Elena Fridman v. NYCB Mortgage Company LLC

EASTERBROOK, Circuit Judge,

dissenting.

Elena Fridman had a mortgage loan from NYCB. Payments were due by the first of each month. On December 14, 2014, or 14 days late, Fridman used NYCB’s web site to request payment from her checking account at Bank of America through Electronic Payment Network, an automated clearing house (ACH). That process usually takes two business days. NYCB told Fridman that her payment would be credited on December 18, two business days hence. (December 14 was a Friday.) Fridman acknowledged this timing, and her payment was posted on December 18. NYCB added a late fee, and in this litigation Fridman maintains that the fee violates 15 U.S.C. § 1639f(a).

Section 1639f(a) provides: “In connection with a consumer credit transaction secured by a consumer’s principal dwelling, no servicer shall fail to credit a payment to the consumer’s loan account as of the date of receipt, except when a delay in crediting does not result in any charge to the consumer or in the reporting of negative information to a consumer reporting agency”. (A “servicer” is the entity responsible for collecting the debt. NYCB handles its own collections and is a “servicer” under the statute.)

NYCB did not receive a “payment” by the end of its 15-day grace period. What happened on December 14 was not “payment” but an electronic instruction directing NYCB to request a transfer from Bank of America (and authorizing Bank of America to remit). Money did not reach NYCB until December 18. On this all agree. Nonetheless, Fridman maintains, the instruction of December 14 should be treated as equivalent to a payment — and, although no statute requires lenders to have grace periods, Fridman wants to combine NYCB’s 15-day forbearance with *781 the statutory requirement that “payment” be credited immediately to produce a conclusion that the late fee is impermissible.

The statute does not define “payment.” A regulation, 12 C.F.R. § 1026.36 (c)(l)(i), tracks the statutory language without adding a definition. My colleagues turn to commentary provided by the staff of the Consumer Financial Protection Bureau. Yet it, too, fails to define “payment.” It does say, however, the “date of receipt” (a term in both the statute and the regulation) is “the date that the payment instrument or other means of payment reaches the mortgage servicer.” 12 C.F.R. Part 1026, Supp. I, pt. 3 § 1026.36(c)(1)® ¶3.

It is not clear to me that we owe this commentary any deference, as opposed to the careful consideration all agencies’ views receive. The Bureau receives leeway when explaining its regulations, see Ford Motor Credit Co. v. Milhollin, 444 U.S. 555 , 100 S.Ct. 790 , 63 L.Ed.2d 22 (1980) (discussing the status of commentary by the Federal Reserve, which formerly administered the Truth in Lending Act), but “date of receipt” is a phrase in the statute. Why should an agency that parrots a statute in a regulation, as the Bureau did, get to make binding rules through “official commentary” that did not go through notice-and-comment rulemaking? See Gonzales v. Oregon, 546 U.S. 243, 257 , 126 S.Ct. 904 , 163 L.Ed.2d 748 (2006) (“the near equivalence of the statute and regulation belies the Government’s argument for ... deference”). Especially when the statute is implemented through litigation rather than administrative adjudication? See Adams Fruit Co. v. Barrett, 494 U.S. 638 , 110 S.Ct. 1384 , 108 L.Ed.2d 585 (1990). Cf. Perez v. Mortgage Bankers Association, No. 13-1041, • — - U.S. -,---n. 4, 135 S.Ct. 1199 , 191 L.Ed.2d 186 (U.S. Mar. 9, 2015), slip op. 10-11 n. 4, 2015 WL 998535 and concurring opinions. But NYCB has not relied on ’ Gonzales or Adams Fruit, and this court is not the right forum to resolve any dispute about the status of Bowles v. Seminole Rock & Sand Co., 325 U.S. 410 , 65 S.Ct. 1215 , 89 L.Ed. 1700 (1945), and its successors (including Ford Motor), so I let this pass. The question remains how a payment instruction should be treated.

An instruction is not a “payment”; NYCB was not paid until December 18. Was it a “payment instrument” such as a check? No; it was not an “instrument” of any kind. The statute, regulation, and commentary all leave “instrument” undefined, and if we turn to the payments articles of the Uniform Commercial Code we do not find any definition equating a payment instruction routed through a clearing house the same as a payment instrument such as a check. Article 4A of the UCC, which covers electronic transfers, speaks of the transaction that Fridman initiated on December 14 as a “payment order” for a “funds transfer” and never as an “instrument” (a word used in the Article on checks). Similarly, an instruction to start the process of obtaining funds from a depositary bank does not sound like a “means of payment”; if this procedure has a “means,” it is the entirety of the ACH’s operation, which did not produce a payment until NYCB received its credit on December 18.

The majority’s tour, op. 777-78, through state statutes and federal opinions shows the power of electronic databases. It is linguistically possible to use “instrument” as one statute in each of Kansas and Michigan does, but this doesn’t show that such a usage is normal (what of the other 48 states and the UCC?; what of all the other statutes in Kansas and Michigan?) or appropriate for this particular federal regulatory system. And if you look closely at the language quoted from the Kansas and *782 Michigan statutes, you see that they contrast “orders” for the payment of money with “instruments”; these are different ideas.

Because “payment,” “instrument,” and “means of payment” are not defined, my colleagues turn to another sentence of the staffs commentary:

If the consumer elects to have payment made by a third-party payor such as a financial institution, through a preauthorized payment or telephone bill-payment arrangement, payment is received when the mortgage servicer receives the third-party payor’s check or other transfer medium, such as an electronic fund transfer.

This ought to clinch the case for NYCB, because it says that “payment is received when the mortgage servicer receives the third-party payor’s check or other transfer medium, such as an electronic fund transfer.” It shows that the staff thinks “electronic fund transfer” different from an “instrument” and that the lender must credit the payment when it “receives the third-party payor’s ... transfer medium”— when the process is finished, not when it is initiated — which in this case means December 18. This is why the district court granted summary judgment in NYCB’s favor.

But my colleagues do not read the sentence this way. Instead they say that a third-party transfer is credited on the date of receipt only when the payment instruction was issued by the borrower directly to the third party (here, to Bank of America). If the payment instruction is routed through the lender or servicer, my colleagues conclude, then this sentence of the staff commentary is irrelevant.

I don’t follow this. The staffs language does not specify a difference according to who receives the payment instruction. The sentence asks when the third party’s payment reaches the lender. How the transaction begins is neither here nor there. The phrase “preauthorized payments,” on which my colleagues rely (op. 779-80), does not do the trick. Whether the process starts with the lender or the borrower’s bank, the payment is “preauthorized” in the sense that the authorization precedes the credit. A customer could authorize a payment two days, a month, or a year in advance, but all are “preauthorized.”

Now let us suppose that everything I have said is wrong, and that the staff commentary not only trumps the statute but also treats a payment order as an “instrument” or “means of payment.” The best analogy for that point of view would be to equate a payment instruction with the use of a debit card, which might be called a “means of payment” (though the debit card also produces an immediate transfer, unlike the delay built into the ACH system). Is a lender required to accept a debit card, or for that matter a payment order, on a par with cash? The statute does not say — but the regulation does.

Section 1026.36(c)(l)(iii) says that a servicer may require customers to pay using a menu of ways that it specifies. Thus NYCB is entitled to reject debit and credit cards. In the absence of a written policy specifying acceptable ways to pay, a servicer can reject anything other than cash, money orders, or negotiable instruments (of which checks are examples). Staff commentary on § 1026.36(c)(l)(iii) at ¶ 3. So NYCB need not accept as statutory (and regulatory) “payment” orders that leave it with the burden of using an ACH to obtain funds from the customer’s bank. The regulation recognizes, however, that a servicer may permit a method not on its authorized list (or the staffs default list). If it does that, it may defer giving credit *783 for as long as five days. 12 C.F.R. § 1026.36 (c)(l)(iii).

As far as I can see, NYCB has not put transfer via ACH on a list of approved payments. In' other words, it accepts a payment order as a means of producing a payment, but not as a payment. Before being allowed to enter the payment instruction on NYCB’s web site, Fridman had to check a box acknowledging that a funds transfer through an ACH would not qualify as immediate payment. This brought it within the scope of § 1026.36(c)(1)(iii) and allowed NYCB to wait as long as five days before giving credit. NYCB credited Fridman’s account in two business days — indeed, promised credit in two business days even if the ACH took longer. Fridman therefore cannot complain about the late charge.

My colleagues express concern that, if a lender need not treat an ACH order as a statutory “payment” until it receives the funds from the depositary bank, it may be tempted to delay the start of the collection process in order to run up late fees. Op. 779-80. That’s not a risk for NYCB, which promises credit in two business days no matter how long the ACH process takes. And I do not think it likely for any other servicer. Playing games would put its reputation at risk. Users of the Internet proclaim their grievances loudly, and many sites rate merchants based on users’ observations.

The majority’s understanding can lead to bad consequences too — worse, and more likely, than the possibility that concerns my colleagues. One thing a lender may do in response to today’s decision is refuse to accept payment orders. Then a borrower such as Fridman would either have to write a paper check, taking all risk of delay in the mails, or go to her own bank’s web site to cause it to make a funds transfer (something that, the majority acknowledges, would allow the lender to defer credit until the money arrives).

A second thing a lender could do would be to reduce or eliminate grace periods. NYCB now gives its customers 15 days past the deadline to make payments without incurring charges. Under NYCB’s procedures, a borrower who wants to use an ACH collection must act within the first 13 of those days to avoid a late fee. If as my colleagues hold a lender must give the borrower credit the same day a payment order is received, that turns 15 grace days to 17 (or 19 with weekends). The lender can cut the time back to 15 by reducing the grace period to 13 or 11 days. But that’s hard to remember. A reduction to 10, 7, or zero would be more likely. Customers would lose.

Consequences, good or bad, are the province of Congress and the Bureau. Our job is to interpret the statutory and regulatory language. Instead of stretching that language in a way that may induce lenders to reduce or eliminate grace periods, or stop facilitating ACH transfers, we should read the statute and regulation to mean what they say: lenders must give credit when they receive payment. NYCB gave Fridman credit the day it received payment. It has complied with the statute.