standards and make more of the kinds of unsafe loans that triggered the foreclosure crisis.
To be sure, disparate impact theory and its statistical applications are effective weapons in the hands of the current crop of civil-rights advocates gunning for banks at Justice.
Under the theory, Justice doesn't have to prove overt lending discrimination. It just has to produce stats showing lending decisions disfavor a sample of minorities. It's up to defendants to demonstrate decisions were "justified by business necessity." Even then, prosecutors can come back and argue they could have made "less discriminatory" decisions.
However, courts increasingly frown on discrimination cases built on statistics. By those methods, whites theoretically could claim disparate impact because their prime loan approval rates are typically lower than those of Asians. And legal experts agree that if the theory ever gets before the Supreme Court in a lending-bias case, justices would strike it down.
In November, Perez assured bank officials gathered in Baltimore that disparate impact was a "legitimate tool" for enforcing fair lending. He also claimed it has been "widely accepted by the courts" in related cases.
At the same time, however, Perez was working behind the scenes to derail a Supreme Court challenge to the theory. The high bench had agreed to hear the landmark case that month.
But in early February, just as justices were set to hear the case, the petitioner abruptly withdrew it. Perez has admitted personally calling the Minnesota plaintiff and convincing him to drop the case out of fear the government could lose a potent tool for fighting "predatory lending" and "economic injustice.
Congress is investigating the interference from Justice, which wasn't a party to the case.
Not long after that potential legal obstacle was removed, the administration announced that HUD and the new Consumer Financial Protection Bureau will also now enforce fair lending, housing and other regulations based on disparate-impact theory.