Case Study 37.2 — Why Net Tangible Benefit Exists: serial refinancing, and the tests written to stop it

Type: real regulatory history (Tier 1 for the statutes and their existence; Tier 2 for any current threshold) plus a clearly labeled composite illustrating the mechanics Complementary angle to Case Study 37.1: where the first case study is about a market that disappeared, this one is about a practice that had to be legislated against — and about the specific places where the legislation still does not reach.

Two honesty notes. First, no enforcement penalty amounts, company names, or case citations appear here. Those exist in the public record and you should read the record rather than a summary of it. Second, the extended worked example in the middle of this case study is constructed — it is built from documented patterns and its arithmetic is exact, but it is not a real household's file and it is labeled at every appearance.

Background: a product that can be sold to the same person forever

Most consumer credit products are sold once per need. You finance a car when you need a car. You borrow for tuition when there is tuition.

A refinance has no such natural frequency. The same household, in the same house, with the same mortgage, can be refinanced again and again. Each transaction generates origination revenue. Each one can be made to feel beneficial, because the payment goes down and the borrower has been trained by an entire industry to evaluate this decision on the payment. And each one — this is the part that does the damage — leaves the household measurably worse off through mechanisms that are invisible to a payment comparison.

That combination is unusual and dangerous: unlimited repeatability, revenue on each repetition, and a harm the customer cannot see. Whenever those three appear together in consumer finance, the outcome has been the same, and it was the same here.

The regulatory record, in outline

Four markers, each verifiable, none of which this case study will quantify from memory.

1994 — the Home Ownership and Equity Protection Act

HOEPA amended the Truth in Lending Act to impose heightened disclosure requirements and substantive restrictions on high-cost mortgages. The abuses that prompted it were concentrated in the refinance market and centered on equity stripping: repeated, fee-laden refinances of homeowners — frequently older homeowners with substantial equity and modest incomes — in which the fees were financed into the balance and the equity gradually left the house.

HOEPA is the first federal statutory recognition that a refinance can be a mechanism of harm rather than a benefit. Everything in §37.4 descends from it.

2010 — Dodd-Frank, the CFPB, ATR, and the compensation rule

The Dodd-Frank Wall Street Reform and Consumer Protection Act created the Consumer Financial Protection Bureau, established the Ability-to-Repay requirement (a refinance is a new extension of credit and gets the full analysis — Chapter 24), and produced the Loan Originator Compensation rule, which prohibits compensating an originator based on the terms of a transaction. Chapter 26 works compensation properly.

The compensation rule matters here for a specific reason. It removes one incentive — the incentive to steer a borrower toward a higher rate or a worse structure because you personally earn more on it. It does not remove the incentive to do a marginal transaction at all, because an originator paid per closed loan earns more from a marginal refinance than from declining it. Chapter 26 is explicit about this gap. §37.7's prevention discipline exists to fill it.

2018 — the VA refinance protections

Federal legislation enacted in 2018 — the Economic Growth, Regulatory Relief, and Consumer Protection Act — added protections aimed squarely at repeated refinancing of VA-guaranteed loans, following a documented pattern of veterans being refinanced at unusually high frequency. Three requirements, and they are worth knowing structurally even though you must verify their current form:

• Seasoning. The loan being refinanced must have been outstanding for a defined period, measured from its first payment due date, with a minimum number of consecutive payments made.
• Recoupment. All fees and costs of the refinance must be recouped from the payment reduction within a defined number of months.
• A minimum rate improvement, differing depending on whether the borrower is going fixed-to-fixed or adjustable-to-fixed.

Separately, Ginnie Mae imposed seasoning requirements before rapidly refinanced VA loans could be pooled into its securities. That is the quiet lever and it is worth understanding: a loan that cannot be pooled is a loan the lender is stuck holding, which changes the economics of the practice without requiring anyone to prove intent. Chapter 28 explains why an investor cares about a loan that prepays too fast.

And the CFPB has brought enforcement actions concerning deceptive advertising of VA refinances — mailers built to appear as though they came from a government agency, quoting savings that could not be delivered on the terms shown. Chapter 17 covers VA lending; the pattern is the point here.

Verify the current form of every one of those requirements. Thresholds, seasoning periods, and recoupment windows have all been revised and will be again.

The composite: three refinances in six years

[constructed teaching example — built from documented patterns; not a real household's file]

A household buys a home and takes a conventional thirty-year fixed loan of \$220,000 at 7.500%. P&I is \$1,538.27. Over the next six years, an originator they like calls them three times, each time with a lower rate, each time with "no money out of pocket." Every transaction is legal, every one is disclosed correctly, and every one is presented with a truthful payment comparison.

Read the fourth column downward. The balance goes up every time. Twenty-four payments of principal, then \$5,900 of costs added back, and the household ends each cycle owing more than they did at the start of it.

What six years produced

Cash paid. Twenty-four payments at each of the three payment levels:

CASH OUT THE DOOR, MONTHS 1-72          [constructed teaching example]
  24 x $1,538.27  =  $36,918.48
  24 x $1,474.89  =  $35,397.36
  24 x $1,409.01  =  $33,816.24
  ─────────────────────────────
                     $106,132.08

Debt owed at month 72: \$223,670.00 — **\$3,670.00 more than the \$220,000 they originally borrowed.** Six years, one hundred six thousand dollars, and the mortgage grew.

The counterfactual. Had they simply kept the original 7.500% loan, at month 72 they would have paid $72 \times \$1{,}538.27 = \$110{,}755.44$ and would owe \$205,211.18.

**The serial refinancer is \$13,835.46 worse off**, having "saved" \$4,623.36 in payments and having improved their note rate by 137.5 basis points.

The \$18,458.82 of additional debt decomposes exactly:

$$\$17{,}700.00 \ (\text{three rounds of financed costs}) + \$758.82 \ (\text{foregone amortization}) = \$18{,}458.82$$

And the term. The original loan retired at month 360. After the third refinance it retires at month 432 — six additional years and seventy-two additional payments, on a house they will have been paying for thirty-six years.

The point of the composite

Every individual transaction in that table passes a payment comparison. Every one lowered the rate. Not one of them was a lie. And the household lost thirteen thousand eight hundred dollars and six years, because all four harm mechanisms in §37.7 ran simultaneously and none of them appears in a payment comparison: costs financed and refinanced, amortization reset three times, the term extended by six years, and equity drained out of the house.

What it shows — including where the rules do not reach

Here is the part that makes this a case study rather than an anecdote, and it is uncomfortable.

Apply the VA-style tests to the composite and they work. Take a recoupment requirement of thirty-six months and a fifty-basis-point minimum improvement:

All three fail. The tests, where they apply, do their job — and notice which test does the work. A seasoning requirement would have caught none of them: the transactions were twenty-four months apart, comfortably past any seasoning window. It is the recoupment test that stops this pattern, because recoupment is the only one of the three that compares the cost of the transaction to the benefit of it.

Now notice what the composite is. It is a conventional loan. The VA requirements do not apply to it. Neither does the FHA Streamline benefit test. Unless the household's state imposes a statutory net tangible benefit test — a number do, with real variation in what they require and what they provide as a remedy — there is no federal recoupment requirement standing between this household and three refinances that cost them fourteen thousand dollars.

The transaction is ability-to-repay compliant. The disclosures are accurate. The compensation is structured lawfully. Everything is in order, and the household is worse off.

That is the honest limit of the regulatory answer, and it is the reason §37.7 is written as a personal discipline rather than a compliance checklist. On most refinance files in the United States, the recoupment test is whichever one the loan officer decides to run.

Outcome

The direction of the regulatory record is consistent and it is worth stating plainly: every time an identifiable segment of the refinance market has been industrialized, rules have followed — in 1994 for high-cost refinances, in 2010 for the incentive structure, in 2018 for VA loans, and continuously at the state level. Each round of rules has been narrower than the harm it addressed, because rules are written against documented patterns and patterns are documented after the fact.

The practical outcome for a working originator is unglamorous. Government refinance programs now carry explicit, quantified benefit tests. Investors impose their own. Some states impose theirs. And the largest category of refinance transactions — conventional, rate-and-term, on a willing borrower with a compliant file — is governed by professional judgment.

The lesson

Net tangible benefit is not a compliance concept that happens to be ethical. It is an ethical concept that has been partially codified, and the uncodified part is the part you will spend your career in.

The composite above is the argument for §37.7's five habits, and each habit maps to one of the four harm mechanisms:

• Ask when the current loan closed, every time, first. Catches the frequency.
• Ask what they paid last time, and whether it was financed. Catches the compounding costs.
• Compute the recoupment yourself, honestly, using §37.5's method. Catches everything the payment comparison hides.
• Ask where they are on the mortgage insurance schedule. Catches the reset that §37.10 shows can exceed every closing cost in the file.
• Decline the ones that do not clear.

The last one is the only one that costs anything, which is why it is the only one that matters.

And there is a commercial argument, which is not a consolation prize. A loan officer who will decline a marginal refinance is a loan officer whose borrowers believe them when they say a refinance is good. The household in that composite would have taken a fourth call. The originator who instead said "you refinanced twenty-two months ago, you financed \$5,900 of costs, and forty basis points does not recoup that in your lifetime in this house — do not do this" would have lost one transaction and gained a client who refers people for thirty years. Chapter 38 turns that into arithmetic.

Discussion questions

1. Three of the composite's refinances fail a thirty-six-month recoupment test and none of them fails a seasoning test. Explain why recoupment is the more powerful test, and construct a transaction that would pass recoupment while still being a bad idea for the borrower. (Hint: mortgage insurance.)

2. The composite is conventional and therefore outside the VA requirements. Should there be a federal recoupment requirement on conventional rate-and-term refinances? Argue both sides, including the cost to borrowers who would be denied a refinance they genuinely wanted.

3. The Loan Originator Compensation rule prohibits compensation based on a transaction's terms, which removes one incentive but not the incentive to originate a marginal transaction at all. Design a compensation structure that would address the second incentive. Then name the way your structure could be gamed, and who it would hurt.

4. Every transaction in the composite was accurately disclosed. The borrower received a Loan Estimate and a Closing Disclosure with correct figures each time. Why did accurate disclosure fail to protect them? What would a disclosure that did protect them have to show, and why is that a hard document to design?

5. Rewrite the composite's Refinance 3 conversation as it should have happened. The borrower called you, they have heard rates are down, and they want to refinance. Under 300 words, and you may not simply refuse — you have to leave them better informed than you found them.

6. A colleague argues that the composite is unfair because the household could have paid the costs at closing instead of financing them. Test the argument: recompute the month-72 net position assuming all three sets of costs were paid in cash. Does the household come out ahead of the keep-the-original-loan column? What does your answer say about whether financed costs are the root problem or a symptom of it?