Lenovo at all-time highs: the AI trade goes multi-polar

What happened

On 13 August, Lenovo Group Ltd reported results for the first quarter of its 2026/27 fiscal year. Revenue reached an all-time quarterly high of $26.9bn, up 43% year-on-year, the fastest growth in five years and the strongest quarter in the group's history. Adjusted net income rose 176% to $1.1bn, crossing the billion-dollar mark for the first time. AI-related revenue grew 60% to $9.3bn and now represents 35% of the group's top line. Management raised its outlook and is targeting $100bn in revenue this fiscal year.

Shares surged as much as 22% in Hong Kong to a record high, extending year-to-date gains to roughly 280% and making Lenovo the top performer on the Hang Seng China Enterprise Index.

Growth of this magnitude from a company long viewed as a low-margin PC assembler deserves an explanation that goes beyond the quarter itself.

Impact on our Investment Case

Hybrid AI: the deployment model the market underestimated

For three years, the consensus view held that AI value would accrue almost entirely to the US hyperscaler cloud and the chip vendors supplying it. We argued in our June note that the landscape was fragmenting into multiple deployment vectors. Lenovo's quarter is the clearest validation of that view so far, because its hybrid AI strategy spans every vector at once.

• On-device. Lenovo is embedding AI across its PC families and introduced Qira, a cross-device personal AI system, in 2026. As capable models shrink to fit local hardware, the AI PC replacement cycle turns an installed base of hundreds of millions of machines into an upgrade opportunity. That demand flows to device makers rather than to cloud providers.
• Private and sovereign infrastructure. The Infrastructure Solutions Group delivered record revenue and has become the group's growth and profitability engine. It sells AI-optimized servers for training and, increasingly, for inference, workloads that run where the customer's data resides rather than in a US hyperscaler's region. Lenovo's recent deployment of sovereign AI infrastructure in Saudi Arabia, aligned with Vision 2030, is a concrete example of the sovereign procurement wave we flagged as a catalyst: governments and state-linked enterprises buying compute they control.
• The inference shift. As AI economics move from training frontier models to running them at scale, the hardware mix moves away from a single-vendor GPU monoculture and toward heterogeneous fleets deployed across clouds, private data centers, and devices. Lenovo sells the deployment layer and benefits regardless of which silicon vendor wins.

The distinction we drew during the SaaSpocalypse, between what AI replaces and what AI requires, applies here in hardware form: Lenovo sells infrastructure that AI deployment requires wherever that deployment lands.

China: the re-rating is underway

Lenovo's move also illustrates the broader Chinese technology undervaluation story we laid out in our April note. A mega-cap, globally scrutinized company does not nearly quadruple in eight months in a market that was pricing it correctly.

The ingredients we identified then remain in place. China produces roughly one-third of the world's engineering graduates and spends about 2.7% of GDP on R&D. DeepSeek proved the innovation is real; Lenovo's $9.3bn of quarterly AI revenue proves it monetizes. The valuation gap persists: even after its run, Lenovo trades at a trailing P/E in the low-to-mid 20s, a fraction of what US AI hardware names command for comparable or slower growth. Across the universe, Chinese technology still trades at a structural discount that its earnings trajectory no longer justifies.

How we were positioned

Our China Technology strategy does not pick stocks by conviction: as detailed in our methodology note, a machine-learning model selects around 35 names from a proprietary universe of over 1'000 Chinese technology stocks to replicate the broad index, and rebalancing is rule-based. At the 12 May 2026 rebalancing, the model initiated Lenovo as a new position at 6.7% of the portfolio, immediately the second-largest holding, at HKD 12.95. There was no analyst call behind the trade and no earnings preview. Nine days later, Lenovo reported its fiscal Q4 and jumped more than 15%; the August quarter added the second leg. With no trades in between, appreciation alone carried the weight to 16%, past Tencent, to the top of the portfolio.

Lenovo is not an isolated hit either. Since its launch on 4 February 2025, the strategy has returned +79.4% (NAV 179.4 as of 17 August 2026), several times the return of the large-cap China ETFs over the same period. The model makes no forecasts. It holds a representative sample of the universe, so when a company or a specific subsector gains footprint in the Chinese technology landscape, the portfolio comes to reflect it.

Our Takeaway

Lenovo's record quarter and all-time high are best read as a data point in two larger trends: AI value creation moving beyond the US cloud into devices, edge, and sovereign infrastructure, and Chinese technology closing a discount its fundamentals stopped justifying some time ago. Both trends have room to run, and neither depends on any single company's next print.

Re-ratings of this kind rarely stop at one name. Lenovo was demonstrably cheap in May; the more useful question is how many of its peers still are. Our answer is embedded in the portfolio itself: a universe of over 1,000 Chinese technology companies, most of them absent from the large-cap benchmarks, in which the same combination of accelerating earnings and depressed multiples can be found repeatedly.

At the portfolio level, no action is required from us, which is the point. A 16% single-name weight in a replication portfolio is a drift artifact, not a target, and the latest rebalancing trimmed the position back toward its model weight, applying the same discipline that sized the entry and locking in part of the gain without any attempt to time the top.

Our systematic process was positioned ahead of both catalysts not because it saw them coming, but because it is built to mirror the universe: when Chinese technology's center of gravity moves, the portfolio moves with it. That construction, rather than any forecast, is what placed Lenovo at the top.