2026-05-15 · 2026-05 / week-1
Baidu Still Trades Legacy Search, Not the Return Stack
Baidu Still Trades Legacy Search, Not the Return Stack
Summary: BIDU last traded at $143.29 with a $40.21 billion market cap at 08:16 Singapore time on May 15, 2026 via the OpenAI finance tool. The market still seems to price Baidu mainly as a China internet incumbent with AI spending risk. The official filings now show something harder to dismiss: a new $5.0 billion repurchase program, a first-ever dividend policy, RMB294.1 billion of total cash and investments as of December 31, 2025, and an AI-powered business that already reached RMB40.0 billion of 2025 revenue, up 48% year over year. Q1 2026 results print before the U.S. open on May 18, 2026. Baidu buyback and dividend policy, February 4, 2026 Baidu fourth quarter and fiscal year 2025 results, February 26, 2026 Baidu Q1 2026 earnings date, April 23, 2026
Opportunity Ranking
| Rank | Idea | Discovery Lane | Why It May Be Best Now | Evidence Freshness | Catalyst Window | Asymmetry | Main Reason to Reject |
|---|---|---|---|---|---|---|---|
| 1 | Baidu still trades legacy search, not the return stack | Broader Asia / China ADR / capital return / AI transition | BIDU trades at $143.29 even after the board authorized a new $5.0 billion buyback, adopted its first dividend policy, and disclosed $42.06 billion of total cash and investments at year-end 2025. AI-powered business revenue already reached RMB40.0 billion in 2025, and Q1 prints on May 18, 2026. |
Official releases dated February 4, February 26, and April 23, 2026, plus live market snapshot checked May 15, 2026 at 08:16 Singapore time. | Q1 results on May 18, 2026, later first-dividend declaration in 2026, ongoing buybacks, and Kunlunxin spin-off progress. | The return stack is large, the balance sheet is heavier than the market cap suggests, and the AI transition is already visible in reported revenue mix. | Selected. |
| 2 | GlobalFoundries still trades the cycle, not the optical stack | U.S. large-cap semiconductor / photonics / first dividend | GFS trades at $73.84 after Q1 adjusted free cash flow of $233 million, a first-ever $0.12 quarterly dividend, a new framework to return up to 50% of trailing adjusted free cash flow after investments, and a co-packaged optics launch aimed at AI scale-up architectures. |
Official releases dated March 11, May 5, and May 7, 2026, plus live market snapshot checked May 15, 2026 at 08:15 Singapore time. | Q2 follow-through, first dividend payment on July 14, 2026, and further optical-program evidence. | The product mix is better than the old foundry label. | The stock has already rerated materially, and the cash-return floor is real but still modest. |
| 3 | Fnac Darty still prices EU clearance risk after the bid has opened | Europe / UK lane / public offer / event-driven spread | Fnac Darty last closed at EUR35.20 on May 7, 2026 versus an official EUR36.00 cash offer that opened on May 12, 2026 after AMF approval and key foreign-investment and foreign-subsidy clearances. | Official offer-opening release dated May 11, 2026 and delayed quote page crawled May 15, 2026. | Offer opened May 12, 2026; close expected in H2 2026 after European Commission merger clearance. | The regulatory path is cleaner than it was in January. | The spread is now too narrow to beat the selected setup on upside per unit of inference. |
| 4 | Daikin has a live Japanese ASR, but the fresh price gap is thinner now | Japan industrial / buyback / capital allocation | Daikin last traded at JPY24,550 on May 8, 2026 after the company authorized a JPY350 billion fully committed share repurchase. | Official Daikin own-share acquisition release dated May 12, 2026 and delayed quote page crawled May 15, 2026. | Buyback execution started immediately after the board resolution. | The corporate bid is real and unusually large. | The buyback is now obvious, the stock already moved, and the rerating case is more patient than the selected setup. |
Selected opportunity: Baidu still trades legacy search, not the return stack.
Why this one now: The Q1 print is three days away, the shareholder-return policy is newly explicit, and the balance-sheet math is far more supportive than the current market narrative implies.
What should surprise the reader: Baidu ended 2025 with $42.06 billion of total cash and investments against a current market cap of $40.21 billion. That gross cash is not distributable one-for-one because the company also carried $22.80 billion of total liabilities and still needs capital for the business. Even after that caveat, the market is still valuing a company with a $5.0 billion buyback, an inaugural dividend policy, and an AI-powered business that already accounts for 43% of Baidu General Business revenue as if it were mainly a tired search franchise.
The Setup
Baidu has become easy to summarize lazily.
The bearish shorthand goes like this: China macro is messy, legacy search is ex-growth, AI costs money, Apollo Go is interesting but still peripheral, and governance always deserves a discount. That shorthand is not invented. It is incomplete.
Fact: on February 4, 2026, Baidu approved a new $5.0 billion buyback and its first-ever dividend policy. The board said the first dividend is expected to be declared in 2026. Baidu buyback and dividend policy, February 4, 2026
Fact: on February 26, 2026, management disclosed that Baidu Core AI-powered Business revenue reached RMB11.3 billion in Q4 2025 and RMB40.0 billion in full-year 2025, up 48% year over year. AI Cloud Infra alone reached RMB19.8 billion in 2025 revenue, up 34% year over year. Baidu fourth quarter and fiscal year 2025 results, February 26, 2026
Fact: the same results release said total cash and investments were RMB294.1 billion, or $42.06 billion, at December 31, 2025. Baidu fourth quarter and fiscal year 2025 results, February 26, 2026
Inference: the market is still putting more weight on the old category than on the emerging one. The price still looks anchored to legacy search and China discounting, not to a balance sheet that can fund both AI transition and shareholder returns.
The Market Price
| Market Level | Current Reading | Source / Timestamp | Why It Matters |
|---|---|---|---|
BIDU ADS price |
$143.29 | OpenAI finance tool, latest trade at 08:16 Singapore time on May 15, 2026 | Current entry reference. |
| Market capitalization | $40.21 billion | OpenAI finance tool, same timestamp | Lets us size the new buyback and compare the equity value against the balance sheet. |
| New repurchase authorization | $5.0 billion | Baidu buyback and dividend-policy release, February 4, 2026 | Equal to roughly 12.4% of current market cap. |
| Dividend policy | First-ever dividend policy, with first payment expected in 2026 | Baidu buyback and dividend-policy release, February 4, 2026 | Adds a second capital-return leg beyond buybacks. |
| Total cash and investments | RMB294.1 billion / $42.06 billion | Baidu fiscal-year 2025 results, December 31, 2025 balance-sheet date | Gross cash and investments are larger than the current market cap. |
| Total liabilities | RMB144.2 billion / $22.80 billion | Baidu fiscal-year 2025 results, December 31, 2025 balance-sheet date | The gross-cash headline needs this liability check to stay honest. |
| FY2025 non-GAAP net income attributable to Baidu | RMB18.9 billion / $2.71 billion | Baidu fiscal-year 2025 results, February 26, 2026 | Shows the business remains materially profitable even during the transition. |
| FY2025 non-GAAP diluted EPS | RMB53.41 / $7.64 per ADS | Baidu fiscal-year 2025 results, February 26, 2026 | Gives a direct earning-power anchor for the ADR. |
| Q4 2025 AI-powered Business revenue | RMB11.3 billion | Baidu fiscal-year 2025 results, February 26, 2026 | AI transition is already visible in reported revenue. |
| FY2025 AI-powered Business revenue | RMB40.0 billion, up 48% YoY | Baidu fiscal-year 2025 results, February 26, 2026 | This is not a concept slide. It is a real revenue line. |
| FY2025 AI Cloud Infra revenue | RMB19.8 billion, up 34% YoY | Baidu fiscal-year 2025 results, February 26, 2026 | The cloud piece is scaling faster than the legacy narrative implies. |
| Apollo Go Q4 2025 fully driverless rides | 3.4 million | Baidu fiscal-year 2025 results, February 26, 2026 | Reinforces that the optionality is not just in language models. |
| Apollo Go cumulative rides as of February 2026 | More than 20 million | Baidu fiscal-year 2025 results, February 26, 2026 | Shows real operating scale in a business the market still tends to dismiss. |
| Q1 2026 results date | May 18, 2026 before the U.S. market opens | Baidu Q1 2026 earnings-date release, April 23, 2026 | Near-dated catalyst. |
The Positioning
The cleanest positioning evidence here is corporate, not speculative.
Baidu itself is the bid.
The board authorized $5.0 billion of repurchases and adopted a dividend policy for the first time. That does not prove the stock is cheap. It does prove management now wants the market to think about shareholder returns in a more disciplined way than it did before. Baidu buyback and dividend policy, February 4, 2026
I do not have a clean live read on short interest, borrow cost, or options skew in BIDU for this run. I also do not have a reliable real-time view of mainland or southbound positioning. This is not a squeeze thesis. It is a balance-sheet, capital-return, and category-reclassification thesis.
The Catalyst
Three catalysts matter.
First, Baidu reports Q1 2026 results before the U.S. open on May 18, 2026. That is close enough to matter now, and it is the first hard test of whether AI-powered Business and AI Cloud Infra kept their momentum into 2026. Baidu Q1 2026 earnings date, April 23, 2026
Second, the first dividend declaration is expected later in 2026. The policy is not yet a cash payment, but once the board moves from framework to amount, the market loses one excuse for treating the return language as ornamental. Baidu buyback and dividend policy, February 4, 2026
Third, Kunlunxin's spin-off and separate listing are still described by management as progressing. That is not the main thesis, but it is additional value-unlock optionality inside a stock the market still tends to summarize too crudely. Baidu fourth quarter and fiscal year 2025 results, February 26, 2026
The Gap
The market appears to be pricing Baidu mainly as a discounted China internet incumbent with expensive AI ambition.
That frame misses two things.
First, the business mix is changing faster than the stock label. Baidu's AI-powered Business already accounted for 43% of Baidu General Business revenue in Q4 2025, and AI Cloud Infra subscription-based revenue from AI accelerator infrastructure rose 143% year over year in the quarter. Baidu fourth quarter and fiscal year 2025 results, February 26, 2026
Second, the capital-return math is now too large to treat as background noise. A $5.0 billion authorization is about 12.4% of the current market cap. The first dividend policy gives the board another instrument to surface value if the stock stays misread. Baidu buyback and dividend policy, February 4, 2026
The most useful way to express the valuation disagreement is not with gross cash alone. Gross cash is lazy. A cleaner bridge is this:
- current market cap is about $40.21 billion;
- total cash and investments were $42.06 billion at year-end 2025;
- total liabilities were $22.80 billion and redeemable noncontrolling interests were $1.88 billion at the same date;
- FY2025 non-GAAP net income attributable to Baidu was $2.71 billion.
Inference: using the latest reported balance sheet, the market is valuing Baidu's operating equity at roughly $22.8 billion, or about 8.4x FY2025 non-GAAP earnings. That is not a proof of cheapness. It is a much lower multiple than the current narrative tone implies.
The Payoff Map
One possible expression is long BIDU common ADRs.
That is the cleanest wrapper because the thesis needs both the balance sheet and the Q1 catalyst. Common stock captures any rerating from better-than-feared Q1 evidence, any execution of the buyback, and any later dividend declaration without forcing a tight expiry view.
An alternative expression is the Hong Kong line, 9888, for investors who prefer local-market execution. I did not verify a live options chain during this run, so options are secondary rather than primary.
Price Target and Probability Map
| Scenario | Probability | Target / Level | Return / Payoff | Time Horizon | Conditions Required | Evidence Quality |
|---|---|---|---|---|---|---|
| Top Case | 30% | BIDU $178.00 |
+24.2% | 3 to 9 months | Q1 confirms AI-powered Business momentum, the board moves toward a first dividend, buybacks become more visible, and the market begins to value the operating business above a single-digit ex-net-cash earnings multiple. | Medium |
| Base Case | 50% | BIDU $164.00 |
+14.5% | 3 to 9 months | Q1 is merely solid rather than spectacular, the return stack stays credible, and the market gives modest credit to the balance sheet and AI mix shift. | Medium / High |
| Bottom Case | 20% | BIDU $120.00 |
-16.3% | 3 to 9 months | Q1 disappoints, AI monetization progress looks slower, governance and China risk take over the tape again, and management offers little concrete follow-through on returns. | Medium |
| Invalidation / Stop Condition | n/a | Sustained break below BIDU $120.00 |
n/a | n/a | The thesis is broken if the operating transition stalls, cash flow weakens, and the board fails to convert return policy into observable capital return. | Medium |
Probability-weighted expected value: approximately +11.3%, using the scenario returns above. This excludes any upside from a faster-than-expected Kunlunxin value unlock or a larger-than-expected dividend.
Current market price / level: BIDU $143.29.
Timestamp: 08:16 Singapore time on May 15, 2026.
Primary instrument: BIDU ADSs.
Alternative expressions considered: Hong Kong-listed 9888, or long-dated call spreads only after a live options-chain check.
Confidence: Medium.
What Could Go Wrong
The strongest counterargument is that the market is not missing the return stack. It is discounting a China platform company whose AI transition is still expensive, politically exposed, and hard to trust on a Western multiple.
That counterargument has force.
Baidu's full-year 2025 total revenue still fell 3% year over year, and GAAP operating loss for the year was RMB5.8 billion because the company also recognized a RMB16.2 billion long-lived-asset impairment. The transition is real, but it is not clean. Baidu fourth quarter and fiscal year 2025 results, February 26, 2026
There is also governance risk. The dividend policy is only a policy until the board states an amount. The buyback is authorized until December 31, 2028, which means the timing of actual execution still matters. Baidu buyback and dividend policy, February 4, 2026
Finally, Apollo Go and AI Cloud Infra can keep scaling operationally without immediately closing the valuation gap if investors keep assigning a structural China discount to the whole asset.
What Would Prove This Wrong
This thesis fails if the operating transition stops becoming more legible.
It is wrong if one or more of the following happens:
- Q1 shows AI-powered Business momentum flattening or AI Cloud Infra slowing sharply;
- operating cash flow turns decisively weaker again instead of stabilizing;
- management declines to provide any credible first-dividend path after creating the policy;
- or the stock cannot hold above $120.00 even after a clean Q1 and visible repurchase activity.
That combination would tell you the market is not underpricing a return stack. It is rationally discounting a harder transition.
Bottom Line
Baidu is no longer just a search-and-policy stock, even if the tape still treats it that way. The board has already put a $5.0 billion buyback and a first-ever dividend policy on the record. The AI-powered business has already reached RMB40.0 billion of annual revenue. The balance sheet is far heavier than the current equity label implies. With Q1 results due on May 18, 2026, the market is about to test whether it still wants to price Baidu as legacy search first and everything else second.
Best trade strategy: Long BIDU common ADRs. Options are secondary only if a later live chain check shows liquid long-dated structures at acceptable prices.
Sources
- Baidu announces new share repurchase program and dividend policy, February 4, 2026
- Baidu announces fourth quarter and fiscal year 2025 results, February 26, 2026
- Baidu to report first-quarter 2026 financial results on May 18, 2026, April 23, 2026
- GlobalFoundries announces launch of public secondary offering and concurrent share repurchase, March 11, 2026
- GlobalFoundries reports first quarter 2026 financial results, May 5, 2026
- GlobalFoundries outlines long-term growth roadmap and announces first-ever dividend at 2026 Investor Day, May 7, 2026
- Fnac Darty: opening on May 12 of the takeover bid initiated by EP Group, May 11, 2026
- Daikin notice regarding the acquisition of own shares, May 12, 2026
- StockAnalysis delayed quote page for Daikin Industries
- StockAnalysis delayed historical quote page for Fnac Darty
- Live market snapshots checked during this run via the OpenAI finance tool for
BIDUandGFS