Amazon at $3 Trillion: The AWS Rally Looks Real, But the Stock Has Run Too Far, Too Fast

Amazon hit a $3 trillion market capitalization on Monday - the fifth company in history to join that club. The stock jumped 5.3% to clear $284, capping a five-day run that added more than 23%. The catalyst was a second-quarter earnings report that delivered exactly what the AI infrastructure trade has been begging for: AWS revenue of $42.2 billion, up 37% year over year, the fastest growth in 18 quarters. Management told investors demand is so strong that even a revised $220 billion capital expenditure budget won't be enough to keep up.

The headlines are already recycling old fear. Bezos is selling shares again, the story goes. Except the Bezos selling cycle began with a prearranged trading plan... adopted in March - a prearranged trading program designed specifically to signal that sales are routine, not a reflection of conviction. Bezos sold... roughly $736.7 million in shares in July 2025, part of a 25 million shares... ending May 29, 2026 program. He sold $13.5 billion... in 2024 under a prior plan. This is how he fund his space exploration company, Blue Origin and his philanthropy. It's been happening for years. The market should stop treating it as new information.

The real question isn't whether Bezos is cashing out. It's whether a stock that just exploded through a psychological milestone on a wall of AI euphoria has already priced in everything that could go right.

The AWS story is legit - that's the problem

I'm not here to argue that AWS growth is a mirage. It's not. The 37% acceleration in cloud revenue, the $496 billion in contracted backlog that jumped $132 billion in a single quarter, the fact that Jassy said demand... for 2028 is striking - this is genuine infrastructure demand. AWS operating margin... 39.4%, generating $16.6 billion in segment operating income for the quarter. Amazon's custom silicon and AI product businesses each independently crossed a $25 billion annualized run rate, growing at triple-digit percentages.

The problem for investors isn't that the story is wrong. It's that the story is now priced into a stock that trades at 43.4 times forward earnings.

The trailing P/E is misleading

On the surface, Amazon's trailing P/E of 22.6x looks reasonable. It's below Microsoft at 27.1x, in line with Meta at 22.1x, and above Alphabet at 18.7x. But that trailing multiple is distorted by a one-time bomb. Q2 net income... $62.6 billion, or $5.75 per share - includes a $53.4 billion, primarily from its investments in AI lab Anthropic pre-tax gain. Without that, the GAAP earnings picture looks dramatically different. Adjusted EPS came in at $1.97... versus $1.82 consensus, but the gap between GAAP and adjusted is almost entirely that single inflection.

Once the one-time gain fades from the trailing window, that 22.6x P/E evaporates. What's left is the forward multiple of 43.4x, which is what the market is actually paying for future earnings. Amazon's revenue growth rate sits at roughly 15.8% year over year. A forward P/E of 43.4x for mid-teens revenue growth is expensive by any GARP standard. The PEG ratio - forward P/E divided by growth rate - works out above 2.7x, well north of the 1.0x threshold that signals good-growth-at-a-reasonable-price.

Free cash flow is bleeding

Then there's the cash flow problem. Amazon's trailing-twelve-month free cash flow is deeply negative - down to an outflow of roughly $7.6 billion after printing $18.2 billion a year earlier. The driver is capital expenditure. Amazon expects to spend $220 billion this year, up from the $200 billion the company projected in February. Jassy attributed the increase to rising memory prices, not an expansion of buildout scope, but $220 billion is still the largest single-year capex commitment among the Big Four hyperscalers. Microsoft $110 billion to $120 billion. Alphabet $175 billion to $185 billion. Meta $125 billion to $145 billion. Amazon's Q2 capex alone was $54.2 billion... a 68% year-over-year increase.

This isn't irresponsible spending. The $496 billion backlog justifies the buildout. But free cash flow negativity at this scale means Amazon is not self-funding its AI infrastructure play. The company carries $54.4 billion in total debt against $78.2 billion in cash, giving a manageable net debt position, but that balance sheet cushion will shrink as capex runs through 2026 and 2027. If AWS growth slows even moderately while capex stays elevated, the cash flow gap could pressure margins for years.

The technicals are flashing caution

Price action doesn't care about your long-term thesis. And right now, the charts are stretched. Amazon's RSI sits at 72, well into overbought territory. The stock closed at $284, roughly 15% above its 50-day moving average of $247 and 21% above its 200-day average of $235. A gap that wide, that fast, typically precedes a mean-reversion pullback. The five-day surge of 22.75% on 90.8 million shares of volume - $25.8 billion in turnover - is a euphoric blow-off, not a measured accumulation pattern.

AInvest's aggregate signal still labels AMZN a Buy, which reflects the broader analyst consensus around AWS momentum. But consensus doesn't protect you from entry risk. A stock can be a long-term winner and a terrible place to open a new position.

So what's the call?

I'm not selling the AWS thesis. The backlog, the growth acceleration, the AI demand Jassy described - this is the kind of secular infrastructure build that took years to price into Microsoft and Alphabet during their own cloud cycles. Amazon's moat in cloud distribution, enterprise lock-in, and custom silicon (Trainium and Graviton) remains intact. The Anthropic partnership, even if the accounting gain was one-time, signals a structural alliance that strengthens Amazon's AI positioning.

But the risk/reward at $284 is not compelling. A forward P/E of 43.4x, negative free cash flow, an RSI at 72, and a 15% deviation from the 50-day average - these are conditions where the market has arguably baked in a flawless execution scenario through 2027. The stock needs AWS to keep accelerating, capex to convert into revenue without margin compression, and the macro to cooperate. One quarter of AWS growth back toward the high-20s would be treated as deceleration at this valuation.

I would reassess this as a buy if the stock pulls back toward its 50-day moving average in the $245-$250 zone, or if RSI resets below 50, signaling that the euphoria has cooled and a more measured entry is available. Don't chase the rally here. The setup for Amazon remains constructive over a medium-term horizon, but the better risk/reward is on weakness, not at all-time highs fueled by a one-time earnings inflection and recycled insider selling headlines.

Chasing $284 Amazon would be buying the peak of the narrative cycle. Patience is the harder discipline, and it pays better.