Who Controls Wahaha? The Succession Fight No Court Can Finish

Zong Fuli inherited 29.4% of Wahaha. She also inherited the expectation that the founder's daughter would run the founder's company, which is a culturally powerful asset and, regrettably, does not appear on a cap table.

The other shareholders did appear there. A vehicle controlled by Hangzhou's Shangcheng district government held 46%. The employee shareholding vehicle held 24.6%. Zong had the largest private block and less than a third of the company.³ She became chairwoman after Zong Qinghou died on 25 February 2024. Then, in September 2025, she resigned as chair, director and legal representative after a fight over the Wahaha trademarks. The shareholders and board approved her departure.⁶

She did not disappear. Her Hongsheng group remained central to producing and selling Wahaha drinks, and she still owned the 29.4% block. Meanwhile, three other children of Zong Qinghou were pursuing claims concerning offshore trust assets held through a British Virgin Islands company. A Hong Kong court preserved the assets. Proceedings in Hangzhou were left to decide the underlying claims. The trademark sat with the Wahaha group in Hangzhou. The production network sat substantially with Hongsheng. A possible purchase of the state block required government approvals. Everyone had leverage and nobody had the whole file.

This is the cleanest available answer to a question founders usually ask too late: who controls the company after the founder dies?

At Wahaha, the answer is different for every important verb.

Wahaha's ownership structure looks arithmetically stable. The state-backed holder has 46%, Zong Fuli has 29.4%, and the employee vehicle has 24.6%. Together, the two non-state blocks make 54%.³ That arithmetic helps explain how Zong Qinghou could operate with more authority than his personal stake suggested. It does not mean his daughter automatically controlled the employee vote, the board or the state shareholder.

Founder control often contains a large undocumented component. Long tenure decides which disagreement becomes a meeting and which one becomes an instruction. Suppliers know whose call matters. Executives have been selected inside the founder's orbit. An institutional shareholder may tolerate practices under one person that it reviews under the successor. None of this passes through probate.

Zong Fuli encountered the difference almost immediately. In July 2024, she briefly tendered a resignation from management after shareholders questioned the reasonableness of her management, then returned after discussions.¹⁰ The second resignation, tendered on 12 September 2025, held.⁶ By then the argument had moved from who occupied the chair to who could use and move the brand.

Calling her “the heir” answered a family question. Wahaha's articles, shareholder approvals and asset ownership were answering corporate questions, with less interest in the family branding exercise.

This distinction also separates Wahaha from the succession mechanisms in the LG inheritance agreement. LG's family signed the voting stake into the chosen heir's hands. At Wahaha, the founder's personal block never contained all the authority people had associated with him. Zong Fuli inherited the percentage and had to renegotiate the rest.

On 12 February 2025, Wahaha said that 387 Wahaha-series trademarks were in the process of being transferred from Hangzhou Wahaha Group to Hangzhou Wahaha Food Co. The company's explanation was historical. When the food company was formed, Wahaha Group had contributed the marks and acquired 39% of that company, while some marks were bought separately. Old disputes had prevented the registration from being completed, and the new application was meant to finish the paperwork.⁴

The ownership of Hangzhou Wahaha Food made that paperwork consequential. Zong Fuli controlled 51% of the food company. Moving the marks would therefore move the brand from a company where she held 29.4% into one where she held a majority. Later reporting said the state shareholder opposed the transfer and stopped it.⁵

There are two fair readings. One is the company's stated view that a decades-old asset contribution needed its registry entry. The other is the governance view: completing that entry would relocate the most recognisable asset in the system across a control boundary. Historical housekeeping becomes exciting when the broom is worth the company.

The episode exposes why “the company owns the brand” is not a complete diligence answer. Which company? Who approves a transfer? Who licences the mark to the production entities? Can a shareholder veto use as well as disposal? Does an old contribution agreement compel a new registration? At Wahaha, use of the mark under the current ownership structure was reported to require unanimous shareholder consent, and as of November 2025 Hongsheng had no formal licence from Wahaha Group.⁷⁵

A trademark is an intellectual-property right. In a consumer company it is also the shortest path between factory output and customer trust. Zong could direct much of the production and sales network through Hongsheng, yet still need a company she did not control to authorise the three characters on the bottle.

Hongsheng began as a Wahaha contract manufacturer and became the group's core production and sales operator under Zong Fuli. Reporting in May 2026 described Wahaha and Hongsheng as one team carrying two nameplates, with Hongsheng responsible for production and sales.⁵⁸ Employees, contracts, research staff and distributor relationships had moved toward entities in the Hongsheng network.

That gave Zong practical leverage outside Wahaha's registered cap table. A shareholder vote could block a trademark transfer. It could not make syrup move through a filling line or persuade a distributor to renew with the entity the board preferred. Operating control is made of contracts and people long before anyone gives it a tidy name.

Zong's attempted answer to the trademark veto was a replacement brand. Hongsheng filed applications for “Wa Xiao Zong” and related names. In September 2025 a notice to distributors proposed replacing Wahaha with Wa Xiao Zong for the 2026 sales year, citing legal risk from unresolved historical issues. The target reportedly attached to the new brand was 30 billion yuan in sales.⁷

The pivot did not hold. By May 2026 no Wa Xiao Zong product had appeared and distributors were still selling Wahaha. Hongsheng instead restarted KELLYONE, a brand named after Zong Fuli's English name, with a fruit soda sold in parts of eastern China. The packaging put Hongsheng in view, and the route to market leaned on direct retail and e-commerce rather than Wahaha's national distribution system.⁸

The corporate plan was to replace a contested mark with an uncontested one. Consumers, demonstrating their traditional hostility to elegant legal structures, still had to want the drink.

This is why the operating pivot matters more than the abandoned name. Zong could build a parallel legal vehicle and put a registered mark on it. She could not transfer Wahaha's accumulated recognition by board resolution. Hongsheng gave her the machinery to compete with the company in which she remained a 29.4% shareholder, while that company still supplied the brand the machinery was best placed to monetise.

The offshore dispute starts with Jian Hao Ventures Limited, a BVI company. Zong Fuli became its sole registered shareholder on 2 February 2024. Her father remained its sole director until his death, after which she replaced him. On 31 May 2024, Jian Hao's HSBC Hong Kong account held net assets of US$1,799,062,412.25, mainly bonds and other fixed-income assets, with cash and time deposits.²

Three other children of Zong Qinghou, Jacky Zong, Jessie Jieli Zong and Jerry Jisheng Zong, brought proceedings against Zong Fuli. The Hong Kong judgments identify them as Zong Qinghou's children with Du Jianying, a former Wahaha executive. Zong Fuli is his daughter with Shi Youzhen.²

Their case relies on three documents. The first is a set of handwritten instructions attributed to Zong Qinghou directing the creation of three offshore trusts, each with US$700 million. The second is a letter of entrustment that Zong Fuli confirmed in writing on 2 February 2024. The third is a family agreement signed on 14 March 2024, after Zong Qinghou's death. Under the court's preliminary reading, the three plaintiffs agreed to recognise the wills and not obstruct estate administration; Zong Fuli agreed to establish the trusts.²

The plaintiffs allege that she failed to establish them, withheld account information and made unauthorised withdrawals. Zong disputes the validity of the handwritten instructions, says the account movements included currency and portfolio changes, loan repayment and fund capital calls, and argues that the plaintiffs have no proprietary interest in the account principal. She also says the account was below the US$2.1 billion required to fund the trusts and that she has a say in their terms.¹ Those are contested positions awaiting merits adjudication.

The strange part is not that a family agreement became litigation. The strange part is the document's design. The supposed trusts required US$2.1 billion. The identified account held about US$1.8 billion. The person asked to establish them was also the registered owner and later director of the company holding the money, and her counterparties had agreed not to challenge other estate arrangements. The settlement contained its next dispute as an unfunded line item.

For founders, this is the exits and liquidity lesson inside the governance case. A beneficial promise is only as operable as the asset, trustee, funding instruction and enforcement route behind it. “Set up a trust later” is not a trust. It is a future project assigned to a family member whose incentives may change when the settlor can no longer chair the meeting.

On 1 August 2025, Hong Kong's Court of First Instance restrained Zong Fuli and Jian Hao from withdrawing or encumbering the HSBC assets. It also required disclosure of the current balance, movements after 2 February 2024, and the location or proceeds of transferred assets. The orders supported the proceedings in Hangzhou.²

On 21 July 2026, the Court of Appeal rejected the defendants' renewed application for leave to appeal, or an extension of time if permission was unnecessary. It held that the five proposed grounds were not reasonably arguable, refused a stay and ordered HK$250,000 in costs.¹

That is a meaningful procedural loss for Zong. It is not a final ruling that the three plaintiffs own US$1.8 billion. The Court of Appeal expressly treated the enforceability of the documents and the alleged trust relationship as serious issues to be tried. The preservation order keeps the relevant assets available if Hangzhou eventually orders performance or transfer.¹

The legal mechanism matters. Section 21M of Hong Kong's High Court Ordinance permits interim relief in aid of proceedings outside Hong Kong. The Court of Appeal rejected the argument that the order was really a Mareva injunction requiring proof of a good arguable case plus a real risk of dissipation. This order preserved the property that could be affected by specific-performance or transfer relief in Hangzhou. For that purpose, the lower threshold of a serious issue to be tried could apply.¹

So the courts are doing different jobs. Hangzhou handles the underlying agreement and requested performance. Hong Kong controls what may happen to assets located within its reach while that case runs. BVI company law supplies the entity holding the account. The family chose “friendly consultation” in its agreement and gave disputes to Hangzhou. Friendly consultation lasted until the first originating summons.

Lay the structure out by asset and the succession stops looking like one fight. Former and retired employees have also challenged share-repurchase arrangements, so the registered employee vehicle should not be treated as thousands of direct holders who necessarily vote as one.¹¹

In March 2026, Caixin reported that Zong was working to acquire the state-owned 46% block for more than 2 billion yuan, with government approvals still in process.⁹ No reliable public report located for this article confirms that the transfer had completed by 24 August 2026. Even that apparent solution sits inside the thesis. Buying the largest block may settle shareholder arithmetic. It does not by itself decide employee claims, offshore trust ownership, trademark licences or whether a replacement brand can sell.

The structure has no clean jurisdiction because its founder did not build one. He accumulated authority across a mixed-ownership company, related operating entities, personal relationships and offshore assets. That worked while the same person stood at the centre and could reconcile the parts. After his death, each part reverted to its legal owner and its own dispute process.

This is the opposite of the Murdoch succession settlement, where the family eventually put economic interests, the voting block, the manager appointment right and liquidity into disclosed layers. Wahaha had layers too. It just waited until litigation to label them.

Start with control-bearing assets, not family members. Put the operating company, holding companies, intellectual property, voting agreements, major contracts, licences, bank accounts, trusts and shareholder vehicles in separate rows. For each row, record five things:

1. The legal holder. Name the exact entity. “The group” is how trademarks end up owned by a company the chief executive cannot command.
2. The human decision maker. Identify who signs, votes, appoints, removes and instructs. Beneficial ownership and authority may live in different people.
3. Every veto. Include state approvals, unanimous shareholder matters, trustees, lenders, employee vehicles, class votes and contractual counterparties.
4. The governing law and enforcement forum. A BVI company, Hong Kong account and Hangzhou agreement are three legal facts. Putting the same surname beside each one does not merge them.
5. The funding source. If a trust, tax bill or family buyout requires cash, show the amount, date and account. The Hanmi succession fight shows what happens when a payment calendar owns the cap table.

Then run the map twice. First, assume everyone cooperates. Second, assume the chosen heir and one major stakeholder disagree within six months of the founder's death. If the second map depends on the founder making a phone call, you have found an asset the estate plan does not control.

Have corporate counsel, trusts and estates counsel, intellectual-property counsel and advisers in every relevant jurisdiction reconcile the same page. A perfect will cannot transfer a trademark held by somebody else. A perfect cap table cannot compel an offshore trust. A court order over a bank account cannot appoint the operating company's board.

Wahaha's fight is still live. The 29.4% block is registered. The chair has changed. Hongsheng operates. The brand remains valuable. The Hong Kong account remains preserved. Hangzhou still has merits to decide, and the possible state-block transfer still needs its own approvals.

That is already enough of a ruling for founders. Choose an heir if you like. Then do the less ceremonial work of making sure the things you expect them to control are held in structures that will listen to them after you are gone.