Bilibili Finally Makes Money. That Is Exactly Why the Stock Keeps Falling.

The consensus headline writes itself: Bilibili, the Chinese video community that burned through years of losses, has turned the corner. In the second quarter it booked net profit of RMB339 million, up 55% from a year earlier. Gross margin hit 37.2%, its 16th straight quarter of improvement. After a decade of "when will it make money?" the answer is finally "now." Everyone is right about that.

The problem is what the market was paying for all along. Because profitability is arriving at the same moment the stock is sliding toward a 52-week low — down roughly 40% over the past four months, trading near $15 after comfortably clearing $36 earlier in the cycle. The good news is real. The market problem begins one step later. Profit is up because the company around it is getting smaller at the top.

The profit is the product of a shrinking machine

Look at what "profitable" is being built from. Revenue grew just 8% year over year last quarter, a sharp deceleration from the roughly 20% clip a year earlier. The engine that once made this stock exciting — mobile games, the high-margin, hit-driven business where one title can carry a quarter — fell 14% year over year. That was the fourth consecutive quarter of double-digit contraction. Management blames a hard comparison against an older hit, but the pattern is broad: the game-development and publishing pipeline has not delivered the new money-printers that made optionality the reason to own Bilibili in the first place.

The growth carrying the company now is advertising, up 28% and roughly 40% of revenue. That is a fine business, but it is not the same business. Ads are lower-margin and cheaper to scale competitively, and the version Bilibili is selling runs on rising AI spending — research and development jumped 16% last quarter, largely on server depreciation tied to AI infrastructure. Profit growth is coming from margin discipline and a surging ad line while the founding revenue engine shrinks and the cost base gets heavier.

The hidden premise everyone stopped checking

The bull case for Bilibili never rested on this quarter's earnings. It rested on a promise: games would reaccelerate. That is the hidden premise. Management now says growth there will not return until the fourth quarter — another quarter of waiting for the old story to come back. Every quarter that passes without the rebound forces the market to pay for a slower, capital-hungrier company at the price that once made sense only for a hit-driven one.

This is where the "wrong metric" cuts hardest. Headline net profit is the number fans celebrate, and it genuinely rose 55%. But GAAP profit can grow on cost discipline while the forward-looking engine shrinks. The celebrated numerator is answering an easier question than the one investors actually need, which is: where does the next dollar of profitable revenue come from? Right now the only leg growing fast is the one saddled with the biggest new cost — AI.

What you are actually paying — and what it assumes

Here is the uncomfortable arithmetic. Even after a 40% drawdown, Bilibili trades at roughly 109 times forward earnings, against a trailing multiple near 29 times. The gap between those two numbers is the market's expectation, still embedded in the price, that earnings grow dramatically in the next four quarters. The company is not priced as a steady margin-grind platform. It remains priced as a company about to reaccelerate — precisely the rebound that has now failed to show up four quarters running.

Put the obvious peer next to it and the premium looks stranger. NetEase, a profitable Chinese games and media business growing real revenue on real earnings, trades at roughly 16 times trailing earnings with an EV/EBITDA around 8 and a dividend yield near 2.4%. Bilibili, whose core games franchise is contracting, sits at a higher earnings multiple and a richer enterprise multiple. It is not the cheap breakdown story the falling price implies. It is an expensive breakdown story whose premise keeps failing.

To be fair to the bulls: this is not a business in obvious collapse. Bilibili still holds roughly RMB24.3 billion — about $3.6 billion — in cash and short-term investments, giving an enterprise value far below its market cap and plenty of room to wait out the pipeline. Daily active users rose 7% to about 117 million, engagement at 113 minutes a day, and it has been buying back stock. The platform is not dying. It is just no longer what the multiple was built to pay for.

What would make the crowd capitulate

A contrarian argument has to be able to lose. On the bull side, the failure condition is legible: games return to year-over-year growth in the fourth quarter as promised, ad margins hold while AI capex is absorbed, or value-added services reaccelerate from its 5% crawl. Any of those would let the reacceleration narrative drag the multiple back up, and the deep cash cushion makes a catastrophic outcome unlikely. On the bear side, the risk is the opposite — every quarter that pushes the games rebound further out compresses the valuation further, because the stock is still carrying a forward multiple that assumes the rebound is near.

That is the honest shape of the trade. The market is not confused, and it is not punishing a profitable company for being profitable. It is re-pricing Bilibili from a game-hit optionality story into a slower, ad-dependent, capex-hungry platform — and the current price still embeds the optimistic version. The company may keep winning the accounting while shareholders lose the multiple. Being with the crowd that keeps saying "profitability finally!" protects a narrative; it does not protect a portfolio. Watch the fourth quarter the way you would watch a landmine: the difference between the story you were sold and the story that trades is whether games come back — and whether anyone still cares about games by the time they do.