Amazon Stock Price Prediction: $340 Bull vs $190 Bear (AMZN)

Most Amazon stock price prediction pieces treat AMZN as a straight race between AWS growth and the AI capex bill, and that framing misses the asset now moving Amazon’s earnings more than either. At 30 June, Amazon carried its Anthropic stake at $190.4 billion, roughly 1.5 times its $128.9 billion of long-term debt and about 7% of a $2.73 trillion market value, according to its Form 10-Q. A $50.5 billion mark-up on that stake is why second-quarter net income of $62.6 billion was more than double operating income of $27.5 billion. From the 14 September close of $253.54, our 12-month call is a $340 bull case (+34.1%), a $290 base case (+14.4%) and a $190 bear case (-25.1%). The shares closed 10.7% below their 3 August record close of $284.02, and they sit just 1.4% above the lowest price target on a 42-analyst panel.

The reason that matters for a price target is the direction of two balance-sheet lines. Trailing free cash flow swung from an $18.2 billion inflow to a $7.6 billion outflow in a year, while Amazon raised $81.9 billion of long-term debt over the same twelve months and then priced another £4.25 billion of sterling notes on 9 September. Put the filings side by side and the picture is plain: the operating business is funding the AI build with borrowed money, and a private-company valuation carried at Level 3 fair value is doing the heavy lifting in GAAP earnings. If Anthropic’s reported October Nasdaq listing goes ahead, that mark becomes a public price, subject to a lock-up. If the listing slips or prices lower, the cushion under a debt-funded capex cycle gets thinner. That hinge runs through all three scenarios, and it is why we raised our 24 August bull case by $10 to $340 while keeping the bear case within $5 of where it was.

What Is Actually Driving Amazon Stock Right Now

The operating story is strong, and it is accelerating. Net sales rose 20% to $200.6 billion in the second quarter, operating income climbed 43% to $27.5 billion, and AWS grew 36.7%, its fastest pace in 18 quarters. AWS added more than $4.6 billion of revenue in a single quarter and now runs at a $169 billion annualised rate. Its segment margin reached 39.3%. Advertising grew 26%. Management said the AI business and the in-house chips business (Trainium and Graviton) each passed a $25 billion annual run rate.

The market rewarded it. AMZN jumped 15.3% on 31 July, from $235.50 to $271.58, and kept climbing to the record close on 3 August. What has happened since is the part a price target has to explain: a 10.7% slide in six weeks with no change to the numbers.

The explanation is in the cash flow statement. Having tracked Amazon’s capex line through the last six quarterly releases, the speed of the step-up is what stands out. Trailing net purchases of property and equipment rose from $88.0 billion in the twelve months to March 2025 to $169.0 billion in the twelve months to June 2026, a 92% increase in five quarters. Cash capex was $53.1 billion in Q2 alone.

A useful way to think about AWS right now is a utility building power stations before the customers are wired in. The contracts exist: the backlog is about $496 billion. The cash, though, arrives only once each station is commissioned and switched on. Jassy described exactly that sequence on the call: servers and networking equipment take “a little less than three years to break even,” data centres last 30 years or more, and in the meantime Amazon will “encounter free cash flow headwinds until these data centers come online.” For scale, Oracle reported a larger $664 billion of remaining performance obligations last week, and its free cash flow also turned negative, as FinanceFeeds reported on Oracle’s Q1. Backlog is no longer a differentiator on its own. Conversion speed is.

Amazon’s guidance leaves room for both readings. Third-quarter sales are guided to $197 billion to $202 billion, up 9% to 12%, or nearly 400 basis points faster excluding the Prime Day timing shift, with operating income of $22.5 billion to $26.5 billion against $17.4 billion a year ago. The capex number is the one that moved.

“We now believe we will spend approximately $220 billion in cash CapEx in 2026. The higher cost of memory pushing this number up from our prior estimate of about $200 billion. Even at that amount, we 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,” said Andy Jassy, President and CEO at Amazon, on the second-quarter earnings call.

How Amazon, Its AI Partners and the Market Are Responding

Amazon’s own response has been to double down on the two labs that anchor AWS demand, and to lock in funding for it. The 10-Q shows that in Q2, AWS and Anthropic expanded their existing commitment by more than $100 billion over ten years. Amazon put $5.0 billion into Anthropic’s Series G preferred stock and a further $5.0 billion into Series H. It also agreed to a financing facility of up to $20 billion, now $15 billion after the Series H draw, which becomes available to Anthropic as AWS hits compute delivery milestones. On the OpenAI side, Amazon holds $28.7 billion of OpenAI Series C preferred stock, and AWS expanded its existing $38 billion commitment with OpenAI by $100 billion over eight years.

The funding came from the bond market. Amazon’s long-term debt nearly doubled in six months, and on 9 September it priced £4.25 billion of sterling notes in four tranches, maturing from 2029 to 2045, with coupons from 5.200% to 6.650%. The notes carry A1, AA and AA- ratings from Moody’s, S&P and Fitch, and settled on 14 September. Investment-grade credit markets are clearly still open to Amazon. The question for equity holders is how much debt the capex plan will need in total.

The third response is strategic. In April Amazon agreed to buy satellite operator Globalstar for $90.00 in cash or 0.3210 Amazon shares per Globalstar share, according to the S-4 registration. The deal adds spectrum and direct-to-device capability to Amazon Leo, which Jassy said is “close to 400 satellites in orbit, enough to begin initial satellite internet service this year.”

The market’s response has been less generous. Over the weekend of 12 and 13 September, the chief executives of Anthropic, OpenAI and xAI publicly backed a deliberate slowing of frontier-model development, and AI-linked stocks sold off on Monday, 14 September. FinanceFeeds laid out what a slowdown called from inside the industry does to $700 billion of capex, with Amazon’s roughly $200 billion-plus budget the largest line in it. AMZN closed 1.3% lower that day at $253.54. Retail commentary picked up the same thread, with CFA-run YouTube channels framing the build as bond-funded.

Management’s counter is that the returns are already visible in AWS profitability. “You’re seeing, despite the large investments, AWS margins have continued to remain strong, and we’re up 650 basis points year-over-year. 520 basis points if you exclude the derivative accounting gain,” said Brian Olsavsky, Senior Vice President and CFO at Amazon, on the Q2 call.

Market Data: Valuation, Targets and the Chart

At $253.54 and 10,786,313,572 shares outstanding (10-Q cover, 22 July), Amazon is valued at about $2.73 trillion. The stock is up 9.8% in 2026 from a $230.82 year-end close, and its twelve-month closing range runs from $198.79 on 13 February to $284.02 on 3 August.

Consensus earnings need care. The S&P Global consensus on stockanalysis.com puts 2026 EPS at $12.88 and 2027 EPS at $10.40, a 19% decline. That decline is not a forecast of a weaker business. It reflects $62.8 billion of Anthropic mark-ups booked in the first half of 2026 that analysts do not assume will repeat. The same panel expects revenue to grow from $828.3 billion in 2026 to $946.7 billion in 2027. On the 2027 figure, AMZN trades at 24.4 times earnings.

Now combine two numbers that are rarely put together. The $190.4 billion Anthropic carrying value works out to about $17.65 per Amazon share. Strip it out and the rest of Amazon trades near $236, or roughly 22.7 times 2027 consensus, for a company growing its cloud arm at 37%. That is the case for the bull scenario. The case against is that the 2027 estimate already assumes capex turns into operating profit on schedule, and 2027 is the first year in which Barclays has modelled negative free cash flow across the hyperscalers, as CNBC reported and the FinanceFeeds capex analysis above sets out.

The Street sits well above us on the base case. The 42-target panel averages $332.45, and the lowest target, $250, is only 1.4% below the current price. A stock that trades at the Street’s floor while the average implies 31% upside is telling you the market doubts the timeline, not the thesis. Our $290 base case is 12.8% below that average. That is deliberate: we give the free-cash-flow gap a longer runway than the consensus does.

Regulatory and Structural Tension

Three pressure points sit outside the income statement, and each one can move the stock.

The Globalstar approvals. The Hart-Scott-Rodino waiting period on the Globalstar deal expired on 17 July 2026, so the US antitrust waiting period has passed. The deal still needs the US Federal Communications Commission, plus France’s ANFR and ARCEP and two French ministries, and Globalstar has to obtain government authorisations for its C-3 satellite system. Amazon and Globalstar expect completion in 2027. There is also a letter agreement with Apple: if Globalstar misses C-3 milestones, a “Customer payment” accrues and reduces what Globalstar holders receive. For Amazon Leo this means the spectrum is contracted but not yet delivered.

What the Anthropic mark can and cannot do. The 10-Q describes the stake as a Level 3 fair-value measurement that uses “discounts for lack of marketability.” It is subject to an ownership cap, and Amazon says it expects “to be subject to a customary lock-up period following an IPO.” A public listing would replace an internal model with a market price, which helps credibility. It would not turn the stake into cash for months. Any gap between the IPO price and the implied carrying value will flow straight through “Other income (expense), net” and into reported EPS.

Circularity. The $20 billion facility is drawn in new Anthropic convertible notes, and it becomes available as AWS delivers compute under a contract that sits inside AWS backlog. In practice, Amazon is partly financing a customer whose spending shows up as AWS revenue. That arrangement is disclosed and legal. It is also the kind of vendor-financing loop that credit analysts discount first when the cycle turns.

Power is the physical constraint behind all of it. Jassy reiterated on the call that Amazon expects “to have double the power capacity by the end of 2027 that we had in 2025,” which is why long-dated supply deals such as Vistra’s 20-year nuclear contracts with Amazon and Meta now matter to the AWS margin outlook.

Amazon Stock Price Prediction: Bull, Base and Bear

Base case, $290 (+14.4%), 50%. AWS growth holds in the mid-30s, capex lands near $220 billion, and free cash flow stays negative into 2027 without getting worse. Anthropic lists, and the public price roughly confirms the carrying value. The stock gets back above its $284.02 record close but trades at about 28 times 2027 earnings, below the Street’s average target because the market keeps charging for the borrowing.

Bull case, $340 (+34.1%), 25%. Q3 shows AWS growth at or above 37% and backlog stepping up from $496 billion. Anthropic prices at or near the reported $2 trillion, which re-rates the stake well above $190.4 billion, and the market capitalises it despite the lock-up. Capex guidance holds at $220 billion rather than rising again. $340 sits 2.3% above the consensus average and 19.7% above the record close.

Bear case, $190 (-25.1%), 25%. The slowdown debate turns into slower lab spending, AWS growth drops below 30%, and Amazon raises capex again anyway because memory costs keep climbing. The Anthropic listing slips or prices below the implied mark, which reverses part of this year’s gains through the income statement. The multiple compresses to about 18 times 2027 earnings, below the $196.00 52-week low. Weighting the three scenarios gives $277.50, or 9.4% above spot.

What would change the call. A Q3 report, due in late October, with AWS growth above 37% and backlog above $550 billion would push the base case towards $320. A capex increase above $220 billion without matching backlog growth, or a postponed Anthropic IPO, would pull it towards $240. Other mega-cap calls follow the same logic, including our Microsoft stock prediction. The listing itself is the next date to watch, and FinanceFeeds is tracking Anthropic’s reported October Nasdaq plan.

FAQ

What is the Amazon stock price prediction for the next 12 months?

FinanceFeeds’ Amazon stock price prediction from the 14 September 2026 close of $253.54 is a $290 base case (+14.4%), a $340 bull case (+34.1%) and a $190 bear case (-25.1%), weighted 50/25/25. The probability-weighted value is $277.50. The swing factors are AWS growth in the Q3 report, 2026 capex against the $220 billion guide, and whether Anthropic’s reported IPO confirms the $190.4 billion carrying value of Amazon’s stake.

Why was Amazon’s Q2 net income bigger than its operating income?

Amazon booked $53.4 billion of pre-tax non-operating income in Q2 2026, mainly a $50.5 billion upward revaluation of its Anthropic nonvoting preferred stock. That lifted net income to $62.6 billion, or $5.75 per diluted share, against operating income of $27.5 billion. The gain is an accounting mark based on Anthropic’s funding rounds, not cash, which is why 2027 consensus EPS of $10.40 sits below 2026’s $12.88.

How much is Amazon’s Anthropic stake worth?

At 30 June 2026 Amazon carried its Anthropic investment at about $190.4 billion: $92.5 billion of nonvoting preferred stock and $97.9 billion of convertible notes at estimated fair value, according to its 10-Q. That is about $17.65 per Amazon share and roughly 7% of the company’s market value. An IPO would convert the holdings to nonvoting common stock, subject to an ownership cap and a lock-up.

Is Amazon’s free cash flow negative?

Yes. Trailing twelve-month free cash flow was an outflow of $7.6 billion at the end of Q2 2026, compared with an inflow of $18.2 billion a year earlier. Net capex rose by $66.1 billion year on year, mostly for AI infrastructure. Amazon guides 2026 cash capex to about $220 billion, and long-term debt rose to $128.9 billion from $65.6 billion at the end of 2025.

What do Wall Street analysts expect for AMZN stock?

The 42 price targets tracked by S&P Global Market Intelligence on stockanalysis.com average $332.45, with a median of $330, a low of $250 and a high of $400, as of 3 September 2026. The consensus rating is Strong Buy. The current price of $253.54 is just 1.4% above the lowest target, so the market is pricing more execution risk than most analysts are.

When will Amazon’s Globalstar acquisition close?

Amazon and Globalstar expect the merger to complete in 2027. The US HSR waiting period expired on 17 July 2026, but the deal still needs approval from the FCC and French regulators, including ANFR and ARCEP, plus authorisations for Globalstar’s C-3 system.

Disclaimer: This article is analysis, not investment advice. Price scenarios are illustrative, based on public filings and market data available on 15 September 2026, and are not a recommendation to buy or sell any security. Equity investments carry risk, including the loss of capital.