Following the initial proposal of "building a new form of smart economy" in the 2026 Government Work Report, the capital market's immediate response was to identify companies capable of translating AI technology into tangible productivity. GEEKPLUS-W (02590) is positioned precisely at this intersection—benefiting both from top-level policy support for the "smart economy" and the fundamental confirmation of its own profit inflection point. However, an intriguing phenomenon persists: a significant gap remains between the company's stock performance and its fundamentals. Based on the closing price on March 9th, the company's price-to-sales ratio stands at just over 9 times, substantially lower than the valuation levels of over 30 times for peers such as UBTECH and Dji. The question is, how long can this divergence last? The answer depends on when the market completes its three-stage cognitive reassessment of GEEKPLUS-W.

The first reassessment involves a shift from "loss-making expansion" to "profit replication." GEEKPLUS-W recently released a landmark report: new orders in 2025 reached 4.137 billion yuan, a year-on-year increase of 31.7%; adjusted net profit is projected to be between 25 million and 45 million yuan, officially turning positive. This is rare among the new generation of smart robotics companies—while most are still testing investor patience with the question of "when will they become profitable," GEEKPLUS-W has already crossed that threshold. The significance of achieving profitability extends beyond improved financial figures; it validates the business model. When a company can achieve self-sufficiency without relying on continuous financing, its operational philosophy and commercialization capabilities are fully affirmed. Industry insiders note that after a significant prior correction, GEEKPLUS-W has entered an "attractive valuation range." This reflects a repricing based on its profit certainty—the market is finally beginning to believe that this company can not only tell a compelling story but also generate real earnings.

The second reassessment is a transition from "warehouse robotics" to "king of embodied AI scenarios." In February of this year, GEEKPLUS-W launched Gino1, the world's first general-purpose humanoid robot for warehouse scenarios, which entered mass production immediately upon release. While some investors simplistically interpreted this as "chasing a trend," the crucial aspect is the strategic positioning it represents. Market enthusiasm for humanoid robots is undeniable, but most related companies remain in the laboratory stage, lacking real-world data feedback and commercial validation. GEEKPLUS-W's differentiation lies in the fact that it is not starting from scratch with humanoid robots; instead, it is a natural extension of its embodied AI capabilities, built upon a decade of experience serving over 850 clients globally. Warehousing represents the scenario closest to commercialization for embodied AI—featuring relatively structured environments, clearly defined tasks, and clear customer willingness to pay. The company's unmanned picking workstation passed the proof-of-concept validation with a Fortune 500 company within just three months of its launch, and Gino1 was mass-production-ready upon release. This ability to "generate revenue immediately upon deployment" is precisely the scarce quality the market most desires, yet finds hardest to locate, within the humanoid robotics sector. As the market gradually realizes that GEEKPLUS-W is not merely riding a conceptual wave but has genuinely established a closed loop from technology R&D and product validation to commercial delivery, its valuation logic will be upgraded from that of a "warehouse robotics manufacturer" to the "king of embodied AI scenarios."

The third reassessment entails a shift from an "overseas-focused enterprise" to a "global infrastructure" player. Nearly 80% of GEEKPLUS-W's revenue originates from markets outside Mainland China, with operations spanning over 40 countries and regions. Previously, this was simplistically understood as "strong export capability." However, the 2026 Government Work Report's emphasis on "promoting the expansion and upgrade of the cross-border e-commerce + overseas warehouse model" has infused this narrative with new meaning. Overseas warehouses have transitioned from being an "optional configuration" to a "strategic stronghold" for cross-border fulfillment. Supporting the expansion and upgrade of these warehouses necessitates intelligentization, an unavoidable step. For instance, GEEKPLUS-W deployed over 500 robots at a warehouse in Kentucky, USA, for WINIT, resulting in a 215% increase in storage capacity and a 250% improvement in picking efficiency. Behind these figures lies the reality that GEEKPLUS-W is becoming a critical infrastructure carrier supporting the global expansion of Chinese supply chains. When a company attains this "infrastructure attribute," its valuation logic undergoes another transformation: it is no longer merely an equipment supplier but an indispensable technological node within the entire industrial chain. This is a key reason why Morgan Stanley is optimistic about GEEKPLUS-W's opportunity to increase its global market share and has set a target price of HK$37.

In conclusion, China Merchants Securities and Huatai Securities have given target prices of HK$35 and HK$36.39 for GEEKPLUS-W, respectively, implying an upside potential of over 80% from the current share price. This gap can be interpreted as a pricing discrepancy caused by a lag in market cognition. However, with the profit inflection point confirmed, the embodied AI business gaining traction, and policy tailwinds for overseas warehouses being released, these three reassessments are occurring simultaneously. Once the market completes the repricing of these narratives, the valuation correction for GEEKPLUS-W may not be gradual but rather a significant leap. After all, with favorable policy winds blowing, a company positioned at the forefront of this trend is unlikely to remain undervalued indefinitely.