Why has growth in L2++ demand not yet translated into incremental business for Horizon?

  • China's passenger-vehicle market recorded a year-on-year sales decline of 25% in 2026's third quarter, but combined sales of L2+ and L2++ models still grew by about 6% year on year during the same period; L2++ penetration rose to 32.4% in September 2026. This shows that the adoption of advanced driver-assistance systems is still progressing, but industry expansion does not mean Horizon is securing orders at the same pace.
  • Horizon held market shares of 45% and 37% in L2 and L2+, respectively, but only 3% in the faster-growing L2++. Bernstein said conversion of design wins and sales of models equipped with HSD remain below expectations, and a breakthrough in L2++ has yet to materialize.

Which competitive pressures and customer developments are limiting share gains?

  • Automakers' in-house chip development is expanding: Nio, XPeng and Li Auto have deployed their own chips; BYD expects to begin production of its own chip in 2027; and Chery has also announced an in-house development plan. Meanwhile, Nvidia still accounted for 53% of L2++ vehicle installations in 2026's third quarter, while suppliers including Qualcomm and SemiDrive, which is associated with Momenta, are also adding to the range of choices. Automakers' vertical integration, established leaders and new entrants are all fragmenting the share available to win.
  • Horizon's key L2++ customers remain Chery and Changan; after Li Auto shifted to in-house chips, Horizon's installed TOPS declined quarter on quarter. The J6P ramp-up has also been slower than Bernstein originally expected: in 2026's third quarter, only one new model, the Changan Qiyuan Q06, adopted it. Whether BYD and Volkswagen can contribute a meaningful share is key to Horizon expanding its position in L2++.

Why have profit forecasts been cut substantially despite continued revenue growth?

  • Bernstein forecasts Horizon's 2027 revenue at RMB 8.026 billion, down 15.9% from its previous forecast of 9.541 billion, and 3.4% below the market consensus of 8.31 billion. Its revenue forecast for 2028 is RMB 12.45 billion, down 18.8% from its previous forecast. Growth remains fast, but both the forecast growth rate and scale are below those in the original model.
  • The change is more pronounced on the profit side: Bernstein forecasts operating profit of only RMB 40 million for 2028, versus its original forecast of RMB 1.48 billion. The report also notes that some of the company's business is shifting from Product Solutions to the less visible Licensing & Services model; together with slower L2++ share gains, this means revenue growth has not supported the profit expectations set previously.

What valuation underpins the price-target cut?

  • Bernstein lowered its rating from Outperform to Market Perform and cut its price target from HK$8 to HK$4.5 per share. The valuation multiple was reduced from 8.1x EV/sales to 4.4x, applied to blended sales of RMB 11.8 billion for 2027 and 2028; the valuation base was previously RMB 11.5 billion. The price-target cut mainly reflects its reassessment of L2++ share growth and the competitive environment.
  • Bernstein believes that whether Horizon can secure a meaningful supply share from BYD and Volkswagen will determine whether it can establish a more substantial position in the L2++ business. However, BYD plans to develop its own chips and Volkswagen is pursuing a multi-supplier strategy, so the available evidence does not support a view that volumes will ramp up quickly.

Related topics