Investment ResearchTopic research, evidence curves, and cycle signals
HomeSign inEN

Research library · updated 2026-06-15 · public

Leaderdrive / 绿的谐波 S4-to-S5 Evidence Debt Artifact v1

Date: 2026-06-15 Owner: Hugo / Genius Team Agent: Finance / Charlie AGT-002 Status: public-safe research artifact Visibility: PUBLIC Primary use: morning review, slide 10 speaker notes, /robotics/supply-chain case-study module

Related files:

  • knowledge/robotics-research-harness.md
  • knowledge/robotics-sharing-backlog.md
  • knowledge/robotics-supply-chain.md
  • knowledge/robotics-sources.md
  • knowledge/robotics-leaderdrive-public-case-v1.md
  • knowledge/robotics-leaderdrive-morning-review-public-packaging-guardrail-v1.md
  • knowledge/robotics-public-claim-bank-v1.md

Public-safety note:

  • Public-safe: yes, if used as evidence map / case study only.
  • No Hugo private portfolio data.
  • No buy / sell / hold language for 688017.SH.
  • No named-customer inference unless disclosed by primary source.

0. One-line answer

截至 2026-06-15，绿的谐波 / Leaderdrive 最有用的公开表达不是“humanoid winner”，而是“供应链 S4 证据债务案例”：2025 年报披露收入 RMB 570.714m（+47.31% YoY）、归母净利润 RMB 124.367m（+121.42% YoY）、谐波减速器销量 425,158 台（+72.48% YoY）、“工业及具身智能机器人零部件”收入 RMB 422.528m（+52.61% YoY，毛利率 34.88%）；2026 Q1 披露收入 RMB 140.134m（+42.96% YoY）、归母净利润 RMB 32.634m（+61.17% YoY）。这些是强 S4 filing-backed supply-chain signals，但 reviewed public sources 仍未披露具名 humanoid OEM、合同金额、订单 volume、humanoid-only revenue/margin、single-robot value content 或 customer ROI/payback。🟢 2025 annual report and 2026 Q1 report disclosed 2026-04-23; 🟠 Charlie evidence-debt classification as of 2026-06-15.

1. What this artifact adds

已有文件已经能支持一个 sanitized public case。这个 v1 的增量是把 case study 改写成“证据债务”工具，方便早晨 review 时决定：哪些证据已经能进 slide/site，哪些必须留在 S5 missing bucket。

新增结构：

  1. 把 revenue / volume / product breadth / price-action caveat 拆成独立 evidence units。
  2. 给每个 evidence unit 标注 debt severity。
  3. 明确 S4→S5 upgrade trigger，而不是把报表增长直接等同为 humanoid thesis completion。
  4. 给 Codex 一个短模块：三张卡 + do-not-say list + footer。

2. Source verification snapshot

2026-06-15 quick reachability check:

SourceURL / sourceRetrieval result on 2026-06-15Source gradeVerification limitation
绿的谐波 2025 annual reportSSE static PDF URL captured in existing source listHTTP 200; server returned HTML wrapper / first 3,872 bytes sampled🟢 primary filing source, previously extractedThis run verifies URL reachability, not full PDF text extraction. Numeric anchors rely on previously extracted annual-report notes.
绿的谐波 2026 Q1 reportSSE static PDF URL captured in existing source listHTTP 200; server returned HTML wrapper / first 3,872 bytes sampled🟢 primary filing source, previously extractedSame limitation.
绿的谐波 2026-06-09 abnormal-trading announcementSSE static PDF URL captured in existing source listHTTP 200; server returned HTML wrapper / first 3,872 bytes sampled🟢 primary filing source, previously extractedSame limitation.
Leaderdrive official sitehttps://www.leaderdrive.com/HTTP 200; first 4,096 bytes sampled🟢 company officialReachability only; not a customer/order proof.
Leaderdrive product pagehttps://www.leaderdrive.com/product/HTTP 200; first 4,096 bytes sampled🟢 company officialProduct availability / fit does not equal production order.

Method: Python urllib.request with browser-like user-agent in /home/ubuntu/personal-ai/workspace/finance, run 2026-06-15. 🟠 method note.

3. Evidence units vs evidence debt

Evidence unitWhat is provenCurrent signalEvidence debtDebt severityInterpretation
2025 revenue / profit accelerationRevenue RMB 570.714m, +47.31% YoY; attributable net profit RMB 124.367m, +121.42% YoY. 🟢 2025 annual report, disclosed 2026-04-23S4 filing-backed operating accelerationHow much of growth is humanoid-specific vs industrial robot / general embodied-intelligence / cycle recovery.HighStronger than concept-stock language, but not yet humanoid-only validation.
Robot / embodied-intelligence reporting line“工业及具身智能机器人零部件” revenue RMB 422.528m, +52.61% YoY, gross margin 34.88%. 🟢 2025 annual reportS4 reporting-line signalHumanoid-only split, named customer, order duration, backlog and margin by end-market.HighThe line is valuable because it is filing-backed; it must not be compressed into pure humanoid revenue.
Harmonic reducer volumeHarmonic reducer production 433,655 units (+72.27% YoY), sales 425,158 units (+72.48% YoY), inventory 18,238 units (-6.34% YoY). 🟢 2025 annual reportS4 volume signalWhich portion maps to humanoid / embodied-intelligence customers, ASP by use case, customer concentration and repeat demand.Medium-highVolume growth supports supply-chain acceleration; it does not identify the downstream robot platforms.
2026 Q1 follow-throughQ1 revenue RMB 140.134m (+42.96% YoY), attributable net profit RMB 32.634m (+61.17% YoY). 🟢 2026 Q1 report, disclosed 2026-04-23S4 continuation signalWhether growth continues through 2026 with margin stability and clean cash conversion.MediumUseful continuation evidence, but still early and not customer-specific.
Product breadth / architecture hedgePublic materials reference harmonic reducers, micro reducers, rotary actuators, frameless torque motors, planetary roller screws / linear transmission, planetary joint modules and DD motors. 🟢 annual report / official pages / prior evidence filesS3/S4 architecture-fit signalRevenue by newer actuator / PRS / linear products; validated humanoid BOM attach; qualification status.MediumProduct breadth reduces single-architecture risk, but product display is not order proof.
Abnormal-trading caveat2026-06-04 to 2026-06-08 share price +36.00% vs SSE Composite -3.05% and automation-equipment index +2.13%; company said no undisclosed major matter. 🟢 abnormal-trading announcement, 2026-06-09S3 anti-hype / risk-control signalNone for order proof; this is mainly a boundary source.MediumPrevents public packaging from implying price action equals new customer/order evidence.

4. Current classification

Current stage: S4 supply-chain validation; not S5 scaled humanoid commercialization proof. 🟠 as of 2026-06-15.

Why S4:

  • Financial statement evidence exists: revenue, profit, robot-parts reporting line, unit shipments and Q1 follow-through. 🟢
  • Product architecture relevance exists across reducer / actuator / PRS / linear / joint-module surfaces. 🟢/🟠
  • Abnormal-trading announcement gives a useful anti-hype boundary. 🟢

Why not S5:

  • No reviewed primary source discloses named humanoid OEM customer. 🟢/🟠
  • No reviewed primary source discloses contract value, order volume, backlog duration or humanoid-specific revenue/margin. 🟢/🟠
  • No customer-side source discloses downstream robot deployment ROI/payback tied to Leaderdrive modules. 🟢/🟠
  • No public source validates single-robot value content or actual BOM attach in a production humanoid platform. 🟢/🟠

5. Evidence-debt priority order

P0 — would change the case fastest

  1. Named humanoid OEM design win with supply scope, volume, contract period or order amount. 🟢 needed.

    • Why P0: It converts product fit + reporting-line evidence into customer validation.
    • Current status: absent from reviewed public filings / source packets as of 2026-06-15. 🟢/🟠

  2. Humanoid-specific revenue / backlog / margin disclosed separately from broader industrial + embodied-intelligence components. 🟢 needed.

    • Why P0: It prevents over-reading RMB 422.528m as pure humanoid revenue.
    • Current status: absent from reviewed public filings as of 2026-06-15. 🟢/🟠

  3. Segment margin durability while robot-related revenue grows above ~30% YoY. 🟢/🟠 threshold.

    • Why P0: The 34.88% reported gross margin is a key baseline; humanoid ramp matters only if value capture survives ASP pressure, competition and OEM bargaining.
    • Current status: one strong 2025 baseline plus 2026 Q1 continuation, but not enough multi-period humanoid-specific data. 🟢/🟠

  4. Validated BOM / teardown / customer-side confirmation that Leaderdrive modules are in production humanoid platforms. 🟢/🟡 needed.

    • Why P0: It ties product breadth to actual platform attach.
    • Current status: absent in reviewed public-safe source set. 🟢/🟠

P1 — strengthens / weakens S4 quality

  1. Revenue growth accompanied by stable inventory, receivables and operating cash flow quality. 🟢 needed.
  2. New actuator / PRS / linear products becoming material revenue contributors. 🟢 needed.
  3. Repeat customer orders or longer supply agreements, even if customer names remain partly undisclosed. 🟢/🟠 needed.
  4. Competitor ASP / margin evidence showing whether harmonic reducers are commoditizing. 🟡/🟠 useful.

P2 — useful but not thesis-changing alone

  1. More product-exhibition photos without revenue/customer link. 🟢 for product existence / 🟠 for commercialization.
  2. Sell-side / social supply-chain lists without primary customer confirmation. 🟡/🔴.
  3. Stock-price movement without disclosed operating update. 🔴 as proof; 🟢 only as market-context.

6. Public slide / site module draft

Suggested title

绿的谐波：强报表信号，不等于已验证 humanoid winner

Three-card layout

  1. What is proven 🟢

    • 2025 revenue RMB 570.714m, +47.31%; net profit RMB 124.367m, +121.42%.
    • “工业及具身智能机器人零部件” revenue RMB 422.528m, +52.61%, GM 34.88%.
    • Harmonic reducer sales 425,158 units, +72.48%.
    • 2026 Q1 revenue RMB 140.134m, +42.96%; net profit RMB 32.634m, +61.17%.

  2. Why it matters 🟢/🟠

    • This is filing-backed revenue / profit / unit-volume evidence, stronger than concept-stock narrative.
    • Product breadth covers reducer, micro reducer, rotary actuator, frameless motor, PRS / linear actuator, planetary joint module and DD motor surfaces.
    • Current public grade: S4 supply-chain validation.

  3. What remains missing before S5 🟢/🟠

    • Named humanoid OEM customer.
    • Contract value / backlog / order volume.
    • Humanoid-specific revenue and gross margin.
    • Single-robot value content / validated BOM attach.
    • Downstream customer ROI/payback and repeat deployment economics.
    • Margin durability if volume scales.

Required footer: Evidence map only. No trade recommendation. No private portfolio data. Product fit ≠ customer proof; stock-price move ≠ order evidence.

7. Signal vs noise

Signal

  • Filing-backed revenue, profit, reporting-line and unit-volume acceleration. 🟢 S4.
  • Q1 follow-through after 2025 acceleration. 🟢 S4 continuation.
  • Product expansion beyond discrete harmonic reducers. 🟢/🟠 S3/S4 architecture hedge.
  • Abnormal-trading announcement saying no undisclosed major matter after +36.00% three-trading-day move. 🟢 S3 caution.

Noise unless upgraded

  • “Leaderdrive supplies Tesla / Figure / Unitree” without primary customer disclosure, filing evidence, teardown or identifiable production-part evidence. 🔴
  • “RMB 422.528m is humanoid revenue.” Incorrect compression: the filing line is industrial + embodied-intelligence robot components, not humanoid-only. 🟢/🟠
  • “Share price +36% proves new order.” Contradicted as order evidence by the abnormal-trading disclosure boundary. 🟢/🔴
  • “Every humanoid requires fixed N harmonic reducers.” Only usable as an estimate with explicit method. 🟠/🔴
  • “Product display means large order.” Product fit is not revenue, customer, margin or repeat order proof. 🟢/🟠

8. Common misconceptions

  1. Misconception: “绿的谐波已经证明 humanoid 商业化。”

    • Correction: 它证明的是 component supplier 的 S4 filing-backed acceleration；下游 humanoid customer economics 仍未披露。🟢/🟠

  2. Misconception: “RMB 422.528m 就是 humanoid revenue。”

    • Correction: 披露口径是“工业及具身智能机器人零部件”，不是 humanoid-only revenue。🟢

  3. Misconception: “供应链收入比 OEM 部署证据更确定。”

    • Correction: 供应链收入可以更早进报表，但仍面对 OEM 议价、ASP 下降、客户集中、垂直整合、架构替代和 inventory/receivables 质量风险。🟠

  4. Misconception: “股价上涨证明订单。”

    • Correction: 2026-06-09 abnormal-trading announcement 说不存在应披露而未披露的重大事项；股价是市场信号，不是订单证据。🟢

  5. Misconception: “谐波减速器是唯一价值池。”

    • Correction: humanoid value may migrate among reducers, integrated actuators, PRS / linear actuators, motors, sensors, batteries, compute, model/software and deployment service. 🟠

9. Think Deeper questions

  1. 如果 humanoid 放量但低价 OEM 强压 ASP，上游供应商是受益于 volume，还是损失在 margin？
  2. “工业及具身智能机器人零部件”这类混合口径，最低要拆到什么程度才足以支撑 S5 thesis：customer、use-case、humanoid-only revenue、还是 margin？
  3. 如果 integrated actuator 取代部分 discrete harmonic reducer，Leaderdrive 的产品扩张会增加 per-robot value content，还是只是防守？
  4. 供应链公司先进入 revenue validation，OEM 先进入 deployment KPI validation；两类证据应该如何互相校验？
  5. 若股价已经先行反映主题，而 filings 仍无具名客户，该如何在 public field guide 里表达“强信号 + 强边界”？

10. Do-not-say list for Codex

  1. Do not say 688017.SH is buy / sell / hold. 🔴
  2. Do not include Hugo private portfolio weight, entry logic, or private rationale. 🔴
  3. Do not infer Tesla / Figure / Unitree / AgiBot / UBTECH as customers without source-backed confirmation. 🔴
  4. Do not call “工业及具身智能机器人零部件” pure humanoid revenue. 🔴
  5. Do not treat share-price movement as customer/order evidence. 🔴
  6. Do not use estimated reducer counts or BOM values as facts unless labeled 🟠 estimate with method. 🔴
  7. Do not turn the case study into a valuation pitch.

11. Source list

Primary sources 🟢:

Internal synthesis / cross-reference 🟠:

  • knowledge/robotics-leaderdrive-public-case-v1.md
  • knowledge/robotics-leaderdrive-morning-review-public-packaging-guardrail-v1.md
  • knowledge/robotics-public-claim-bank-v1.md
  • knowledge/robotics-supply-chain.md

12. Public-safe flag

Public-safe: yes, if used as an evidence-debt map and not as a recommendation.

Exclude:

  • Hugo private portfolio data.
  • Buy / sell / hold language.
  • Named-customer claims not present in primary sources.
  • Paid-report excerpts or private channel checks.
  • Any implication that S4 filing-backed supplier validation equals S5 scaled humanoid commercialization proof.