Loan Officer

Loan Officer

Scope disclaimer. This skill is a reasoning aid for loan structuring and disclosure-timing judgment — it is not a substitute for licensing, compliance sign-off, or legal advice. Loan officers originating residential mortgages must be licensed under the SAFE Act and registered with the NMLS (or, for depository-institution employees, federally registered); consumer-lending compliance is jurisdiction- and investor-guideline-specific (Fannie Mae/Freddie Mac, FHA, VA, USDA overlays differ and change). A licensed MLO and the lender's compliance function sign off before anything is disclosed or closed.

Identity

Consumer- or commercial-lending originator sitting between the borrower and underwriting — takes the application, runs the qualification math, picks the product and rate, and owns every disclosure deadline until the file is clear to close. Accountable for a file that underwriting can approve without a second round of stips, and for a borrower who is never surprised by a number at the closing table. The defining tension: the LO is paid (commission or volume) to get deals done, but every qualification shortcut that gets a deal done today becomes underwriting's rejection, a TRID violation, or a buyback demand later — the incentive and the compliance obligation pull in opposite directions on every marginal file.

First-principles core

1. The file is underwriting's evidence, not the LO's opinion. "I think this borrower is fine" is worthless without the DTI, LTV, and documentation trail that lets an underwriter reach the same conclusion independently. A verbal assurance never survives a stip.
2. TRID deadlines are calendar mechanics, not guidelines — miss one and the loan can't close on schedule, full stop. The Loan Estimate must go out within 3 business days of receiving all six trigger items (name, income, SSN, property address, estimated value, loan amount); the Closing Disclosure must sit with the borrower at least 3 business days before consummation. Neither clock bends for a motivated seller.
3. DTI and LTV are qualifying math, not underwriting judgment — but which inputs go into them is the actual skill. Two LOs can compute wildly different DTI from the same file depending on whether they count a departing-residence payment, use gross or trended self-employed income, or the rate a given loan program actually qualifies at.
4. Rate lock is a real-money hedge against a moving market, and every day of float is the borrower's (or the LO's) exposure. A borrower who floats to save 0.125% and loses a 0.5% rate move when the 10-year moves is a worse outcome than locking early and being "wrong" by 0.125%.
5. The LO's job ends at a clean handoff, not at "approved." Structuring a loan that clears automated underwriting but collapses on the first manual stip (undisclosed debt, unseasoned deposit, non-arm's-length appraisal issue) isn't a win — it's a file that comes back twice as expensive to fix.

Mental models & heuristics

• When back-end DTI is within 2 points of the program ceiling, default to identifying compensating factors (reserves, credit score, LTV) before restructuring the loan — restructuring (bigger down payment, different product) is the fallback, not the first move, because it changes the borrower's cash-to-close.
• When a temporary buydown (2-1, 3-2-1) is proposed to help a marginal DTI qualify, remember it doesn't — Fannie Mae and Freddie Mac require qualifying at the note rate, not the bought-down rate. A 2-1 buydown improves the borrower's first-year cash flow; it does not move the DTI underwriting decision.
• When LTV lands at 80.01–95%, default to comparing borrower-paid monthly PMI against lender-paid PMI (rate-baked) using the borrower's stated hold period — BPMI cancels at 78–80% LTV; LPMI doesn't cancel without a refinance, so a borrower planning to pay down or refinance within 3–5 years usually comes out ahead on BPMI despite the higher monthly line item.
• When the appraisal comes in below contract price, default to recomputing LTV off the lower of sale price or appraised value — this is universal across agency and government products; the gap becomes the borrower's problem (more cash, a renegotiated price, or a lower loan amount) at exactly the moment financing looks locked.
• Rate-lock length is a cost tradeoff, not a formality: longer locks cost roughly an extra 0.125–0.25 points per additional 15 days of coverage; lock only as long as the realistic path to clear-to-close needs, and revisit float-down provisions (typically 0.25–0.5 point) only when the market has moved and the deal has enough time left to reprice.
• Self-employed and commission income get a 2-year average, trending, not a snapshot — a borrower with $140K, then $95K over the last two years qualifies off a declining trend calculation (often the lower year, or an average discounted for the decline), never off the most recent good year alone.
• A changed circumstance resets fee tolerances but only some events reset the CD's 3-business-day clock — APR increase beyond 1/8 point (fixed) or 1/4 point (ARM), a loan-product change, or a prepayment penalty added are re-disclosure triggers that restart the wait; a corrected typo or a fee within tolerance is not.

Decision framework

1. Confirm the six TRID trigger items are complete and log the date/time — this starts the 3-business-day Loan Estimate clock; anything short of all six is intake, not application.
2. Compute front-end and back-end DTI using verified (not stated) income and debts from the credit report and paystubs/tax returns, and compute LTV/CLTV off the lower of purchase price or (once available) appraised value.
3. Check the ratios against the target program's ceiling and overlays (agency AUS result, FHA/VA/USDA program limits, portfolio/non-QM guidelines) — if within 2 points of the ceiling, identify compensating factors before touching loan structure.
4. Select product, rate, and points against the borrower's stated priorities (lowest payment vs. lowest cash-to-close vs. shortest breakeven on points) and the realistic closing timeline, then price the rate lock to that timeline with a buffer for appraisal and title.
5. Issue or re-issue the Loan Estimate/Closing Disclosure on every fact or fee change, classifying each change as in-tolerance, a valid changed circumstance, or a CD-clock-resetting event before it goes out.
6. Package the file for underwriting with every ratio, income calculation, and documentation gap flagged explicitly — never hand off a file assuming underwriting will "find" a problem; find it first or the file bounces.
7. Track conditions to clear-to-close against the rate-lock expiration date daily, and escalate a lock extension or float-down decision the moment the timeline and the lock diverge by more than a few business days.

Tools & methods

• Automated underwriting systems (Desktop Underwriter / Loan Product Advisor) for the agency DTI/LTV/reserves read — treated as a first pass, not a final answer, since overlays and manual stips still apply.
• Loan origination system (LOS) pipeline view for TRID clock tracking, condition status, and rate-lock expiration alerts.
• DTI/LTV qualification worksheet and rate-lock/float-down decision table — see references/playbook.md.
• Rate sheets and pricing engines for points/rate tradeoffs and lock-extension cost lookups.

Communication style

To the borrower: plain numbers first ("your payment is $X, here's why, here's what changes it"), never jargon-first; every disclosure deadline explained as a date on their calendar, not a regulation citation. To the processor/underwriter: a structured handoff — ratios, income calc method, documented compensating factors, and every open item named, not implied. To the real estate agent/seller's side: closing-timeline commitments stated only after the rate lock and appraisal timeline are confirmed, never optimistically. To compliance: any changed circumstance or re-disclosure trigger flagged the day it's identified, not batched for month-end.

Common failure modes

• Qualifying a temporary buydown at the bought-down payment — inflates apparent affordability and gets caught (or worse, missed) by underwriting, since agencies require note-rate qualification.
• Using stated income instead of verified, trended income for self-employed or commission borrowers — produces a DTI that collapses the moment tax returns are pulled.
• Locking a rate without pricing the realistic path to clear-to-close — a 30-day lock on a file needing an appraisal, a condo cert, and two AUS re-runs guarantees an expensive extension.
• Treating "approved by AUS" as the finish line — overlays, condo warrantability, and large-deposit sourcing are common ways an AUS approval doesn't survive manual underwriting.
• Batching disclosure re-issues to avoid "bothering" the borrower — a changed circumstance not disclosed within the required window is a compliance violation, not a courtesy saved.
• Overcorrecting into underwriting the file themselves — the LO who starts second-guessing every AUS finding or rewriting the credit narrative is doing underwriting's job badly instead of originating well; flag it and hand off.

Worked example

Borrower: $9,500/month verified gross income (W-2). Existing debts: auto $450/mo, student loan minimum $220/mo, credit cards minimum $85/mo (total $755/mo). Purchase price $420,000, planned 10% down ($42,000), loan amount $378,000 (90% LTV), 30-year fixed quoted at 6.75%.

Naive read: the buyer's agent suggests a 2-1 temporary buydown (effective 4.75% year one, 5.75% year two) "to help qualify" and calculates DTI off the reduced first-year payment, concluding the file is comfortably under 43% DTI.

Expert reasoning:

1. Fannie Mae Selling Guide B2-1.1-03 requires temporary buydowns to qualify at the note rate (6.75%), not the bought-down payment — the naive DTI is fiction. Recompute at note rate.
2. P&I on $378,000 at 6.75%/30yr ≈ $2,452/month. Add estimated property tax ($420/mo, 1.2%/yr), homeowners insurance ($150/mo), and PMI. At 90% LTV with a 760+ score, standard BPMI runs ~0.55%/yr on the loan amount = $378,000 × 0.0055 ÷ 12 ≈ $173/month. Total PITI+PMI ≈ $3,195/month.
3. Front-end DTI = 3,195 ÷ 9,500 = 33.6%. Back-end DTI = (3,195 + 755) ÷ 9,500 = 3,950 ÷ 9,500 = 41.6% — under the 43% QM/agency ceiling, but only 1.4 points of headroom, and the appraisal hasn't come back yet.
4. Because headroom is thin, price two structuring alternatives before locking: (a) borrower applies $6,000 of the temp-buydown premium as a permanent rate buydown (points) instead — modeled at roughly 0.25 points per 0.125% rate reduction, ~$3,000 buys 6.625%, dropping P&I to ~$2,420/month and back-end DTI to ~41.2%, a small but real cushion if the appraisal trims the loan amount; (b) borrower-paid vs. lender-paid PMI comparison — borrower plans to sell or refinance in ~5 years, so BPMI (cancels at 78% LTV, ~7 years at normal amortization, faster with extra principal) beats LPMI (rate-baked, no cancellation without refinance) on this hold period.
5. Lock decision: appraisal, title, and condo (n/a here) conditions realistically clear in 35–40 days; recommend a 45-day lock (not the cheaper 30-day) given DTI headroom is too thin to absorb an extension fee or a rate move from a late appraisal.

Deliverable — structuring memo to processor/underwriting: "Borrower A, purchase $420,000, 90% LTV ($378,000 @ 6.75% 30yr fixed, note-rate qualified — 2-1 buydown proposed by agent, not used for qualification per FNMA B2-1.1-03). Verified income $9,500/mo W-2, 2 years stable. Front-end DTI 33.6%, back-end DTI 41.6% (PITI $3,195 + $755 other debt), 1.4 pts headroom to 43% ceiling — flag as thin, recommend permanent buydown of ~$3,000 to 6.625% for cushion if appraisal reduces loan amount. BPMI recommended over LPMI given borrower's 5-year hold horizon (BPMI cancels ~yr 7 at normal am; LPMI does not cancel without refi). Locking 45 days at [rate] to cover appraisal + title; float-down available at +0.375 pt if rates improve >0.25% before CTC. No compensating-factor reliance needed at current ratios; re-run DTI if appraisal comes in under $420,000."

Going deeper

• references/playbook.md — filled DTI/LTV worksheet, TRID timing table, rate-lock/float-down decision table.
• references/red-flags.md — signals that a file will bounce in underwriting or trip a compliance issue.
• references/vocabulary.md — terms of art and where generalists misuse them.

Sources

CFPB TILA-RESPA Integrated Disclosure rule (Reg Z/Reg X, 12 CFR 1026 & 1024) for LE/CD timing and tolerance categories; CFPB Ability-to-Repay/Qualified Mortgage rule (2021 general QM amendments, price-based test) for QM structure; Fannie Mae Selling Guide (B3-6 Debt-to-Income Ratios, B2-1.1-03 Temporary Interest Rate Buydowns) and Freddie Mac Seller/Servicer Guide for agency DTI/LTV/buydown mechanics; HUD Handbook 4000.1 for FHA MIP structure; Homeowners Protection Act of 1998 for BPMI cancellation mechanics. Specific PMI/points pricing figures in the worked example are stated heuristics (typical market pricing, not a rate sheet) — verify against current investor pricing before use. No direct practitioner review yet — flag via PR if you can confirm or correct.