Amazon stock has been one of the more notable underperformers in the Magnificent Seven this year, weighed down by a capital expenditure announcement in February that caught investors off guard and has yet to fully recover.
The question on Wall Street is whether the selloff reflects a rational reassessment of the company’s financial trajectory or a misreading of how to evaluate what Amazon is actually building.
Evercore ISI analyst Mark Mahaney is in the misreading camp.
He has a price target that implies roughly 50 percent upside from where Amazon shares currently trade, and the core of his argument is that investors are looking at the spending numbers without adequately accounting for the demand numbers behind them.
Amazon’s representative confirmed the news in a statement and no additional details were provided.
The current share price reflects a stock that was already a relative laggard, Amazon gained just 5.2 percent in 2025, the worst performance in the Magnificent Seven, significantly below the S&P 500’s 16.4 percent climb that year.
The stock has continued to slide into 2026 following the earnings report.
What Happened To Amazon In February?
Amazon reported its fourth-quarter 2025 results on February 5. The headline numbers were broadly solid. Revenue came in at $213.4 billion, up 13.6 percent year over year.
Amazon Web Services posted $35.6 billion in Q4 revenue, up 24 percent, its fastest growth rate in 13 quarters, with an annualized run rate of roughly $142 billion.
The company’s North America and International retail segments both hit record operating margins, adjusted for one-time items.
What upended the stock was the guidance. Amazon announced capital expenditures of approximately $200 billion for 2026, against analyst expectations that had clustered around $146 billion to $150 billion.
The gap between $146 billion expected and $200 billion announced is not a rounding error, it is a $54 billion surprise that fundamentally changes the near-term free cash flow picture.
The market’s reaction was sharp. Shares fell approximately 8 percent in premarket trading the morning after the announcement, with additional losses during the regular session.
The stock has remained under pressure since. Amazon’s 2025 capital expenditures were already $131.8 billion, up 58.8 percent from $83 billion in 2024, and free cash flow had already compressed dramatically as a result, falling from $32.9 billion in 2024 to $7.7 billion in 2025, with capex consuming nearly 95 percent of operating cash flow.
The $200 billion 2026 guidance implies free cash flow likely goes negative, some analysts estimate it reaches approximately negative $18 billion, before recovering.
CEO Andy Jassy’s explanation for the spending was framed as demand-driven rather than speculative.
“We are monetizing capacity as fast as we can install it,” he told investors on the earnings call, noting the investment is “predominantly in AWS, because we have very high demand.”
CFO Brian Olsavsky declined to specify a free cash flow floor or payback timeline, saying only that Amazon sees “long strong return on invested capital” and that demand signals justify the pace.
Why Investors Could Be Getting It Wrong
The bull case on Amazon at current prices centers on a specific claim: that the market is pricing the capex without adequately pricing the backlog that underpins it.
Amazon’s AWS contracted backlog stood at $244 billion at the end of 2025, up 40 percent year over year.
That number represents signed contracts from enterprise customers committing to future cloud spending, it is not a pipeline figure or a projection.
It is money that is contractually owed. UBS analyst Stephen Ju projects that performance obligations and backlog could reach approximately $400 billion by the end of 2026, as enterprises accelerate their AI-driven cloud commitments.
The argument Mahaney and others are making is that spending $200 billion to build capacity against $244 billion in contracted backlog, with more accumulating, looks different than spending $200 billion speculatively. The ratio matters.
If the capex is demand-driven and the backlog conversion is reliable, then the free cash flow compression in 2026 is a timing phenomenon, not a structural one.
Jefferies made this point explicitly, calling the negative free cash flow situation “a timing issue, not a structural or permanent one.”
UBS’s Ju describes Amazon as a “coiled spring,” a phrase he has used repeatedly, arguing that neither UBS nor the broader Street is adequately pricing in a scenario where AWS revenue doubles by 2028 following the step-up in capital intensity.
He projects AWS growth rates doubling from 19 percent in 2025 to 38 percent in 2026, with 2027 GAAP earnings per share reaching approximately $14.
At roughly 14 times those earnings, the stock would be trading at a discount to megacap peers despite what he characterizes as a superior earnings growth profile. His current price target is $301.
Jefferies pointed to a separate valuation argument. Amazon’s stock is currently trading at approximately 11 times 2026 EV/EBITDA, against a peer average of roughly 14 times.
The firm’s analysts described this as “too cheap for what it is,” noting that Amazon has matched Walmart in retail gross merchandise volume while delivering higher growth and margins, yet trades at roughly 10 turns below Walmart on the same metric.
The Counterargument
The skeptical case is not without substance, and even some of the bulls acknowledge it.
Mahaney himself flagged that 2026 is likely a negative free cash flow year, and that this could leave the stock range-bound until investors can see a clearer free cash flow rebound in 2027 or more obvious revenue acceleration.
His price target of approximately $285 is down from a prior target of $335, reflecting real concessions to the bear case even from a bull.
The core concern among skeptics is not that the demand is fabricated, the backlog figures are real, but that the margin of error on a $200 billion capex bet is narrow.
If AWS growth decelerates below expectations, if the backlog-to-revenue conversion slows, or if the AI infrastructure build-out encounters unexpected costs, the free cash flow recovery projected for 2027 could slip further.
At that point, the stock’s multiple would need to compress further to reflect a longer payback period on the investment.
AWS is also growing more slowly than its primary competitors. Google Cloud grew 48 percent in the most recent comparable period.
Microsoft Azure grew 39 percent. AWS grew 24 percent, which is fast in absolute terms and a reacceleration from prior quarters, but still meaningfully below peers in a market where relative growth rates matter for investor sentiment.
That gap has contributed to Amazon being viewed as an “underappreciated” AI winner rather than a leading one.
DA Davidson analyst Gil Luria characterized the overall investor posture toward big tech AI spending as “very healthy skepticism,” adding that investors are “going to wait to see the returns that the companies are promising them in order to increase the price they’re willing to pay for these stocks.” That caution is not irrational.
Where Wall Street Currently Stands
Of the 67 sell-side analysts who cover Amazon, 96 percent rate it a Strong Buy or Buy.
The average price target is approximately $280 to $295, implying 27 to 35 percent upside from current levels depending on where shares are trading. Morgan Stanley maintains an Overweight rating with a $300 target.
JPMorgan’s Doug Anmuth estimates 30 percent upside and cited Amazon’s newly announced $38 billion, seven-year cloud services deal with OpenAI as a potential additional revenue driver.
Bank of America maintains a Buy with a $275 target.
Retail investors tracked by prediction market Polymarket are more skeptical than the analyst consensus, a not uncommon pattern in situations where the professional view has become heavily one-sided.
The 50 percent upside figure attached to Mahaney’s thesis should be understood in context. It is a twelve-to-twenty-four month forward projection based on a specific scenario.
AWS growth reaccelerates as contracted backlog converts to revenue, Trainium AI chips gain enterprise adoption and improve margins, advertising revenue continues to grow at high rates, and free cash flow inflects sharply positive by late 2027.
Each of those assumptions is individually defensible, and collectively they could produce the result Mahaney describes.
They could also fail to materialize on schedule. A $200 billion capital commitment made in an environment of rapid AI development carries execution risk.
The technology landscape is shifting faster than analysts can model it. The free cash flow picture for 2026 is genuinely negative.
The stock’s recent underperformance against peers is not entirely explained by misreading, some of it reflects real uncertainty about the payback timeline on the largest infrastructure investment in Amazon’s history.
Whether the market is looking at Amazon’s AI spending all wrong, or whether the market has correctly identified a two-year window of compressed returns before the thesis plays out, is the central question for Amazon investors in 2026.
Analysts are nearly unanimous on the direction. The timing is where the disagreement lives.
