Loan Officer

I start by setting expectations in the first call: what documents I'll need, roughly how long each stage takes, and what happens if their situation changes midway through. For a typical mortgage or personal loan file, that means pulling payslips, tax returns, bank statements, and a credit pull within the first week, then walking the applicant through pre-qualification so they know their likely rate and loan amount before they commit time to the process. I check in proactively rather than waiting for them to chase me, usually every three to five working days once the file is with underwriting, because silence is what makes borrowers anxious and start calling the branch. If underwriting comes back with a condition, like needing an updated bank statement or a letter explaining a large deposit, I explain exactly why it's needed rather than just forwarding a form request. By completion, the borrower should never feel surprised by a number or a requirement. That consistency is what turns a stressful process into a referral source.

I avoid leading with the APR number itself and start with what it means in monthly terms, because most borrowers think in terms of what hits their bank account, not percentages. For a mortgage, I'll walk through principal, interest, and insurance separately using their actual numbers rather than a generic example, then show them the amortisation schedule so they can see how much goes to principal in year one versus year ten. When there's an arrangement fee or points involved, I explain the trade-off directly: paying points now to lower the rate only makes sense if they plan to keep the loan past the break-even point, and I'll calculate that break-even point with them on the spot. I also flag anything that changes after completion, like an adjustable rate reset or mortgage insurance dropping off once they hit a certain loan-to-value, so there are no surprises eighteen months later. If a borrower nods without asking a follow-up question, I usually ask them to repeat the number back to me in their own words.

I keep every file tagged by stage, whether that's application, processing, underwriting, conditional approval, or clear to complete, inside our loan origination system, and I review the full pipeline first thing every morning before I take new calls. Files that have been sitting without movement for more than three working days get flagged first, because those are the ones most likely to fall apart or push past a rate lock deadline. I also separate my time mentally between prospecting for new applications and servicing files already in the pipeline, usually blocking mornings for existing files and afternoons for new business, otherwise the loudest deadline always wins and quieter files get neglected. For anything nearing a completion date, I confirm with the solicitor and the borrower forty eight hours ahead that all conditions are cleared, because a missed condition discovered the day before completion damages trust with a referral partner. At any point I can tell you exactly how many files I have in each stage and which ones need attention today.

I start with debt-to-income ratio because it's the single number that tells you the most, comparing gross monthly income against all recurring debt obligations plus the new proposed payment. For most conventional products I want to see that back-end DTI under 43 percent, though I'll look closer at the full picture if someone is right at the edge, since a strong credit score and significant reserves can offset a slightly higher ratio. Beyond the ratio, I look at income stability: is this a salaried role, or self-employment income that needs two years of tax returns averaged out because it can swing significantly year to year. I also check how they're actually using existing credit, not just their score, because someone maxing out revolving credit every month tells a different story than someone with high utilisation but a clean payment history. Employment gaps, recent large deposits without explanation, and a debt load that's grown faster than income are things I always flag for a closer look before I move a file forward.

I had an applicant who wanted to refinance to pull cash out for a home renovation, but their debt-to-income ratio came back at 52 percent once I included a car loan they hadn't mentioned on the initial call. I couldn't get the file to work under any of our products at that debt load. Rather than sending a form adverse action letter as the first thing they heard from me, I called them directly, explained exactly which numbers were driving the decision, and walked through what would need to change: paying down the car loan, waiting for a bonus that was coming in three months, or reducing the cash-out amount they were requesting. I also told them plainly that a smaller cash-out amount at a slightly lower loan-to-value would likely qualify right away, and I ran that scenario for them on the spot. They came back four months later after paying down the car loan and we closed the original amount they wanted. Being upfront about why and specific about the path back is what keeps the relationship alive after a decline.

A borrower called me two days before their scheduled completion, furious because underwriting had come back with a new condition asking for an explanation of a five-figure deposit that had shown up in their bank statement the month before. He felt like we were moving the goalposts at the last minute and threatened to walk away from the deal. I let him finish without interrupting, then explained that the condition was non-negotiable on our end, since any auditor would flag it regardless of who was underwriting the file, and that the fastest way through it was a simple letter plus one supporting document, like a gift letter or a payslip showing a bonus. I offered to draft the letter template myself so he only had to fill in the specifics, and I called the underwriter directly to ask if we could review it same day given the completion date. We cleared the condition within twenty four hours and completed on schedule. He apologised afterward, and I think what actually calmed him down was that I gave him something concrete to do instead of just an explanation.

Near the end of a quarter I was short of my origination goal, and I had a file where the valuation came back lower than the purchase price, which meant the loan-to-value no longer worked for the product we'd originally quoted. There was pressure, mostly self-imposed, to find a way to make the numbers fit so the deal wouldn't fall through and I wouldn't miss my target. I went back to the borrower and their estate agent with the actual numbers instead of looking for a workaround, and we restructured the loan with a larger deposit and a different rate. It completed two weeks later than planned, which meant it landed in the following month's numbers instead of the current quarter, but the file was clean and would hold up under any audit. I'd rather miss a monthly target with a file I can defend than hit it with one that gets flagged in a post-completion review, because that costs the branch far more in the long run, in both compliance risk and reputation with investors who buy the loans.

For a conventional loan file I'm working from Fannie Mae or Freddie Mac guidelines depending on which agency we're selling to, and the core documentation set is two years of tax documents, thirty days of payslips, two months of bank statements for all accounts being used for deposit or reserves, and a credit report pulled within the last few weeks of underwriting. I verify employment directly with the employer close to completion, not just at application, because a job change or redundancy between application and completion changes the entire file. For self-employed applicants, I'm pulling two years of full tax returns including schedules, and often a year-to-date profit and loss statement plus business bank statements to confirm the income is stable rather than a one-time spike. I also check that the source of any large deposit is documented with a paper trail, not just a verbal explanation, because that's one of the first things a post-completion quality control review looks at. Every condition I clear gets documented in the file with the supporting document attached, so the file tells its own story if it's ever audited.

Income and employment fraud is the most common thing I watch for: payslips with formatting that doesn't match a standard payroll provider, an employer that doesn't answer a verification call or only has a mobile number listed, or income that's suspiciously round and doesn't match the tax return. On bank statements, I look for large deposits right before completion that get explained as gifts but don't have a matching withdrawal from the gift-giver's account, which is a classic sign of an undisclosed loan being dressed up as a gift. I'm also alert to straw buyer patterns, where the person signing the loan doesn't seem to match the profile of who will actually live in or benefit from the property, and to occupancy fraud, where someone claims a purchase is owner-occupied to get a better rate when it's actually an investment property. Identity documents that look altered, or a national insurance number that doesn't match the name in a database check, get escalated immediately rather than worked around. When something doesn't add up, I document the inconsistency and get a second set of eyes before moving the file forward.

Front-end DTI is the proposed housing payment, principal, interest, taxes, insurance, and any service charges, divided by gross monthly income, and I generally want that under 28 to 31 percent depending on the product. Back-end DTI adds in all other recurring debt: car payments, student loans, credit card minimums, and other instalment debt, divided by the same gross monthly income figure, and most conventional guidelines cap that around 43 to 45 percent, though automated underwriting systems will sometimes approve higher ratios if the borrower has strong compensating factors like a high credit score or significant reserves. I always calculate both ratios manually before trusting the automated underwriting output, because a system will sometimes miscount an instalment debt that's close to being paid off, which the guidelines let you exclude if there are ten or fewer payments remaining. I also recalculate DTI immediately if new debt shows up on an updated credit pull close to completion, since a new car loan taken out mid-process can push a borderline file over the edge and kill the deal at the worst possible time.