For most of the past three months, the market had one story about Amazon.

A great business buried under a single number: $200 billion.

Two hundred billion dollars in AI infrastructure spending. No payback timeline from the CFO. The stock down 18% from its May high while AWS posted its fastest growth in four years.

Then this week, AMZN touched its 20-day exponential moving average near $243. For momentum investors, that is a buy signal. For a lot of readers, the same question arrived: is this the entry?

I have a question first.

Not whether Amazon is a good business. That answer is public and clear.

The harder question is what you are actually buying at the 20 EMA, sixteen days before the Q2 earnings print on July 30 that will answer everything the market is currently waiting to hear.

That is what This Week's Focus is here to examine.

Missed Monday's edition on SK Hynix?

This MarketBeat piece argues Micron is moving aggressively because the high-bandwidth memory shortage is not going away soon. The company raised its 10-year U.S. investment plan to $250 billion, targeting more fabrication capacity and HBM technology, while analysts still rate the stock a consensus Buy with nearly 30% expected upside. In plain English, Micron is trying to turn AI memory scarcity into a long-term competitive advantage, not just a cyclical windfall.

The Fool’s point is that Trump Accounts are now live, but investors should understand the guardrails. At launch, contributions are invested in the State Street SPDR Portfolio S&P 500 ETF, giving eligible children simple broad-market exposure rather than a menu of individual stocks or speculative assets. The bigger takeaway is that the account is designed more like a long-term starter investment vehicle for children than a trading account.

Fed Chair Kevin Warsh used his first congressional testimony to stress that the Fed has “no tolerance” for persistently high inflation and is still committed to bringing inflation back to its 2% target. He also pointed to the AI investment boom as a potential productivity driver, but gave no clear hint on the Fed’s next rate move, which matters because policymakers remain split ahead of the July 28 to 29 meeting. The market takeaway is simple: Warsh wants investors to believe inflation will be beaten, but policy uncertainty is not gone yet.

AI is moving fast and most people are falling behind. 

The Rundown AI keeps you ahead of the curve. 

It's a free AI newsletter that keeps you up-to-date on the latest AI news, and teaches you how to apply it in just 5 minutes a day.

Plus, complete the quiz after signing up and they’ll recommend the best AI tools, guides, and courses — tailored to your needs.

Sign up to start learning.

Amazon (AMZN) has pulled back roughly 18% from its May 2026 high near $278.56 and is now sitting near its 20-day exponential moving average at approximately $243.

The 20 EMA at a key support level after a meaningful pullback is a buy signal. Any technical trader will say that without hesitation.

But I want to answer a different question.

Are you buying the chart, or are you buying the business? Because those are two different trades, two different theses, two different conviction requirements, and two very different reactions to the Q2 earnings print sixteen days from now.

Forty percent faster than fifteen quarters ago.

AWS posted $37.6 billion in revenue in Q1 2026, up 28% year over year, the fastest growth rate the division has seen since 2022. The AI revenue run rate inside AWS crossed $15 billion annualized, growing 260 times larger than AWS's total revenue run rate was three years after the service launched.

And then there is the number most retail investors have not seen.

AWS had a signed contract backlog of $364 billion as of Q1 2026. That figure excludes a separately announced deal with Anthropic valued at over $100 billion. This is not speculative future revenue. This is revenue Amazon has already contracted to deliver.

Because a $364 billion backlog means the demand is not in question. The demand is already signed. The debate is about the timing of the payback, not the existence of the business.

The $200 billion capex number is not the problem. The answer that came after it is.

Amazon told investors to expect negative free cash flow in 2026 as a result of the record capital spending. When analysts pressed CFO Brian Olsavsky on the FCF floor and the payback timeline, his answer was: Amazon sees "long strong return on invested capital" and demand signals justify the pace of spending.

A philosophy, not a number.

A market paying 35x earnings has zero patience for philosophy where a number was supposed to be. Bernstein and Benchmark trimmed price targets on capex sustainability concerns. Morgan Stanley held Overweight with a $300 target, arguing AWS growth could exceed 30%.

That gap between Bernstein and Morgan Stanley is the entire investment decision right now. Both firms have the same data. They are making different assumptions about what $200 billion of spending produces, and when.

The 18% pullback is not the market saying Amazon's business has deteriorated. It is the market saying: give us the number, not the philosophy.

The stock has been grinding sideways to lower since May, and the price action has a specific shape to it.

Every time buyers push Amazon back toward $255, they hit the 50-day SMA sitting as overhead supply. The stock touches that level, slows, and drifts back. That pattern has repeated three times since June, and it tells you institutional sellers are using the 50-day as an exit while buyers lack the conviction to push through it.

The 20 EMA at approximately $243 is where the stock is finding its footing this week. The 200-day SMA at approximately $233 is the floor below. Sitting between those two levels at $247 is not a stock in a trend. It is a stock waiting for a catalyst.

The catalyst is July 30.

What I am watching heading into that date is whether Amazon can reclaim and hold $255 before the print. A pre-earnings move through the 50-day SMA tells you institutional buyers are positioning with conviction. A continued drift below it tells you the market is waiting for the data before committing size.

This is the number that resolves the entire investment debate.

Watch for AWS revenue growth relative to Q1's 28% benchmark. Because if AWS accelerates toward 30% or beyond in Q2, the $200 billion capex story shifts from "uncertain bet" to "infrastructure paying back at the speed of demand." Analysts expect Q2 total revenue of approximately $196 billion and EPS of $1.82. An AWS beat with any management color on FCF trajectory changes the stock from a technical bounce to a structural re-rating. A miss hands the bears another quarter to work with.

The Q2 call transcript will tell you more than the reported numbers will.

Watch specifically for any change in how Brian Olsavsky discusses free cash flow. Because the Q1 "long strong return on invested capital" with no specific timeline was the sentence that sent Amazon from $278 to $247. A Q2 call that introduces a specific FCF floor, a payback year, or a signal that capex decelerates in 2027 removes the primary overhang on the stock. A Q2 call that repeats the same philosophy gives sellers exactly what they need to maintain the 50-day SMA as overhead resistance. That is the line, not the slide deck, that moves Amazon's multiple.

The 50-day SMA at approximately $255 is the technical tell before earnings arrive.

Watch whether Amazon pushes through $255 and holds it in the next two weeks. A stock that climbs from the 20 EMA through the 50-day SMA before an earnings print tells you institutional investors are positioning with conviction. A stock that bounces off the 20 EMA but stalls below the 50-day tells you the buy-side is waiting for confirmation rather than anticipating it. One of those setups justifies adding before July 30. The other justifies watching from the sideline.

The Anthropic deal valued at over $100 billion is not yet in the $364 billion backlog figure.

Watch for any early commentary on Anthropic-related workloads ramping through AWS in Q2. Because that deal represents a separate tranche of committed demand that has not yet entered the revenue recognition pipeline. If Q2 shows early Anthropic workload activity on AWS, the timeline from backlog to income statement accelerates materially. That is what converts the $200 billion capex story from faith-based investing to math-based investing. I looked at Amazon's AWS AI trajectory earlier this year when the run rate first crossed $10 billion — the pace of that expansion suggests the Anthropic revenue will matter to the income statement sooner than the market is pricing.

Microsoft and Google both report Q2 earnings before Amazon does.

Watch what Azure and GCP say about AI cloud workload momentum. Because AWS competes directly with both for enterprise AI training and inference, and the hyperscaler peers' commentary will set the narrative context before Amazon's July 30 print. If Azure reports accelerating AI revenue and raises its cloud outlook, the setup for AWS entering earnings is constructive. If either shows deceleration, markets will apply that narrative to Amazon before it can respond with its own data. These are the data points to watch before the data point that matters most.

I have followed Amazon long enough to know how this story ends, even when the middle part looks uncertain.

The business is exceptional. A $364 billion AWS backlog. Twenty-eight percent cloud growth at the fastest pace in fifteen quarters. A $15 billion AI revenue run rate that barely existed three years ago. And a Q2 guide that beat consensus by $5 to $10 billion.

That is not a business in trouble. That is a business building the infrastructure that will define AI economics for the next decade.

The 20 EMA is not the reason to buy Amazon. The $364 billion backlog is.

If you are buying the business, the 20 EMA at $243 is a reasonable entry and the Q2 print on July 30 is a risk you are accepting, not one you are avoiding. Size the position for the conviction level you have in the AWS thesis, not the conviction level you have in the chart.

If you are trying to trade the technical bounce, know what you are actually doing. The 50-day SMA at $255 is your first resistance, July 30 is your binary event, and a Q2 call that repeats "long strong ROIC" without a number sends you back toward $233.

Buy the business. Use the chart to find the entry. The investors who confuse those two decisions are the ones who sell a $364 billion backlog at the 50-day SMA.

Subscribe to The AI Report, the free 5-minute daily AI brief for 400,000+ business leaders, and you’ll get 37 Claude prompts free in your welcome email. They’re organised by the 8 situations every manager faces. You get both: the newsletter and the prompts.

Subscribe & Get the Prompts

Technical signals ask the question. Fundamentals answer it.

The 20 EMA brought Amazon back into focus this week. The $364 billion AWS backlog is the reason to stay.

Investors who treat a chart signal as a reason to buy are working with half the information. The other half is what the business is actually doing behind the price action.

Know why you are buying. The chart is the entry. The business is the thesis.

They are not the same thing, and the investors who know the difference are the ones who hold through the noise.

If this analysis changed how you see AMZN this week, forward it to one reader who is still trading the headline.