People

A Founder Holds 91% of the Votes — and Just Lent the Company $120M

Yatsen is a founder-controlled vehicle in which the gap between economic ownership and voting control is unusually wide and getting wider, not narrower. Chairman-CEO Jinfeng Huang directs 90.7% of the votes from 34.3% of the economics [1], up from 63.8% of the votes at the November 2020 IPO [2]. Two co-founders have left, dual-class supervoting Class B shares (20 votes each [3]) have aggregated into the CEO's vehicle, and a multi-year share repurchase program has shrunk Class A float, concentrating his vote share further. In March 2026 the CEO put his own pocket on the other side of a $120M convertible-note-and-warrant private placement [4]; the company's largest outside holder, Hillhouse, formally objected and forced a renegotiation before joining the first tranche [5] [6].

Yatsen's people are credentialed and competent — Harvard MBAs, ex-Procter and Gamble, ex-Estée Lauder, ex-Vipshop CFO, ex-Goldman Sachs economist, ex-Sina CFO. The structure they sit inside is not.

Governance Grade

C-

CEO Voting Power

90.7%

CEO Economic Stake

34.3%

Independent Directors

3 of 5

Top alignment concern: The CEO's personal participation in a $120M convertible note and warrant issuance to the company — at a $4.63 ADS conversion price and a 4% IRR repurchase floor [4] — is a related-party financing approved under a board where independent directors sit on the audit committee but cannot outvote the founder. A material shareholder publicly objected before the deal closed [5].

Control vs. alignment — the wedge keeps widening

Yatsen has two share classes: Class A (1 vote) and Class B (20 votes). The 20-vote uplift was inserted immediately prior to the IPO — the prospectus warned investors that Class B would step up from 10 votes per share to 20 votes [3]. All 600,572,880 Class B shares now sit in one entity, Veritas Vision Holding Limited (formerly Slumdunk Holding Limited), of which Jinfeng Huang is sole director [7]. Two co-founders who once held 28% of the votes between them — Yuwen Chen (former COO) and Jianhua Lyu (former CSO) [8] — have exited the company and the cap table; their Class B blocks have either been converted to Class A or consolidated into Huang's vehicle.

The wedge has widened, not narrowed, since the IPO:

At IPO Huang held 25.3% of the economics and 63.8% of the votes [2]. Today he holds 34.3% of the economics and 90.7% of the votes [1]. His vote share rose by 27 percentage points without buying a single new Class B share — it rose because the Class A float shrank under the repurchase program, two co-founders left, and the dual-class structure mechanically amplified everything that remained. A buyback in this structure is not a return of capital to public holders so much as a transfer of voting power to the founder.

The board has formally acknowledged the consequence: Yatsen is a "controlled company" under NYSE rules, which lets it skip the requirement that a majority of directors be independent and the requirement that the nominating and compensation committees be fully independent [9].

The March 2026 self-deal — a CEO who is also the company's largest creditor

On March 11, 2026 Yatsen signed a Note Purchase Agreement with an investment vehicle affiliated with Trustar Capital and Mr. Jinfeng Huang, the Company's founder, Chairman of the Board of Directors and Chief Executive Officer, for ~US$120M of RMB-denominated convertible senior notes plus warrants [4]. The economics, lifted from the subsequent-events note:

No Results

The conversion price is now well above market — the ADS closed at $3.17 on June 18, 2026 versus the $4.63 conversion strike — but the CEO's vehicle has a 4% IRR floor on the principal if conversion is not attractive. That is, the CEO gets a 4% return whatever happens to the stock, with optional upside if the equity recovers. Outside Class A holders sit on the dilution risk and have no floor.

Hillhouse, the largest outside holder at 13.8% economic / 1.9% voting [1], did not accept this quietly. The FY2025 annual report itself flags that "a significant shareholder of ours raised objections to, and concerns regarding, the consummation of the transaction as contemplated under the Note Purchase Agreement, and sought to participate in the transaction" [5]. On the Q1 FY2026 call, Huang confirmed that the first tranche closed on May 21, 2026 "in addition to myself" — i.e., with a new institutional investor (Hillhouse, per the company's own press release) joining the deal [6] [10].

That outcome — a major outside holder having to publicly contest a related-party financing to be included on equal terms — tells you what the board's "independent" majority is and is not. The audit committee, which is supposed to review and approve all related-party transactions [11], let a financing structured around the CEO's pocket reach an executed Note Purchase Agreement before the friction with Hillhouse became public.

Read the convertible deal as a governance test. A controlled-company audit committee approved a financing where the CEO is on both sides of the table. The largest outside shareholder had to publicly object to get in. This is the structural risk Class A holders priced when they bought ADSs — now realized.

The convertible note is the headline-grabbing related party transaction, but it is not the longest-running one. Yatsen has been quietly compounding its purchases from "companies over which we exercise significant control" — chiefly its Cosmax joint venture, Yatsen Biological Technology (Guangzhou) [12], and Guangzhou Intelligent Logistics — at a far faster rate than the business itself has grown.

Yatsen purchased RMB 15.2M of inventories and services from significantly-influenced companies in 2020, growing to RMB 372.0M (US$53.2M) in 2025 — a 24× increase over six years [13] [14]. Revenue over the same period roughly flattened (RMB ~3.5–4.3B). In parallel, in 2022–2024 the company sold inventory worth RMB 11.4M, RMB 20.1M and RMB 10.4M to a company controlled by the CEO personally [15] — small amounts, but the principle is what matters: the CEO is a customer, a supplier-partner, and (now) a creditor.

None of this is necessarily abusive. Cosmax is a legitimate contract manufacturer; supply-chain consolidation through an investee can be defensible economics. But the trend line — 25× growth in related-party purchases over five years, set against persistent operating losses — is one an audit committee should be challenging, not ratifying. The 20-F discloses the totals; it does not disclose whether independent benchmarks were used or whether margins paid to related parties were tested. That gap is itself a disclosure choice.

What management is paid — and what we can't see

Yatsen reports as a foreign private issuer. It discloses an aggregate compensation figure for the executive officer group plus an aggregate figure for independent directors, and individual NEO salary and bonus tables are not required and not disclosed. The reported cash-and-benefits aggregate has run between US$0.4M and US$1.4M for the whole executive officer group across five years:

The all-in cash run-rate for the executive officer group is around US$1.1M in FY2025 (RMB 7.5M cash + RMB 0.3M benefits) plus US$0.2M to all independent directors [16]. That is strikingly small in absolute terms for an NYSE-listed company doing ~RMB 4.3B of revenue. The reported cash compensation has also been mostly falling through the loss-making years, dropping from FY2024 to FY2025 even as the company swung from a RMB 708M to a RMB 81M net loss [17] [16].

The catch — and it is a large catch — is that the real compensation here is equity. The 2018 Share Option Plan capped issuance at 249.2M Class A shares (~12% of basic share count today), and the 2022 Share Incentive Plan adds 1.5% of issued and outstanding shares per year [16]. Independent directors Bonnie Yi Zhang and Jiming Ha each received 653,480-share grants on January 1, 2025 priced at US$0.025 per share, with ten-year exercise terms [18] — penny-strike grants that are effectively restricted-share awards. Newly appointed independent director Alan Hao Zong received a 600,000-share grant at the same US$0.025 strike on the day he joined the board, February 28, 2026 [18].

Penny-strike options to independent directors are common in Chinese ADR governance, but they are not innocuous: a director whose unrealized at-market equity rises in lockstep with management compensation has a different incentive on a related-party transaction vote than a director paid in cash. Combined with the controlled-company status, this matters.

The board — credentialed, formally independent, structurally captured

The board is five people: two executives (Huang, Yang) and three independents (Bonnie Yi Zhang, Jiming Ha, Alan Hao Zong). Chair and CEO are the same person (Huang), there is no lead independent director disclosed, and the founding CEO is also the named administrator of the 2022 Share Incentive Plan [19].

No Results

(Board composition per [20] and [21].)

Two of the three independents have been here from year one (Zhang and Ha joined in 2020–2021 [20]). Sidney Xuande Huang (no relation), the prior audit chair, departed in 2025 and was replaced by Alan Hao Zong in February 2026 — meaningful because Sidney Xuande Huang was the only independent the FY2024 filing called out as an "audit committee financial expert" alongside Bonnie Yi Zhang [22]. The FY2025 filing names only Zhang as the financial expert.

The expertise mix maps onto the actual governance load:

No Results

The matrix is honest but reveals the thinness: only one director (Alan Hao Zong, joined Feb 2026) has direct beauty/consumer-product operating expertise outside of management itself. Chief Scientific Officer Jing Cheng — 17 years at Estée Lauder, ex-VP APAC research and development — sits at the executive level [20] but does not vote on the board.

The structural problem isn't the people, it's the math. With Huang holding 90.7% of votes, the three independents cannot outvote the executive directors on any matter put to shareholders. A controlled-company structure was disclosed at the IPO [23] — but it has tightened, not loosened, in the years since.

Insider behavior — the buyback funds the CEO's increase in control

There is no public-record evidence of Huang or Yang trading ADSs personally on the open market in size; what the disclosures show instead is change at the entity level. Huang's beneficial holding via Veritas Vision / Yellow Bee fell modestly from 644.1M total ordinary shares as of February 28, 2025 [24] to 643.7M as of February 28, 2026 [1] — essentially flat. CFO Donghao Yang's position rose from 67.2M to 82.7M shares over the same window, taking him from 3.7% to 4.4% economic [24] [1] — an additional ~15.4M shares accumulated through option vesting.

Hillhouse, meanwhile, has been trimming: 300.6M shares (16.4%) in FY2024 → 258.5M shares (13.8%) in FY2025 [24] [1]. Banyan Partners, a 4.9% holder in FY2024 [24], is no longer listed among 5%-plus principal shareholders in FY2025 [1]. The cap table is consolidating: outside institutions down, insider stakes (relatively) up. As of February 28, 2025 the company had repurchased 39,781,505 ADSs for an aggregate US$199.9M [25] under the share repurchase program.

That repurchase-driven Class A shrinkage is the mechanism by which Huang's vote share has continued to creep up despite his roughly flat absolute holding.

Litigation and governance overhang

A securities class action filed September 23, 2022 in S.D.N.Y. (Maeshiro v. Yatsen Holding Limited, No. 1:22-cv-08165) alleging the IPO registration statement contained materially false or misleading statements was granted dismissal on July 22, 2024 [26]. Plaintiffs sought leave to file a second amended complaint in August 2024; that motion remained pending in the FY2024 filing — a clean resolution would be a tail-risk reduction.

The other resolved overhang: SEC PCAOB delisting risk has receded — that was a 2021–2023 China-ADR concern that no longer dominates this filing. The company published a 2023 ESG report (per its September 2024 press release), but the corporate-governance scoring those reports rely on cannot fix what the dual-class math has built.

What would move the grade

No Results

The one thing that would most quickly move the verdict in the right direction is straightforward: appoint a Lead Independent Director with the authority to call executive sessions and approve related-party transactions, and publish a related-party-transaction policy with benchmark margins for the Yatsen Biological / Cosmax arrangement. Neither requires changing the dual-class structure. Both would signal the audit committee can act as a brake.

The one thing that would most quickly move it in the wrong direction is equally straightforward: a second tranche of the Trustar/Huang convertible that does not include outside-shareholder participation on equivalent terms.

References

Generated Jun 19, 2026 10:16 UTC