Why IMBs Don’t Need A CRA Cramdown

Expanding Community Reinvestment Act to IMBs looks good on paper, but is it?

Ryan Kingsley
Why IMBs Don’t Need A CRA Cramdown

You’ve seen it in your neighborhood. The marquee grocery store has closed. With regional variations, a Dollar General, Dollar Tree, or Family Dollar opens in its place.

The message Stop & Shop or Publix send by closing is clear: profitability exists elsewhere.

This scenario has been mirrored in the mortgage lending landscape since the financial crisis of 2007-2008. On a national scale, banks have strategically receded from residential lending and Main Street. In their stead, independent mortgage banks (IMBs) have emerged, broadening the homeownership opportunities for lower and moderate-income (LMI) individuals and minorities.

In April, the Housing Finance Policy Center of the Urban Institute, a Washington, D.C.–based think tank that conducts economic and social policy research, released, “An Assessment of Lending to LMI and Minority Neighborhoods and Borrowers.” The report showed how in the context of a widening homeownership gap and proposed reforms to the Community Reinvestment Act (CRA), IMBs are dramatically outperforming banks in residential lending to LMI and minority borrowers.

All of which, surprisingly, has even the Urban Institute questioning the groundswell movement to subject independent mortgage banks to the strictures of the Community Reinvestment Act.

Reputational Risk

For the years the Urban Institute examined, “IMBs accounted for approximately 60% of all mortgage originations, including 75% of originations backed by Fannie Mae, Freddie Mac, the Federal Housing Administration, and other government agencies.” In fact, 35.3% of all IMB lending was found to be through government channels, versus 17.8% for banks and 11.1% for credit unions.

Laurie Goodman, an Urban Institute fellow, founder of the Housing Finance Policy Center, and one of the assessment’s authors, says this performance gap stems in part from banks originating a lower volume of Federal Housing Administration (FHA) loans, which disproportionately go to LMI neighborhoods and borrowers.

Furthermore, while banks view mortgages as an auxiliary business that complements their front-and-center retail offerings, IMBs are wholly in the mortgage banking business, she says. “I’m not sure nonbanks should be subject to CRA because they’re already doing more, they’re in one line of business. It seems like it’s a solution in search of a problem.”

Banks, she added, aren’t looking for the same challenges IMBs are. “They basically make loans to anyone that fits within the box. By contrast, the banks are in a lot of different businesses. … The result of that is if there’s one hint of reputational risk, the banks run,” she says.

Reputational risk because, Goodman points out, the banks aren’t really bearing the credit risk for government-insured loans – the government is.

The nonbank (IMBs and credit unions) origination share for all agency loans was 83% as of March 2023, per the Urban Institute’s monthly chartbook for April. The nonbank origination share for Ginnie Mae, Fannie Mae, and Freddie Mac stood at 93%, 78%, and 77%, respectively.   

To put those numbers in the context of banks’ withdrawal from mortgages, Ginne Mae issuances by nonbanks skyrocketed from 12% to 90% from 2010 to 2021, per Ginnie Mae data.

The assessment is limited to owner-occupied, residential purchase loans, and examines five-year American Community Survey data from 2015-2019, alongside 2021 Home Mortgage Disclosure Act (HMDA) data.

The Mortgage Bankers Association (MBA) commissioned the report to compare lending patterns of entities subject to CRA requirements, namely banks and thrifts, with those of lenders not subject to the statute, IMBs and credit unions.

– Laurie Goodman,

an Urban Institute fellow

No Appetite

The reasons banks have retreated from mortgage lending, especially FHA programs, mostly stem from the 2007-2008 financial crisis, says Clifford Rossi, director of the Smith Enterprise Risk Consortium at the University of Maryland and a former banking industry executive. Many banks no longer had an appetite for volatile assets such as mortgage servicing rights. Banks were also getting burned on the origination side through non-traditional mortgages, such as alt-A and subprime loans.

“Another big reason the banks got out,” says Rossi, “particularly around FHA lending, but even more broadly, in the years following the financial crisis there was an awful lot of uproar by banks in terms of the lack of transparency of repurchase demands that were being made by the GSEs [Government Sponsored Enterprises], private mortgage insurance companies, and also by the FHA.”

Mortgage repurchases occur when buyers of mortgage-backed securities, such as Fannie Mae or Freddie Mac, determine there are defects in how a loan was made, leading them to demand a repurchase by the lender.

“We know that the vast majority of loans to disadvantaged borrowers are done through the FHA loan program through the Department of HUD. We know that factually, we know it mathematically, and we know it statistically,” says Taylor Stork, chief operating officer of Developer’s Mortgage Co. and president of the Community Home Lenders of America, a national association of small and mid-sized community-based mortgage lenders.” We also know that 90% of those loans are made by IMBs. IMBs are by far and away the first lender and the lender of choice for the very borrowers that a new CRA (Community Reinvestment Act) program is intended to support in the first place.”

Clifford Rossi,

director of the Smith Enterprise Risk Consortium at the University of Maryland

In the years following the financial crisis, lenders accused of improperly certifying mortgage loans as eligible for FHA insurance, for example, could be held liable for making false claims to the government. Under the False Claims Act, litigation was brought by the Department of Justice (DOJ) against some of the country’s largest lenders, including Bank of America, JPMorgan Chase, and Citigroup. DOJ and Citigroup reached a $7 billion settlement on mortgage fraud claims that the bank backed bad mortgages during the lead-up to the financial crisis. Bank of America and JPMorgan Chase settled similar lawsuits for $16.65 billion and $13 billion, respectively.

After such hefty penalties were levied, Rossi noticed a prevailing attitude among banks – that there was a lack of consistency how FHA and the agencies audit for defects in the way banks underwrite or value loans from a collateral standpoint. In response, banks pulled back on FHA lending to avoid costly fines and unpredictable repurchase demands.

Rising FHA insurance premiums and the jacked-up cost of servicing delinquent FHA loans also have eaten away at the profitability of government-backed loans for banks. Rossi acknowledges February’s .30% reduction in FHA insurance premiums, “but historically, FHA premiums had been increasing, and so that caused the shift away from FHA in terms of the profitability the banks could get from these loans as well.”

Goodman echoes Rossi’s assessment, saying that “banks feel like a lot of them received fines as a result of bad lending during the financial crisis, and that the rules of engagement aren’t as clear as they could be, particularly for servicing loans.”

The Cat Flap Of Exclusive Overlays

The Urban Institute report underscores the reality that IMBs use the widest possible credit box in their lending decisions. IMBs are originating for borrowers with higher debt-to-income ratios (DTIs) and lower FICO credit scores while remaining within FHA and agency lending guidelines.

“Another way to think about it is the GSEs and Ginnie Mae each have a credit box … these loans with this debt-to-income ratio, this loan-to-value ratio, this credit score fits within my credit box. What the banks often do is put overlays on top of that, so they’re not lending to the full extent of the GSE or FHA credit box. They don’t want to deal with lower credit score borrowers, for instance, in many cases,” Goodman says.

– Taylor Stork, president,

Community Home Lenders of America

Stork says support and structure need to be created for borrowers instead of for loans.

“I believe that we need to focus on the people instead of the package,” he says. “When we talk about lending to people, we’re talking about lending in a community and helping build the American Dream. When we look at loans, we’re talking about assets. Assets can be problematic on a bank’s balance sheet when they have basis risk and the assets don’t match the deposits and things like Silicon Valley Bank happen.”

But, the risk that lower-FICO and higher-DTI borrowers represent bears out in the real lending environment, says Rossi, who sees what Goodman calls “reputational risk” as the market-facing measure of an institution’s executive and operational risk management. “There is no question that the delinquency and default rates of FHA loans are higher because the cohort that’s served by that market are riskier borrow profiles than what you see in either Fannie Mae or Freddie Mac,” Rossi says.

A former chief risk officer at Countrywide and chief credit officer at Washington Mutual, Rossi was managing director and chief risk officer of Citigroup’s lending division when the financial crisis hit – front and center to witness how expanding access to homeownership at the cost of sound lending practices can harm more than help higher-risk borrowers.

“The tighter lending standards, from the banks’ perspective, are prudent lending standards. They’re making a decision based on their risk appetite,” Rossi says.

As of March 2023, there was a nearly 30-point difference between banks’ and nonbanks’ median DTIs for Ginnie Mae originations, and a 20-point difference between banks’ and nonbanks’ Fannie Mae and Freddie Mac originations. According to the Urban Institute’s monthly chartbook for April, the median FICO credit score for nonbanks’ Ginnie Mae originations was 676, and 754 for Fannie and Freddie. Meanwhile, banks’ median credit score for Ginnie Mae originations was nearly 20 points higher, 698, and 756 for Fannie and Freddie.

‘Playing Within The Confines’

The extent to which widening the credit box to expand homeownership is worth the increased risk of default and delinquency – a risk most thinly capitalized IMBs have too little liquidity to offset should repurchase demands mount – is up for regulators to decide. But Rossi, who built and studied a statistical default model using historical loan level data from Fannie and Freddie for the years 2000-2010, thinks the historical repurchase rate is a strong indicator of how much risk GSEs are taking on.   

In a 2017 study he conducted, “Managing Mortgage Product Development Risk,” Rossi found that for mortgages with loan-to-value (LTV) ratios greater than 80%, the default risk of high repurchase rate originators (a proxy for poor loan manufacturing processes) was 1.4 times that of other lenders. The findings show how companies with poor loan manufacturing processes experienced greater credit risk on the more questionable loans.

“The IMBs are playing within the confines of the structure,” Rossi acknowledges, “let’s say the agency rules or even the FHA rules that are set out for them. But by widening their credit lending criteria, they’re playing at the higher end of that risk spectrum and not in the kind of middle- to lower-end range.”

Tighter lending requirements as part of a strategic shift away from residential lending has turned banks into second chairs when it comes to expanding access to homeownership. Functionally, banks’ overlays act like cat flaps that let in the cat but keep out the dog, erecting barriers to homeownership that disproportionately impact LMI and minority borrowers.

“Clearly, minority borrowers have lower FICO scores than their white counterparts. There’s actually been a lot of work on that. So, if you’re putting a FICO overlay on it, you’re disproportionately squeezing out non-white borrowers,” Goodman says.

IMBs Doing Their Part

Aconcern for mortgage industry leaders as federal regulators pursue CRA modernization is whether IMBs will be brought under a mandate. Only Illinois, Massachusetts, and New York have adopted state-level CRA frameworks that cover nonbank mortgage lenders, though California, Maryland, and Pennsylvania have considered adopting similar versions.

The MBA, for its part, believes the Urban Institute’s research serves as a caution to federal and state policymakers who think shrinking the IMB market share is a good policy objective on its own and provides more solid evidence that IMBs already lead the mortgage market in sustainable lending to LMI borrowers and communities. The MBA said in a statement that proposals that apply CRA mandates to IMBs are “ineffective and misguided, as IMBs do not have deposits to reinvest and do not have access to direct government support.”

A CRA-like mandate for IMBs could also conflict with Iicensing requirements, creating an undue burden for businesses and borrowers. While banks face heavier federal regulations, IMBs are regulated on a state-by-state basis. Every state has its own specific requirements, such as distance regulations. This means, Stork says, “whether it’s Illinois or New York or somewhere else, IMBs tend to operate in the communities where they serve the borrowers. We do loans where we live and where we work. It’s just the nature of the model.”

In New York, for example, the physical location of an IMB’s branch office must be licensed, approved by the state, and staffed by a licensed loan officer. That ensures that a loan officer isn’t making loans in Albany out of a branch office 150 miles away in Queens or Oceanside, NY.

Besides, he explains, “When you add new legislation or regulatory controls on top of that – particularly when you add regulatory controls to an industry that is already outpacing the controls that are meant to manage them in the first place – when you add those regulatory controls what you’re actually doing is making it more difficult for the IMBs to operate.

Adding operational complexity with layers of redundant regulation only increases origination costs that are passed on to borrowers. Due in part to a drop in volume, in the first quarter of 2023 the cost to originate a mortgage climbed to $13,171, according to MBA industry analysis, up from $10,637 in the first quarter of 2022.

But, IMBs’ business model isn’t just built on a profit motive. Rather, what Taylor calls an “existence motive” drives IMBs to help any borrower they can, even the higher risk ones.

A Wider Credit Box

Partially explaining their deployment of a wider credit box than banks, IMBs don’t get to be picky with the loans they close because, in Stork’s eyes, “all loans are precious. All loans are the things that ensure that the income statement is healthy and the IMB is able to continue to operate. When you unwrap both the regulatory requirements as well as the traditional business requirements, IMBs are functionally required to be profitable.”

To the Urban Institute’s Goodman, the more central question is how to measure whether any institution is doing what they should be doing in this market. One benchmark is market performance on the whole.

“But again,” Goodman explains, “nonbanks are most of the market, and they’re doing much more than banks. The other benchmark, which we used for the report, was the share of homeowners in the community.”

That benchmark has its own issues because of what Goodman says is an obvious age bias in these numbers.

“You might expect there should be more Hispanic borrowers approved than there are homeowners in that community just because the Hispanic community is younger, so they’re just aging into the age group where they buy mortgages,” she says.


CRA Modernization: A Bank Challenge

The Urban Institute’s findings underscore the extent to which IMBs are positioned to continue expanding access to mortgage credit at a time when federal regulators are in the process of modernizing the Community Reinvestment Act, and banks’ abilities to meet the credit needs of their communities stretch far beyond mortgage lending.

The CRA, signed into law in 1977, provides a framework for federal regulators to address discriminatory lending practices. Toothless by design, the CRA has never contained blanket criteria for evaluating the performance of financial institutions, nor does it outline specific penalties for non-compliance. Rather, banks’ efforts to meet the credit needs of their communities have always been examined on a bank-by-bank, community-by-community basis. Compliance or noncompliance is only taken into account by regulating agencies in the case of proposed mergers, acquisitions, and branching.

Federal regulators are in the process of modernizing the CRA; for example, tailoring evaluations and data collection to bank size and type. While delivering remarks at National Community Reinvestment Coalition’s Just Economy Conference in 2021, Federal Reserve Chairman Jerome Powell voiced support for a modernized CRA that includes non-depository institutions, such as IMBs. “Like activities should have like regulation,” he said.

Jason Keller, associate director over Fair and Responsible Banking and the Community Reinvestment Act for Wolters Kluwer, an information services provider, supports CRA modernization, saying the process is about establishing new criteria for community development activities and establishing a new confirmation process allowing institutions to seek credit approval in advance of CRA-related activities.

Keller says the proposed changes make CRA more proactive than its current reactive focus. “CRA is a retroactive regulation. It looks backward; it doesn’t look forward. It doesn’t give you credit for things that haven’t happened yet, per se. It only gives you credit for what actually has happened. So, this notion of a modernized CRA is a game changer in that institutions will be able to work with their regulator to understand what they may be given credit for in advance, and that’s where I think a modernized CRA is going to change the landscape of this country going forward.”

When the CRA was implemented, most mortgage credit was provided by banks and thrifts, all of which operated out of branches, many with limited geographic reach.

“Mortgages are a relatively small line of bank business [now],” says Laurie Goodman, founder of the Urban Institute’s Housing Finance Policy Center. “The CRA covers much more than that. Historically, the metrics used for CRA have been very soft and not well defined. They’re trying to change that at the federal level.”

Pledges numerous banks have made to increase lending to marginalized borrowers seem at odds with their withdrawal from FHA lending. Since the financial crisis, even the country’s largest banks have largely withdrawn from the very business which the CRA was designed, in part, to help them conduct more fairly. Published reports show that in 2007, Wells Fargo, Bank of America and JPMorgan originated 19% of all U.S. mortgages. By 2021, that share had shrunk to 4%.

But for banks to do more may be even trickier as their retreat from the mortgage industry has occurred in tandem with what the National Community Reinvestment Coalition (NCRC) terms “The Great Consolidation” – the dramatic merger and decline of banking institutions and physical bank branches.

Closed To Homes

From 2017-2021, a period overlapping that of the Urban Institute’s lending pattern assessment, 7,425 bank branches closed their doors – 9% of all locations across the U.S. One-third of these closures occurred in LMI and/or majority-minority neighborhoods where access to branches, says the NCRC, is crucial to ending inequities in access to financial services, particularly credit.

Some 4,000 of the 7,425 branch closures occurred since March 2020, says NCRC, as the COVID-19 pandemic hastened customers’ adoption of online and mobile banking services, driving the closure rate to double.

Goodman says bank consolidation and branch closures have certainly helped IMBs gain market share and lend to more LMI and minority borrowers, to the extent that people have weaker relationships with their bank. “Unlike most banks and thrifts, which have a geographic footprint and a defined community, most IMBs operate widely through various delivery channels, such as traditional retail branches, wholesale lending through mortgage brokers, and consumer direct through call centers,” the report’s authors write.

Current calls for CRA modernization echo a desire to make banks more agile at meeting the credit needs of their entire communities, given the rise of digital banking. Yet, importantly, this need not include residential lending.

Keller has found that in his 25 years of working with banks, they want to be strong community partners. “They don’t comply with CRA because they have to – they comply with CRA because they want to. … They want to be able to find things that their competitors are not doing.”

This article was originally published in the Mortgage Banker Magazine August 2023 issue.
Ryan Kingsley
Published on
Aug 14, 2023
Mortgage Banker Magazine
If You Build It, They Will Come

Ratings will go a long way toward growing the market for home equity-backed securities

Ryan Kingsley
Mortgage Banker Magazine
Investors May Be Losing Interest In The Single-Family Rental Market

Rise in interest rates, drop in values have them pulling back in certain regions

Xander Snyder
Mortgage Banker Magazine
A Top-Tier Team

Longtime pros launch new company, see early success

Erica Drzewiecki
Mortgage Banker Magazine
Rate Table Roulette

Barrier to entry is steep in this direct-to-consumer advertorial method

Sarah Wolak
Mortgage Banker Magazine
Capital Markets Are Driven By …

GSEs, private institution originations, shifts in demand play leading roles

Rob Chrisman
Mortgage Banker Magazine
Lending A Hand To Appraiser Trainees

How lenders can fix the appraiser shortage

Katie Jensen
Connect with your local mortgage community.

Meet your your colleagues, both national and local, by attending an event in your area.

Become a subscriber.

Discover the story of your success. Subscribe to the nation’s longest running mortgage magazine.