Coastline Retail Group

A listed file arrives pre-blessed — audited, board-approved, analyst-covered — and the deference that invites is the specific trap of this genre. The bank's job is not to re-audit; every number here may be perfectly compliant. The job is to triangulate: profit is a narrative constructed from accounting policies, cash is what happened, and when the two diverge by this much, one of them is doing rhetorical work. Line up the three footnotes. A provision release adds NZ$3.2m of profit that no customer paid for. The rebate timing change pulls next year's supplier income into this year's P&L. The refit capitalisation moves NZ$4.1m of cost off the income statement and onto the balance sheet. Each is defensible alone. Together, they assemble the +9%.

Do the subtraction and the year reads differently: strip the release and the timing shift and underlying NPAT lands somewhere near NZ$8–9m — down roughly 20%, not up 9% — which is precisely the story the cash flow statement has been telling all along. Conversion at 62% against a 90–110% norm, from 118% last year: cash has no accounting policies, and it says the consumer year was as soft for Coastline as for everyone else. Then read the consent request in that light. A dividend lift from 60% to 75%, in a down year, funded increasingly by the facility rather than by earnings, is not a capital-management detail. It is the payload of the narrative — the board buying the market's belief in the +9% with the bank's liquidity.

Why not A: the covenants being clean is not comfort, because the covenants are measured on reported EBITDA — they inherit every one of the three adjustments. A covenant denominated in the borrower's own narrative currency cannot warn you about the narrative. Why not C or D: repricing or re-underwriting a four-bank club over a quality-of-earnings read is a coordination war the facts don't yet justify — the company is not distressed, it is flattering. Proportionality matters: the response should be exactly as large as the lever the bank actually holds.

Which is why the senior banker's read is B. The facility renews — the core credit is a real business having an ordinary bad year. The dividend consent is declined, politely and specifically: consent revisited when cash conversion holds above 90% for two consecutive halves, with the triangulation tabled as the reason. The consent right exists for exactly this moment; it is the one lever in the document that responds to distribution behaviour, and using it costs the relationship far less than a limit cut while saying something a limit cut cannot say — that the bank reads footnotes. One practical note for the room: club banks talk. A precise, documented read, shared calmly, tends to raise the whole club's floor — and the bank that circulated it becomes the one whose consent gets negotiated with first next time. The accounts are the company's rhetoric. The cash flow is its testimony. Lend to the testimony.