AMZN Q4 FY25 Earnings Review — Bear

Date: 2026-04-01 Quarter: Q4 FY25 (ended Dec 31, 2025) | Reported: Feb 5, 2026 Source: Scout brief at briefs/AMZN_earnings-review_2026-04-01/


Prior Beliefs

I haven't published a formal AMZN earnings review before, but here is where I stood going into Q4 FY25, based on my portfolio notes through Oct 2024:

  1. Revenue growth: Expected consolidated growth in the 11-13% range, consistent with Q2/Q3 FY25 trajectory. Holiday seasonality should lift absolute dollars significantly.
  2. AWS: Expected continued acceleration from the ~19% run rate in Q2/Q3 FY25 toward low-20s. The custom chip narrative and AI demand signals were strong.
  3. Margins: Expected GAAP op margins around 11%, possibly compressed by continued capex ramp and seasonal mix. AWS margins to remain in the 33-36% range.
  4. Guidance: Expected Q1 FY26 revenue guide implying 10-12% YoY growth (conservative, per Amazon's habit). Expected another significant capex number for FY26.
  5. Valuation: At ~$2.2T market cap, P/S around 2.6x on run-rate revenue — reasonable for a diversified mega-cap with an accelerating cloud/AI business. My conviction in Aug 2024 was that AMZN was "quite undervalued" and by Oct 2024 it was my "only large position."

Updated Beliefs — Results vs Expectations

The Numbers

Metric Prior Belief Actual Verdict
Revenue $205-210B range (11-13% YoY) $213.4B (+13.6% YoY) Better — above my high end
AWS Revenue Low-20s% growth (~$34B) $35.6B (+24% YoY) Significantly better — fastest in 13 quarters
Op Margin [GAAP] ~11% 11.7% (reported), ~12.8% adjusted for $2.4B charges Better — adjusted is very strong
Q1 FY26 Revenue Guide 10-12% YoY implied $176B mid (+13% YoY) Better — higher than expected
FY26 CapEx Large number, $150B+ ~$200B Higher than expected — the debate-worthy number

Multi-Quarter Revenue Trajectory

Q1 FY24 Q2 FY24 Q3 FY24 Q4 FY24 Q1 FY25 Q2 FY25 Q3 FY25 Q4 FY25
Revenue ($B) $143.3 $148.0 $158.9 $187.8 $155.7 $167.7 $180.2 $213.4
YoY % 12.5% 10.1% 11.0% 10.5% 8.7% 13.3% 13.4% 13.6%
Beat vs Guide -- +1.0% +1.7% +1.5% +1.6% +3.8% +1.9% +1.9%

Assessment: Revenue growth re-accelerated from the Q1 FY25 trough (8.7%) to 13.6% in Q4. Four consecutive quarters of acceleration. At Amazon's scale — $213B in a single quarter, $717B for FY25 — this is genuinely impressive. The numbers match the theory.

AWS Deep Dive

Q3 FY24 Q4 FY24 Q1 FY25 Q2 FY25 Q3 FY25 Q4 FY25
AWS Revenue ($B) $27.5 $28.8 $29.3 $30.9 $33.0 $35.6
AWS YoY % ~19% ~17% ~17% ~19% ~19% 24%
AWS Op Income ($B) $10.4 $10.6 $11.5 $10.2 $11.4 $12.5
AWS Op Margin 38.1% 36.9% 39.5% 32.9% 34.6% 35.0%

AWS accelerated from 17% to 24% over the past year. At $142B annualized run rate, adding $7B YoY in a single quarter is substantial. The backlog of $244B (+40% YoY) is the strongest leading indicator I've seen from Amazon. I have to note: 24% growth at $142B ARR is structurally different from competitors growing faster on a much smaller base — Jassy made this point explicitly, and the numbers support it.

AWS margins at 35% are down from the 39.5% peak in Q1 FY25, but that's expected given the intensity of AI infrastructure buildout. The CFO was transparent: "margins will fluctuate." As long as the 32-39% range holds while growth accelerates, the margin story is intact.

Custom chips are the underappreciated story: Trainium + Graviton combined >$10B ARR, growing triple-digit % YoY. Trainium2 has 1.4M chips landed. Trainium3 already in production. This is Amazon building its own cost advantage underneath the cloud business — exactly the kind of structural moat I value.

Segment Profitability

Segment Q4 FY25 Revenue Q4 FY25 Op Income Op Margin YoY Margin Change
North America $127.1B $11.5B 9.0% +1.0pp
International $50.7B $1.0B 2.1% -0.9pp (charges)
AWS $35.6B $12.5B 35.0% +0.4pp (adjusted basis)

North America at 9.0% op margin is a record. That's significant — this is a retail business generating $11.5B of operating income in a single quarter. The fulfillment efficiency story (regional network, robotics, same-day economics) is showing up in the numbers.

International margin slipped to 2.1%, but $1.1B of the weakness was the Italy tax settlement and other charges. Ex-charges, margins expanded YoY. The company is investing in quick commerce (Amazon Now) internationally, which is the right long-term move even if it suppresses near-term margins.

Advertising — The Quiet Powerhouse

Q3 FY24 Q4 FY24 Q1 FY25 Q2 FY25 Q3 FY25 Q4 FY25
Ad Revenue ($B) $14.3 $17.3 $13.9 $15.7 $17.7 $21.3
YoY % ~19% ~27% ~18% +23% +24% +23%

Advertising is a $68.6B FY25 business growing 20%+ and it is nearly pure margin. Prime Video ads reaching 315M global viewers (up from 200M in early 2024) extends the TAM significantly. This business alone would be worth a meaningful chunk of the market cap.

The CapEx Question — Where I Get More Cautious

Period CapEx (net) % of Revenue FCF
FY2023 ~$48B 7.5% $32.2B
FY2024 $77.7B 12.2% $38.2B
FY2025 $128.3B 17.9% $11.2B
FY2026E ~$200B ~28% (est) ?

Let me be clear: the $200B FY26 capex guidance is the single most important number in this report.

FCF went from $38.2B in FY2024 to 11.2BinFY2025—a71139.5B TTM, +20%), so the underlying business is generating cash. The compression is entirely capex-driven.

Two top-tier analysts (Mahaney at Evercore, Anmuth at JPM) asked management for a FCF floor or financial guardrails on the capex program. Both were deflected. Jassy's answer was essentially: "This is an extraordinarily unusual opportunity." He's probably right. But "we're spending because the opportunity is enormous" without a numerical floor is the kind of statement that can age poorly.

I've seen this movie before — not with Amazon specifically, but with companies that say "the opportunity justifies the spend." Sometimes it does. Sometimes it doesn't. The difference is Amazon has $123B in cash + securities, $57B net cash, and a retail + advertising business generating massive operating cash flow underneath the cloud investment. The balance sheet can support this.

But I want to see the returns. The $244B backlog is encouraging. AWS growing 24% is encouraging. Custom chips at $10B+ ARR is encouraging. I don't need to see FCF normalize immediately, but I need to see the capex turning into revenue acceleration — and so far, it is. If AWS growth starts decelerating while capex stays at $200B, that's when the thesis would weaken.

Valuation

Metric Value
Market Cap ~$2,200B
P/S (TTM) 3.1x
P/E (TTM GAAP) 28.4x
P/Op Income 27.5x
PEG (P/S / Rev Growth%) 0.23x

At 3.1x P/S with 13.6% consolidated revenue growth, the PEG of 0.23x looks attractive. But revenue growth isn't the whole story — this is a business with an accelerating 24% growth cloud segment, a 23% growth advertising segment, and an improving retail margin profile.

If I value AWS separately at, say, 10x revenue (1.4T), advertisingat15xrevenue(1.0T), and the retail/subscription business at 1x revenue (~$500B), I get to roughly $2.9T in parts value — above the current $2.2T market cap. I could be wrong on any of those multiples, but the directional case is clear.

The P/E of 28x on GAAP earnings understates earnings power because $128B in capex is running through the cash flow statement but not the income statement (depreciation lags capex by years). The operating earnings power of this business is higher than reported GAAP suggests.


Key Observations

  1. AWS growth acceleration is the headline and it's real. 24% at $142B ARR, with a $244B backlog growing 40% YoY. This isn't a one-quarter spike — it's four consecutive quarters of acceleration from the 17% trough. The demand signals (backlog, custom chip subscription, Bedrock growth) all corroborate.

  2. The AI infrastructure story is differentiated. Trainium + Graviton at >$10B combined ARR, Bedrock at multibillion-dollar ARR with 60% QoQ customer spend growth, Amazon Connect at $1B ARR — these are concrete proof points, not vaporware. The vertical integration (custom chips + cloud platform + enterprise services) creates a structural cost advantage that gets wider over time.

  3. Retail is better than people think. North America 9.0% op margin (record), paid units +12%, same-day delivery reaching ~100M customers, Rufus AI assistant used by 300M+ customers. This isn't a stagnant retail business subsidizing cloud — it's a margin-expanding logistics platform.

  4. The capex cycle is the risk to own. $200B in FY26 is ~28% of estimated revenue. FCF will remain compressed or potentially negative on a quarterly basis. Management refused twice to provide a FCF floor. I'm comfortable with this for now because the demand is clearly there, but I'm watching the ratio of capex to AWS revenue growth intently.

  5. Special charges are noise but worth flagging. Q3 had $4.3B, Q4 had $2.4B — that's $6.7B in two quarters. Italy tax, FTC settlement, severance, store impairments. These are one-time in nature but they make the GAAP numbers harder to read. Adjusted Q4 op income was $27.4B, not $25.0B.


Prior Beliefs vs Updated Beliefs Summary

Belief Prior Updated
Consolidated growth trajectory Steady 11-13% Accelerating — four quarters of improvement, now 13.6%
AWS momentum Low-20s achievable Exceeded — 24%, fastest in 13 quarters
Margin profile ~11% op margin, stable Slightly better — 11.7% reported, ~12.8% adjusted
Valuation "Quite undervalued" (Aug 2024) Reasonable — 3.1x P/S, 28x P/E, earnings power understated
Risk profile Manageable Elevated but manageable — $200B capex is unprecedented, no FCF floor committed

Thesis

Strengthening.

Amazon is firing on all cylinders in the segments that matter most for long-term value creation: AWS accelerating into the AI infrastructure cycle, advertising growing 20%+ as a near-pure-margin business, and retail margins expanding to record levels. The $244B backlog gives multi-year visibility. The custom chip program creates structural cost advantages.

The capex cycle is the variable to watch. At 200BforFY26, Amazonismakingthelargestcapitalinvestmentincorporatehistory, andmanagementexplicitlydeclinedtosetfinancialguardrails.Thebalancesheet(57B net cash, $140B TTM operating cash flow) supports it, and the demand indicators (backlog, chip subscription, revenue acceleration) justify it. But I want to see continued AWS acceleration and early ROIC proof points through FY26.

If AWS growth re-decelerates toward 20% while capex stays at $200B, my thesis weakens. If AWS growth holds 24%+ and capex converts to revenue, the stock is meaningfully undervalued at $2.2T.


Position Action

Hold / Opportunistic Add on weakness. This was a strong quarter. The numbers exceeded my expectations across every major dimension. I don't need to chase it at current prices, but I'd add on any pullback to the 180 − 190range(implied 1.9-2.0T market cap). The capex uncertainty provides periodic opportunities for the patient.

I could be wrong about the capex cycle paying off. Shows what I know — I sold Amazon in March 2024 "because I already own them through the ETFs" and had to buy back at higher prices by August. The lesson: don't get cute with dominant companies. If the business keeps executing like this, own it and be patient.


Bear