The signal is loud and clear on-chain: Tencent is burning capital at a pace that would make most DeFi ponzis blush.
*Daiwa just slashed their price target—from HKD 700 to HKD 670—but here's the kicker: they raised their AI CapEx forecast by a staggering 67%.* We're talking about a projected ¥181 billion (roughly $25 billion USD) by 2026. That's not a tweak. That's a declaration of war.
Let me paint you the scene. I've been in this industry since the Fomo3D code audit days. I've seen capital get thrown around like confetti at a bull market party. But this? This is different. We're not looking at a company speculating on JPEGs or yield farms. We're watching one of the world's largest internet behemoths pivot from a 'traffic monetization' machine into a 'compute capital' empire. The code didn't foresee this level of commitment to infrastructure.
Daiwa's report is a window into a brutal reality: Tencent is sacrificing near-term profit for the long-term throne of AI infrastructure. They predict a 1-6% earnings dent. That's the price of admission to the next era of the internet.
Let's dissect the core of this. The narrative is classic 'short-term pain for long-term gain,' but the numbers are so aggressive they scream a paradigm shift.
From 'Selling Water Cups' to 'Selling Water'
Remember how we used to analyze DeFi protocols? You'd look at TVL, at the tokenomics. With Tencent, you have to look at their product matrix: WeChat (the super-app), Gaming (the cash cow), Tencent Cloud (the B2B play).
The old model was simple: extract value from user attention (ads, games) and enterprise server rent (cloud). But that's a 'water cup' business—finite, competitive, and low-margin.
Daiwa's report, albeit from a traditional finance lens, is screaming that Tencent is building an AI 'water pipeline'. The massive CapEx isn't just for training bigger models. It's for inference—the ability to serve AI to billions of users in real-time.
We didn't see how quickly the narrative would flip. The assumption was that AI inference would be commoditized. But Tencent is betting that scale creates a new kind of moat. They are not just buying GPUs; they are building an AI operating system for their entire ecosystem.
The Unreported Angle: The 'Profit Buffer' Danger
Everyone is talking about the CapEx number. The contrarian angle is the stress test on their core businesses.
I've been in rooms where the vibe shifts from 'bullish' to 'toxic' in seconds. For Tencent, their 'Luna' moment is the potential failure of their Gaming and Advertising revenue to act as a sufficient profit buffer.
Daiwa's report tacitly assumes that Gaming (despite a high base) and Ads (despite macro headwinds) will continue to generate a massive cash flow that can be funneled into the AI fire. But what if they don't?
On-chain, we call this a 'liquidity crunch.' If Gaming revenues slip, Tencent's AI spending spree could become a self-cannibalizing loop. They would have to either slow the investment (losing the AI arms race) or start burning their balance sheet (a massive red flag for the stock).
The code didn't predict this vulnerability. The risk isn't that Tencent can't build AI; it's that they might be building it on a foundation of sand if their core cash engines falter.
The 'Insider-Access' Signal: The Chip Supply Chain
I had a private dinner with a hardware supply chain analyst last month in Toronto’s King West district. The vibe was clear: 'China has found its workaround.'
Daiwa mentions 'improved chip supply' as a key driver for the increased CapEx. This is the most bullish signal they are implicitly giving. It suggests that Tencent (and likely Alibaba/Bytedance) have secured alternative supply lines for high-performance GPUs (like NVIDIA's H100/B200) or are successfully scaling with domestic alternatives like Huawei's Ascend.
This is the 'black box' of the thesis. If the chip supply narrative is true, Tencent has a 2-3 year runway to build its AI lead without fear of a sanctions hammer. If it's a bluff, they are building a castle on a foundation of geopolitical quicksand.
The Strategic Logic: 'First Mover' vs. 'Efficient Mover'
In the DeFi world, we learned that being a first mover is overrated. Being capital-efficient and capturing audience is what matters. Uniswap didn't win because it was first; it won because it was simple and community-driven.
Tencent is betting on being an 'Efficient Mover' —using its WeChat ecosystem (the community) as a distribution channel to flood the market with AI-powered products.
Let's break down their three value pyramids:
- The 'Killer Utility' (SaaS Evolution): This is their API play. They will sell 'Mixun' (their LLM) inference-as-a-service. It's like Amazon Web Services (AWS) in 2006, but for AI. The cost? High. The value? Potentially infinite if they lock in enterprise customers who will become dependent on their model for their own operations.
- The 'Product Reskin' (Gaming & Ads): This is the low-hanging fruit. AI will be injected into their games (smarter NPCs, dynamic levels) and their ad engine (hyper-personalization). In the short-term, this is how they will soothe investors. Daiwa sees this as the 'safe' bet.
- The 'WeChat AI Agent' (The Super-Platform): This is the moonshot. Imagine WeChat transforming from a 'service aggregator' to a 'service generator.' You can just tell the app to 'book me a dinner and a car,' and an AI agent orchestrates the entire process via mini-programs. This is the end-game. If Tencent can pull this off, the CapEx will be the best money they ever spent.
The Contrarian Take: The 'Code Audit' Trap
I spent years auditing smart contracts. The biggest risk isn't a bug in the code; it's a flaw in the assumption.
The assumption here is that 'scale' is the answer. But in the AI world, we are seeing the rise of smaller, more efficient models (like Mistral's or the new open-source Mixture-of-Experts models). If the industry pivots away from massive, centralized compute farms towards more distributed, efficient architectures, Tencent’s ‘Golden Palace’ of GPU farms could suddenly become a depreciating liability.
The Empathy Layer: The Human Cost
Every massive pivot has a human cost. I saw it during the Terra collapse—the emotional burnout. For Tencent, the cost is not just the $25 billion. It's the organizational trauma of transforming a 100,000-person company from a 'social media company' to an 'AI industrial conglomerate.' Engineers who were proud of building WeChat will now be writing tools for GPU orchestration. This shift in identity is harder than any financial metric.
The Takeaway: The Clock is Ticking
Daiwa's report is a data point, not a promise. It reveals that Tencent has decided to buy a ticket to the AI party at almost any cost. The key question isn't 'Is this good or bad?' The key question is 'Can they execute?'
We need to watch two signals:
- The Monetization Pulse: Can their Tencent Cloud AI API show significant revenue by H2 2026 as Daiwa predicts? If not, the narrative breaks.
- The 'Liquidity' of their Core Business: If gaming or advertising revenue starts to decline meaningfully, the CapEx story becomes a liability.
The code didn't predict this paradigm shift. But the narrative is clear: Tencent is building the new oil pipeline. The only question is whether the oil will be worth more than the pipe.
-Ben White, Crypto News Editor-in-Chief Signal from the chain: Your move, Alibaba.
