Case Study 26.2 — Three Lawful Clauses That Together Made an Honest Originator Worse

A labeled composite. The compensation plan analyzed below is constructed, assembled from structures that are individually common in the industry. It is not any particular company's plan and no company is described. All figures are illustrative and none should be quoted as a benchmark. The public enforcement pattern referenced at the end is real; specific terms and amounts are deliberately not stated here and should be read in the agencies' own published releases.

Case Study 26.1 examined a compensation structure that was prohibited. This one examines three clauses that are almost certainly lawful — and that, stacked, produce a plan under which a conscientious originator's own money argues against their client three separate times.

That is the more useful lesson, because you are far more likely to encounter this than a yield spread premium.

The plan

COMPENSATION PLAN — COMPOSITE                     [constructed teaching example]

  1. COMPENSATION.  110.0 basis points of the TOTAL loan amount.

  2. VOLUME TIER.   Monthly funded volume of $1,500,000.00 or more is
                    compensated at 135.0 basis points, applied RETROACTIVELY
                    to all volume funded in that calendar month.

  3. EARLY PAYOFF.  Compensation is recovered in full on any loan that pays
                    off within 180 days of the first payment date. Recovery
                    is charged to the individual originator and offset
                    against future compensation.

  4. DRAW.          $4,500.00 monthly, recoverable.

Read those four clauses against §26.3 and you will find no obvious violation. Compensation is a percentage of the amount of credit extended, which is expressly permitted. The tier is based on volume, which is not a term of any individual transaction. The chargeback is triggered by the consumer's own prepayment behavior, not by anything the originator set. The draw is a financing arrangement.

Now watch what the plan does to a person.

Incentive one: the thirty-first of the month

A normal file on this plan, at the market's average loan amount of \$365,750.00, pays

$$\$365{,}750.00 \times 0.0110 = \$4{,}023.25$$

Now take an originator sitting at four funded files on the twenty-eighth of the month:

Check: \$1,463,000.00 × 0.0110 = \$16,093.00. \$1,828,750.00 × 0.0135 = \$24,688.125, rounded to \$24,688.13.

The fifth file is worth \$8,595.13** — \$24,688.13 − \$16,093.00 — which is more than twice** what a file is worth on any ordinary day of the month. And it is worth that only if it funds by the thirty-first.

Nothing about this violates the compensation rule. But consider the originator's position on the twenty-ninth with a file that is clear to close pending a verbal verification of employment, a borrower who wants one more day to review the Closing Disclosure, and a title company that is booked. Every friction on that file now carries an \$8,595.13 price tag, payable to the originator, for making it go away.

The plan has not asked anyone to do anything wrong. It has simply made doing something wrong worth about two months of a normal person's mortgage payment.

What actually breaks. Not fraud, usually. What breaks is smaller and harder to see: a borrower pressed to sign before they have finished reading; a condition cleared with a document that is "basically" the right one; a verbal verification made on the twenty-ninth for a note dated the thirty-first when the window was going to be tight; a closing scheduled at a time convenient for the calendar rather than the family. Each of those is a small compromise, and the compensation rule has nothing to say about any of them.

Incentive two: the originator who cannot recommend a refinance

Clause 3 is the one that inverts the relationship.

Suppose rates fall. The Linden Street borrowers are at 6.625% with a principal and interest payment of \$2,341.94. If a comparable loan were available at 5.625%, the payment on the same amount would be about \$2,105.46** — a difference of **\$236.48 a month, or \$2,837.76 a year. (Illustrative: ignores the amortized balance, the costs of refinancing, and the break-even analysis Chapter 13 teaches, all of which the borrower would need before deciding anything.)

If that happens inside the 180-day window, the originator loses their entire compensation on the original file: \$4,023.25, recovered from future earnings, on top of a draw balance they may still owe.

So the plan has arranged for the originator's own money to sit directly across the table from \$2,837.76 a year of their client's money. And the borrower, who has no idea any of this exists, is going to call the person who helped them last time and ask whether they should refinance.

There is no lawful way for that conversation to be affected by the chargeback, and there is no realistic way to believe it never is. This is exactly the phenomenon §26.3 described: compensation systems shape behavior in people who would sincerely deny being influenced by them.

Is it a rule violation? Probably not on its face. The chargeback is triggered by the consumer's prepayment, not by a term the originator selected. But note the shape of it: it reduces compensation on a specific transaction after the fact. That is the shape the rule is suspicious of, which is why many compliance departments will not charge early payoffs to individual originators at all — they absorb them at the branch or company level, where the incentive lands on a manager rather than on the person holding the client relationship. Ask which one your plan does, and ask in writing.

Incentive three: the quiet product tilt

Clause 1 pays on the total loan amount. On a conventional loan that is the same as the base loan amount. On an FHA loan with a financed upfront mortgage insurance premium it is not.

Take the Harlow Street file — base loan \$207,475.00**, total loan **\$211,105.81 after \$3,630.81 of financed UFMIP:

Thirty-nine dollars and ninety-three cents. Trivially small, and worth taking seriously anyway, for two reasons.

It is expressly permitted. Financed UFMIP is part of the amount of credit extended, which is the one basis the rule carves out. There is no proxy analysis to run; the carve-out ends the inquiry.

And it still points in a direction. The rule cannot neutralize every incentive. What it does instead is put a second rule — anti-steering, §26.5 — behind the first one, precisely because the compensation rule alone will always leave residue. A plan can be fully compliant and still tilt.

The professional response is not outrage. It is the discipline the safe harbor describes: present the real options, in writing, for each type of transaction the borrower expressed interest in, and let the arithmetic decide. Chapter 13 chose conventional over FHA on the Linden Street file for reasons that had nothing to do with anyone's compensation, and it wrote down why. That written comparison is the thing that protects the originator as much as the borrower.

What it shows

Compliance is a floor, not a design goal. All four clauses are defensible. The plan as a whole is a machine for producing bad judgment on the twenty-ninth of the month and silence when a client asks about refinancing. "Is it legal?" and "does it point me at my client's interests?" are different questions and a good originator asks both.

Stacking matters. No single clause here is alarming. The volume cliff alone is manageable if you plan your month. The chargeback alone is survivable if the company absorbs it. Paying on the total loan amount alone is worth forty dollars. Together they produce a person whose financial interests conflict with their client's at closing, at refinance, and at product selection.

The proxy test cannot see any of this. Run both prongs on all four clauses and every one passes. The test was built to detect compensation that varies with a loan term. It was never built to detect compensation that varies with time pressure, or client retention, or plan stacking. Do not mistake a passed test for a clean structure.

This is where fair lending arrives. There is a documented public record of federal enforcement in the 2010s concerning discretionary pricing — arrangements in which originators or brokers had latitude to set a borrower's price and were compensated in a way that rewarded higher pricing. The theory in those matters was that discretion plus a compensation interest produced outcomes that differed across protected classes even without any originator intending it. Chapter 25 owns that analysis and the disparate-impact framework it rests on. Read the agencies' own published releases for the terms of any particular matter; the pattern is what matters here. A compensation structure that survives Regulation Z can still be a fair-lending problem, and the two analyses are entirely independent.

The fixes

If you are being offered this plan, four questions and one calculation:

1. Is the tier marginal or retroactive? A marginal tier removes the cliff entirely and costs the company very little. Ask why it is retroactive.
2. Who absorbs the early payoff? If the answer is "the originator," ask whether compliance has reviewed that as applied to individual originators, and ask for the review in writing.
3. Does "total loan amount" include financed mortgage insurance? Get it defined in the document, not in an email.
4. Is the draw recoverable, and what is owed on separation? §26.7 has the arithmetic; a recoverable draw plus an early-payoff chargeback can produce a negative balance at exactly the moment you least want one.

And the calculation: run your realistic monthly volume against both the marginal and retroactive structures for a full year. If the retroactive tier is only worth a few thousand dollars annually to you, you are being paid a few thousand dollars to accept a monthly cliff. Decide that on purpose.

Discussion questions

1. Run the two-prong proxy test on all four clauses of the composite plan. Show your work. Then explain in two sentences why the test's silence is not a defense of the plan.

2. Rewrite clause 2 so that it preserves the company's interest in production volume and eliminates the month-end cliff. Compute what your rewrite would pay on a five-file, \$1,828,750.00 month and compare it to \$24,688.13.

3. A borrower calls eight months after closing and asks whether they should refinance. The 180-day window has passed, so no chargeback applies. Does that make the advice you give trustworthy? What would you need to know about a plan before you could say yes?

4. The chapter argues that "is it legal?" and "does it point me at my client's interests?" are different questions. Give an example from an earlier chapter where the answers diverge, and say which question the reader of that chapter was being taught to ask.

5. The composite pays 110 basis points of the total loan amount. Design a plan that removes all three incentives identified here while paying an originator the same total compensation across a normal year. State what your design costs the company in administrative complexity, and whether that cost is worth it.

6. Compensation discretion has a documented relationship to fair-lending outcomes. Without re-deriving Chapter 25's framework, explain why a compensation plan with no discretion at all is a fair-lending asset as well as a Regulation Z one.