Amazon +20% This Month! Anthropic $25B Mega Deal: Break $300 Like Google’s Re-Rating?
After Monday’s close, Amazon announced it will invest up to an additional $25 billion into Anthropic, while Anthropic simultaneously committed to spending $100 billion on AWS over the next decade. $Amazon.com(AMZN)$ rose more than 2.4% pre-market to $254. With earnings season approaching (April 29), one core question is now on the table:
This time, will Amazon become the next “Google-style re-rating”?
Over the past two weeks, Amazon has rebounded strongly by 18%, turning its year-to-date performance positive at +7.5%, completely shaking off its label as “one of the worst-performing Mag 7 stocks” at the start of the year. Now, the weakest names are Microsoft and Tesla — Amazon is no longer on that list.
1. What does Anthropic mega deal really mean for Amazon — AI Trade or AI Bubble?
Amazon’s total investment in Anthropic has now reached around $33 billion (previously $8B + this $25B), making it the largest single external bet in the global AI model race.
In return, Anthropic commits to spending over $100 billion on AWS over the next 10 years, and locking in 5 gigawatts of compute capacity for training and deploying Claude — including nearly 1GW of Trainium2/Trainium3 capacity coming online by the end of this year.
AWS customers can now access the full native Anthropic Claude interface directly within the AWS console, without additional contracts or billing relationships.
-
Claude Opus 4.7 is already live on Bedrock, with SWE-bench Pro score of 64.3% and Verified score of 87.6%
-
The Mythos model achieves 87% accuracy under a 3M token context window
Is This a “Circular Trade”?
Yes, some people are concerned. Is this similar to last year’s OpenAI–NVIDIA dynamic?
Amazon invests in Anthropic, and Anthropic uses that money to buy AWS — it looks like money moving from one hand to the other.
But ultimately, the market will judge based on one thing: Whether AWS revenue actually accelerates in a meaningful way
2. Amazon’s two investment thesis — Which has more upside?
AI Compute: $15B Run Rate, Still Supply-Constrained
AWS AI revenue has already surpassed a $15 billion annual run rate. Goldman describes it as “supply-constrained” — Amazon doesn’t lack customers, it lacks the ability to build data centers fast enough.
Its in-house chips Trainium and Inferentia now represent a $20 billion annualized scale, meaning Amazon is no longer fully dependent on NVIDIA in the AI compute stack, and is building its own moat.
Anthropic + Bedrock: Becoming Standard in Enterprise Procurement
This is currently the most direct commercialization path.
Now, 1 in 3 U.S. companies is paying for Anthropic models, meaning this has moved from “tech experimentation” to a fixed line item in enterprise IT budgets.
For AWS, this is one of the stickiest layers of recurring service revenue.
3. Can It Break $300? Referencing $Alphabet(GOOG)$ Re-Rating Playbook
Google’s re-rating this year was driven by a key shift: The market regained confidence in Gemini and Cloud AI commercialization, moving from a bearish narrative of “lagging OpenAI” to a bullish narrative of “ads + cloud dual engines.”
That shift led to a meaningful valuation expansion. Amazon is now at a similar narrative inflection point.
Over the past year, concerns focused on: AWS growth, AI capex, Anthropic investment returns.
But the recent price action suggests that, as AI revenue continues to be validated, the market is starting to assign Amazon a higher AI infrastructure premium.
Key Difference vs Google
Google has clearer AI monetization pathways(e.g., AI Overview ads, paid Gemini users)
Amazon’s AI revenue is currently more reflected in AWS compute consumption, and its “last mile” monetization is less clear.
That’s why its rally hasn’t been as strong as Google’s.
From the current pre-market price of $254, there is about 18% upside to $300.
Discussion
-
Anthropic’s $100B AWS commitment — do you see this as real money or an AI bubble?
-
Can Amazon replicate Google’s AI re-rating and break $300 after earnings?
-
In the AI infrastructure space, who really has pricing power now — AMZN, NVDA, or MSFT?
Leave your comments to win tiger coins~
r SG users only, welcome to open a CBA today and enjoy access to a trading limit of up to SGD 20,000 with unlimited trading on SG, HK, and US stocks, as well as ETFs.
🎉Cash Boost Account Now Supports 35,000+ Stocks & ETFs – Greater Flexibility Now
-
Find out more here.
Complete your first Cash Boost Account trade with a trade amount of ≥ SGD1000* to get SGD 688 stock vouchers*! The trade can be executed using any payment type available under the Cash Boost Account: Cash, CPF, SRS, or CDP.
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.
2. AMZN > US$300 is possible, but earnings must show AWS acceleration + AI monetisation + margin discipline. Without clear conversion of AI demand into profit, re-rating may stall.
3. Pricing power now:
NVDA: strongest (chips + ecosystem, high margins)
MSFT: strong via enterprise bundling (Azure + Copilot)
AMZN: improving, but still proving ROI
Summary: NVDA leads pricing, MSFT leads monetisation, AMZN is the upside play.
I’m more bullish on the Anthropic + Bedrock layer than pure compute. Compute is capital-heavy, but enterprise lock-in is the real moat. As companies embed Claude via Amazon Web Services, switching costs rise — similar to $Microsoft(MSFT)$ ’s model.
On $300, a re-rating like $Alphabet(GOOGL)$ is possible but needs clearer monetization. If AWS growth re-accelerates, sentiment can shift. For now, Nvidia still holds the strongest pricing power. I also think the next key catalyst will be whether AI revenue starts contributing meaningfully to AWS margin expansion.
@Tiger_comments @TigerStars @TigerClub @Tiger_SG
2. Amazon can sign a deal with google if necessary to obtain technology
3. There is limited purchasing power among the biggest consumers of technology, with suppliers with greater power
Nvidia (NVDA) is the only player with absolute pricing power right now. While MSFT and AMZN control the platforms, they are currently "taxed" by Nvidia's hardware margins. Nvidia dictates the pace of the entire industry because its software stack (CUDA) creates a moat that cloud providers cannot yet bypass. Among the cloud giants, Microsoft holds more pricing power than Amazon because of its direct "software-as-a-service" monetization through Copilot, whereas Amazon is still in the capital-intensive phase of building out its hardware-based AI moat.
Amazon can break the 300 barrier, but only if it proves that its "AI-to-Retail" integration is driving margin expansion. Unlike Google, which re-rated on search dominance, Amazon needs to show that AI is significantly lowering its logistics costs while simultaneously accelerating AWS revenue back toward 20% growth. If AWS margins expand because of custom silicon adoption, the market will shift from valuing Amazon as a retailer to a high-margin AI utility, making 300 a realistic target.
This commitment is a strategic accounting maneuver rather than "real money" in a traditional sense. It functions as a circular credit system where Amazon provides capital and infrastructure, and Anthropic commits to spending that same capital back on AWS services and Trainium chips. While it signals a massive bet on AI longevity, the 100B dollar figure is an "AI bubble" valuation of future compute hours designed to lock Anthropic into the Amazon ecosystem and inflate AWS's long-term backlog.
Or openai committing to hynix for huge amounts of dram, amounting to 40% of global supply.
Everyone is trying to cash in on the market's insatiable hunger and the big names know that any news will result in more buying.
The fact that openai had to shutter sora shows the run is obviously unsustainable, which shocks me that openai isn't pushing to IPO quickly to make the big bucks before the bubble pops... cash in on the brand recognition before it's too late.
Again, I pity the bag holders who buy in and refuse to take profit.
All these investments, just feel like stunts to keep the AI gold rush going on, further feeding the bubble.
The value of AI as we know it, is simply not what is shown on the market, investors are jumping at any and almost every excuse that has to do with AI, best to keep taking profits on the way to the pop, or better yet hedge against it.
This pop will be spectacular, I wish it would wipe out Musk, but I believe he has left investors holding on to the figurative bomb and will come out relatively ok [Glance] [Facepalm]
In the market any business used to only have two ways of funding themselves: debt or equity issue. Circular funding is a new innovation. In fact in some cases it is a grey area and may even be illegal in some countries. It is quite opaque and not shown in the books completely . This has become a new model in the AI world started by Open AI and could be a reason for future demise of the AI environment.
Like Google, Amazon is being re evaluated not just as a retailer but as an AI infrastructure play. Analysts at Jefferies and Barclays have already set USD 300 price targets, citing AWS re-acceleration and the efficiency of in-house AI chips.
Earnings Outlook: For the April earnings, TD Cowen predicts revenue & operating income to exceed consensus due to accelerating AWS growth.
If Amazon meets these projections and its forward P/E expands to the mid 30s, AMZN could climb toward USD 312.
I believe that Amazon is the most undervalued of the Magnificent 7. Its valuation is at its lowest point in nearly 2 decades.
It is time to start buying Amazon.
@Tiger_SG @Tiger_comments @TigerStars @TigerClub @CaptainTiger
$NVIDIA(NVDA)$ still holds the ultimate pricing power at the hardware layer. Demand for GPUs continues to outpace supply, allowing NVIDIA to maintain huge operating margins of over 60%.
$Amazon.com(AMZN)$ is rapidly gaining ecosystem pricing power. It backs both Anthropic & OpenAI & ensures that regardless of which model wins, the traffic stays on AWS. Its custom Trainium3 chips is 50% cheaper than Nvidia.
$Microsoft(MSFT)$ holds integrated software pricing power. Its deep control over the entire stack - from custom chips to Office 365 Copilot, allows it to scale margins through software rather than just infrastructure rental.
Time will tell who will ultimately hold the key to pricing power.
@Tiger_SG @Tiger_comments @TigerStars @TigerClub @CaptainTiger
The Circular Economics: $Amazon.com(AMZN)$ is investing an additional USD 5 billion with an option for USD 20 billion more into Anthropic.
In return Anthropic pledges USD 100 billion in AWS spending. Effectively Amazon is paying to acquire a USD 100 customer for its proprietary Trainium & Graviton chips.
The Verdict: By backing both Anthropic and OpenAI, Amazon ensures that regardless of which model wins, the traffic stays on AWS. Its custom Trainium3 chips offer a 50% cost reduction over NVIDIA , giving it a unique price performance lever.
The Anthropic deal is real money as it locks in future revenue but a bubble risk if end user demand for AI models fails to materialise.
@Tiger_SG @Tiger_comments @TigerStars @TigerClub @CaptainTiger
但如果AI收入只是增长、利润却被资本开支吃掉，那股价空间会被压住。
能不能上300美元？我觉得关键不在这笔投资本身，而在财报能不能证明两件事：
第一，AI真的在拉动AWS增速，而不是只是在烧CapEx
第二，自研芯片（Trainium）有没有开始改善利润结构
但亚马逊的问题也很现实：它赚的是“用量的钱”，而不是“产品溢价的钱”。对比Alphabet，人家可以直接把AI变成广告收入或订阅，路径更清晰；AWS目前还是偏底层。
先讲结论：这不是简单的AI泡沫，更像是一种“锁需求”的基础设施打法。
这不是泡沫问题，而是“谁先把AI从成本变成利润”的问题。