Credit Analyst: The First 90 Days

The interview tested whether you could think like a lender for forty minutes. The first ninety days test whether you can do it on a Tuesday afternoon with three files open, a relationship manager waiting, and a senior who has read the same numbers you have and wants to know what you make of them. Nothing in the quarter is a test of knowledge — the desk assumes you can read a balance sheet. It is a test of whether you can be handed a file and come back with a view, in the order the desk reads in, in the words the desk uses, in a paper someone else can act on.

Three things get measured without anyone saying so. Whether you read the whole file before you form a view, or form the view from the request and read for confirmation. Whether your first paper reaches a conclusion in its first sentence or in its last. And whether, when a senior changes one fact, your answer moves for a reason or simply moves. The temperament the panel was listening for — the site calls the traces the Four Signals — is now being watched for in the work.

The systems, the templates, the sign-on to the credit workflow: all of it matters, none of it is the job. The job in the first two weeks is to learn how this desk reads, and the fastest way is to read what it has already written. Ask for ten recent papers — a mix of approvals, declines and reviews — and read them for the order they go in, the number each one turned on, and how the conclusion is phrased. Every desk has a house reading order; it is usually the one in Credit Analysis for Beginners with local variations, and the variations are what you are there to learn.

Learn the desk's words. A clean-down, a borrowing base, headroom — the glossary gives the desk's meaning, and a new analyst who uses them precisely is read as someone who has done this before. And learn the shelf: the facilities the bank actually writes, one file at a time, are How Banks Lend. You will be asked to read a request against the product it fits, and you cannot do that without knowing what each product is for.

New analysts get annual reviews first, and it is not because reviews are easy. A review is a file with a known ending. Last year's paper said what the desk expected the business to do; the numbers since say what it did; the gap between the two is the education, and it is free. Read the old paper, form your own view of the year before you read the relationship manager's commentary, and then compare. Where you agree, you have learned the desk's reading. Where you differ, one of you has missed something, and finding out which is the fastest learning in banking.

Three habits belong here. Read cash before profit — the statement a desk reads first is the cash flow, and Cash Flow Analysis for Business Lending explains why a profit that did not arrive as cash is a loan the borrower has already made. Make the three statements agree before you believe any of them — a debtor figure that grew faster than sales is a question, not a line item. And find the one number the file turns on. Every file has one; on the site's Issue 01 it is the cash conversion cycle lengthening under 50% revenue growth, and the whole decision follows from it.

The paper is the seat's product, and the first one will be rewritten. What gets rewritten is rarely the arithmetic. It is a conclusion that arrives in the last paragraph instead of the first; ratios reported without a sentence saying what they mean on this file; risks listed without what is left of each one once the structure has dealt with it; and conditions that do not answer a risk — asked for because they are always asked for. How to Write a Credit Paper sets out the structure and a complete example on a published file; read it before the first paper, not after the first rewrite.

Take the rewrite as data, not as a verdict. The corrections a senior makes to a first paper are the desk's reading order made visible, and an analyst who reads the tracked changes as carefully as the file writes a second paper that is mostly left alone.

At some point in the third month you will carry a file into a room — a two-signature approval, a credit meeting, a committee — and the file will not be the test. The question will be. Committees and sanctioners almost never ask whether the paper is complete; they change one fact and watch whether your recommendation moves for a reason. What if the top customer leaves? What if the season does not clear? What if the payout is a dollar lower? The analyst who can say “then the cover falls to here, and the condition on page two is there for that” is ready. The one who says “I would need to look at that” has described a paper, not read it.

Prepare for the room by asking yourself, before it, what would reverse your view — and checking whether the file already answers it. That single habit is most of what the independent read exists to do, and an analyst who does it to their own paper before a sanctioner does it for them has understood the seat.

Asking for more information instead of reading what is there. A new analyst's reflex is a list of questions for the relationship manager. Most of them are answered in the file, and the senior who sees the list knows it was written before the file was read.

Reporting instead of reading. Twelve ratios, each correct, none with a sentence saying what it means for whether the money comes back. The desk can compute ratios; it hired you to read them.

Hedging the conclusion. “On balance, subject to the considerations above, the request appears broadly supportable.” A committee reads that as an analyst who does not want to be wrong, and will not let them be right either.

Treating the covenants as the read. A file that is inside every covenant can be going wrong, and the early-warning read is mostly behaviour before numbers: the pack that comes late, the tone that changes. Covenants are the last thing to move.

Being right and unread. A correct view in a paper nobody can find it in. The first sentence is where the answer lives; everything after it is evidence.

The reps the quarter demands can be built before it starts. The case drills hand you a file and four options and hold the senior read back until you have committed — the same shape as being handed a file on the desk. Read Issue 04 and Issue 01 together: the same request, opposite answers, and the reading order that separates them. Each published file now carries an appendix of three years of statements and, after you answer, the numbers the senior read used — which is the annual-review exercise of section 03 in miniature. The learning map sets every module a credit course covers beside the guide that works it and the file where it decides something; the seat's interview, which you have already passed, is here; and where the seat leads is the career path.