Morgan Stanley: Horizon's Shift to IP Licensing is Not a Zero-Sum Game; Maintains Overweight Rating

Overview

This research note provides an interpretation of meeting minutes from Morgan Stanley regarding Horizon Robotics (9660.HK). The core argument clarifies market concerns that "automaker in-house chip development will squeeze out third-party suppliers." The institution believes Horizon is adapting to this industry paradigm shift by transitioning to an IP licensing business model, which is not a zero-sum game. Although macroeconomic weakness in the auto market may drag down short-term growth rates, the high margin characteristics of IP licensing and the synergy effects of the software ecosystem support its medium-term growth logic.

Core views

Business Model Transformation in Response to In-House Trends: As more Original Equipment Manufacturers (OEMs) pursue in-house chip development, Horizon has launched a licensing business model akin to semiconductor IP companies. This model includes upfront licensing fees and ongoing royalties based on shipment volumes. Management emphasized that the gross margin of this licensing model approaches 100%, significantly higher than the traditional chip sales level of approximately 40-50%. While management did not disclose the current revenue contribution of this specific model, related projects are already underway. Coexistence Logic of In-House and Outsourced Development: The institution points out that while automaker in-house chip development is a future trend, it does not mean third-party suppliers will be eliminated. Instead, design-ins based on Horizon's BPU architecture may drive demand for its high-end Intelligent Driving Software (HSD). Furthermore, due to a lack of economies of scale, the inability to keep up with rapidly iterating chip specifications, and the complexity of hardware-software integration, not all automakers possess the capability or economic rationality to develop in-house chips. Therefore, third-party suppliers will remain relevant under different cooperation models. Short-Term Performance Guidance: Management stated that sales volumes for the J6M chip were not negatively impacted this year. However, considering the broader automotive market weakness, Horizon's year-over-year growth rate in the first half of 2026 may fall short of the previously set 60% target.

Analysis framework

The institution adopted an analytical perspective of "Business Model Restructuring" and "Industry Chain Division of Labor." First, by comparing the gross margin differences between traditional chip sales and the IP licensing model (40-50% vs. ~100%), they quantified the potential improvement in profitability quality of the new business model. Second, applying an industry chain competitive landscape analysis, they identified two paths in the intelligent driving sector: "full-stack in-house development" versus "professional division of labor," arguing that due to technical thresholds and economies of scale limitations, most automakers still rely on third-party core IP and software, thereby alleviating market concerns over share loss. Finally, by incorporating the macro automotive environment to adjust short-term sales targets, the research reflects an integrated approach combining top-down (macro) and bottom-up (company micro) methodologies.

Methodology notes

• Company Fundamentals and Financial Framework

  Impact of Business Model Transformation on Gross Margin Structure

  The research note analyzes potential changes in the company's profitability quality by comparing gross margin differences between different business models (selling hardware vs. selling IP). This helps investors understand why profit ends may still outperform expectations even if revenue scales face constraints.

• Competition and Strategy FrameworkMoat / competitive advantage

  Technology Architecture Lock-in Effect

  The note mentions that automaker in-house chip development might increase demand for Horizon's software because its BPU architecture has formed a certain ecological barrier. The characteristic of "hardware replaceable, software difficult to migrate" constitutes the company's competitive advantage.

• Valuation MethodologyDCF Cash Flow Discount

  Probability-Weighted DCF Valuation

  The institution uses a probability-weighted cash flow discount model for valuation, setting probability distributions of 25% Bull Case, 50% Base Case, and 25% Bear Case to reflect uncertainties such as intensifying competition in the intelligent driving market and geopolitical risks.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Horizon Robotics (9660.HK)

  Beneficiary: Resolves risks of automaker in-house development through IP licensing and expands software revenue

  Strengths

  BPU IP architecture possesses high ecosystem stickiness; IP licensing model offers extremely high gross margins; Sales of mature products like J6M are robust

  Weaknesses

  Short-term growth may underperform expectations due to macro auto market weakness; Revenue proportion of new business model remains unclear

  Comparison

  Compared to pure hardware chip manufacturers, Horizon possesses stronger anti-cyclical capabilities and pricing power through combined hardware/software solutions and IP licensing

  Risks

  Automaker in-house development progresses faster than expected; ADAS adoption slows down; Geopolitical risks

Key data

• Stock RatingOverweightRating Maintained
• Target PriceHK$10.00Implies approximately 89% upside from current price
• Current Stock PriceHK$5.29Closing price as of May 29, 2026
• IP Licensing Model Gross Margin~100%Significantly higher than the 40-50% for traditional chip sales
• First Half 2026 Growth Expectation< 60%Due to weak auto market conditions, may fall short of the originally targeted 60% year-over-year growth rate
• DCF Assumption - WACC12.2%Includes 1.9 Beta and 3% long-term growth rate assumptions

Impact & implications

The research note believes that Horizon's transition to IP licensing will change the market's valuation logic for the company as a pure hardware supplier. If this model is successfully rolled out, the company will obtain a more stable and high-margin recurring revenue stream. For the industry, this means the intelligent driving supply chain will shift from a simple "buy-sell relationship" to a deeper "technology licensing and ecosystem collaboration." Investors should monitor the progress of Horizon's ecosystem building around the BPU architecture and its implementation cases with leading automakers under the new cooperation models.

Risks

• China ADAS/Advanced Driver Assistance Systems adoption rate is slower than expected
• Supply chain disruption risks
• Automakers successfully succeed in in-house hardware design, reducing external dependence
• Key Chinese customers face pressures on vehicle sales
• Intensifying competition in intelligent driving business and potential geopolitical headwinds

What to watch

• Actual growth rate of China ADAS/AD adoption
• Delays or failures in automakers' in-house hardware design efforts
• Whether Horizon succeeds in expanding its key customer base
• Changes in market access for other solution providers in China