How S.2155 (the Bank Lobbyist Act) Facilitates Discriminatory Lending

If you think it's ridiculous that the CDC can't gather data on gun violence, consider the financial regulatory world's equivalent:  S.2155, formally known as the Economic Growth, Regulatory Relief, and Consumer Protection Act, but better (and properly) known as the Bank Lobbyist Act.  S.2155 is going to facilitate discriminatory lending. Let me say that again.  S.2155 is legislation that will facilitate discriminatory lending. This bill functionally exempts 85% of US banks and credit unions from fair lending laws in the mortgage market.  Support for this bill should be a real mark of shame for its sponsors. 

There's a lot of bad and whacky stuff in S.2155.  Just start with its title. Why would anyone would want to throw more lighter fluid on an already supercharged economy? And isn't it practically daring the Fed to raise rates to offset whatever effect the bill has?  Likewise, why do US banks possibly need any regulatory relief when they are at record profit levels--for both megabanks and community banks?  This ain't an industry that's hurting. 

Turning tot he substance, S.2155 gets it wrong time and time again.  Since 2010, megabanks (that is banks with over $50 billion in assets) have been subject to "enhanced prudential standards"--that is extra safeguards to make sure that they do not fail.  S.2155 raises that the coverage threshold from $50 billion to $250 billion.  Exempt institutions include bit players in the economy such as Deutsche Bank.  Is it really wise to exempt large institutions from the additional post-crisis regulation?  Firms like Countrywide and Washington Mutual were large enough to cause major economic problems, but they would be exempt from additional regulation under S.2155. 

The bad policy judgments in S.2155 aren't just limited to megabanks, however. S.2155 also sets up a lot of community bank failures. One of S.2155's moves is to exempt mortgage loans held in portfolio from the Dodd-Frank Act's "ability to repay" requirement on the theory that if banks have to eat their own cooking they'll be smart about it.  I've got an S&L crisis that says otherwise. Here's what will happen: a number of community banks are going to start making long-term, fixed-rate mortgage loans in geographically concentrated areas. And then some of those areas are going to have local economic downturns right when interest rates start to rise. The result will be a bunch of rapidly decapitalized community banks that will fail. If they were dry cleaners or sandwich shops, this would just be the market at work--businesses that make bad decisions fail.  But when banks fail there are serious collateral consequences.  Sometimes those are global economic crises.  And sometimes it's much more local pain, such as a rural community that is left without a local bank, or any bank at all. The ability to make portfolio loans without complying with the ability-to-repay requirement isn't going to be a make-or-break matter for any community bank's bottom line, but it is going to set up a bunch of them to fail, which will, of course, then be blamed on other regulation.   

But the worst thing in S.2155 is what it would do regarding discriminatory lending.  Discriminatory lending is arguably the very worst activity that occurs in finance.  It's worse than outright theft because it isn't neutral as among victims. So what would S.2155 do? It would impair the key tool for policing discriminatory lending, the Home Mortgage Disclosure Act.  HMDA is a data collection act.  It requires mortgage lenders to report certain information about loan applications and loans funded. That data is the central tool for monitoring for discriminatory lending in the housing market, which is the largest consumer finance market by an order of magnitude, and by far the most important consumer finance market in terms of wealth-building.  

To illustrate, consider a recent study by the Center for Investigative Reporting that found that in 61 metro areas around the country (that's basically all large metro markets) people of color were significantly more likely to be denied a conventional home purchase loan than their white counterparts, even after controlling for income.  In Philadelphia, for example, a black family was 2.7 times more likely to be denied a mortgage than an equivalent white family.  In other words, discriminatory lending appears to be very much live and well.  (Perhaps no surprise given that we've got a President who settled a housing discrimination suit early in his career.)  

Now the American Bankers Association has criticized the CIR study as "defective," noting that it does not control for credit scores or debt-to-income ratios or loan-to-value ratios. That's a fair enough criticism, although it does not disprove the CIR study.  But there's no question that without controlling for factors like credit scores, DTIs, or LTVs, it would be difficult to prove a discriminatory lending case in court.  

Here's the irony: the American Bankers Association supports S.2155. And what would S.2155 do?  It would exempt depositories and credit unions that make less than 500 loans per year—85% of banks and credit unions—from having to report credit scores, DTIs, and LTVs (and certain other data) under HMDA. The bill's sponsors cast this as a reduction in regulatory burdens, but in real terms it means an exemption for 85% of banks and credit unions from the Fair Housing Act and Equal Credit Opportunity Act (at least as applied to mortgage lending). That's shocking. Without credit scores, DTIs, and LTVs, we're unlikely to see any fair lending enforcement in the housing market against small institutions.  

Why on earth should fair lending laws only apply to large institutions? Where do you think discriminatory lending is most likely? At small institutions that rely on face-to-face underwriting determinations and the loan officer's "judgment" or large ones that are doing more algorithmic underwriting? It's the face-to-face underwriting that worries me more, not least because the compliance departments at larger lenders are likely self-testing to ensure that their algorithms are not discriminatory. Without HMDA data, it is near impossible to test for, much less prosecute discriminatory lending by an institution. And that's how S.2155 is the equivalent of telling the CDC not to track gun violence. Take away the data and it's hard to regulation. 

S.2155 isn't just a get-out-of-jail free card for small institutions that engage in discriminatory lending. It also impairs the overall usability of HMDA data. There will be many counties in which there is no HMDA reporting in critical categories whatsoever. That makes it more difficult to use HMDA data even to police discriminatory lending at larger institutions.  Shame on the sponsors of S.2155 for backing legislation that guts fair lending laws.  

[Updated and corrected 1.23.18]

Comments

[Mike S]

Is this right?

It looks to me like Sec. 104 of S.2155 exempts the <500 mortgage institutions from some of the financial data reporting requirements (i.e. 12 USC 2803(b)(5) and (6)), but not the requirement to report the number of mortgages and applications by race and gender (which is found at 12 USC 2803(b)(4)).

Or is there another section of S.2155 I'm missing?

[Adam Levitin]

Hi Mike. I was actually correcting the post as you commented. You're right. It's not a blanket HMDA exemption, but without data on the financial terms of the loans it's difficult to make a Fair Housing Act or ECOA case about discriminatory lending that is going to survive Iqbal/Twombly review.

[td bank hours]

Wow...! Thanks for sharing your idea. I will definitely bookmark your blog and I will come back in the future.

[sms shayari]

Hi Mike. I was actually correcting the post as you commented. You're right. It's not a blanket HMDA exemption, but without data on the financial terms of the loans it's difficult to make a Fair Housing Act or ECOA case about discriminatory lending that is going to survive Iqbal/Twombly review.