Amazon (AMZN) shares exploded 15.32% to close at $271.58 on July 31, 2026, following the release of its second-quarter earnings report. The stock has now climbed 17.66% year-to-date, significantly outpacing the SPDR S&P 500 ETF (SPY), which is up roughly 9.55% over the same period. While the headline numbers scream success, a deeper dive into the financials reveals a complex picture that has Morgan Stanley analysts flagging a major blind spot in the prevailing bull thesis.
Headline Numbers vs. Underlying Reality
On the surface, Amazon’s Q2 2026 results were spectacular. Revenue surged 20% year-over-year to $200.6 billion, while net income skyrocketed 243.95% to $62.6 billion. However, the company disclosed a critical caveat: Q2 net income included $53.4 billion in non-operating pre-tax other income, primarily driven by a theoretical valuation increase in its stake in Anthropic, a private artificial intelligence company.
This accounting gain creates a distorted view of operational profitability. Because Anthropic is not publicly traded, its valuation is subjective and illiquid. As the article notes, “the valuation may or may not be of practical value.” Stripping out this one-time paper gain paints a far less rosy picture of Amazon’s core earnings power.
The Capital Expenditure Tsunami
Perhaps the most alarming data point for long-term investors is the trajectory of capital expenditures (CapEx). During the earnings call, CEO Andy Jassy revealed the company now expects to spend approximately $220 billion in cash CapEx in 2026, up from a prior estimate of $200 billion. Jassy explicitly stated that even at this elevated level, Amazon “will still not have enough capacity to meet all the demand we have in 2026, and I believe this dynamic will also be true in 2027, too.”
This spending spree is already crushing free cash flow (FCF). Q2 2026 FCF swung to a negative $7.6 billion, a dramatic reversal from positive $1.2 billion in Q1 2026 and a robust $18.2 billion in Q2 2025. Morgan Stanley’s team, led by analyst Brian Nowak, has subsequently raised their 2027 and 2028 CapEx estimates to $325 billion and $350 billion, respectively.
Funding the AI Arms Race: OpenAI and Anthropic
Amazon is not just building its own infrastructure; it is bankrolling the competition. The company’s Form 10-Q confirmed it invested the final $21.3 billion tranche into OpenAI (Anthropic’s primary rival) in late July, bringing its total commitment to massive levels. This creates a paradoxical dynamic: Amazon spends hundreds of billions on CapEx to serve AI demand, while simultaneously funding the very companies—OpenAI and Anthropic—that drive that demand but remain unprofitable.
Morgan Stanley’s bull case relies heavily on AWS’s “Rule of 70” (a variation of the Rule of 40 where growth plus margin exceeds 70%), citing AWS’s 36.8% year-over-year growth acceleration. However, the analysis arguably ignores that the two primary drivers of this AWS growth—OpenAI and Anthropic—are capital-intensive, loss-making entities propped up by Amazon’s own balance sheet and potentially unsustainable credit markets.
Credit Market Signals and the Meta Threat
The fragility of the AI funding chain is evident in credit markets. SoftBank, a major OpenAI investor, has reportedly struggled for months to secure a $10 billion loan collateralized by its OpenAI shares, with talks stalling even at a reduced $6 billion ask. Meanwhile, The Wall Street Journal reports that Nvidia (NVDA) is in talks to provide a roughly $250 billion backstop for OpenAI’s data center projects, which would be built by SoftBank Energy. This suggests private equity and debt markets are tapping out, forcing strategic partners like Nvidia to underwrite the infrastructure build-out.
Adding competitive pressure, Bloomberg reported that Meta Platforms (META) is developing a cloud business to sell its excess AI compute capacity. CEO Mark Zuckerberg hinted at this on their Q2 call, stating, “We’re getting a lot of offers for compute at a significant premium over what we paid for it.” If Meta enters the hyperscaler market aggressively, a price war could erode AWS margins and invalidate the “Rule of 70” thesis.
Analyst Consensus and Downside Risks
Despite these structural concerns, Wall Street remains overwhelmingly bullish. According to MarketBeat, 56 of 59 analysts rate AMZN a Buy, with an average price target of $322.12. Morgan Stanley’s base case target sits at $335 (25x multiple), but their bear case sits at a jarring $215.
Key downside risks identified by Morgan Stanley include:
- Investments step up and continue for longer than expected.
- Merchant margins worsen more than anticipated.
- AWS revenue decelerates and/or margins decline.
Conclusion: A Verdict with a Blind Spot
Amazon’s Q2 report was a masterclass in revenue growth, but the quality of earnings is deteriorating. The market is currently pricing in perpetual AI-driven growth for AWS while ignoring the circular funding mechanism propping it up and the looming threat of new, well-capitalized entrants like Meta. For investors, the 15% pop may represent a peak in optimism rather than a floor for future gains. The “blind spot” isn’t just a minor oversight—it’s the structural foundation of the bull case.
Frequently Asked Questions (FAQ)
Why did Amazon’s net income jump 244% in Q2 2026?
The surge was largely driven by a $53.4 billion non-operating pre-tax gain from the theoretical revaluation of its investment in Anthropic, a private AI company. Excluding this paper gain, operational net income growth was significantly lower.
What is the “Rule of 70” cited by Morgan Stanley regarding AWS?
The Rule of 70 is a variation of the SaaS “Rule of 40.” It posits that a healthy growth company’s revenue growth rate plus its profit margin should exceed 40% (or 70% in this aggressive variation). Morgan Stanley argues AWS hits 70, justifying a premium valuation, but critics note this metric ignores the source of that growth (heavily subsidized AI training workloads).
How does Meta’s potential cloud entry threaten Amazon?
Meta has built massive AI infrastructure for its own needs (Reels, LLMs) and reportedly has excess capacity. If they launch a hyperscaler cloud offering to monetize this spare compute—potentially at lower margins than AWS—it could trigger a price war in cloud infrastructure, compressing AWS operating margins and slowing revenue growth.
