Read the request before the numbers, and read it as a structure, not a number. A working capital facility has a currency the same way it has a tenor — it follows the cash conversion cycle it funds. Aoraki invoices three-quarters of its sales in USD and pays almost all of its costs in NZD. For that business, a USD drawing line against USD receivables is not financial engineering — it's the textbook natural hedge, and arguably something the company should have had years ago: borrow the currency your customers owe you, repay from their remittances, and the exchange rate mostly cancels out of the loan. So the request contains a legitimate core. The work is figuring out what the rest of it is.
Open the file and this is not Issue 01. Revenue is up 28% over two years and the margin held — the growth is real and it's priced properly. Conduct is clean. What moves together is quieter: the USD share of sales went from 55% to 75%; export DSO stretched 21 days as the two new Greater China distributors came on at open-account terms; and hedge cover shortened from nine months to four. Read as one sentence: more of the company's cash is in USD, it arrives later, and less of it is insured. Operating cash flow is still positive but down 45% — not because the business weakened, but because the cycle lengthened. A borrower in this position asking to draw USD is being rational. A borrower in this position who has quietly stopped hedging while asking for "currency flexibility" may have started running treasury as a profit centre — which is the polite name for taking a view.
Then the tell. "Multi-currency, mix to be confirmed" is not a working capital specification — it is an option, and banks do not grant options silently inside working capital lines; they name them and they price them. In this corridor, an unnamed third currency usually resolves to one of three things: an offshore related party that needs funding — the "distributor" that is really the family's Shanghai entity; a customer who has moved settlement from USD to RMB on longer terms, which the MD would rather not headline; or a funding arbitrage — drawing a cheaper currency against no matching flow, a carry trade wearing a working capital jacket. None of these is automatically fatal. Every one of them is a different conversation, a different structure, and a different price. That is why the currency schedule is not paperwork. The currency schedule is the credit decision.
The senior banker's read is B. C — refusing to move until the full map arrives — treats the whole request as contaminated by the missing schedule, but the file already verifies the USD leg: the aged ledger, the distributor contracts, the remittance history are all on the table. Approve what the evidence supports. Restructure the line as an NZD core sized to local costs, plus a USD sub-limit capped at 80% of eligible USD receivables, drawings matched to invoice tenor, with a minimum hedge-cover covenant so the insurance can't quietly lapse again. Any third currency is available on one condition: a currency-mapped cash flow and named counterparties. Run the arithmetic and the irony surfaces — sized off the disclosed flows, the structure supports very nearly the NZ$6m equivalent he asked for. The size was never the problem. The silence was.
Note what B is not. It is not a decline wearing conditions, and it is not calling the MD's bluff about the other bank — if a competitor wants to write an unpriced currency option inside a working capital line, it can have the trade. B answers the legitimate need in full, takes the disclosed part of the request out of the auction, and makes the undisclosed part impossible to fund in silence. The MD's reaction on Friday is the last diagnostic: a borrower who wanted the hedge takes the structure; a borrower who wanted the option keeps negotiating for vagueness. Either way the bank has learned the thing the file couldn't say. When a borrower won't name the currency, the silence is the disclosure — the structure just makes it legible.