While we have switched away from US stocks, our main investing theme has not changed. That is, we continue to believe in the future of artificial intelligence (AI), including generative AI (GenAI) and agentic AI — and, more specifically, in enterprises that embrace technology and AI tools in their core businesses, be it to transform existing processes to boost operational efficiency, productivity and competitiveness, drive innovation or create new products-demand-markets and so on. This being the case, we recently added Alibaba Group Holding and Trip.com Group to our portfolio. As with our previous investment, Tencent Holdings, these Chinese-based consumer tech companies are trading at more attractive valuations compared with their US counterparts. All are deeply invested in AI to drive their businesses.
The release of DeepSeek’s open-source, cost-effective reasoning model this year has been a major catalyst for investments, adoption and expansion of the AI ecosystem in China. Enterprises — both public and private, large and small — across a broad swath of the economy have jumped onto the bandwagon, integrating AI into their operations.
Alibaba
Alibaba Cloud is one direct beneficiary of this trend. It is the leading cloud service provider in China, with an estimated market share of about 36%, ahead of Huawei Cloud (19%) and Tencent Cloud (15%). Cloud business growth is accelerating and expected to continue to do so. Case in point: Revenue growth for its Cloud Intelligence segment jumped from 13% year on year in 4Q2024 to 18% in the latest 1Q2025. Alibaba expects further acceleration in the current quarter and sees cloud — which accounted for about 12% of total revenue in the financial year ended March 2025 (before intersegment elimination) — as the new engine for long-term growth. It intends to spend at least RMB380 billion (roughly US$53 billion or RM224 billion) over the next three years to expand cloud computing and AI infrastructure. The company also expects gradually improving margins with scale, increasing cloud and AI-related product adoption as well as customers switching from project-based to subscription-based packages (AI-as-a-Service). For example, its Lingma AI coding assistant has seen strong adoption among enterprise customers since its launch last year, registering robust revenue growth.
See also: The coming of a perfect financial storm?
Equally important, Alibaba’s in-house developed LLM (large language model), the Qwen suite of models, serves not only as the backbone for the cloud infrastructure, but also powers AI and GenAI tools in services and applications across its core businesses. For example, customer management revenue for Taobao and Tmall grew 12% y-o-y in 1Q2025, underpinned by higher take rate (percentage of transaction value that the platforms retain), thanks to software service fees and, in particular, the increasing penetration of Quanzhantui.
Quanzhantui is a digital advertising and marketing solution that helps merchants boost product visibility, sales and operational efficiency — by using AI-GenAI to analyse user behaviour, preferences and recommend products to the right shoppers; auto generate product descriptions and custom visuals, create and adjust ads, help in dynamic ad placements and bidding strategies; automate campaign setup, budget allocation; and help sellers manage promotions across Alibaba’s e-commerce platforms. In short, AI tools are used to enhance merchant productivity, increase advertising ROI (return on investment) and drive monetisation through personalised marketing. Taobao and Tmall also integrate AI chatbots into merchant dashboards to improve user experience. AI is also widely integrated into Alibaba’s other core businesses, including Amap (a leading provider of location-based solutions), logistics arm Cainiao and on-demand delivery platform, Ele.me, to enhance user experience and optimise operational and cost efficiencies.
To be sure, Alibaba’s final-quarter earnings results for FY ended March 2025 came in slightly below market expectations, owing to persistent weakness in domestic consumer demand. Revenue was up 6% y-o-y while operating income and adjusted Ebitda (earnings before interest, taxes, depreciation and amortisation) increased 24% y-o-y and 5% y-o-y respectively. Taobao-Tmall remain the biggest revenue (about 45% of total revenue) and earnings generators. Nonetheless, given Alibaba’s comparatively low prevailing valuations against global peers — PE (price-earnings) multiples of 12 times, EV (enterprise value)/Ebitda of just over 6 times plus net cash of RMB366 billion — this short-term consumer demand weakness is likely to have been already priced in. Alibaba’s long-term prospects, we think, is upbeat. More critically, it is from the turnaround of this weakness that we expect the next upcycle — that is, rising domestic demand in China in the near term.
See also: Rubber glove companies — to buy or not to buy
Tencent
In this respect, Tencent has comparatively better increasing AI adoption-driven short-term earnings visibility. The company has, over the years, developed a formidable flywheel model that reinforces its core businesses. For 1Q2025, revenue was up 13% y-o-y while gross profit and operating profit rose 20% and 18% respectively. Operating margins also increased to 39%, from 37%. Its shares, too, are trading at attractive valuations, at 17 times estimated earnings; and forward EV/Ebitda is under 15 times. Tencent has net cash totalling RMB90 billion. The company is also investing heavily in cloud and AI and, similar to Alibaba, integrating AI-related tools into Tencent Cloud and its ecosystem to further drive monetisation and profitability.
For example, AI-powered features are available to help enterprises-users automate and improve content creation, such as illustrate posts with visuals generated based on the texts or easily enhancing videos with preset templates in the Weixin (WeChat) ecosystem. GenAI tools can be used to quickly generate ad copy, visuals and customised campaign variations for small businesses to advertise more efficiently. Coding assistants using natural language can help save development time for mini programmes and increasingly intelligent and interactive digital human-virtual influencers can facilitate live-streaming activities while AI bots handle customer service for merchants — all low-cost and scalable.
The company recently highlighted the potential of incorporating AI agents into Weixin. Yuanbao, Tencent’s AI assistant, is now usable directly in Weixin chat as a new “friend” (contact), the conversations powered by its Hunyuan LLMs and enhanced by DeepSeek. The chatbot assistant can answer queries, analyse, translate and summarise posts, articles and images, conduct searches and so on. This could soon be expanded to multi-step executions, for instance, trigger mini programmes; find, compare and buy products; recommend and make restaurant reservations or food delivery; and plan itineraries and book trips — the AI agents learning and predicting user preferences using Tencent’s extensive database of chat, content interest and search histories.
Basically, it will make the connections seamless across its core businesses — between conversations, search, mini-programmes, official accounts, social media and transactions. This will further improve user experience, engagement and stickiness — and, in the process, drive ad revenues (with personalised and improved targeting) and in-app transactions (for which Tencent takes a cut). Transactions also flow through its WeChat Pay system, boosting monetisation touchpoints.
The Hunyuan-Game is an AI toolset — trained on its massive gaming IP library and user behaviour data — that helps its own game studios as well as third-party game designers enhance and shorten game development time for faster launches and at lower costs. For instance, AI could create smart NPCs (non-playable characters) to converse naturally with players and respond with adaptive dialogues, dynamic plots and branching storylines for more immersive and interactive gameplay. More advanced AI and agentic AI tools could help users, not just developers, create and monetise their own characters, accessories and games with little or no coding expertise. Gaming is the single-largest revenue generator for the company, and AI-powered creative tools will help Tencent retain its competitive edge over peers.
Trip.com
For more stories about where money flows, click here for Capital Section
Our most recent addition to the Absolute Returns Portfolio is Trip.com, which runs the largest online travel booking app in China. Its Ctrip and Qunar brands cater for the domestic market, commanding an estimated 50% market share. Trip.com is its international booking platform and Skyscanner is its real-time price comparison site for airlines and bookings. We believe the company has robust long-term prospects, underpinned by both domestic as well as inbound-outbound international traffic. Geographically, China accounted for 86% of total revenue last year — primarily from accommodation reservations and transportation ticketing.
The travel industry has a long growth runway, as consumers are increasingly switching from spending on goods to experiences. This is a global trend. The China travel industry, especially, has huge untapped growth prospects — the country is vast, offering a wide range of experiences for both domestic and foreign travellers. Meanwhile, less than 15% of the population are estimated to hold a passport currently. Outbound bookings in 2024 have recovered to more than 120% of pre-Covid-19 pandemic levels. Inbound bookings doubled on-the-year, no doubt spurred by the Chinese government’s visa-free entry agreements with multiple Asean countries, including Malaysia. Asia-Pacific makes up some 70% of the company’s international platform revenue, which in turn accounted for about 10% of total revenue last year. The Trip.com platform is still in the red but expected to break even by 2026/27 on the back of robust growth in regional travel demand.
The company has been rapidly integrating AI-related tools into its operations to enhance user experience, as well as to improve efficiency, supply chain quality and monetisation opportunities. For instance, its enhanced TripGenie chatbot is powered by LLMs and can decipher complex user requests in addition to standard customer service calls; offer personalised recommendations and comprehensive travel plans; and handle bookings, refunds and changes — cutting costs and speeding up response times. It can further recommend tourist spots and shopping at destination using location-based services, translate foreign languages with cultural context, and even provide booking links, images or city maps — all within the human-like conversational interface.
TripGenie itself has gone through several iterations since its debut in February 2023, mirroring the rapid advancements in GenAI, evolving from a reactive tool to proactive AI assistant that can anticipate and solve travel challenges. The chatbot now taps into location-based services, user behaviour as well as chat, browsing and bookings histories to offer real-time recommendations — before the query is asked, and even when the user is not travelling. For example, alerting users to local festivals aligned with their interests or suggesting flights based on past bookings.
Trip.com, of course, has also applied technology to streamline its own operations. Gross margins have held steady over the past decade while net margin and profits have recovered well above pre-pandemic levels on the back of record-high revenue last year. The stock is currently trading at 18 times estimated earnings — compared to peer Booking Holdings’ 24 times — and forward EV/Ebitda of 14 times. It is also sitting on total net cash of roughly RMB36 billion.
Tencent, Alibaba and Trip.com are all strong Chinese companies with big cash piles on their balance sheets, commanding market shares in their respective industries, have good growth prospects and are continuously improving productivity by embracing the latest technologies in their businesses. Indeed, many apply technologies that have surpassed those of the West. And, yet, all inevitably trade at discounted valuations to global peers. What is the reason for this apparent systemic undervaluation? We will explore this conundrum in a future article.
The Absolute Returns Portfolio fell 0.4% for the week ended June 4, paring total returns since inception to 24.2%. The top three gainers were SPDR Gold (+2.3%), Tencent (+1.1%) and US Steel Corp (+0.8%); and the big losers included Berkshire Hathaway (-2.4%), CrowdStrike Holdings (-1.8%) and Trip.com (-1.6%).
The Malaysian Portfolio also ended the week 0.4% lower. Shares in United Plantations fell 3.3% while Insas Bhd – Warrants C and Kim Loong were unchanged. Total portfolio returns now stand at 184.9% since inception. This portfolio is outperforming the benchmark FBM KLCI, which is down 17.6% over the same period, by a long, long way.
Tong’s AI Portfolio, on the other hand, bucked the downtrend, gaining 1% for the week and lifting total portfolio returns since inception marginally back into positive territory. The big gainers were Workday (+4.0%), SAP (+3.8%) and Datadog (+2.9%); and the three losing stocks were RoboSense Technology Co (-5.1%), Alibaba (-1.0%) and ServiceNow (-0.8%).
Disclaimer: This is a personal portfolio for information purposes only and does not constitute a recommendation or solicitation or expression of views to influence readers to buy/sell stocks, including the particular stocks mentioned herein. It does not take into account an individual investor's particular financial situation, investment objectives, investment horizon, risk profile and/ or risk preference. Our shareholders, directors and employees may have positions in or may be materially interested in any of the stocks. We may also have or have had dealings with or may provide or have provided content services to the companies mentioned in the reports.
