A couple months ago I was complaining about the lack of inspiration for my Musings, suggesting that CFPB is basically implementing (albeit with questionable durability), my preferred deregulatory guidance, leaving me uninspired. Well, thanks to current world events, a new movie on Homer’s The Odyssey, the CFPB and the Trump Administration , I’m offering this cynical take (highlighting my penchant for metaphorical flourish)[1] about the latest outrage from our government complicating mortgage lending. As the LLMs like to say, I hope this “lands”.
The EO and CFPB Statement
On May 19, the President signed Executive Order 14406, “Restoring Integrity to America’s Financial System,” (the “EO”) which, among other things, directed the CFPB to “consider clarifying” that potential deportation and the associated loss of wages are factors bearing on a borrower’s ability to repay (“ATR”).[2] The CFPB jumped right on it, delivering its (effective upon publication) Statement on Ability To Repay and Immigration Status in the Federal Register, (the “Statement”) just 2½ weeks later.
“Big whoop” you might say: isn’t the possibility someone might get deported obviously relevant to their ability to repay a loan? Of course it is, and suggesting otherwise is only likely to lead to bad borrower outcomes, but whether immigration status is or isn’t relevant to ATR isn’t the issue that’s got me riled up enough to write this Musing.
Regulatory whiplash
First, it’s important to place this ATR immigration status underwriting issue in context. In October 2023, the (Rohit Chopra led) CFPB and (Joe Biden led) Justice Department jointly warned creditors in a statement published in the Federal Register that unnecessary or overbroad reliance on immigration status in credit decisioning “may run afoul”[3] of ECOA’s antidiscrimination provisions. In other words Biden’s fair lending cops said, “Watch out lenders, we might bring a fair lending action against you if you deny loans based on risk of deportation.” Keep in mind that was at a time when fair lending enforcers religiously championed disparate impact in alleging evidence of discrimination. I didn’t squawk about that statement then, frankly, because it was already fully inside the Venn diagram of where fair lending enforcement was at that time.
In January of this year,[4] however, the CFPB officially withdrew that Chopra/Biden DOJ statement. Then, in April this year, the CFPB finalized its overhaul of ECOA’s Regulation B, removing disparate impact liability from that fair lending regulation effective July 21, 2026.[5] Whatever you think of that Reg. B rule (my thoughts were expressed in my last Musing), it essentially eliminates the fair lending disparate impact risk from underwriting based on immigration status under ECOA. As I see it, that Reg B change should have been the CFPB’s “quit while you’re ahead moment,” but, alas, our polarized politics over the past 20 years or so no longer allows us to have nice things. To summarize, we went from a fair lending enforcement message of “consider immigration status at your peril” to “fail to consider at your peril,” in three years with a brief stop at “do what you think is right”.
That’s what I mean by regulatory whiplash. While the two positions can be squared on paper as permitting consideration of immigration status tied to repayment risk, risk management/compliance officers only hear the clear message that the enforcement tenor has flipped, and they need to respond to that.
CFPB’s Statement analyzed
The CFPB’s Statement is premised on TILA’s existing ATR framework, requiring creditors to make a “reasonable and good faith determination” of repayment ability based on, among other things, current or reasonably expected income and current employment status. In fact, old TILA official commentary says that if information in the application or the creditor’s records indicates a post-consummation change in repayment ability is anticipated, the creditor must consider it.[6]
Meanwhile, Reg B has always permitted creditors to consider immigration status to the extent necessary to ascertain their rights and remedies regarding repayment. If the Statement said only that “income continuity matters,” and “a work authorization expiring before the first payment date is information in the file”, it would be unremarkable (and wholly uninspiring to write about) because lenders have been underwriting visa holders’ work authorizations for decades.
But the CFPB Statement did two things that got me going. First, it says a creditor “may regard” an applicant who is neither lawfully present nor work-authorized as subject to removal: not because of anything in the borrower’s application, but “in light of the Administration’s stated policy” of removing all persons unlawfully present, (citing another Executive Order).[7] Again, TILA, however, just asks lenders to determine what can be reasonably anticipated from the consumer’s application or records.
Rather than focusing just on the file, the Statement permissively enables lenders to consider what can reasonably be anticipated from the President’s enforcement priorities in light of immigration status. Although permissive, how’s an examiner (or court assessing ATR) going to look at a lender who chooses not to assess that risk? Lenders are experts at interpreting a loan file, but assessing deportation risk of a particular borrower under shifting enforcement priorities is above most of their pay grades.[8]
Second, consistent with the EO’s direction about risk factors, the CFPB Statement identifies reliance on an Individual Taxpayer Identification Number (ITIN) as a risk indicator that a borrower “may not be lawfully present.” At a minimum, that seems to require a potentially costly enhanced due diligence for BSA/AML[9] purposes for lenders seeking to comply with the Statement. But, set aside that many ITIN holders are lawfully present. ITIN lending is a relatively small mostly non-QM market that has operated for years on precisely the kind of individualized underwriting the Statement claims to want. But the Statement fails to tell lenders exactly how to risk assess deportation of an ITIN borrower that is not lawfully present in its ATR assessment.
Navigating the Strait of Homer(z)
Two things came together to inspire my metaphor for the pickle lenders face. First, the word “strait” has been in the news a lot, and second, I’ve seen a few ads for a new Christopher Nolan movie about The Odyssey coming in July starring Matt Damon. For those of you who missed out on the joy of reading Great Books like Homer’s The Odyssey[10] in high school,[11] On his epic journey, Odysseus, the Greek hero, must sail the strait between Scylla, the six-headed monster who picked off sailors one at a time, and Charybdis, the whirlpool that swallowed ships whole. Mortgage lenders are (metaphorically) in the same Homeric strait right now with this immigration status issue, and, like Odysseus you have to figure out how to navigate “damned if you do/damned if you don’t” hazards.
Charybdis, the unsurvivable whirlpool in this metaphor is, of course, fair lending; specifically, the Fair Housing Act. The CFPB’s recent Reg. B rule removed disparate impact from ECOA, but the FHA version survives with Inclusive Communities[12] and its robust causality charade. But a lender that reads the Statement to require adoption of a policy to weigh deportation risk in underwriting has constructed a (facially neutral) practice that correlates almost perfectly with national origin. That would make robust causality for national origin discrimination easy to show by private plaintiffs and state attorneys general alike notwithstanding any Reg. B changes. Moreover, a categorical deportation-risk policy invites disparate treatment claims on national origin grounds since immigration status functions as a proxy for national origin. Disparate treatment exposure survives even if Inclusive Communities is overruled tomorrow.
On the other side of this strait and metaphor, Scylla, the survivable risk, is the TILA/ATR exposure. TILA lets a borrower raise an ATR violation defensively in foreclosure without regard to the statute of limitations (life of loan) so this defense never expires. Now imagine the future lawsuit: an ITIN borrower in a non-QM loan defaults, and the foreclosure answer arrives with the CFPB’s Statement attached as Exhibit A.[13] “The government itself said my deportation risk bore on my ability to repay and the lender didn’t account for it, so the ATR determination wasn’t reasonable”.[14] So the borrowers this policy was essentially designed to price out of the credit market are now armed to avoid paying back the lenders who serve them.[15]
ATR cases are by their nature, unique. A lender that performed and documented an individualized income analysis should have a genuine reasonable/good-faith record to defend the claim. QM loans, of course, carry their safe harbor (another sea reference?) with them, so an ATR case is likely to only impact non-QM ITIN paper on a loan-by-loan basis.
Ultimately, Homer’s navigational advice[16] makes sense: hug the cliff (where Scylla resides) by engaging in individualized ATR analysis and refrain from categorical status-based policies (keeping as much distance from the fair lending whirlpool as the strait allows). Scylla might take six sailors, but won’t get the whole ship. Although costly to implement, individualized, file-based review also happens to be the approach to underwriting which maximizes accuracy (i.e., lower defaults), unlike status or identity-based policies that sacrifice risk assessment to politics leading to more bad consumer outcomes as suggested by the University of Chicago empirical data I referenced in my last Musing.
Lighten up, Levy
The fall of Chevron deference was hailed by many in the Trump Administration (and me too) as key to a deregulatory agenda. After Loper Bright, an agency’s interpretation receives only the weight its persuasiveness earns, and under Skidmore, that persuasiveness depends substantially on consistency and expertise.
So, considering the impact of Loper Bright and Skidmore on these flip-flopping and obviously politically motivated regulatory pronouncements, maybe I should just get a grip. Perhaps both the Biden DOJ/CFPB 2023 statement and the CFPB’s Statement were just performative signaling with little legal impact and I shouldn’t get so worked up about them.[17] Moreover, the Statement’s “Regulatory Matters” section expressly says it “does not have the force or effect of law” and it “has no legally binding effect”.[18] Likewise, the Biden DOJ/CFPB October 2023 statement had a similar disclaimer.[19]
Unfortunately, I don’t think you can wave these things away by just saying executive branch policy statements carry about the same force of law as a press release. Their impact is still significant for enforcement, and lenders and their compliance leaders are going to respond to that signal regardless of what a court may ultimately have to say about it.
[1] This PlainsCapital capital markets guy seems to have a virtually insatiable penchant for metaphor in his digestible, timely and fun to read interest rate updates.
[2] Loyal readers will recall that, while embarrassing it is needed, the law to require mortgage lenders to make sure borrowers can repay their loans is one part of the Dodd Frank Act I still appreciate (unlike the LO Comp Rule).
[3] “May run afoul” reminds me of this movie I watched with my kids when they were little.
[4] With the Trump administration now in charge (if you are keeping score).
[5] The CFPB’s Reg. B rule has been challenged in NFHA v. CFPB (D.D.C., filed May 27, 2026).
[6]TILA Comment 43(c)(1)-2 “if the application or records considered at or before consummation indicate there will be a change in a consumer’s repayment ability after consummation (for example, if a consumer’s application states that the consumer plans to retire within 12 months without obtaining new employment or that the consumer will transition from full-time to part-time employment), the creditor must consider that information under the rule.”
[7]91 Fed. Reg. at 34608, citing Exec. Order No. 14159, “Protecting the American People Against Invasion,” 90 Fed. Reg. 8443 (Jan. 29, 2025).
[8] I understand that the MBA is considering providing lenders with some guidance about the implications. I’m not sure where they will get immigration law expertise either, but I can refer them to good counsel.
[9] Bank Secrecy Act/Anti-Money Laundering: if you know, you know.
[10] If you didn’t learn about metaphors from reading the Odyssey, you didn’t read it.
[11] I mean, why read anything now when ChatGPT can summarize it in paragraph. Personally, I found the Cliff’s Notes a lot to read too.
[12]Texas Dep’t of Housing & Community Affairs v. Inclusive Communities Project, Inc., 576 U.S. 519 (2015).
[13] From the Statement, “failure to account for such a reasonably expected change in income may not comply with a creditor’s obligation to reasonably assess a borrower’s ability to repay.”
[14]A defense lawyer will have responses, but every contested ITIN foreclosure just got a whole lot messier.
[15] Nothing seems to inspire my writing more than a “no good deed goes unpunished” bureaucratic solution that is worse than the problem.
[16] Reminder, this publication is not legal advice.
[18]Notwithstanding the disclaimers, OMB’s Office of Information and Regulatory Affairs designated the Statement a “significant regulatory action” under EO 12866, and CFPB submitted it to Congress under the Congressional Review Act “in an abundance of caution”. More whiplash. Please make up your mind.
[19]88 Fed. Reg. at 71845 n.1 (“This document is for informational purposes only. It does not impose any legal requirements . . . and it is not enforceable.”).