--- title: "HSAI: Worried About ASP Collapse? 'Tesla Castoff' Goes for Volume---" type: "Topics" locale: "en" url: "https://longbridge.com/en/topics/39486324.md" description: "FY26 shipment guidance raised sharply to 3.0-3.5 mn units" datetime: "2026-03-24T15:12:29.000Z" locales: - [en](https://longbridge.com/en/topics/39486324.md) - [zh-CN](https://longbridge.com/zh-CN/topics/39486324.md) - [zh-HK](https://longbridge.com/zh-HK/topics/39486324.md) author: "[Dolphin Research](https://longbridge.com/en/news/dolphin.md)" --- > Supported Languages: [简体中文](https://longbridge.com/zh-CN/topics/39486324.md) | [繁體中文](https://longbridge.com/zh-HK/topics/39486324.md) # HSAI: Worried About ASP Collapse? 'Tesla Castoff' Goes for Volume--- Overall, $Hesai(HSAI.US) released its Q4 2025 results after the Hong Kong close on Mar 24 Beijing time. The print again showed classic price-down, volume-up dynamics, with shipments surging on lower ASPs. Even so, the company defended GPM with solid cost control and issued upbeat 2026 shipment guidance, as follows: **1) Revenue at the low end of guidance on weaker ASPs**: Q4 revenue was about Rmb 1 bn (+39% YoY), below the Rmb 1.08 bn consensus and at the low end of prior guidance (Rmb 1–1.2 bn). The key driver was continued declines in lidar ASPs. **2) Shipments beat again:** Total lidar shipments reached 631k units in Q4 (+184% YoY), above the 619k consensus and ahead of the prior 600k guide. Within this: a. Robotics lidar shipments were 80k units, well above the 50k expected. This likely reflects a higher mix of the lower-priced JT series. b. ADAS lidar shipments were 550k units (+185% YoY), likely driven by the ramp of the low-priced ATX. Management guided that ATX would account for 80% of Q4 shipments. **3) ASPs continued to fall faster:** Q4 lidar ASP dropped to about Rmb 1,557, down 51% YoY and 13% QoQ from Rmb 1,790. This was well below the market expectation of Rmb 1,718, making it the most direct reason revenue missed expectations. Dolphin Research sees three main drivers of the ASP decline: a. Product mix shift: ATX, a lower-priced model, continued to gain share (estimated to rise from 70% in Q3 to 80% in Q4). The sub-Rmb 1,000-class ATX is rapidly replacing the much higher-priced legacy AT128. b. Pricing strategy: ATX is priced around US$200 at list, and the company offered discounts to key customers in Q4. This pushed the realized ASP lower. c. Robotics mix pressure: Robotics lidar shipments jumped, with a high mix of the low-priced JT series. This structurally dragged down blended ASPs. **4) Scale and tech-driven cost-down supported a resilient GPM**: Despite rapid ASP compression, Hesai delivered about 41% GPM in Q4, above the 39.7% consensus and the 40% company guide. The upside was driven by: a. Fixed-cost dilution from surging volumes: lidar sales reached 630k units in Q4, up 43% QoQ. This improved manufacturing leverage. b. Ongoing cost-down via a platformized product line, ASIC design, and supply-chain and manufacturing optimization. These initiatives helped offset pricing pressure. c. Services revenue, which carries higher margins, recovered. This mix benefit supported overall profitability. **5) Core profit slightly below, with disciplined opex control overall:** GAAP net income was Rmb 150 mn, near the top end of the Rmb 70–170 mn guide. This mainly reflected interest income and other income totaling about Rmb 120 mn, including roughly US$6.4 mn (about Rmb 45 mn) received from Ouster as an IP arbitration settlement. Excluding these items, core OP (GP minus core opex) was about Rmb 40 mn, slightly below the Rmb 60 mn expected. The shortfall was driven by higher R&D and sales expenses QoQ, but overall opex remained well managed. **Dolphin Research key takeaways:** Hesai continued to execute on the volume-up, price-down playbook, with tech-led cost reductions driving penetration. While Q4 revenue looked muted as lidar ASPs kept falling, the company protected a 40%+ GPM through strong cost control and scale. **Looking to 2026, Hesai sharply raised shipment guidance, underscoring lidar’s democratization in autos and accelerating penetration:** **a. Full-year shipment guide is strong with a very full order book:** The company raised 2026 shipment guidance to 3.0–3.5 mn units (from 2.0–3.0 mn), implying +85% to +116% YoY, well above the 2.66 mn market view. To meet demand, annual capacity is planned to double from 2.0 mn units in 2025 to over 4.0 mn in 2026. **ADAS:** The company has won orders from China’s top 10 OEMs, covering 40+ brands and 160+ models. The revamped ATX has over 6 mn units of backlog (SOP in Apr 2026, powered by the in-house FMC500 SoC), demonstrating strong scalability, and management guides ADAS shipments could double to 2.77 mn units. **Robotics:** JT series shipments exceeded 200k units in the first year, with additional intent orders for lawn-mowing robots from Dreame and Mova in the tens of millions. Hesai also has near-full coverage in Robotaxi/Robovan, and robotics lidar shipments could also double to 480k units. **b. Q1 seasonally solid, with mix upgrades to stabilize ASPs:** Q1 2026 net revenue is guided at Rmb 650–700 mn (+24% to +33% YoY). Given Q1 is typically a seasonal trough for autos, with roughly 12% of annual shipments historically, Q1 shipments are implied at 400–450k units. Dolphin Research expects that while ATX pricing will step down from about US$200 in 2025 to US$150 in 2026 (about 25% YoY down), Q1 blended ASP is estimated at Rmb 1,548. This would be basically flat QoQ vs. Rmb 1,557 in Q4. The stabilization should be driven by a better mix: ADAS shipment cadence slows in Q1, while higher-priced, higher-margin premium robotics lidar shipments are set to reach about 100k units. That would lift mix to 22%–25% (up roughly 13ppts QoQ) and provide a solid offset. **Dolphin Research views Hesai as a high-certainty name benefiting from rising smart-driving penetration and a second-leg expansion in robotics:** **1) Clear competitive landscape with Hesai as the No. 1 in passenger vehicles** While Hesai’s share has fluctuated since Jun last year due to Huawei’s blind-spot lidar ramp, Huawei’s lidar is largely bundled as an integrated HW/SW stack and priced higher, mainly serving its Aito and HI ecosystems. In the independent Tier-1 market, the real contest remains among Hesai, RoboSense, and Tudatong. Backed by strong ATX volumes, Hesai’s Q4 share in passenger-vehicle lidar reached 31.3%. That was about 13ppts ahead of No. 2 RoboSense, keeping Hesai firmly in the lead. **2) 2026 guide beats by a wide margin, easing customer-loss concerns, with volume-up, price-down intact** Investors had worried that 2026 ADAS growth might slow, even hinting at potential large-customer attrition. The new 3.0–3.5 mn shipment guide dispels these concerns, with the bulk of the upside coming from four areas: **1) ATX entering the sub-Rmb 1,000 era, penetrating further into lower price bands:** Changan Qiyuan Q05 and Leapmotor A10 are among the first Rmb 100k-class models to adopt lidar. Dolphin Research expects penetration to accelerate into sub-Rmb 100k segments in H1 2026. By customer, the 2026 incremental volumes are expected to be led by Xiaomi (sole-sourced, about 550k units), Leapmotor (essentially sole-sourced, about 600k), Li Auto (sole-sourced, about 450–500k), BYD (roughly half of the business, about 300–350k), Geely (about 500k), and Great Wall (about 200k). **2) L3 to drive a step-up in per-vehicle sensor count:** Regulatory milestones for L3 autonomy expected in 2026 could be a key catalyst. L3 perception needs multiply, with the stack moving from a single main lidar at L2 to one main lidar (ATX/ETX) plus multiple blind-spot lidars (FTX). This will lift per-vehicle lidar value from roughly US$200 to US$500–1,000. With higher utilization at L3, customers should also be more tolerant on price. **3) Intl markets approaching a breakthrough, strengthened by the NVIDIA ecosystem:** Hesai has completed C-sample development of a long-range lidar for a leading European OEM, with overseas mass production targeted by end-2026. NVIDIA’s Drive Hyperion full-stack platform, covering HW/SW and pre-trained AI models, boosts integration efficiency and accelerates OEM adoption abroad. Given lower price sensitivity and a preference for premium products among overseas customers, exports can help offset domestic ASP pressure. This geographic mix shift should support profit resilience. **4) Robotics as a second growth curve, with Hesai still dominant in Robotaxi:** The potential TAM for robotics is roughly twice that of autonomous driving, and robotics ASPs and margins are generally higher than ADAS. This helps Hesai counter price pressure in passenger vehicles and maintain high overall margins. Consumer/service robot orders are strong: leveraging JT series designed for compact, high-performance scenarios, Hesai has exclusive supply deals for lawn-mowing robots with Dreame and its brand Mova. The order pipeline exceeds 10 mn units, highlighting powerful shipment potential. Robotaxi/Robovan leadership remains intact: in L4 autonomous markets, Hesai stays on top. The company holds 60%–70% share in global Robotaxi, with supply agreements across top players in North America, Asia, and Europe. In Robovan for urban logistics, Hesai also ranks No. 1 in GGI lidar nominations. It is the exclusive lidar supplier to global leaders such as Neolix and DoorDash. **On pricing, Dolphin Research expects 2026 lidar ASP to fall another ~21% YoY to Rmb 1,460:** The pressure mainly comes from a higher mix of lower-priced ADAS units, including ATX moving to about US$150 (about 25% YoY down). Volume-linked discounts for large customers and annual OEM price cuts add to the headwinds. These will be partially offset by L3 multi-lidar configurations, ramp of the premium ETX, higher-margin overseas orders, and the profitable robotics business. The net result should be manageable ASP pressure. **Despite ongoing ASP pressure, Hesai expects overall GPM to remain resilient in 2026, driven by tech cost-down and scale:** a. Localization and high integration of the main control chip: the in-house FMC500 SoC is based on RISC-V and highly integrates MCU, FPGA, and ADC. A single-chip solution greatly reduces the core chip cost, which accounts for about 40% of BOM. b. In-house SPAD integration technology slated for 2026 mass production. This should further reduce costs. c. Massive scale effects at 3.0–3.5 mn units and highly automated manufacturing. Fixed costs will be spread over much larger volumes. **On opex:** Total operating expenses are expected to grow in the mid-teens, driven by about US$200 mn of forward-looking R&D investment. These funds will expand the frontier product roadmap on both perception (eyes) and actuation (muscles). Excluding new-business spending, core opex for the legacy businesses should be flat to slightly down. This reflects strong cost discipline, operating leverage, and deep internal use of AI, which has already delivered quantifiable cost savings in the tens of millions of Rmb. **At the current share price, Dolphin Research estimates 2026 revenue of Rmb 4.85 bn (+60% YoY) and net income of Rmb 700–800 mn (+60%–70% YoY), implying 2026 P/S of about 5.2x and P/E of about 32–36x. With net income still growing rapidly in 2026, this valuation does not look expensive.** **Over a longer horizon, the lidar industry exhibits a strong Moore-like trajectory, with point-cloud density rising and costs falling exponentially, underpinned by high barriers from ecosystem lock-in and scale.** Industry consolidation should continue, reinforcing the strong-get-stronger dynamic, with Hesai firmly the No. 1 in China. From a TAM perspective, the current ~Rmb 25.6 bn market cap largely reflects only the domestic passenger-vehicle ADAS base, which could reach Rmb 40 bn by 2030 on a conservative view. Hesai’s expected overseas breakthroughs and the sizable upside optionality in robotics suggest further upside remains. Execution on both fronts can unlock that elasticity. Please refer to [**Hesai: Tesla’s discard cannot stop lidar’s grand comeback**](https://longportapp.cn/zh-CN/topics/33959548?app_id=longbridge&utm_source=longbridge_app_share&channel=t33959548&invite-code=4NOXYT&locale=zh-CN&community_badge=1&profile_following_followers_activities=1) for detailed valuation work. ![32](https://pub.pbkrs.com/cms/2026/0/LvM3vsZg6cGwwMos8yJXvidh81NdHneD.jpg?x-oss-process=style/lg) **Previous articles:** Hesai in-depth: [**4x Hesai: why the lidar dropped by Tesla is shining again**](https://longportapp.cn/zh-CN/topics/31812806?app_id=longbridge&utm_source=longbridge_app_share&channel=t31812806&invite-code=4NOXYT&locale=zh-CN&community_badge=1&profile_following_followers_activities=1) [**Hesai: Tesla’s discard?**](https://longportapp.cn/zh-CN/topics/33959548?app_id=longbridge&utm_source=longbridge_app_share&channel=t33959548&invite-code=4NOXYT&locale=zh-CN&community_badge=1&profile_following_followers_activities=1) **The grand comeback of lidar** **Hesai earnings** [**Hesai: strong shipment growth for the eyes of smart driving, can it soar again**](https://longportapp.cn/zh-CN/topics/33018953?app_id=longbridge&utm_source=longbridge_app_share&channel=t33018953&invite-code=4NOXYT&locale=zh-CN&community_badge=1&profile_following_followers_activities=1) ### Related Stocks - [Hesai Group (HSAI.US)](https://longbridge.com/en/quote/HSAI.US.md) - [Ouster, Inc. (OUST.US)](https://longbridge.com/en/quote/OUST.US.md) - [HESAI-W (02525.HK)](https://longbridge.com/en/quote/02525.HK.md) - [Ouster A - CW25 (OUSTW.US)](https://longbridge.com/en/quote/OUSTW.US.md) ## Comments (3) - **新用户_6lI22d · 2026-03-24T15:30:04.000Z · 👍 1**: When will Tesla be able to integrate LiDAR solutions? - **过尽千帆便是** (2026-03-24T15:55:01.000Z): Musk is very stubborn. - **新用户_6lI22d** (2026-03-24T16:01:48.000Z): I see BYD uses RoboSense.