Regulatory arbitrage of the day, CRA edition

By Felix Salmon
August 13, 2010
National People's Action have a fascinating report out today about America's big four banks -- Citi, JPM, Wells Fargo, and Bank of America -- and how they all seem to be able to easily obtain "outstanding" ratings on their CRA exams.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

National People’s Action have a fascinating report out today about America’s big four banks — Citi, JPM, Wells Fargo, and Bank of America — and how they all seem to be able to easily obtain “outstanding” ratings on their CRA exams.

The CRA, of course, is the Community Reinvestment Act, and it exists to ensure that America’s largest banks are doing a good job of providing the same (and not higher-priced) products in poor areas as they do in rich ones. Regulators have been examining fewer banks of late under the CFR, but one thing remains constant: the number of “outstanding” ratings is always very small as a percentage of the whole. And yet all four of the big banks always seem to be able to get that rating. How come?

It turns out they’re using two tricks, neither of which is available to most smaller banks. First, they do most of their lending to poor people outside what’s known as their key “full-scope assessment areas”, on which they’re mainly judged. Taking the four big banks as a whole, just 19.2% of their high-cost loans to low and middle-income borrowers take place in these assessment areas.

wells.tiffAnd secondly, they use subsidiaries and affiliates to do their high-cost lending to poorer Americans, which aren’t included in the CRA exam. These subsidiaries account for just 17.1% of the loan volume for the big four banks, but 45.5% of the high-cost loans.

In other words, if you get a mortgage from Citimortgage or Citifinancial rather than from Citibank, you’re not going to get noticed in Citi’s CRA exam. And at Wells Fargo, the list of affiliate mortgage lenders goes on for the best part of three pages. A snippet, just to give you an idea, is at right.

Add it all up, and it’s pretty obvious that the way that the CRA is administered has signally failed to keep pace with the way that banks lend. As the report says:

The intention of the Act was to cover the mortgage lending industry. In the mid-1970s that meant depository banks originating mortgages from a network of branches. As a result, the CRA exam conducted by a banking regulator grades only that lending that takes place in a bank’s predetermined “Assessment Areas” that are based on where the bank has physical branches.

People don’t get their mortgage from their local bank branch any more, and it’s silly that the CRA is still predicated on the idea that they do. If the CRA is to have any meaning going forwards, it has to assess the actual lending that these banks do, rather than a tiny subset thereof. Let’s hope the Consumer Financial Protection Bureau is significantly savvier than what’s on show in this report.

Update: AABender1, in the comments, says that the CRA is not, as the report implies, the main tool by which the US government tries to prevent discriminatory lending, redlining, and the like. It’s a good point, which I’m a bit annoyed that I missed. But insofar as a CRA “outstanding” rating has any value at all, it should probably take such things into account.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Actually, this is a fairly unintelligible and specious report.

First it completely misses the point of CRA. CRA was enacted to ensure that (i) deposits taken in one community were lent back into that same community, and (ii) those deposits were lent back at least proportionately to the low/mod income areas and/or people in that community.

The point was to prohibit banks from opening branches in one area and shuttling off those savings to be invested in another.

Second, CRA’s stated purpose is not to examine safety and soundness, predatory practices, or “high cost” lending. It looks at where monies are deposited (assessment area) and where monies are lent (lending and investment tests). Period.

Third, it is Fair Lending compliance (along with other consumer compliance regs like Reg B) and those compliance exams that review what the National People’s Action report discusses. That is why “high cost loan” data is included in a bank’s HMDA reporting. And fair lending and consumer compliance exams can include affiliate loans under certain circumstances. Moreover, the Home Equity Protection Act (HOEPA) of 1994 mandated that the Fed could enforce housing-related consumer protection standards to non-bank and affiliates alike. (They didn’t.)

As a result of HOEPA non-enforcement, more than half of the sub-prime, predatory lending happened outside the regulated banking and banking affiliate system.

So, they pick the wrong target, confuse the policy weapon of choice, and end up with some very wrong-headed prescriptions.

Posted by AABender1 | Report as abusive

This looks like a list of correspondent lenders to me. Calling WF the ‘parent company’ of these is misleading at best. Granted, they probably sell 90% of their production to Wells, but still. It’s like calling McDonalds the ‘parent company’ of Ripon Pickle Co., Inc.

Posted by Shnaps | Report as abusive

I wonder about the extent to which this is an unintended effect of CRA itself. At the time it was written, the assumption was that the core retail activities of a bank would take place through its branches. So it imposed standards based on the locations of those branches – to ensure that local deposits would be used to fund local lending.

When big national banks began to pursue subprime lending as a potential profit center, they might have opened branches in low- and moderate-income neighborhoods. But that would have extended their full-scope assessment areas, making it harder to hit CRA targets. Even worse, it would have undermined the whole point of the expansion – to muscle in on the highly-lucrative market for subprime products. They quite simply could not have originated a sufficient supply of subprime loans to meet demand while also complying with CRA. So they worked instead through affiliates and subsidiaries, which, unfettered by CRA, were free to steer a great many of their customers who had the credit to qualify for prime products into subprime loans. Problem solved.

Of course, this has a variety of perverse effects. For one thing, it provided a strong disincentive for banks to extend their services to the unbanked via the construction of new branches. For another, it placed much of the lending out of the regulatory spotlight, if not out of the regulatory purview entirely. And it set up a separate-but-not-equal lending architecture, in which the same financial behemoths had two parallel lending systems: one for residents of fairly homogenous and relatively affluent communities, and another for more heterogeneous areas, including their affluent and successful residents.

It’s neither an argument for more or less regulation. I think it’s an argument for more unified financial regulation. Crafting specific regulatory acts and agencies for specific functions has its benefits, but it never keeps pace with change. Having regulators with broad purview, who can rapidly retask to account for unfolding change, makes a good deal more sense. Now if only we had a reliable way to actually make them do that…

Posted by Cynic | Report as abusive