Chief Financial Officer China

Choices that cost a quarter to protect the channel

Revenue, 2025: RMB 8.62bn; 54 per cent chilled milk and yoghurt; 44 per cent of sales through distributors
Distributor receivable days, December 2023: 27 days; Down from 71 days in March 2021
Finance colleagues led: 260; Group, plant and regional finance, treasury, tax and investor relations

Decisions & consequences

Shortening distributor credit to fit the life of the milk

When I became CFO in April 2021, Cuiling's receivables from distributors stood at RMB 820 million, or 71 days of sales, and RMB 214 million was more than ninety days overdue. Regional sales heads granted credit, usually forty-five to sixty days, for chilled milk that lasted seven to fifteen days. In practice we were financing our distributors' other businesses, and in 2020 two Jiangxi distributors had failed, costing us RMB 58 million in write-offs.

I proposed moving credit decisions to a monthly committee chaired by finance, scoring every distributor, shortening terms to fifteen days for chilled products and thirty for ambient, offering a 1.5 per cent discount for prepayment and stopping shipments to any account more than sixty days overdue. Our vice-president of sales opposed it. He argued that national competitors demanded prepayment and that credit was the main reason distributors in smaller cities chose a regional brand. He also expected to lose distributors in Jiangxi and Anhui, where several owners had no bank finance. He was right about both.

The president backed the proposal on condition that we phased it over eighteen months, starting with chilled products. Over 2021 and 2022 about 240 distributors left or were consolidated into neighbouring ones, and our count fell from 1,380 to 1,140. Jiangxi sales fell 9 per cent in 2022, and we spent RMB 26 million on additional direct deliveries to shops that had lost their distributor.

By December 2023 distributor receivables were RMB 270 million, or 27 days, and balances more than ninety days overdue had fallen to RMB 29 million. The share of distributor sales paid in advance rose from 21 to 46 per cent. Our expected-credit-loss provision is now built from distributor scores, and the Audit Committee receives the credit committee's minutes every quarter.

What changed most was the conversation with distributors. Credit became something negotiated on evidence rather than a reward for volume, and the stronger distributors, who were already paying on time, gained share from the weaker ones.

I would do one thing differently. We introduced supply-chain finance through a partner bank for smaller distributors only in 2023, after the losses in Jiangxi had happened. Had I arranged it before shortening terms, some of those distributors could have stayed.

Walking away from the biggest livestream anchors

In 2022 Cuiling's e-commerce channel reported RMB 1.62 billion of gross merchandise value, and the team celebrated it. Sixty-three per cent came from sessions with a handful of large anchors who charged fixed placement fees of up to RMB 300,000 a session plus commissions of 18 to 25 per cent, and who expected prices lower than any other channel. After refunds, coupons, commissions and fulfilment, the channel lost RMB 96 million that year. Refunds on anchor sessions ran at 21 per cent, against 6 per cent on our own. Distributors made RMB 37 million of price-support claims when online prices undercut their shops.

I recommended ending fixed placement fees, setting a floor below which no online price could go, and building our own livestream studio. Our head of e-commerce disagreed. She argued that anchors reached younger city households who never walked into the shops our distributors served, and that walking away would hand them to competitors. Our marketing director agreed with her. Both were partly right, and I said so in my paper to the executive committee.

The president approved the change from January 2023. We kept commission-only anchors for new products, where trial mattered, and invested RMB 18 million in a studio with fourteen hosts broadcasting six hours a day. Gross merchandise value fell to RMB 1.08 billion in 2023, and the share of first-time buyers in our online sales fell from 41 to 29 per cent. Two analysts asked on our results call whether we had given up on the channel.

By 2025 gross merchandise value had recovered to RMB 1.43 billion, net revenue was RMB 1.21 billion and e-commerce made a contribution of RMB 58 million. Anchors provided 18 per cent of online sales. Price-support claims fell to RMB 4 million in 2024. Refunds are now estimated by session type when goods ship, and the board sees the lowest price any channel paid each quarter.

What changed for the enterprise was that online and offline teams began planning promotions on one calendar, against one floor price.

I would do one thing differently. I underestimated the loss of new customers. I should have kept a small, measured anchor budget for reaching households we could not reach ourselves, and judged it on repeat purchases rather than on the first order.

Keeping our contract farms through the raw-milk surplus

Between 2021 and 2022, when raw milk was scarce and expensive, Cuiling signed five-year agreements with 46 contract farms in Zhejiang, Anhui and Jiangsu. The agreements set annual volumes and a floor price of RMB 3.85 a kilogram. From 2023 herd expansion across the country outpaced demand and raw-milk prices fell steeply. By late 2024 our region's spot price was about RMB 3.30, and we were receiving more milk than our chilled and ambient products could use.

Two courses were proposed. Our head of milk procurement wanted to invoke the agreements' market-disruption clause, cut contract intake by a fifth and buy cheaper spot milk instead. The president leaned towards it, and one independent director asked why shareholders should pay more than the market for an input. I recommended honouring the volumes, drying our surplus into powder at rented spray-drying capacity, and reducing imported whole milk powder instead. Contract farms were our most reliable source of tested milk outside our own ranches, several had borrowed to expand at our request, and we would need them when the cycle turned.

The board accepted my recommendation in March 2023 and asked for a quarterly report on its cost. Over three years we paid about RMB 128 million above spot and lost about RMB 82 million on drying surplus milk and writing the powder down: roughly RMB 210 million in all. Net profit fell 14 per cent in 2024. Imported powder fell from 18,000 tonnes in 2022 to 5,500 tonnes in 2025, and our dollar purchases from USD 71 million to USD 20 million.

When the agreements came up for renewal in 2025, 44 of the 46 farms renewed on new terms, with prices set against a published regional raw-milk price within a band, so that neither side carries the whole of the next cycle.

Procurement, ranch planning and powder purchasing now share one three-year supply balance, reviewed by the board annually.

I would do two things differently. Our forward contracts for imported powder ran until mid-2024, so for a year we were drying our own milk while still receiving powder from overseas. I should have shortened those commitments in 2022. And I should have proposed index-linked pricing when the agreements were first signed, rather than after they had cost us money.

Advising the board to defer a Hong Kong listing

In March 2025, as many mainland companies were preparing Hong Kong listings, our chairman proposed that Cuiling issue H-shares and raise about HK$2.5 billion. Most of the money would fund the purchase of 60 per cent of a New Zealand processor of milk powder and butter; the rest would provide offshore working capital for imports. The case was serious: foreign currency, international shareholders and visibility in Southeast Asia.

My analysis was less encouraging. Our advisers' valuation range implied an H-share price 25 to 35 per cent below our A-share price, so raising the money would dilute existing shareholders by about 9 per cent. Our dollar purchases had fallen to USD 20 million a year, so our need for offshore currency was small. The New Zealand processor would have earned less than our cost of capital at 2025 powder prices and would have added a New Zealand-dollar exposure we did not have. Listing would also cost about RMB 22 million a year and require a filing with the CSRC before the offering.

I recommended deferring the listing and negotiating a five-year supply agreement with the processor instead, with the option of a minority stake later. The chairman disagreed, and so did one independent director, who valued the discipline international investors bring. We were giving up a market window and an investor base that would have challenged us usefully.

The board decided in August 2025 to defer the listing. It approved the supply agreement, signed that November, and asked me to bring an annual capital-structure review that would reopen the question if one of three conditions were met: an offshore acquisition above RMB 1 billion, dollar purchases above USD 60 million a year, or a valuation gap between our A-share and an H-share price narrower than 15 per cent.

The supply agreement has delivered powder about 4 per cent below spot in its first eleven months; the first review is due in December 2026.

I would do two things differently. I would have prepared the groundwork for a filing in parallel, so that the option to list stayed cheap to exercise. And I would have given more weight in my paper to the governance argument, which I answered with figures when it deserved a proposal; we now publish fuller English disclosure and meet foreign holders of our A-shares.

Discuss a leadership challenge