
IPO
Traded ValueHorizon Robotics Analysis

1. Company Overview and Industry Positioning
Horizon Robotics was founded in 2015 and is a leading provider of intelligent driving computing solutions in China, focusing on advanced driver assistance systems (ADAS) and high-level autonomous driving (AD) solutions. It went public on the Hong Kong Stock Exchange in October 2024, raising HKD 5.4 billion. The company is positioned as a Tier 2 supplier, offering full-stack intelligent driving solutions to automakers through a combination of hardware and software (chips + algorithms). Its customers include 27 OEMs (42 brands) such as BYD, Li Auto, Volkswagen, and Chery, with its solutions deployed in over 285 vehicle models.
2. Core Business and Technological Advantages
(1) Product Portfolio and Technological Barriers
Journey Series Chips:
Journey 6 Series (launched in 2024): Includes 6 models (J6B/J6L/J6E/J6M/J6H/J6P), covering scenarios from low-level ADAS to high-level urban NOA:
J6P Flagship Chip: Computing power of 560 TOPS, supports BEV/Transformer end-to-end models, performance comparable to NVIDIA Thor, with a cost approximately RMB 1,000 lower per vehicle than NVIDIA Orin.
J6M (Urban NOA): CPU computing power of 137 kDMIPS, surpassing NVIDIA Orin N, already used in BYD's "Eye of Heaven" C solution.
Technical Architecture: Self-developed BPU Nash architecture, optimized for Transformer, with computing power efficiency reaching 85% (NVIDIA only 30-40%), achieving extreme energy efficiency through hardware-software synergy.
Solutions:
Horizon Mono (L2 active safety), Pilot (highway NOA), SuperDrive (full-scenario urban NOA), mass production scheduled for September 2025.
(2) Ecosystem and Business Model
Open Toolchain: Tiangang Kaiwu (algorithm development), Tage (middleware), and Aidi (software platform) lower the development threshold for automakers, forming an ecosystem moat similar to CUDA.
Revenue Model: Hardware sales (chips/domain controllers) + software licensing services (increasing proportion of IP licensing), with a gross margin of 70.5% in 2023.
3. Financial Performance and Growth Potential
| Metric | 2023A | 2024H1 | 2025E | 2026E |
|---|---|---|---|---|
| Revenue (RMB billion) | 15.52 | 9.35 | 34.76–36.10 | 53.40–56.97 |
| YoY Growth | 71.3% | 151.6% | 51% | 58% |
| Adjusted Net Profit | Narrowed loss | -RMB 0.8B | -RMB 1.314B | -RMB 0.774B |
| Gross Margin | 70.5% | - | Maintained >70% | Optimizing |
Drivers:
In 2024, intelligent driving solutions shipped 2.9 million units, with cumulative deliveries reaching 7.7 million units, capturing 33.97% of China's independent brand ADAS market (ranked first).
By 2025, L2+ intelligent driving penetration is expected to reach 20%, with NOA functionality extending to vehicles priced as low as RMB 150,000, and Horizon Robotics is strengthening partnerships with leading automakers.
4. Market Competitiveness Analysis
| Comparison Dimension | Horizon Robotics' Advantages | Main Competitors |
|---|---|---|
| Technical Approach | Full-stack self-developed hardware-software integration (algorithm-defined hardware) | NVIDIA: Hardware + basic software, lacks algorithms |
| Computing Efficiency | BPU utilization 85% vs. NVIDIA 30-40% | Mobileye: Closed ecosystem, slow iteration |
| Cost | J6 solution RMB 1,000 cheaper per vehicle than NVIDIA | Momenta: Pure software, relies on third-party hardware |
| Ecosystem Openness | Toolchain supports rapid development by automakers | Huawei: Full-stack but deeply tied to automakers, limited options |
Conclusion: Horizon Robotics, with its open ecosystem + extreme cost-performance ratio, has become the "greatest common divisor" for automakers' intelligent driving democratization.
5. Industry Trends and Catalysts
Policy and Demand: China's L2+ penetration rate rose from 38% in 2023 to 55.8% in 2024, with NOA pre-installation reaching 8.62%; global AD market size is projected to reach RMB 998.2 billion by 2030 (RMB 403.2 billion in China).
Technological Inflection Point: By 2025, end-to-end large models will enter mass production (e.g., HSD system), driving intelligent driving from "usable" to "user-friendly."
Tesla FSD in China: Intensifies competition but accelerates iteration of local solutions, with Horizon's J6P performance comparable to FSD hardware.
6. Risk Warnings
Profit Pressure: High R&D investment (2023 R&D expense ratio >50%), may continue to incur losses before 2026.
Intensified Competition: High-end solutions like NVIDIA Thor and Huawei ADS 3.0 squeeze the market.
Supply Chain Risks: Automotive-grade chips rely on TSMC for production, geopolitical factors may impact capacity.
7. Investment Value and Recommendations
Short-term (1 year):
Buy Range: HKD 6.8–7.2 (2025 PS 18-19x), current price HKD 7.05 near valuation midpoint.
Target Price: HKD 8.5 (+20% upside), catalyzed by 2025 urban NOA mass production + Chery partnership.
Medium-to-long-term (2026+):
Inflection Signal: Profitability expected in 2027 (projected net profit RMB 668 million), PS drops to 9-11x, market cap outlook HKD 120 billion.
Strategy:
| Price (HKD) | Positioning Advice | Logic | Win Rate |
|---|---|---|---|
| ≤6.8 | Increase to 50% | High valuation safety margin, technical pullback complete | 75% |
| 7.0–7.2 | Build 30% position | Industry beta + new product volume | 65% |
| ≥8.5 | Reduce to 10% | Short-term valuation fully priced in | 80% |
| Stop Loss | <6.0 (-15%) | Technical breakdown or industry penetration below expectations | - |
💡 Key Conclusion: Current price HKD 7.05 is suitable for building a 30% position, increase to 50% if it falls to HKD 6.8, target price HKD 8.5 (win rate 65%-75%), 2025 dynamic PS 19x is reasonable, long-term beneficiary of intelligent driving penetration leap.
The copyright of this article belongs to the original author/organization.
The views expressed herein are solely those of the author and do not reflect the stance of the platform. The content is intended for investment reference purposes only and shall not be considered as investment advice. Please contact us if you have any questions or suggestions regarding the content services provided by the platform.

