Synlait Denies a2 Milk and Fonterra Takeover Talks: Why Buying the Factory Was Never the Fix
Synlait told the New Zealand exchange it is not in talks with a2 Milk or Fonterra, the same day a2 Milk reported China-label infant formula sales down 14% on supply problems at that plant. The deal the market wanted would not have fixed the problem, because the scarce asset in infant formula is a China registration and there are only three per factory.

a2 Milk sold 12.4% more product last year and made less money. Sales for the twelve months to 30 June 2026 reached NZ$1.98bn. Net profit fell 5.8% to NZ$207.5m. The gap sits inside one product line, in one quarter, at one factory the company does not own.
The same day those numbers landed, Synlait Milk filed a notice with the New Zealand exchange. It said it is not in talks with a2 Milk or Fonterra about any deal. Australian press had reported that the two might take Synlait private. a2 Milk said it does not comment on market talk. Fonterra said nothing.
Why the market reached for a takeover
The logic looked clean. a2 Milk owns 19.8% of Synlait. Synlait makes its China-label infant formula. In the fourth quarter a backlog at Synlait met slow customs clearance and extra product testing. Shelves in China ran empty. China-label infant formula sales fell 14% in a year when the rest of the range grew.
Buy the factory, the thinking goes, and you own the problem. That does not work here. Bright Dairy of Shanghai holds 65.25% of Synlait. a2 Milk cannot buy what its co-owner will not sell.
The rest of the year was strong
Take out China-label and the year reads well. Sales in China and other Asian markets rose 11.2% to NZ$1.45bn. Other nutrition lines in that region grew 71%. Australian liquid milk grew 17.2% to NZ$244.9m. US sales rose 28.6% to NZ$179m, and the US arm broke even in the second half.
The ordinary dividend rose 5% to 21 cents a share. a2 Milk added a special dividend of 41.4 cents. It then told the market that first-half profit in the new year will be well down, on flat sales. One bad quarter costs a full year.
The scarce asset is a licence
Here is the part most readers will have missed. China caps infant formula registrations at three per factory. Each one is a recipe a company is allowed to sell there. a2 Milk held one, and it sat inside Synlait.
In August 2025 a2 Milk paid NZ$282m for Yashili New Zealand Dairy and its plant at Pokeno. It added a NZ$100m upgrade on top. That deal came with two more China registrations. Buying Synlait would have added a plant a2 Milk used and a licence it held. The Yashili deal added the one thing money could not otherwise reach.
The plan now shows up in the guidance. a2 Milk is moving English-label a2 Platinum output into its own Pokeno site. It expects that site to break even next year. The company still says it wants to work with Synlait for the long term. It is also building a way to need it less.
Synlait had its own reasons
Synlait lost NZ$80.6m in the first half of its year. Net debt rose 88% to NZ$472.1m on sales of NZ$949m. Chief executive Richard Wyeth left in May after a year in the job. Its quality chief and its supply chain chief had gone before him.
Its shareholders have kept it going. Bright Dairy and a2 Milk backed a NZ$217.8m share issue. Bright Dairy also put in a NZ$130m loan. Synlait sold its North Island sites, including its own Pokeno plant, to Abbott for NZ$307m in April. Its banks want debt down to three times earnings by June 2027.
What buyers should take from this
Supply deals get priced as control questions. In regulated markets they are permission questions. Ask what the target holds that you cannot apply for, wait for or build. In infant formula that is a China registration. In drinks it can be a filling slot on one line. It can be a spring source consent, or a certificate with a two-year queue.
Owning a fifth of your supplier bought a2 Milk information and no control. The outage still happened. The fix was a second site with licences attached, bought a year before anyone needed it. Watch the a2 Milk annual meeting on 19 November for the recovery plan. Watch, too, whether Bright Dairy ever becomes a willing seller. Until it does, the takeover story stays a story.

Global Aseptic Packaging Report 2026
Aseptic volumes across dairy, dairy alternatives, beverages and food in 37 countries — 2025 actuals, historical series and forecasts to 2030. Publishing August/September 2026.
Pre-order registrations are open at 2018 prices.
Pre-order your copyStrategic Insights
📊 Analytics & Strategic Insight
The licence is the asset. The factory is only the wrapper.
The decision most in this industry are avoiding:
👉 Map your permissions before your plants. Most supply reviews count sites and tonnes. Few list the licences, registrations and consents that let those sites sell. That second list is shorter and much harder to replace.
👉 A small holding in a supplier buys a seat and no lever. a2 Milk owned 19.8% of Synlait and still lost a quarter of China sales. A board seat does not move a production queue.
👉 Buy the constraint before you need it. The Pokeno deal closed a year before the outage. Assets that fix a risk are cheapest while the risk is still an idea on a page.
Here's the full context:
→ 2024: Synlait warns on debt. Bright Dairy lends it NZ$130m. a2 Milk pays NZ$24.8m to settle a pricing dispute between them.
→ 2024 to 2025: A NZ$217.8m share issue lifts Bright Dairy from 39.01% to 65.25%. a2 Milk holds near 19.8%.
→ August 2025: a2 Milk buys Yashili New Zealand Dairy and its Pokeno plant for NZ$282m. Two extra China registrations come with it.
→ April 2026: Synlait sells its North Island sites to Abbott for NZ$307m. Its chief executive leaves a month later.
→ Most recent: a2 Milk posts sales up 12.4% with China-label formula down 14%. Synlait tells the exchange there are no talks.
What this means for food and beverage operators and investors:
✅ Single-site risk is licence risk first. If one plant holds your only permit for a market, that market has one point of failure. Spare capacity somewhere else does not help you.
✅ Recovery runs longer than the outage. a2 Milk cleared the backlog and still guided first-half profit well down. Shoppers who switch brands come back on their own timetable.
✅ A distressed supplier is cheap and slow. Synlait has to get debt to three times earnings by June 2027. A supplier fixing its own books is not fixing your service level.
3 moves you can make this week:
1️⃣ List every permit and registration you sell against. Put the issuing body, the expiry date and the site next to each one. Most teams cannot produce that list in a week.
2️⃣ Price your worst quarter. Take your biggest single-site line and model 60 days with no supply. Include the shoppers you never get back.
3️⃣ Call a second source before you need one. Approval takes months in regulated markets. Start the paperwork while your first source is still working.
Take the Next Step
🧭 Facing a decision like this in your own category?
Describe it in a few lines. Selected enquiries receive an initial strategic assessment: direction, likely scope and indicative investment range.
→ Start a project enquiry
Zenith Consulting
Submit your food & beverage project enquiry.
Share your requirements. If there is a strong fit, we’ll come back with an indicative investment range, project timeline and recommended strategic approach.
Reviewed by Zenith Consulting’s senior food & beverage strategy team.
Related analyses
- M&A, Investment & Valuation
Sazerac Buys Au Vodka for a Reported £500m: The Real Price of a Format You Cannot Build
Brown-Forman refused Sazerac's $15bn offer in July. Three weeks later Sazerac agreed to buy an 11-year-old Welsh vodka brand at a reported £500m, which is about six times last year's sales.
Read analysis → - M&A, Investment & Valuation
Danone Just Took a Whole Country Off Its Revenue Line. The Arcor Deal Shows Why Deconsolidation Is Now a Strategy
Danone and Grupo Arcor completed their 50/50 Argentine dairy joint venture on 1 August, moving Danone's entire Argentine dairy business off the reported revenue line and into a single equity-accounted line. Reported sales will fall, the guided growth number will not move, and that is exactly the point.
Read analysis → - Corporate Strategy & Portfolio
Saputo Sells Cathedral City and Davidstow to Lactalis for £988m: How a Lost Bid Turned Into a Sale
Saputo chased Fonterra's consumer brands last year and lost them to Lactalis. This month it sold its own British business to Lactalis for about £988m, eight days after telling investors that British margins were finally rising.
Read analysis →
Share it with your peers
Pass this analysis to colleagues who track the food and beverage market.
Zenith Market Intel
Need a specific food or beverage market report?
Tell us which category, region or question would be useful for your team.
Sister Publication
Also follow our Water Dispense Market Intelligence
Category analyses, operator briefings, and investor signals across the global water dispense market.