Baidu: Still Cheap, But Bull Case Is Getting Harder To Defend (Rating Downgrade)

Baidu (BIDU) is downgraded to a Buy as weakness in its core advertising business and margin declines temper prior bullishness. While AI-native marketing and cloud infrastructure are growing quickly, legacy ad revenue declines are outpacing these gains, leaving sustainability concerns. Apollo Go’s international expansion and ride growth look promising but are not yet large enough to offset the core advertising slowdown.

Analysis

The market is at risk of overpricing AI optionality relative to the actual earnings engine. The core issue is mix: incremental cloud/AI revenue may look high-growth, but it does not offset the margin drag from losing higher-quality ad dollars, so headline growth can coexist with weaker EPS power. If management uses AI-native marketing as a defense mechanism, that is more likely to preserve share than to create net-new demand, which limits upside to valuation expansion.

Second-order, the real beneficiaries are competitors with stronger consumer traffic and better conversion loops, because ad budgets tend to reallocate toward platforms that can prove ROI fastest. Apollo Go is a longer-dated call option, but international scaling introduces licensing, fleet utilization, and capex intensity before it becomes a meaningful profit contributor. That means the near-term narrative can be bullish while the underlying free-cash-flow profile quietly deteriorates.

This is a stock where days and months matter differently: in the next few sessions, AI headlines can still drive reflexive buying; over 1-3 months, ad trend and margin commentary should dominate; over 6-18 months, the question is whether the new businesses can replace the cash-generation role of the legacy franchise. The thesis breaks if core ads stabilize and operating margin inflects higher, or if Apollo Go proves it can scale abroad without heavy subsidy.

Key Decisions for Investors

• Short BIDU on any 5-10% relief rally; use 1-3 month puts if implied vol is not already inflated. Best entry is ahead of earnings/guidance, when the market is most exposed to disappointment on ad trends.
• Pair trade: short BIDU / long TCEHY or BABA to isolate company-specific margin deterioration versus broader China internet beta. The relative trade should work if the market rotates back to fundamentals rather than AI narrative.
• Set a falsification alert on two consecutive quarters of ad revenue stabilization and operating margin expansion; cover shorts if management shows the mix is improving instead of simply growing lower-quality revenue.
• Treat Apollo Go as a long-dated option, not a core valuation driver. Do not pay for the international robotaxi story unless there is evidence of regulatory approvals and unit economics that reduce, rather than increase, capital intensity.