ByteDance's share repurchase at a $400 billion valuation: the unlisted giant is "pricing" itself with cash

On November 13, according to a report from The Information, ByteDance's shareholders received a key proposal: the company plans to repurchase up to $6 billion of shares from existing investors at an approximate valuation of $4 trillion. The repurchase price is set at $241.35 per share, representing a roughly 20% increase from the $200.41 price of the previous round of repurchases a year ago.

The $6 billion repurchase scale is not aggressive for a super giant of ByteDance's size. It is more like a precise liquidity management move rather than a drastic adjustment to the capital structure. But the real key information of this transaction is hidden in the core valuation figure: $4 trillion.

This is the "current price" officially set by ByteDance for itself. In public narratives and private equity transactions, ByteDance's valuation has long exceeded this figure. In February 2026, when General Atlantic sold ByteDance shares in the secondary market, it gave a valuation as high as $5.5 trillion. In May, after evaluation, the company's board of directors concluded that the company is unlikely to go public in the near future, on the grounds that "the valuation will only keep rising". In June, a report from Nikkei Asia further confirmed that ByteDance's stock is valued at more than $6 trillion in the private market, and some investment banking professionals even believe that its trillion-dollar market capitalization is "completely possible".

A clear gap lies between the $4 trillion repurchase price and the $5.5 trillion or even $6 trillion narrative in the secondary market. This gap is the key entry point to understand ByteDance's current capital strategy. It reveals a cruel yet rational reality: at a time when AI is reshaping everything, ByteDance is using huge cash flow to force itself back to the ground from the hustle and bustle of "valuation bubbles".

The underlying logic of repurchase: the "pricing anchor" that replaces IPO

In the normal logic of the capital market, a company's share repurchase is usually interpreted as one of two signals: either the management believes that the stock price is severely undervalued, or the company lacks investment opportunities with higher returns. ByteDance's situation does not fully match either of these two classic explanations.

ByteDance is not "short of money". In the first half of 2026, the company's revenue was about $120 billion, a year-on-year increase of 30%, and its net profit was about $20 billion. Although the net profit saw a single-digit decline due to AI investment, the $20 billion semi-annual profit means its annualized free cash flow remains at an extremely healthy level. Recently, ByteDance even obtained a huge loan of about $300 billion from the banking system, which is the largest single loan in the company's history. For a company that can easily obtain a $300 billion credit line from the banking system, the motivation for using $6 billion to repurchase shares is by no means liquidity management, but pricing power management.

The deeper logic is that ByteDance is using repurchases to replace the pricing function of an IPO. For a listed company, its valuation is determined by daily transactions in the public market; for an unlisted company, its valuation is anchored by a small number of transactions — financing rounds, old share transfers, and employee option repurchases. ByteDance has clearly postponed its IPO plan in 2026. The judgment conveyed by the management to the board of directors in May is that there is still huge room for the company's valuation to rise, and "there is no urgency to go public at this stage".

Not going public means ByteDance gives up the opportunity to raise huge amounts of capital at one time from the public market, but it also avoids the harsh scrutiny of the public market on quarterly profits and the questioning of the return cycle of AI investment. As a substitute, ByteDance uses periodic repurchases to provide liquidity for shareholders, and at the same time uses the repurchase price to convey an "official valuation anchor" to the market.

$4 trillion is the current anchor. It is higher than the price range of the previous round of repurchases a year ago, and also higher than some early transactions in November 2025. But it is deliberately lower than the $5.5 trillion to $6 trillion hyped in the secondary market. This "deliberate underpricing" is in itself a rational strategy: in years when profit margins are under pressure due to AI investment, using a conservative official pricing to manage shareholder expectations, and avoiding pushing the valuation narrative to a height that cannot be fulfilled by financial data. Completing the repurchase with the confidence of $4 trillion essentially declares to all shareholders: this is ByteDance's real chassis with strong anti-risk capabilities.

TikTok's cash cow and the "bottomless pit" of AI

What supports the $4 trillion valuation are two parallel and constantly tugging narratives: one is about the carnival of growth, and the other is about the anxiety of investment.

Growth undoubtedly comes from TikTok. In the first half of 2026, TikTok Shop's global GMV reached $50.3 billion, a year-on-year increase of 92%, and it is expected to exceed $100 billion for the first time in the whole year. The GMV of the US market in the first half of the year was about $11.8 billion, more than double that of the same period last year. TikTok's advertising revenue is also surging. The US advertising revenue is expected to grow from $11.8 billion in 2025 to $13.4 billion in 2026, with the growth rate rising to 27.9%. Overseas revenue has accounted for more than 30% of ByteDance's total revenue, and is still rising strongly.

But AI is drastically changing the profit distribution structure. ByteDance's net profit margin in the first half of 2026 further dropped to 16.7% from 21% in 2024. The reason for the profit decline is clearly pointed to the huge investment in the AI field: independently developing large language models and video models, operating Doubao, expanding AI cloud business, and competing with the Seedance video generation model in the global market. The founder's judgment at the internal meeting is extremely frank. He believes that ByteDance has the ability to compete with the world's top AI companies, but to truly catch up with cutting-edge technology, "the earliest will be 2029 or 2030".

This means that in the next few years in the foreseeable future, AI will continue to exist as ByteDance's "cost center" rather than "profit center". The $4 trillion valuation essentially prices both TikTok's stable cash flow and AI's long-term options. Under this dual narrative, ByteDance's choice to repurchase $6 billion is to convey confidence to shareholders: the company is fully capable of using the huge cash generated by its existing business to absorb the profit impact brought by the AI arms race, and at the same time provide a decent exit for early investors who want to exit and employees who cannot get cash out opportunities.

The $5.5 trillion in the secondary market vs. the $4 trillion repurchase price: who is right?

When General Atlantic sold shares at a valuation of $5.5 trillion in February, the market's pricing of ByteDance included several assumptions that seemed extremely reasonable at the time: the uncertainty of TikTok's overseas business has been greatly dissipated, the annual profit is expected to reach $480 billion and surpass Meta, and the AI narrative is heating up in an all-round way.

But by November, the repurchase price fell back to $4 trillion. This gap cannot be simply attributed to short-term fluctuations in market sentiment. More likely, ByteDance is actively managing expectations through the repurchase price. The $5.5 trillion in the secondary market reflects the scarcity premium and buyer FOMO (fear of missing out) in a small number of transactions; the $4 trillion repurchase price reflects the management's prudent judgment on the financial results that it can deliver in the short and medium term.

When AI investment begins to substantially lower the net profit margin, and when the global macroeconomic environment is full of variables, requiring all shareholders to accept the $5.5 trillion narrative will create unnecessary expectation gaps. Anchoring at $4 trillion is a defensive but more long-termist pricing strategy.

This also explains why ByteDance is still borrowing on a large scale from banks while repurchasing shares. The $300 billion loan plus the $6 billion repurchase constitutes a very clear financial posture: ByteDance is using low-cost funds from the debt market to support the liquidity management of the equity market, while retaining internal cash for continuous investment in AI infrastructure. This is a highly mature capital operation method, rather than the behavior pattern of a startup waiting for an IPO.

Why give up IPO: maintain strategic autonomy and long-term operation

Why not just go for an IPO? This is a question in the minds of many investors.

For ByteDance, not going public is not only a financial choice, but also to maintain the maximum degree of strategic flexibility in the complex global business environment. Becoming a public company means exposing all financial statements, profit details, shareholder structure and core business logic to the microscope of the public and competitors.

In the current complex global competitive environment, a tech giant with hundreds of millions of users around the world will face extremely high compliance costs, interference from market fluctuations and pressure from short-term capital if it chooses to list in the public market. Choosing to "hide" in the primary market and keep the equity structure relatively closed can give ByteDance more room for maneuver and less public intervention when handling overseas business expansion and responding to global regulatory challenges.

At the same time, not going public also allows ByteDance to get rid of "Wall Street myopia". If ByteDance had gone public, its stock price would have suffered a brutal crash in the face of the financial report of declining net profit margin in the first half of the year, and the management might even face a coup from activist investors demanding to cut AI expenses to protect profits. In the private market, the management has absolute strategic determination, and can pour tens of billions of dollars into the computing power base under the pressure of profit.

Employee stock ownership and the soft landing of liquidity

In ByteDance's valuation system, there is also a huge but often ignored group: employees.

ByteDance has more than 100,000 employees, a large number of core R&D, product and operation personnel hold options. In an era when employees of industry giants can easily cash out through stock lifting, the options held by ByteDance's employees once faced the embarrassment of "paper wealth". Under the background that the IPO is far away, if there is a lack of liquidity, these options will depreciate seriously, which will lead to a large loss of core talents.

Therefore, this $6 billion repurchase (as well as the normalized repurchase that ByteDance conducts every year in history) actually provides a kind of "shadow liquidity" for internal employees. At a valuation of $4 trillion, although lower than the price hyped in some secondary markets, it still achieves dozens or even hundreds of times of book returns for early employees.

This gentle liquidity injection ensures that ByteDance's core team remains stable when it treks into the deep water zone of AI-native transformation. It can appease early employees who want to cash out and leave, and enable the company to retain top scientists and engineers who are really fighting at the forefront of AI.

A giant that doesn't need a bell

ByteDance's story is deviating from Silicon Valley's standard narrative template for "successful tech companies". According to that template, a company with revenue exceeding hundreds of billions of dollars, profits of tens of billions of dollars, and hundreds of millions of users around the world, should be standing at the door of IPO, waiting for the bell that symbolizes wealth and power to ring. But ByteDance has chosen another path: not going public, using repurchases to provide liquidity, using conservative official valuations to manage expectations, and using cash flow from existing businesses to fuel an AI competition that will not see the finish line until 2029.

Whether this path can be taken depends on two core variables. First, whether TikTok's growth engine can continue to provide strong enough cash flow to support repurchases, debt repayment and the defense of domestic business, under the background that AI investment continues to erode profits. Second, whether the return on AI investment can begin to appear in a quantifiable way before 2029, instead of becoming a cost black hole that keeps expanding and is infinitely raised by competitors.

The $4 trillion repurchase price is ByteDance's phased answer to these two questions. It is neither a blind fanatical bet on the future nor a negative compromise to reality. It is an unlisted super giant, using its extremely abundant cash, writing a pricing report for itself that does not involve the public market, but also needs to be taken seriously.

In this new business era belonging to AI, the bell can be absent, but the ledger must always be balanced. ByteDance's "internal market making" experiment may provide an atypical survival model for unlisted tech behemoths around the world.

This article is from the WeChat official account "Tech Business", author: Tech Business, published with authorization from 36Kr.