Close 18 Aug 2026
$90.87
Baidu enters the 10–30 day window off its first earnings miss in eight quarters, with online-marketing revenue down 19% and the stock gapped 13% lower to below book value. The AI-cloud engine is real — infrastructure revenue up 50%, GPU cloud up 283% — but it is lower-margin and cash-hungry, and the next print is not until November. Over the next 10–30 days the setup is deep value against negative momentum: an 88.9% idiosyncratic name with a below-book floor but no catalyst to force a re-rate, where the only dated event is a procedural shareholder vote on 26 August.
| Dimension | Finding | Signal |
|---|---|---|
| Revenue growth | TTM revenue −4.2% YoY, a fifth straight quarterly decline; online-marketing (ad) revenue −19% YoY to ¥13.1B. | Bearish |
| Profitability | Gross margin 39.0% versus 51% two years ago; TTM net loss; ROE collapsed from 8.9% to 0.3%. | Bearish |
| Valuation vs peers | P/S 1.64 and P/B 0.79 — the cheapest of the China-internet set, trading below book value. | Cheap |
| Platform KPIs | AI Cloud infra +50%, GPU cloud +283% (fourth straight triple-digit quarter); Apollo Go 23M+ rides, 28 cities. | Bullish |
| Balance sheet | ¥166B ($24.5B) cash & investments, ~$7.9B net cash, debt/equity 0.38, current ratio 1.85. | Bullish |
| Regime state | Quiet Drift (Q4), on the Q3/Q4 boundary; Persistency +0.03 (neutral), Volatility −0.31 (subdued) — but the reading is as of 18 Aug, before the gap. | Neutral |
| Driver exposure | 88.9% idiosyncratic; correlation to the primary market driver is ~0. Company news, not the index, sets the month. | Neutral |
| Key risk in window | Post-miss estimate cuts and continued advertising deterioration, with no offsetting catalyst before November. | Bearish |
| Catalysts in window | Extraordinary general meeting 26 Aug (Hong Kong dual-primary listing approval); next earnings not until mid-November. | Mixed |
| Price action | −12.7% on the Q2 print (8× volume); −18.9% in 30 days; −45% from the January high. | Bearish |
| Overall view (10–30 days) | Deep-value floor versus negative momentum and a catalyst vacuum — high two-way variance, tilted cautious. | Mixed |
Signal reflects the 10–30 day window only. Row tint matches the badge.
Baidu is genuinely cheap — below book, a quarter of its market cap in net cash, and a fast-growing AI-cloud business the market is not paying for. But the thing that broke this quarter, the advertising engine, breaks slowly and on its own schedule, and there is no earnings print to re-set the narrative until November. For the next 10–30 days the value floor limits how far it falls; it does not tell you what turns it around.
Each case states what would have to happen inside the next 10–30 days for it to play out. This section argues; the verdict is section 02.
Green at 8 and above, yellow 6–8, red below 6. The composite row states its weighting. A 5.2 composite says the growth and profitability deterioration outweigh the valuation support for a one-month view — the below-book floor and idiosyncratic volatility are what keep this Mixed rather than outright bearish.
Ordered by what can bite inside the next 10–30 days, not by ultimate severity. Structural risks that cannot resolve inside the window are marked as context.
| Risk | Category | Mechanism | Bites in window? |
|---|---|---|---|
| Estimate revisions | Sentiment | Consensus re-priced down after a 21% EPS miss; targets lag the print. | Likely |
| Advertising decline | Competitive | Online marketing −19% YoY; guided weak into H2. | Possible |
| China macro / consumption | Macro | Ad budgets track consumer demand and stimulus headlines. | Possible |
| RMB / FX swing | Macro | FX and fair-value moves already dented Q2 net income. | Possible |
| ADR delisting / HFCAA | Governance | Forced-delisting tail; slow-moving, policy-driven. | Unlikely |
The risks that can actually move the stock in the next month are all pointed the same way — falling estimates, a weakening ad line, and broken momentum — while the offsetting arguments (deep value, AI growth, robotaxi) are medium-term stories the window will not adjudicate. That asymmetry is why the near-term read is cautious even though the long-term valuation is attractive.
| Quarter | Report date | Revenue | vs est. | EPS (non-GAAP) | vs est. | Reaction |
|---|---|---|---|---|---|---|
| Q2 2026 | 18 Aug 2026 | ¥31.3B | −0.6% | $1.06 | −21% | −12.7% |
| Q1 2026 | 18 May 2026 | ¥32.1B | ~in-line | $1.55 | +2.6% | +1.8% |
| Q4 2025 | 26 Feb 2026 | ¥32.7B | ~in-line | $1.24 | +10.7% | −5.7% |
| Q3 2025 | 18 Nov 2025 | ¥31.2B | ~in-line | $1.23 | +35% | +2.7% |
| Q2 2025 | 20 Aug 2025 | ¥32.7B | ~in-line | $1.49 | +12.9% | −2.6% |
| Q1 2025 | 21 May 2025 | ¥32.5B | ~in-line | $2.27 | +69% | −4.3% |
EPS is non-GAAP diluted per ADS in US$; “vs est.” on EPS from consensus, revenue estimate available only for the latest quarter. Reaction is the report-day close versus the prior close (Baidu reports pre-market). Minus signs are − (U+2212).
| Company | Rev beat rate | EPS beat rate | Guidance |
|---|---|---|---|
| BIDU | ~40% | 8 of 9 | No formal guide |
| Peer median | ~60% | ~7 of 8 | — |
| Item | Value | Comment |
|---|---|---|
| Next report date | mid-Nov 2026 | Outside the 10–30 day window |
| Formal revenue guide | None | Baidu does not guide; flagged ads “under pressure in H2” |
| EGM (HK listing) | 26 Aug 2026 | Inside the window — procedural, not an earnings mover |
The one number that matters here is the calendar: the next earnings print lands in mid-November, outside the window. That means the 18 August miss — the first in eight quarters — is the last hard data the market gets for the whole horizon, and the stock has to trade on estimate revisions and narrative rather than a fresh result. Note the pattern: Baidu sold off on both of its last two beats, so “in line” has not been good enough.
| Period | Source | View | Key point |
|---|---|---|---|
| Aug 2026 | Consensus (32) | Buy | $167 average target — but largely struck before the 18 Aug miss. |
| Jul 2026 | J.P. Morgan | Buy | $205 — AI cloud and robotaxi optionality; top of the range. |
| Jul 2026 | Barclays | Hold | Cut target on advertising-revenue decline; the cautious camp. |
| Jun 2026 | Citi | Buy | $188 — constructive on the AI transition. |
| May 2026 | Susquehanna | Hold | $140 — monetisation concerns on core search. |
Coverage skews Buy on a 12-month horizon, but almost every rating and target here predates the 18 Aug print. Expect the distribution to compress downward as analysts mark to the miss; treat these as stale for the 10–30 day window.
| Date | Insider / actor | Transaction | Shares | Price | Value | Signal read |
|---|---|---|---|---|---|---|
| Ongoing | Baidu, Inc. (company) | Share repurchase program | n/d | — | n/d | Buyback active under a multi-year authorisation — a modest, standing support. |
| Standing | Robin Li (Chair/CEO) | Super-voting control | n/d | — | n/d | Dual-class structure concentrates voting control; no window-specific transaction. |
As a foreign private issuer, Baidu's insiders are exempt from US Section 16 reporting, so routine open-market Form 4 insider transaction data is not available. No individual insider buys or sells are disclosed for the window; the only company-level signal is the standing repurchase program. Share and value fields are not disclosed (n/d).
| Date | Source | Development | In window? |
|---|---|---|---|
| 18 Aug 2026 | Baidu IR | Q2: revenue −4% to ¥31.3B; online marketing −19%; AI Cloud infra +50%, GPU cloud +283%; non-GAAP EPS $1.06 missed $1.35. | Yes |
| 18 Aug 2026 | Benzinga / Quartz | Stock −13%; “advertising cash cow is losing steam” as the AI pivot is put to the test. | Yes |
| 18 Aug 2026 | Earnings call | Apollo Go: 23M+ cumulative rides, 28 cities; driverless testing in Hong Kong, commercial ops in Dubai. | Yes |
| 18 Aug 2026 | Baidu IR | EGM set for 26 Aug to approve the Hong Kong dual-primary listing conversion (expected effective within 2026). | Yes |
| Jul 2026 | Barclays | Price target cut on advertising-revenue decline ahead of the print. | No |
| Jul 2026 | J.P. Morgan / Citi | Buy ratings reiterated at $205 / $188 on AI-cloud and robotaxi optionality. | No |
| Pre-Q2 | IBTimes | Shares had advanced ~4% into the print on AI momentum and dual-listing progress. | No |
Newest first. The final column states whether the event falls inside the next 10–30 days.
Only one dated event falls in the next 10–30 days: the 26 August EGM to approve the Hong Kong dual-primary listing conversion — a liquidity and index-access positive, but procedural and already expected, not an earnings mover. The dual-primary listing itself is expected to become effective within 2026, with timing that may or may not land in the window. There is no earnings print until mid-November. The tape for the horizon will be set by post-earnings drift and analyst revisions, not by a fresh fundamental catalyst.
| Company | Price | Market cap | TTM revenue | P/S TTM | Rev growth | Source view |
|---|---|---|---|---|---|---|
| BIDU | $90.87 | $30.8B | $18.7B | 1.64 | −4.2% | Mixed |
| Alibaba (BABA) | $128.15 | $292.3B | $148.4B | 1.97 | +2.7% | Bullish |
| Tencent (TCEHY) | $56.39 | $506.3B | $116.1B | 4.36 | +11%e | Bullish |
| PDD Holdings (PDD) | $87.27 | $124.2B | $64.1B | 1.94 | +11% | Neutral |
| Peer median | — | $208.3B | $116.1B | 1.97 | +11% | — |
Peer figures from data-provider snapshots; Tencent via its OTC ADR (TCEHY). Baidu is the only name in the set trading below 2× sales and the only one with shrinking revenue — cheap for a reason. Tencent revenue growth marked e is estimated.
| Analyst / source | Current target | Previous | Date | Implied return | Rating | Direction |
|---|---|---|---|---|---|---|
| Barclays | $124 | higher | 13 Jul 2026 | +36.5% | Hold | ▼ Lowered |
| J.P. Morgan | $205 | — | 22 Jul 2026 | +125.6% | Buy | ▶ Maintained |
| Bank of America | $165 | — | 14 Jul 2026 | +81.6% | Buy | ▶ Maintained |
| Citi | $188 | — | 23 Jun 2026 | +106.9% | Buy | ▶ Maintained |
| Susquehanna | $140 | — | 20 May 2026 | +54.1% | Hold | ▶ Maintained |
| Bernstein | n/d | Outperform | 2025 | — | Mkt Perform | ▼ Lowered |
Newest first. Direction is ▲ raised, ▶ maintained, ▼ lowered. Every action here predates the 18 Aug miss, so the “maintained” targets are effectively stale and skewed to be cut.
| Metric | Value |
|---|---|
| Last close | $90.87 |
| Consensus target | $167.09 |
| Median target | $171.34 |
| High target | $215.16 |
| Low target | $92.07 |
| Implied upside to consensus | +83.9% |
| Implied return to low target | +1.3% |
| Analysts contributing | 32 |
The whole target distribution sits above spot — even the lowest target ($92.07) is only ~1% above the close — which means the consensus has far more room to fall toward the price than the price has to fall toward the low. That is the signature of a stale, pre-miss target set.
| Quarter | Revenue | QoQ Δ | YoY Δ | EPS (GAAP dil.) | Gross margin | Adj. EBITDA | Margin | vs est. | Reaction |
|---|---|---|---|---|---|---|---|---|---|
| Q3 2024 | ¥33.6B | — | −2.6% | ¥21.60 | 51.1% | ~¥9.0B | ~27% | Beat | −5.9% |
| Q4 2024 | ¥34.1B | +1.5% | −2.4% | ¥13.27 | 47.2% | ~¥8.4B | ~25% | Beat | −7.5% |
| Q1 2025 | ¥32.5B | −4.7% | +3.0% | ¥21.60 | 46.1% | ~¥7.8B | ~24% | Beat | −4.3% |
| Q2 2025 | ¥32.7B | +0.8% | −3.6% | ¥20.32 | 43.9% | ~¥7.0B | ~21% | Beat | −2.6% |
| Q3 2025 | ¥31.2B | −4.7% | −7.1% | −¥33.84 | 41.3% | ~¥5.5B | ~18% | Beat* | +2.7% |
| Q4 2025 | ¥32.7B | +5.0% | −4.1% | ¥2.52 | 44.2% | ~¥5.8B | ~18% | Beat | −5.7% |
| Q1 2026 | ¥32.1B | −2.0% | −1.2% | ¥8.80 | 38.9% | ~¥6.0B | ~19% | Beat | +1.8% |
| Q2 2026 | ¥31.3B | −2.3% | −4.2% | ¥5.76 | 39.0% | ¥6.2B | 20% | Miss | −12.7% |
EPS is GAAP diluted per ADS in RMB; the Q3'25 loss reflects a large non-operating write-down (non-GAAP EPS was a beat, marked *). Adjusted-EBITDA figures marked ~ are derived from reported margins. The story down the gross-margin column — 51% to 39% in two years — is the whole thesis in one line.
Two series sharing an x axis but not a scale get two stacked bands, never one. Revenue is broadly flat-to-down; the margin line is the signal.
| Metric | 30 Jun 2026 | 31 Dec 2025 | 31 Dec 2024 | Target |
|---|---|---|---|---|
| Current ratio | 1.85 | 1.76 | 2.09 | 1.5–3.0 healthy |
| Quick ratio | 1.57 | 1.53 | 1.75 | ≥1.0 healthy |
| Cash ratioe | 1.87 | 1.34 | 1.57 | Incl. short-term investments |
Column headers are actual reporting dates. Cash ratioe is derived as cash & short-term investments over current liabilities. Liquidity is comfortable and improving on a quick-ratio basis; this is not a balance-sheet-stress story.
| Metric | 30 Jun 2026 | 31 Dec 2025 | 31 Dec 2024 | Target |
|---|---|---|---|---|
| Debt-to-equitye | 0.38 | 0.34 | 0.28 | Lower is safer |
| Debt-to-assetse | 0.24 | 0.22 | 0.19 | <0.5 conservative |
| Net cash position | ¥53B | ¥18B | ¥48B | Cash & ST inv. less total debt |
| Interest coverage | n/m | n/m | n/m | Net creditor — interest income exceeds expense |
Leverage is low and the company is a net creditor, so solvency is not a window risk. Debt has crept up (¥79B to ¥113B) to fund the AI build-out, but it is dwarfed by ¥166B of cash & investments. Ratios marked e are derived from the reported balance sheet.
| Metric | Q2 2026 | TTM | Q2 2025 | FY2025 | FY2024 | Trend |
|---|---|---|---|---|---|---|
| Gross margin | 39.0% | 40.8% | 43.9% | 43.8% | 48.0% | ▼ |
| Operating margin | 9.7% | 6.9% | 10.0% | 8.0% | 14.0% | ▼ |
| Net margin | 7.4% | −3.6% | 22.4% | 3.6% | 14.5% | ▼ |
| Adj. EBITDA margin | 20% | ~19% | ~21% | ~18% | ~24% | ▼ |
| Return on assets | — | 1.25% | — | 1.48% | 3.19% | ▼ |
| Return on equity | — | 0.32% | — | 1.90% | 8.85% | ▼ |
| DuPont (NPM × AT × EM) | — | −3.6 × 0.28 × 1.59 | — | — | — | → |
| Peer comparison (most recent) | Gross | Op margin | Net margin | Adj. EBITDA | ROE | Rank |
|---|---|---|---|---|---|---|
| BIDU | 39.0% | 9.7% | 7.4% | 20% | 0.3% | 4 of 4 |
| Alibaba | 39.8% | 6.2% | 10.4% | ~17% | 9.2% | 3 of 4 |
| PDD | 56.0% | 21.8% | 21.6% | ~26% | 25.4% | 1 of 4 |
| Tencent | 56.7% | 33.3% | 29.9% | ~40% | ~20% | 2 of 4 |
| Peer median | 56.0% | 21.8% | 21.6% | ~26% | ~20% | — |
Net margin uses Q2'26 for BIDU; TTM is a loss on the Q3'25 write-down. Baidu ranks last of four on profitability — its gross margin now sits near Alibaba's and far below PDD/Tencent's 56%+. ROE marked ~ for Tencent is estimated.
| Metric | Current / TTM | Prior year |
|---|---|---|
| Asset turnover | 0.28 | 0.32 |
| EPS growth (non-GAAP YoY) | −29% | +~5% |
| Dividend yield | ~0% | ~0% |
| Platform metric | Q2 2026 | YoY |
|---|---|---|
| Baidu App MAU | 644M | — |
| AI Cloud infra revenue | ¥7.3B | +50% |
| GPU cloud revenue | — | +283% |
| Apollo Go rides (cum.) | 23M+ | rising |
| Company | Asset turnover | Rev / employee | Employees | EPS growth | Div yield | Rank |
|---|---|---|---|---|---|---|
| BIDU | 0.28 | $559K | 33,500 | −29% | ~0% | 4 of 4 |
| Alibaba | ~0.5e | $1,129K | 131,462 | + | low | 3 of 4 |
| PDD | ~0.7e | $2,518K | 25,474 | + | 0% | 1 of 4 |
| Tencent | ~0.4e | $1,001K | 115,927 | + | low | 2 of 4 |
Baidu generates the lowest revenue per employee in the set ($559K vs PDD's $2.5M) and the only negative EPS growth — it is the least efficient and the only one going backwards. Peer asset-turnover figures marked e are estimated.
| Metric | Current | Comment |
|---|---|---|
| P/E TTM | n/a | TTM loss; forward P/E 13.5 |
| Price / book | 0.79 | Below book value |
| Price / sales TTM | 1.64 | Cheapest of the peer set |
| EV / EBITDA | 6.7 | Net-cash-adjusted, low |
| PEG | 2.67 | Growth-adjusted — not cheap on forward growth |
| Peer | P/S TTM | P/E TTM | Fwd P/E |
|---|---|---|---|
| BIDU | 1.64 | n/a | 13.5 |
| Alibaba | 1.97 | 20.1 | — |
| PDD | 1.94 | 9.3 | — |
| Tencent | 4.36 | 14.6 | — |
| Peer median | 1.97 | 14.6 | — |
The fundamentals give the price a hard floor — below book, net cash worth a quarter of the market cap, an EV of just 6.7× EBITDA — but no engine for the next month. On a PEG of 2.7 the stock is not actually cheap against its own (falling) growth, which is why the multiple can stay compressed until margins and the ad line inflect. Over 10–30 days, valuation limits the downside; it does not create the upside.
Persistency on x, Volatility on y, oldest faint to newest bright. Q1 volatile trend, Q2 volatile chop, Q3 quiet range, Q4 quiet drift. Source: Trader workbook, Individual regimes daily. As of 18 Aug 2026 — 1 trading day behind the report date.
| Measure | Current | Mean | Std dev | Min | Max | Percentile | Interpretation |
|---|---|---|---|---|---|---|---|
| Persistency | +0.03 | +0.10 | 0.05 | −0.02 | +0.18 | 12th | Random / Neutral — no exploitable trend structure either way. |
| Volatility | −0.31 | +0.00 | 0.29 | −0.50 | +0.50 | 19th | Subdued Vol — but this reading is as of 18 Aug and does not yet absorb the 13% earnings gap. |
Currently Q4 Quiet Drift, held 18 consecutive periods; Persistency sits within 0.05 of the axis, so the call between Quiet Range (Q3) and Quiet Drift (Q4) is provisional. Correlations, exponents and percentiles are never coloured. Critically, the subdued Volatility reading pre-dates the earnings gap — realised volatility will run well above this over the window.
| Quadrant | Label | % of period |
|---|---|---|
| Q1 | Volatile trend | 46.9 |
| Q2 | Volatile chop | 3.4 |
| Q3 | Quiet range | 2.4 |
| Q4 | Quiet drift | 47.2 |
| Transition | Count | Note |
|---|---|---|
| Q4 → Q1 | 5 | Drift breaking into volatile trend |
| Q1 → Q4 | 5 | Trend settling back to drift |
| Q4 → Q3 / Q3 → Q4 | 2 / 2 | Minor drift/range shuffles |
| Q1 → Q2 / Q2 → Q1 | 1 / 1 | Rare chop episodes |
Baidu has spent ~94% of the period split between Quiet Drift (47%) and Volatile Trend (47%) — it either grinds quietly or breaks out hard, rarely chopping in between. The Persistency series held one value for months at a time, producing the long vertical runs on the trace; that is the data behaving normally. With a fresh catalyst in hand, the historically common Q4→Q1 shift — drift breaking into volatile trend — is the pattern to watch. These statistics describe the stated window and are not predictions.
| Driver | Correlation | R² (%) | Rolling 60d | Rolling min | Rolling max | Stability | Direction |
|---|---|---|---|---|---|---|---|
| Market Driver 1 Primary | −0.017 | 0.0 | −0.097 | −0.415 | 0.350 | Variable | Neutral |
| Market Driver 2 | 0.088 | 0.8 | 0.072 | −0.212 | 0.421 | Variable | Positive |
| Market Driver 3 | 0.046 | 0.2 | −0.056 | −0.245 | 0.269 | Variable | Neutral |
| Market Driver 4 | −0.035 | 0.1 | −0.006 | −0.348 | 0.169 | Variable | Neutral |
| Market Driver 5 | −0.081 | 0.7 | 0.104 | −0.385 | 0.146 | Variable | Negative |
| Sector Driver 1 Primary | −0.333 | 11.1 | −0.385 | −0.686 | 0.061 | Variable | Negative |
| Sector Driver 2 | 0.090 | 0.8 | 0.331 | −0.212 | 0.341 | Variable | Positive |
| Sector Driver 3 | 0.045 | 0.2 | −0.132 | −0.250 | 0.351 | Variable | Neutral |
Daily log returns vs first-differenced Market Driver levels and return-scaled Sector Driver values, 499 observations. Correlations are never coloured. Drivers as of 17 Aug 2026 — 2 trading days behind the report date.
| Driver | Raw beta | Standardised beta | Share of explained variance |
|---|---|---|---|
| Market Driver 1 (primary) | 0.00003 | 0.02 | ~0% |
| Sector Driver 1 (primary) | −0.294 | −0.33 | ~100% |
Baidu is an idiosyncratic name, not a market vehicle. Its correlation to the primary market driver is essentially zero, so the broad index is close to irrelevant; the only systematic tilt is a modest negative link to the primary sector driver. What this means for the next month is simple: Baidu's own news — estimate revisions, ad trends, the EGM — decides the outcome, and an index hedge would not have offset the 13% earnings move.
Market Driver 1 (solid) and Sector Driver 1 (dashed) against Baidu's daily returns. The market-driver correlation oscillates around zero — positive in early 2025, deeply negative by late 2025, back near zero now — while the sector-driver correlation has been persistently negative and currently reads −0.385. Neither is stable; there is no reliable systematic anchor.
Baidu's own daily returns, conditioned on each group's regime over 20 Aug 2024 – 17 Aug 2026. Every group is named, never its proxy ticker. Conditional statistics describe the stated window and are not forecasts — and with 88.9% idiosyncratic variance, this axis is secondary to company news.
| Market regime | Days | % of period | Cumulative return | Ann. return | Ann. vol | Sharpe | Hit rate | Best day | Worst day |
|---|---|---|---|---|---|---|---|---|---|
| Quiet drift | 108 | 21.6 | +18.0% | +47.0% | 47.2 | 1.00 | 49.1 | +10.74 | −8.44 |
| Volatile chop | 186 | 37.3 | +9.8% | +13.5% | 45.7 | 0.29 | 56.5 | +8.70 | −8.56 |
| Volatile trend current | 87 | 17.4 | −1.6% | −4.5% | 52.1 | −0.09 | 47.1 | +10.77 | −10.26 |
| Quiet range | 118 | 23.6 | −9.4% | −19.0% | 45.4 | −0.42 | 47.5 | +14.00 | −7.81 |
| Market regime | Days | % of period | Cumulative return | Ann. return | Ann. vol | Sharpe | Hit rate | Best day | Worst day |
|---|---|---|---|---|---|---|---|---|---|
| Quiet drift | 68 | 13.6 | +29.6% | +161.7% | 49.0 | 3.30 | 52.9 | +10.74 | −8.44 |
| Volatile chop | 192 | 38.5 | +27.1% | +37.0% | 45.9 | 0.81 | 56.3 | +8.70 | −8.56 |
| Quiet range | 150 | 30.1 | −18.3% | −28.8% | 43.6 | −0.66 | 47.3 | +14.00 | −7.81 |
| Volatile trend current | 89 | 17.8 | −14.3% | −35.4% | 53.3 | −0.66 | 44.9 | +10.77 | −10.26 |
| Market regime | Days | % of period | Cumulative return | Ann. return | Ann. vol | Sharpe | Hit rate | Best day | Worst day |
|---|---|---|---|---|---|---|---|---|---|
| Volatile chop current | 234 | 46.9 | +13.9% | +15.1% | 46.7 | 0.32 | 54.7 | +8.70 | −8.56 |
| Quiet range | 232 | 46.5 | +12.4% | +13.5% | 45.1 | 0.30 | 49.1 | +14.00 | −8.44 |
| Volatile trend | 33 | 6.6 | −9.8% | −54.5% | 61.4 | −0.89 | 39.4 | +10.77 | −10.26 |
Best regime for Baidu shaded green, worst red; the current regime is badged. The current US and SP500 regimes (Volatile Trend) sit among Baidu's weaker large-cap buckets, while the current Global regime (Volatile Chop) is its best global bucket — the equity-regime read is mixed, not uniform.
| Group | Current regime | Best regime | Worst regime | Cum. return in current | Sharpe in current | Sharpe spread | Days in current |
|---|---|---|---|---|---|---|---|
| US market | Volatile trend | Quiet drift | Quiet range | −1.6% | −0.09 | 1.42 | 87 |
| SP500 | Volatile trend | Quiet drift | Volatile trend | −14.3% | −0.66 | 3.96 | 89 |
| Global market | Volatile chop | Volatile chop | Volatile trend | +13.9% | 0.32 | 1.21 | 234 |
| Technology | Quiet range | Quiet drift | Quiet range | −32.7% | −0.97 | 23.92 | 171 |
| Financials | Quiet drift | Quiet range | Quiet drift | −23.7% | −1.76 | 3.74 | 36 |
| Energy | Volatile trend | Quiet range | Volatile chop | −7.9% | −0.44 | 3.70 | 94 |
| Utilities | Volatile chop | Quiet range | Volatile trend | +2.3% | 0.10 | 2.53 | 127 |
| Europe | Volatile trend | Volatile chop | Quiet drift | +3.9% | 0.19 | 3.62 | 113 |
| Bonds near | Quiet range | Volatile chop | Volatile trend | −17.9% | −0.44 | 5.38 | 224 |
| Bonds mid | Quiet range | Volatile chop | Quiet drift | +3.9% | 0.08 | 1.03 | 258 |
| Bonds long | Quiet range | Volatile chop | Quiet drift | +6.6% | 0.10 | 4.70 | 346 |
| VIX Mid | Quiet drift | Volatile trend | Quiet drift | −29.4% | −1.09 | 10.27 | 110 |
| VIX Near | Quiet range | Quiet range | Quiet drift | +42.5% | 0.58 | 1.83 | 363 |
| Gold | Volatile chop | Quiet range | Volatile chop | −2.5% | −0.03 | 1.42 | 398 |
All 14 groups. “Cum. return in current” is Baidu's cumulative return during the group's current regime over the period; Sharpe values are performance ratios, not coloured.
Regime-conditional history describes 20 Aug 2024 – 17 Aug 2026, not the future. Buckets under 30 days are not shown; some annualised figures (notably Technology's Quiet-Drift bucket) rest on small samples and should not be relied on. Market-regime series as of 17 Aug 2026 — 2 trading days behind the report date.
| Date | Headline | Sentiment |
|---|---|---|
| 18 Aug 26 | Q2 miss: revenue −4%, ad revenue −19%, non-GAAP EPS $1.06 vs $1.35 expected | Negative |
| 18 Aug 26 | Baidu sinks 13% as soft results put the AI pivot to the test | Negative |
| 18 Aug 26 | Management: advertising to remain “under pressure in the second half” | Negative |
| 18 Aug 26 | AI Cloud infra +50%, GPU cloud +283%; AI business crosses half of Baidu Core | Positive |
| 18 Aug 26 | Apollo Go passes 23M rides; driverless in Hong Kong, commercial ops in Dubai | Positive |
| 18 Aug 26 | Operating cash flow positive a fourth straight quarter (¥3.4B) | Positive |
| 18 Aug 26 | EGM set for 26 Aug to approve Hong Kong dual-primary listing conversion | Positive |
| Jul 26 | Barclays cuts price target on advertising-revenue decline | Negative |
| Jul 26 | J.P. Morgan reiterates Buy, $205 target on AI-cloud / robotaxi optionality | Positive |
| Jun 26 | Citi maintains Buy, $188, constructive on the AI transition | Positive |
| May 26 | Q1 2026 tops estimates ($1.55 vs $1.51) on AI momentum | Positive |
| Pre-Q2 | Shares advanced ~4% into the print on AI momentum and dual-listing progress | Neutral |
Newest first. The narrative flipped hard on 18 Aug: a strong AI/robotaxi story overwhelmed by an advertising miss and a 13% drop.
| Field | Baidu, Inc. | Peer context |
|---|---|---|
| Legal name | Baidu, Inc. | — |
| Exchange / IPO | NASDAQ: BIDU · IPO Aug 2005 | Secondary HKEX 9888, converting to dual-primary |
| Domicile | Cayman Islands (inc.); HQ Beijing, China | China ADR, VIE structure |
| Sector / industry | Communication Services / Interactive Media & Services | — |
| Market cap | $30.8B | Alibaba $292B · Tencent $506B · PDD $124B |
| Employees | 33,500 | Alibaba 131K · Tencent 116K · PDD 25K |
| TTM revenue | ¥127.3B / $18.7B | Alibaba $148B · Tencent $116B |
| Revenue model | Search advertising, AI Cloud, autonomous driving (Apollo Go), iQIYI streaming | Ad + cloud + AI |
| Key differentiators | China's leading search engine; ERNIE LLM; full-stack AI (chips to apps); Apollo Go robotaxi | Only peer with a scaled robotaxi |
| CIK | 0001329099 | — |
| Website | baidu.com | — |
The evidence over the next 10–30 days tilts cautious: a shrinking, high-margin advertising business, a fresh 21% EPS miss that broke an eight-quarter beat streak, a stale consensus that has to be marked down, and no earnings catalyst until November. Against that, a below-book valuation and $7.9B of net cash cap the downside and make a violent oversold bounce possible in a name that is 88.9% idiosyncratic and runs at ~52% volatility. Net: more likely to drift or chop lower than to re-rate inside the window — but high two-way variance, not a one-way bet. The single condition that flips it is a sentiment turn — stabilisation off the below-book floor, helped by the 26 August EGM — rather than any change in the fundamentals, which are not re-tested until the November print.