• Piper Sandler (PIPR) reiterates Overweight rating with a $320 price target, arguing Amazon (AMZN) is best positioned among major tech peers.
  • AWS growth accelerates to 37% in Q2, with a $496 billion contracted backlog highlighting visible demand.
  • Amazon's 2026 ROIC expected around 14%, relatively stable versus historical average, despite massive capex.

AWS Momentum and the ROIC Debate

Piper Sandler's renewed bullishness on Amazon comes as the company's cloud unit shows its fastest growth in roughly four years, with Q2 AWS revenue up 37% year-over-year to $42.2 billion. The analyst firm, which set a $320 price target, argues that Amazon's spending on AI infrastructure is more returns-backed than peers like Meta (META) and Google (GOOG), which face sharper ROIC declines as they invest in frontier AI models.

Piper Sandler's note highlights that Amazon's 2026 ROIC will likely land around 14%, compared with its historical average of approximately 17%. The slight dip is a result of extraordinary capital investment, but the firm believes Amazon's diversified revenue mix and AWS's profitability make the spending narrative credible.

The Numbers Behind the Narrative

Amazon's Q2 results, released in late July, painted a clear picture: revenue reached $200.6 billion, up 20% year-over-year, while operating income surged 43% to $27.5 billion. AWS alone contributed $16.6 billion in operating income, roughly 61% of the total, despite accounting for just 21% of revenue. The cloud unit's contracted backlog ballooned to $496 billion from $364 billion in the prior quarter, offering unprecedented visibility into future demand.

However, the capex surge is palpable: property and equipment purchases hit $54.2 billion in Q2, up from $32.2 billion a year earlier. This drove trailing-twelve-month free cash flow to -$7.6 billion, a stark contrast to +$18.2 billion a year prior. Management has lifted expected 2026 cash capex to approximately $220 billion, citing high memory costs and unmet capacity demand.

"We're seeing a capacity-constrained environment that will persist through 2027," CEO Andy Jassy reportedly said, with meaningful customer reservations already in place for 2028.

AI Infrastructure: A Double-Edged Sword

The broader AI infrastructure cycle is forcing hyperscalers to commit vast sums to data centers and advanced chips. Amazon's bet is that its spending will yield measurable returns through AWS, which is monetizing AI via compute, model hosting, and custom Trainium and Graviton chips. The company also benefits from AI-driven growth in advertising and retail automation.

But critics worry about an overbuilding scenario. "The key question is whether AI demand will generate returns above the cost of capital," noted a portfolio manager. "Amazon's backlog is reassuring, but negative free cash flow means the thesis must be proven through future utilization."

In premarket trading on August 31, AMZN was quoted at $266.43, up 3.97% from the prior session, giving it a market cap of approximately $2.87 trillion. The $320 price target implies roughly 20% upside, though Piper's optimism is shared by most analysts: 31 of 32 tracked ratings are bullish, with an average target of $327.75.

Implications and Risks

The near-term outlook remains strong, with Q3 revenue guided between $197 billion and $202 billion, and operating income expected between $22.5 billion and $26.5 billion. However, investors will closely monitor AWS growth, backlog conversion, and whether the massive capex translates into profitable capacity.

Long-term, Amazon's bull case hinges on AWS becoming the primary monetization layer for enterprise AI. The bear case, meanwhile, centers on industrywide capex outpacing demand, or hardware depreciation and power constraints delaying deployment. With total Big Tech AI investment expected to exceed $700 billion in 2026, the stakes are high.

"Amazon's ability to balance near-term cash flow with long-term growth will be key," said a technology analyst. "If AWS continues its momentum, the spending will look prescient. If not, it could weigh on the stock for years."

This is informational analysis, not personalized investment advice. Analyst targets and forecasts can change quickly.