Credit Analyst

I use the five Cs framework as a working checklist: character, capacity, capital, collateral, and conditions. Character comes from payment history and, for a business, how management has handled past downturns. Capacity is the core of the analysis: I build out three years of historical cash flow and calculate the debt service coverage ratio, looking for at least 1.25x on a stabilised basis. Capital means checking how much of their own money is at risk, since owners with skin in the game behave differently under stress. Collateral gets valued conservatively, usually applying a discount to book value for anything other than cash or receivables. Conditions covers the sector and macro environment: a 1.3x DSCR in a stable sector reads very differently from the same ratio in a business tied to commodity prices. I write up the analysis so someone who has never seen the file can follow the logic to my recommendation.

Debt service coverage ratio is the one I come back to most for cash flow lending, because it answers the question that actually matters: can this borrower pay from operations alone, without leaning on refinancing or asset sales. I pair it with the debt to income ratio for individuals or debt to EBITDA for companies, since a DSCR of 1.3x means something different at 2x leverage than at 6x. Current ratio and quick ratio tell me about short-term liquidity, which matters most for working capital lines. I also look at interest coverage separately from DSCR, since a business can cover interest comfortably but still struggle with principal amortisation on a shorter-term note. No single ratio drives a decision on its own. A DSCR that's looking fine can hide a business that's stretching payables to suppliers, so I always cross-check against the cash flow statement and days payable outstanding.

I go to operating cash flow before anything else, because net income can be flattered by non-cash items or one-off gains that won't recur. I want to see operating cash flow consistently above net income over a multi-year run, which usually signals conservative accounting and real earnings quality. From there I check the trend in working capital, since a growing business that's burning cash to fund receivables and inventory needs a different kind of facility than one that's actually unprofitable. I also separate capital expenditure into maintenance and growth spend where the borrower can break it out, since maintenance capex should come out of my free cash flow calculation before I size any facility. Financing activities tell me whether the business has been propping up operations with new borrowing, which is a flag I follow up on directly with the borrower rather than assuming.

I treat the score as a starting filter that still needs my judgment on top. For consumer lending I use the bureau score to triage volume quickly and flag files that need manual underwriting, typically anything below the lender's automatic approval threshold or with recent delinquencies. For commercial credit I've worked with internal risk rating models that combine financial ratios, industry risk weightings, and management quality scores into a single grade, and I always review the inputs behind the grade rather than accepting it at face value. Models miss context: a temporary dip in revenue from a one-off client loss reads the same to the model as a structural decline, and only a human review catches the difference. I document any override of a model output with the specific reasoning, since that audit trail matters both for compliance and for improving the model over time.

A commercial loan officer brought me a renewal for a restaurant group that had grown from three to eight locations in eighteen months. On paper the DSCR was 1.4x and revenue was up sharply. When I pulled the trailing cash flow apart, growth in same-store sales was flat and almost all the revenue increase came from the new locations, three of which had opened in the last four months and hadn't yet stabilised. Payables had also stretched from 35 to 58 days. I flagged that the coverage ratio was being propped up by unstabilised units and delayed supplier payments, while the underlying performance hadn't actually moved. I recommended a smaller facility with a covenant tied to same-store sales rather than the requested amount. The loan officer pushed back initially, but the credit committee agreed with my read. Two of the new locations underperformed in the following year, and the tighter covenant gave us an early warning instead of a surprise.

My manager wanted to approve a working capital line for a wholesale distributor based on strong historical numbers, but I had concerns about customer concentration: one client represented 40% of receivables. I laid out the exposure in a short memo, showing what happened to the DSCR under a scenario where that client paid 60 days late, which is a real risk in that industry when a large buyer renegotiates terms. My goal was to resize the deal and structure it differently, rather than kill it outright. We agreed on a smaller initial facility with a borrowing base tied to eligible receivables that excluded any single customer above 25% of the total, with a review after two quarters of payment history. My manager credited the structure with catching a real slowdown from that customer six months later, before it became a bigger problem.

A relationship manager needed a term sheet turned around in 24 hours for a competitive deal, well below our usual three-day analysis window. I prioritised the checks that actually drive the decision: I pulled bureau and bank statement data immediately, ran the DSCR and leverage ratios first, and flagged anything materially outside our normal range. I deferred the deeper qualitative write-up, like management background checks and full industry analysis, to run in parallel rather than sequentially, and I was explicit with the relationship manager that the term sheet carried conditions subject to that follow-up. That transparency mattered: it meant the borrower knew the numbers weren't final, and it protected the bank if anything came up later. Nothing did in that case, but I wouldn't compress the process again without flagging exactly what was being deferred and why.

I start with where the sector sits in its cycle: commodities, construction, and hospitality behave very differently from healthcare or essential services when the broader economy turns. I look at typical margins and leverage for the sector so I'm not judging a borrower against the wrong benchmark: a 12% net margin is thin for software but strong for a grocery distributor. I check regulatory exposure specific to the sector, since something like a change in reimbursement rates can hit a healthcare provider's cash flow regardless of how well it's run. I also weigh concentration risk at the sector level across our own portfolio, since a bank overexposed to commercial property in one region carries risk beyond any single borrower's financials. For a borrower in a cyclical sector I stress-test the DSCR against a downturn scenario rather than relying only on trailing twelve-month numbers.

Capital adequacy rules under the Basel framework mean that riskier exposures carry a higher capital charge for the lender, which is why risk-weighted assets factor into how a deal gets priced and sized alongside the borrower's own numbers. I make sure the risk rating I assign is well supported, since an understated rating understates the capital the bank needs to hold against that exposure, and that gets tested in regulatory exams. I also apply fair lending and know-your-customer requirements consistently across files, documenting the same analytical steps for every applicant so a decision can never look like it turned on anything other than the credit merits. For anti-money laundering flags, unusual cash flow patterns or unclear source of funds get escalated rather than explained away, even when the underlying credit numbers look fine. Getting the documentation right protects the decision as much as getting the analysis right.

I wouldn't treat borderline as a binary approve or decline. First I'd identify exactly which metric is weak, for example a DSCR of 1.15x against a 1.25x policy minimum, and whether that's a structural issue or a timing issue, like a large one-off expense in the trailing period. If the underlying business is sound, I'd look at structural levers: a shorter amortisation to reduce annual principal, a covenant requiring a cash sweep above a certain balance, a personal guarantee if one isn't already in place, or additional security to bring loan to value down. I might also propose a smaller facility now with a step-up available after two quarters of clean performance, which lets the bank build a track record with the borrower before extending full exposure. The goal is to find a structure where the risk and the terms match, rather than forcing a borderline file into a standard product it doesn't fit.