How Banks Approve Business Loans

From outside, a loan decision looks like a single event: you asked, the bank answered. From inside, it is a relay — a file moving through stations, each staffed by people asking different questions, none of whom can approve it alone. The design principle behind the whole machine is separation: the person who wants the deal must not be the only person who can commit the bank to it. Once you see that principle, every station makes sense, and so does the time the journey takes.

This page walks the standard journey in a New Zealand or Australian commercial bank; titles vary by institution, the shape travels. It is written for both chairs — the borrower who wants their request to move fast, and the reader entering banking who will spend their first years as one of these stations. The wider machine the journey runs inside is the previous file in this cluster.

The journey starts with the relationship manager — the bank's first filter and, if the deal proceeds, its advocate. A good RM does real underwriting in this conversation: what is the money for, does the amount match the need, is this a request the bank's appetite can hold. Requests die here more often than anywhere else, and the good deaths are fast ones — an RM who says “not as framed, but here is what would work” in week one has saved everyone a month.

What the borrower controls at this station is completeness. Financials, aged debtors, a forecast that connects to the request, and a clear answer to “what does the money do” — arriving with these is the single largest accelerator available, because every station downstream reads the same pack. What banks look for in it is worked from the borrower's side in How Banks Read You.

Behind the RM sits the analyst — the first person to put the bank's view in writing. (What that work actually consists of, from its first question, is set out in credit analysis for beginners.) The financials are spread, the cycle arithmetic is run against the request, the structure is tested against the need, and the result is the credit paper: recommendation first, the facts it stands on, the risks named honestly with what mitigates each. The paper is the file's passport — from here on, most readers will meet the paper, not the borrower.

For the accountant entering banking, this station is where the professional question changes. Accounting asks whether the numbers are right; the paper asks what the numbers mean for a decision — and a technically perfect analysis that reaches no recommendation has not done the job. That shift, practised deliberately, is what the case drills exist to train.

Now the separation principle becomes visible. Banks run delegated lending authorities — a ladder of approval levels, each entitled to commit the bank up to defined size and risk thresholds. Small, clean requests are approved inside the business line; beyond the thresholds, the file crosses to an independent credit function whose reader owns the risk view and nothing else. This is the reader the paper was really written for: someone who has never met the borrower, holds no revenue target, and is professionally rewarded for finding the weakness the advocate missed.

At the top of the ladder sits the credit committee — the bank deciding its largest judgments collectively. Papers are pre-read; the discussion goes to the structural points: the repayment source, the downside, the second way out, what would change the answer. The most common outcome is neither yes nor no but yes-with-conditions: approved, on the terms that make the risk holdable — a covenant package, a security position, a staged drawdown. Being challenged there and holding a position on evidence is a skill the committee question tests in interviews for good reason.

Banks run the independent read in one of two shapes. In the dual-signature model, a business-line approver and a credit approver each sign within their own authority, on their own timetable, and most mid-market files clear this way in days. In the committee model, the paper waits for a scheduled meeting, is pre-read by several people, and is discussed. Larger exposures, weaker grades and anything unusual go to committee whatever the bank; the question is where the line sits.

That line is set by exposure and risk grade together, and it is different at every bank. The same request can sit comfortably inside delegated authority at one institution and a level above it at another — which is most of the answer to why the same loan takes two weeks at one bank and six at the next. The rest of the answer is what always escalates regardless of size: an exception to policy, a client already on the watch list, lending to a related party, a sector the bank has not lent to before. Any one of those moves a file up a level, and a level is a calendar.

What the borrower controls is smaller than the process and larger than it looks: a package that needs no second request for information, and a structure agreed with the banker before the paper is written, so the file arrives at the reader as a proposal rather than a negotiation. A banker who has pre-cleared the shape of a deal with credit informally has removed the most common reason a paper goes back down the ladder.

Approval is an agreement, not a transfer. Between the decision and the drawdown sit documentation — the facility agreement and security documents that turn the approval into enforceable contract — and conditions precedent: the checklist that must be evidenced before funds move. Signed guarantees, perfected registrations, insurances, sometimes the specific contract the lending was predicated on. Alongside them run the legal obligations the bank cannot waive — identity, ownership, anti-money-laundering — which no length of relationship shortens.

The practical advice this station generates is unglamorous and reliable: an approved facility is not available money until the slowest condition is satisfied, so borrowers with a deadline should ask for the conditions list early and work it in parallel. Bankers who manage this handover well are remembered for it; the machinery of what happens after the money moves — reviews, monitoring, the early-warning reads — is the third file in this cluster.