Amazon Stock Price Forecast: AMZN Pulls Back to $268 as $465B AWS Backlog Powers a $368 Price Target

Amazon (AMZN) is trading lower on Thursday, changing hands around $267.98, down roughly 0.80% on the session, or about $2.15, from a previous close of $270.13. The intraday range has been contained between $267.03 and $270.78, and the stock is sitting just beneath its 52-week high of $278.56 against a yearly low of $196.00 — which frames the move correctly: this is a modest pullback within a powerful uptrend, not a breakdown of one. The market capitalization stands at roughly $2.88 trillion, which puts the company within genuine striking distance of the $3 trillion threshold, a level only a tiny handful of names have ever touched. The pullback is happening on a day when much of large-cap tech is actually bid — Nvidia is up nearly 4%, Broadcom close to 4%, Microsoft and Alphabet both modestly higher — so Amazon's red close stands out as stock-specific digestion rather than a sector-wide retreat. The trailing price-to-earnings ratio sits at 32.04, average volume runs near 46.67 million shares, and short interest is strikingly low at 0.66%, which tells you the bearish conviction against this name is minimal even after a run that has carried it up roughly 28% off its March low. For a company of this scale to be trading this close to all-time highs while the broader debate around AI capital spending grows louder by the week is itself the story worth unpacking.

The Q1 Print That Reset the Earnings Bar

The first quarter Amazon delivered was, by almost any measure, a monster. Revenue came in at $181.52 billion, a 16.61% year-over-year improvement and a beat of roughly $4.35 billion over consensus. GAAP earnings per share landed at $2.78, crushing the Wall Street forecast of $1.65 and standing as effectively a quarterly record. Net income reached $30.26 billion, up a striking 76.65% year-over-year, with the net profit margin expanding to 16.67%. Operating income of $23.9 billion came in above the high end of company guidance and produced an operating margin of 13.1% — the highest efficiency mark in Amazon's entire history as a public company. EBITDA hit $42.80 billion, up 31.01%. There is, however, an important asterisk that any serious read of the quarter has to acknowledge: the headline EPS figure was meaningfully distorted by a $16.8 billion non-cash gain tied to Amazon's equity stake in the AI lab Anthropic. Strip that one-time gain out, and the underlying EPS lands closer to $1.5 — which is roughly in line with what consensus was actually modeling. That distinction matters, because it separates the genuine operational outperformance from the accounting noise, and the operational story is strong enough that it does not need the embellishment. The geographic breakdown reinforced the point: North America contributed $104 billion in revenue, up 12% year-over-year, while the International segment reached $39.7 billion, up 19%.

AWS Reaccelerates and the Backlog Becomes the Whole Argument

The single most important development in the quarter was the reacceleration of Amazon Web Services. The cloud division grew 28% year-over-year to $37.6 billion in revenue — the fastest pace in fifteen quarters — and threw off operating income of $14.2 billion, which represented a roughly 50% annual advance and added nearly $3 billion year-over-year. But the revenue growth, impressive as it is, is not actually the headline. The headline is the backlog. AWS is now sitting on a disclosed forward-demand book that runs past $364 billion before layering in Anthropic's incremental $100 billion-plus commitment, and on top of that there are Trainium chip revenue commitments exceeding $225 billion. Taken together, that is somewhere in the neighborhood of $465 billion in pre-sold demand — a figure equivalent to roughly 2.4 times the entire current AWS revenue base. A coverage ratio of that magnitude has simply never been seen at a cloud business of this scale; for context, at the previous cycle peak no hyperscaler disclosed a backlog ratio above 1.7x, and Amazon's ratio has grown 40% year-over-year. The composition is what makes it durable rather than just large. The Anthropic agreement alone secures 5 gigawatts of Trainium capacity over a ten-year span, with Project Rainier scaling from more than 500,000 Trainium2 chips today toward a combined 1 gigawatt of T2/T3 capacity by the end of 2026. That custom-silicon business is already running at a $20 billion-plus revenue rate, growing in the triple digits, with Trainium2 largely sold out — and CEO Andy Jassy has gone on record saying that as a standalone business it would be worth at least $50 billion.

The $200 Billion Capex Question and the Free Cash Flow It Is Eating

Now for the part of the story that genuinely unsettles a portion of the market. Amazon's capital expenditure program for the cycle is an immense $200 billion, and the strain it is placing on free cash flow is severe and unmistakable. Free cash flow for the twelve months ending in March collapsed to just $1.23 billion, a brutal fall from $25.9 billion in the prior-year period, and the quarterly free cash flow figure actually swung negative to roughly -$13.77 billion. Cash from investing ran to -$64.21 billion in the quarter. For value-oriented investors trained to watch the cash line, those numbers are alarming on their face. But the critical distinction — and it is the one the bullish case rests on entirely — is that this is capital expenditure being committed against contracted demand rather than speculative demand. Amazon's management has stated it already holds customer commitments for a substantial portion of the spend, and that it expects the program to yield compelling operating margins and return on invested capital. The honest framing is that the free cash flow situation will likely get worse before it gets better, and that a meaningful chunk of this spend is tied to GPUs and custom accelerators with relatively short depreciation cycles of three to five years, which adds pressure to monetize quickly. This is the heart of the bull-bear divide on the stock: the bears see a 95% year-over-year compression in free cash flow, while the bulls see $465 billion in disclosed forward demand that has effectively de-risked the entire build-out.

How Amazon's Spending Stacks Up Against the Other Hyperscalers

The capex anxiety is not unique to Amazon — it is a sector-wide phenomenon, and the scale of it is staggering. The combined capital expenditure of Amazon, Microsoft, Alphabet, and Meta sat just above $200 billion in 2024, and two years later it is on track to approach $700 billion, an AI infrastructure arms race with no modern precedent. Hyperscaler capex now amounts to roughly 2.2% of U.S. GDP. Combined free cash flow at those four companies slipped to about $200 billion last year from $237 billion in 2024 as the build-out accelerated. But within that group, Amazon's spending arguably has the cleanest justification. Alphabet is set to spend between $180 billion and $190 billion against a cloud business that is only about 37% the size of AWS — meaning Google is committing more absolute dollars to chase a materially smaller revenue base. On the Microsoft side, the recent capex increase has been linked more to higher component prices than to brand-new data center projects. That leaves AWS as the only hyperscaler whose capex line is being committed against an explicitly disclosed multi-year revenue book of business, and the only one with both a mature custom-silicon stack operating at scale and a customer base anchored by both the largest enterprise wallets and a frontier AI lab. There is even a historical parallel some have drawn: AT&T's Bell System kept spending right through the Great Depression because stopping was more dangerous than continuing — a network only holds its value if it keeps expanding — and the modern hyperscaler logic rhymes with that, where capex becomes the last line item to cut rather than the first.

The Advertising Engine That Quietly Became a Top-Four Global Player

Lost beneath the AWS and capex headlines is a business that deserves far more attention than it gets. Amazon's advertising segment has now generated more than $70 billion in revenue over the trailing four quarters, growing at a 24% year-over-year clip. That scale makes it the fourth-largest digital advertising business on the planet, trailing only Google, Meta, and TikTok — a remarkable standing given how differently Amazon's commerce-anchored ad model works compared to those social and search peers. More importantly, advertising is now the highest-margin revenue segment in the entire company, carrying incremental margins in the neighborhood of 80%. Numerous partnerships announced through the quarter are set to extend the segment's reach well beyond Amazon's owned-and-operated inventory, and at its current trajectory the business is on track to clear $90 billion in annual sales by 2027. This is the piece of the Amazon story that genuinely changes the quality of the earnings stream. The old framing of Amazon was a high-margin AWS division bolted onto a thin-margin retail operation. The current reality is AWS running near 40% margins, advertising compounding at 80% incremental margins, and even the core retail business now throwing off margins north of 7% — a fundamentally higher-quality profit mix than at any prior point in the company's history.

Valuation: A Richer Multiple That the Earnings Quality Arguably Earns

There is no getting around the fact that Amazon has become more expensive. The non-GAAP price-to-earnings multiple now sits around 32.4, which is higher than Alphabet at roughly 30.2 and well above Meta at about 20.2. On the surface, that looks like a premium that demands scrutiny. The counterargument is that the multiple is defensible precisely because the underlying earnings are higher quality than they have ever been — the combination of a 40%-margin cloud business, an 80%-incremental-margin advertising engine, and a retail operation finally generating real profit is simply not the same company that carried a similar multiple a decade ago. The forward-looking math from the bullish camp runs roughly as follows: applying a sustained 32x multiple on an estimated 2027 EPS near $11.5 produces a price target of around $368, which would imply something close to a 35% return from current levels. A more conservative framework — assuming roughly $9.87 of normalized non-GAAP EPS for fiscal 2027 and applying a 28x multiple — lands just shy of $280, which is much closer to where the stock trades today and suggests shares are near fair value rather than deeply undervalued. The sell-side has clearly been moving in the bullish direction: over the trailing 90 days there have been 48 EPS upgrades against just three downgrades, a lopsided revision trend that speaks to growing analyst confidence following the Q1 print.

The Technical Picture: A Genuine Breakout Backed by a Golden Cross

The chart has improved markedly, and for a stock sitting near record highs, the technical structure is constructive rather than stretched. Amazon recently broke out to an all-time high, clearing and then negating a previous bearish false breakout from last November. Back in February the stock had formed a symmetrical triangle that actually resolved to the downside, dragging price into a test of a long-term uptrend line — but that test turned into what is best described as a modified island bottom, and the subsequent rally back through the $245 level was the decisive move that flipped the structure bullish. That $245 level now serves as support rather than resistance. The long-term 200-day moving average is rising, which signals the bulls control the primary trend, and in just the last few days a bullish golden cross printed, with the short-term 50-day moving average crossing back above the 200-day. The one note of near-term caution is that the RSI momentum oscillator has been working off technically overbought conditions, so some minor back-and-filling or sideways digestion on the price chart would not be surprising — and Thursday's 0.80% pullback fits comfortably within that expectation. The post-Q1 reaction itself was telling: shares rose just 0.8% in the session that followed the report, the smallest post-earnings move in a year, which suggests the market had already largely priced the strength in.

The Risks That Still Sit Underneath the Bull Case

A complete read of Amazon requires taking the risks seriously, because they are real even if they are not imminent. The most important single variable is AWS growth itself: if the cloud division decelerates below roughly 22% in the coming quarters, the current 32x multiple very likely does not survive, and the stock would face a re-rating lower. The enormous backlog disclosure makes a sharp near-term deceleration unlikely, but it remains the number that matters most. The second risk is the Anthropic relationship, which has become a double-edged exposure — Amazon now carries roughly a $25 billion equity position in the AI lab plus a $100 billion-plus revenue counterparty relationship, which means any stumble in Anthropic's growth trajectory would create a circular problem hitting AWS revenue, the equity stake, and the capex sized against the partnership all at once. Third is the free cash flow situation, which is set to remain negative as the capex ramp continues through 2026 and potentially into 2027, and the short depreciation cycles on GPUs and custom accelerators will keep pressure on the bottom line. Beyond those, cloud competition remains intense with Alphabet, Microsoft, and Oracle all executing well, supply chain risks are rising amid geopolitical tensions including the increasing cost of memory and storage, and rising logistics and fulfillment costs bear watching. The Q2 guidance reflected some of this caution — revenue was guided to a healthy $194 billion to $199 billion, but the profit outlook was more conservative on rising stock-based compensation, satellite project costs, and fuel inflation.

Where the Weight of the Evidence Ultimately Points

Bringing every thread together, Amazon (AMZN) presents as a company in the middle of an enormous, deliberate investment cycle that has been substantially de-risked by $465 billion in disclosed forward demand commitments — and the honest characterization is a Buy with a clear-eyed view of the risks, a stock where the weight of the evidence favors the bulls without pretending the cash flow strain is not real. The case for owning it is substantial: AWS reaccelerating to 28% growth at record 50% segment operating income gains, a backlog coverage ratio of 2.4x that no cloud business of this size has ever shown, an advertising engine compounding at 24% toward $90 billion with 80% incremental margins, an operating margin at an all-time high of 13.1%, a lopsided 48-to-3 sell-side upgrade trend, and a technical structure freshly confirmed by an all-time-high breakout and a golden cross. Set against that, the bear case is not frivolous — free cash flow has compressed 95% year-over-year and swung negative, the $200 billion capex bill is immense, the Anthropic exposure is circular, and the entire thesis hinges on AWS holding above roughly 22% growth. But the distinction that ultimately tips the balance is that this is capital being spent against contracted demand rather than hope, with customer commitments already in hand for a substantial portion of it. The stock trades close to fair value on conservative numbers and offers meaningful upside on the more constructive ones, with $245 established as firm support beneath it. The posture that fits the evidence is constructive ownership heading into the May 20 Annual General Meeting and the July 30 second-quarter report — long the quality and the backlog, fully aware that the AWS growth rate is the one number that decides whether the premium multiple holds.

That's TradingNEWS