Amazon, Apple and Reddit Earnings Show What the AI Trade Rewards—and Punishes

Last updated: July 31, 2026, 11:30 a.m. Eastern Time

Amazon, Apple and Reddit all reported quarterly results that would normally be described as strong. Revenue rose at each company. Profitability exceeded the year-earlier period. Important business lines reached records. Guidance was not uniformly weak. Yet the stocks moved in sharply different directions on July 31: Amazon surged, Apple fell, and Reddit dropped even more steeply.

The split was not a simple verdict on whether the companies “beat” Wall Street estimates. It reflected a more demanding question: what did each report say about the next constraint on growth? Amazon showed that demand for artificial-intelligence infrastructure was accelerating fast enough to make its enormous capital program look commercially grounded. Apple showed excellent current demand but warned that component and manufacturing capacity could limit the September quarter. Reddit showed exceptional advertising economics while reminding investors that an important part of its audience acquisition still depends on volatile search-engine referrals.

That distinction explains why three profitable technology businesses could receive three incompatible market reactions. Amazon’s bottleneck was capacity it could potentially monetize. Apple’s bottleneck was supply that could prevent it from meeting demand. Reddit’s bottleneck was distribution it does not fully control. The numbers mattered, but the character of the bottleneck mattered more.

The reports also offered a compact view of how generative AI is reshaping the technology sector. For Amazon, AI is a demand engine for cloud computing, custom chips and data centers. For Apple, AI is raising component costs and strategic expectations while the company follows a comparatively capital-light infrastructure model. For Reddit, AI is both a commercial tool for advertising and a threat to the referral traffic that brings people from search results to community discussions.

Key Takeaways

• Amazon: Second-quarter net sales rose 20% to $200.6 billion, while AWS revenue increased 37% to $42.2 billion, its fastest growth in 18 quarters. The stock was up about 14.6% at approximately 11:26 a.m. Eastern Time on July 31.
• Apple: Fiscal third-quarter revenue increased 16% to $109.4 billion and iPhone revenue rose almost 22%, but guidance and supply constraints overshadowed the record June quarter. The stock was down about 9.3% at the same market-data snapshot.
• Reddit: Revenue climbed 61% to $805 million, net income reached $253 million and third-quarter guidance exceeded the pre-report consensus cited by major financial outlets. The stock nevertheless fell about 23.6% as investors focused on choppy search referrals and slower U.S. user growth.
• Earnings quality: Amazon’s reported $62.6 billion net income included $53.4 billion of pre-tax non-operating income, primarily related to its Anthropic investment. Apple’s gross margin and earnings per share benefited from tariff refunds. Reddit’s profit, operating cash flow and free cash flow were much more directly tied to its operating performance.
• The broader lesson: Investors are not rejecting AI spending in the abstract. They are distinguishing between spending attached to visible revenue and backlog, spending that pressures cash without enough proof of return, and business models exposed to AI-driven changes in customer acquisition.

Why the Same Earnings Vocabulary Produced Opposite Stock Moves

Earnings coverage often begins with a scorecard: revenue versus consensus, earnings per share versus consensus, and guidance versus consensus. That framework is useful, but it is incomplete. A company can exceed current-quarter estimates and still fall if the report weakens the market’s assumptions about future growth, margins, capital intensity or risk. It can also produce a distorted “beat” when earnings include a large non-operating gain that analysts did not model as part of the underlying business.

Amazon illustrates the first complication in a favorable direction. Its reported earnings per share of $5.75 were not a clean measure of operating performance because the quarter included a $53.4 billion pre-tax non-operating gain, primarily from the company’s Anthropic investment. Yet the stock rose because the operating evidence was strong where investors most needed reassurance: AWS growth accelerated to 37%, AWS operating income rose 64%, the segment maintained a 39.4% operating margin, and management described demand that continued to exceed available capacity.

Apple illustrates the second complication. The company posted a record June quarter, but the market was already looking beyond June. Fiscal fourth-quarter revenue guidance of roughly 9% to 11% growth was below the approximately 12% expectation Reuters cited from LSEG. Management also warned of constraints in advanced chipmaking capacity and increasing memory costs. A quarter can be excellent and the stock can still fall if the next quarter appears less favorable than the valuation had assumed.

Reddit illustrates a third complication: the market can discount even strong guidance when it questions the durability of a key operating input. Reddit’s third-quarter revenue range of $860 million to $870 million exceeded the consensus figures reported before the release. Adjusted EBITDA guidance was also strong. The concern was not that advertisers had suddenly left. It was that U.S. daily active users grew only 6% year over year and declined slightly from the first quarter, while management said search referrals had become choppy and more volatile later in the quarter.

The three reports therefore produced a hierarchy of evidence. Amazon showed demand converting into revenue at a scale large enough to support more investment. Apple showed current demand but less certainty about near-term supply and margin. Reddit showed monetization but less certainty about the traffic funnel. The market rewarded the evidence it considered most durable and penalized the dependencies it considered least controllable.

The Verified Results at a Glance

Sources: company earnings releases, financial statements and shareholder materials. Percentages are year-over-year changes unless otherwise stated. Apple’s fiscal calendar differs from Amazon’s and Reddit’s calendar-quarter reporting.

Amazon Earnings: AWS Acceleration Changed the Conversation

The most important number in Amazon’s report was not the $5.75 in diluted earnings per share. It was the 37% increase in AWS revenue. Amazon Web Services generated $42.2 billion in second-quarter sales, compared with $30.9 billion a year earlier, and reached what the company described as a $169 billion annualized revenue run rate. The growth rate was the fastest in 18 quarters.

That acceleration matters because Amazon entered the report with two linked investor concerns. First, AWS had spent several years growing more slowly than Microsoft Azure and, more recently, Google Cloud. Second, Amazon was committing an extraordinary amount of capital to data centers, chips, networking equipment, power and supporting infrastructure. Investors needed evidence that the spending was creating revenue rather than merely defending market position.

The official Amazon second-quarter earnings release supplied that evidence. AWS sales grew 37%, up from 28% growth in the first quarter, 24% in the fourth quarter of 2025 and 20% in the third quarter of 2025. The sequence shows a genuine acceleration rather than a single favorable comparison.

The combination of faster growth and a higher margin is especially important. AWS operating income increased 64% to $16.6 billion. The segment produced about 60% of Amazon’s consolidated operating income while accounting for roughly 21% of sales. This is why calling Amazon’s report an e-commerce story misses the financial center of gravity. The retail businesses remain essential to scale, customer relationships and advertising inventory, but AWS continues to generate a disproportionate share of operating profit.

Why capacity constraints looked positive for Amazon

Capacity constraints can be good or bad depending on the business and the evidence surrounding them. In Amazon’s case, management said demand continued to exceed available capacity and expected the imbalance to persist into 2027 and 2028. That statement supported the argument that the company’s infrastructure spending was responding to customer commitments rather than being built speculatively without a market.

The quality of that argument depends on what “demand” means. Cloud contracts and remaining performance obligations are not the same as recognized revenue. They can extend over years, and customer usage can vary. Still, Amazon’s reported backlog was approaching $500 billion, according to management commentary covered by Reuters, and the quarter showed that AWS could convert some of that demand into rapid current revenue growth.

Amazon also said its AI business and its custom-chip business had each exceeded a $25 billion annualized revenue run rate, both growing at triple-digit percentages. Run rates are not audited annual revenue and should not be treated as if the company had already booked $25 billion from each business over a completed year. They are useful directional indicators, however, because they show the scale at which Amazon says the businesses were operating during the quarter.

The custom-chip strategy is central to AWS economics. Trainium is designed for AI training and inference, while Graviton serves broader cloud workloads. Proprietary silicon can lower Amazon’s dependence on third-party accelerator suppliers, improve price-performance for customers and allow tighter optimization across chips, networking and software. It also requires sustained research, manufacturing commitments and customer adoption. The quarter did not eliminate those risks, but it strengthened the case that the investment has moved beyond an experimental stage.

Amazon’s operating result was stronger than the EPS headline

Amazon’s consolidated operating income increased 43% to $27.5 billion. That figure is more informative about the quarter’s core business than the reported $62.6 billion of net income because operating income excludes the enormous non-operating gain tied primarily to Anthropic.

The company reported $53.4 billion in pre-tax other income. Subtracting that gain from net income would be wrong because the tax consequences and other non-operating items also matter. A cleaner way to interpret the accounts is to compare operating income and income before taxes. Amazon reported $80.9 billion of pre-tax income, of which $53.4 billion came from net non-operating income. That leaves roughly $27.5 billion associated with operating profit and net interest effects before taxes—very close to the operating result.

This distinction matters because investment gains can reverse. A higher private-company valuation can increase reported income without producing cash. A later valuation decline can reduce earnings. The Anthropic position may be strategically valuable, but it should not be mistaken for recurring profit generated by selling cloud services, advertising or consumer goods.

Amazon’s quarter was nevertheless strong on a core basis. Net service sales rose 24% to $123.0 billion, outpacing the 14% increase in net product sales to $77.6 billion. That mix shift is favorable because services such as AWS, third-party seller services, advertising and subscriptions generally carry different—and often better—economics than first-party retail merchandise.

The free-cash-flow problem did not disappear

The most credible skeptical reading of Amazon’s report begins with cash. Trailing-12-month operating cash flow increased 33% to $161.4 billion, an exceptional figure. Yet Amazon’s measure of free cash flow fell from a positive $18.2 billion a year earlier to an outflow of $7.6 billion. Purchases of property and equipment, net of proceeds from sales and incentives, reached $169.0 billion over the trailing 12 months, up 64%.

Reuters reported that Amazon increased its expected 2026 capital spending to approximately $220 billion, from an earlier plan of about $200 billion. The increase was partly associated with higher memory costs and the continuing build-out of AI infrastructure. A company can grow operating cash flow and still consume free cash when capital expenditure grows faster. That is exactly what happened.

The optimistic interpretation is that Amazon is front-loading infrastructure for contracted or strongly indicated customer demand. If revenue and operating profit continue to accelerate, free cash flow can recover after the most intense building phase. The skeptical interpretation is that AI infrastructure has shorter economic lives, higher power requirements and faster technological obsolescence than traditional data-center assets, forcing Amazon to keep reinvesting before the previous wave has produced an adequate return.

Both interpretations can be true for different parts of the asset base. The quarter supports the demand argument but does not yet settle the return-on-invested-capital question. Investors should therefore avoid treating the July 31 stock rally as proof that the capital program is guaranteed to succeed. It means the evidence improved relative to what the market feared before the report.

Retail was not the headline, but it was not weak

The focus on AWS should not obscure the acceleration in Amazon’s other businesses. North America segment sales increased 16% to $116.2 billion, and operating income rose 21% to $9.1 billion. International sales increased 15% to $42.2 billion, while operating income rose to $1.7 billion from $1.5 billion.

Online-store revenue increased 15% to $70.4 billion. Third-party seller services rose 16% to $46.8 billion. Subscription services increased 12% to $13.7 billion. Those results suggest that Amazon was not relying on cloud alone. The company said same-day and overnight delivery volumes increased more than 40% and that grocery and everyday essentials grew faster than the broader stores business.

Faster delivery can improve customer frequency, especially for low-consideration household purchases. It can also raise costs if inventory placement, routing and fulfillment density are poor. Amazon’s North America operating margin held at 7.9%, the same as in the first quarter and above the 7.5% recorded a year earlier. That stability indicates that faster growth did not require a visible collapse in segment profitability during the quarter.

The international segment’s 4.0% operating margin remained much lower than North America’s but improved from the prior-year period. The international business has historically absorbed investment in logistics, price, selection and market expansion. Sustained positive operating income reduces the risk that retail growth outside North America must be subsidized indefinitely by AWS.

Advertising has become a third profit engine

Amazon advertising-services revenue increased 26% to $19.8 billion, faster than the company’s online stores, third-party seller services and subscriptions. Advertising is strategically powerful because Amazon can place ads near a transaction and use shopping intent to measure performance. It can also extend advertising through Prime Video and other media inventory.

The growth is relevant to the AI story. Amazon said AI-assisted advertising tools contributed to the business. Automated creative generation, bidding, audience selection and campaign measurement can make the platform easier for smaller advertisers to use. That enlarges the addressable market but also exposes Amazon to the same debate facing other ad platforms: whether AI tools create durable incremental demand or simply automate tasks that advertisers were already performing.

At nearly $20 billion in quarterly revenue, advertising is too large to describe as an ancillary business. It helps support retail margins, increases the economic value of customer traffic and gives Amazon another way to monetize Prime engagement. Its rise also means that Amazon competes not only with retailers and cloud providers, but with Alphabet, Meta, Reddit and other digital advertising platforms.

What Amazon’s third-quarter guidance says

Amazon guided third-quarter net sales to $197 billion to $202 billion, representing 9% to 12% growth. Management said the year-over-year comparison was affected by the timing of Prime Day; excluding Prime Day from both periods, growth would be almost four percentage points higher. The guidance also assumed an unfavorable foreign-exchange effect of about 80 basis points.

Operating income guidance of $22.5 billion to $26.5 billion compared with $17.4 billion in the third quarter of 2025. The range implies substantial year-over-year growth but also allows for a sequential decline from the second quarter’s $27.5 billion. That is not unusual given seasonality, investment timing and the variability of Amazon’s cost structure, but it keeps attention on whether AWS acceleration can offset the expense of infrastructure and other initiatives.

The market’s favorable response suggests investors placed more weight on the AWS trajectory than on the sequential operating-income range. That judgment can change quickly. If AWS growth slows while capital spending remains elevated, free cash flow will return to the center of the debate. If growth stays near the current pace and margins remain high, the spending will look more like capacity expansion into a supply-constrained market.

Apple’s Record Quarter Was Not the Market’s Main Question

Apple’s fiscal third quarter looked exceptional when viewed strictly as a report on the three months ended June 27. Revenue rose 16% to $109.4 billion, setting a June-quarter record. Operating income increased 27% to $35.7 billion. Net income rose 27% to $29.8 billion, and diluted earnings per share increased 29% to $2.02. The company generated growth in every reported geographic region, and both the iPhone and Mac businesses produced record June-quarter revenue.

Those figures explain why describing the report as weak would be inaccurate. The negative share-price response was a judgment about the transition from an unusually strong June quarter to a more constrained September quarter. It also reflected the stock’s starting point. Apple entered the release with greater investor confidence than Amazon had enjoyed before its report, leaving less room for results that were strong but accompanied by operational caveats.

Apple’s challenge is not a lack of demand in the conventional sense. The company said its installed base of active devices reached another record, and management described broad geographic growth. The more immediate challenge is whether it can secure enough advanced manufacturing capacity and memory components, maintain premium gross margins, execute its AI product roadmap and support a leadership transition without disrupting the operating model.

iPhone and Mac growth carried the product portfolio

iPhone revenue increased 21.7% to $54.3 billion, accounting for almost half of total company revenue. Mac revenue rose 28.7% to $10.4 billion. Together, those two product categories contributed most of Apple’s incremental revenue compared with the prior-year quarter.

The iPhone result is particularly important because the product remains Apple’s largest direct revenue source and the center of its ecosystem. An iPhone sale can support later Services revenue through app distribution, cloud storage, payments, media subscriptions, warranties and other offerings. Strong device demand therefore affects more than the hardware line in which the sale is recorded.

Mac growth was also meaningful because personal-computer demand is more cyclical and replacement-driven than smartphone demand. New processor generations, operating-system changes, enterprise purchasing and creative workloads can produce bursts of demand, but the category does not normally carry the same installed-base economics as the iPhone. A nearly 29% increase indicates a favorable product and replacement cycle during the quarter.

Not every hardware category advanced. iPad revenue declined 5.9% to $6.2 billion. Wearables, Home and Accessories rose 6.5% to $7.9 billion. The mixed result matters because Apple’s long-term growth case depends partly on finding new device categories and use cases rather than relying indefinitely on increasingly mature smartphone and personal-computer markets.

Services reached a record but did not remove the concern

Services revenue increased 12.1% to $30.7 billion, setting another record. Services accounted for approximately 28% of total revenue and generally carry higher gross margins than hardware. The category helps smooth product cycles, increases customer lifetime value and gives Apple recurring revenue linked to an installed base that management said had reached 2.5 billion active devices.

The Services result was strong in absolute terms but was reported as slightly below market expectations. That gap mattered because investors often assign a higher valuation to recurring, high-margin revenue than to hardware sales. A small shortfall can therefore influence sentiment more than its contribution to a single quarter would suggest.

The category is also broader and less uniform than the label implies. It includes the App Store, advertising, AppleCare, cloud services, digital content, payments and subscription products. Growth can come from a larger installed base, higher engagement, price increases, new subscriptions, advertising demand or changes in commercial terms. Those drivers have different margins and different regulatory risks.

Apple’s control over app distribution and payment rules has attracted scrutiny from competition authorities in multiple jurisdictions. Regulatory changes can affect commission rates, payment flows and the degree to which developers can steer customers outside Apple’s systems. The June-quarter result did not resolve those structural questions. It did show that the Services engine remained large and growing even as the company prepared for a more demanding period of AI investment and product execution.

The tariff refund improved the quarter but was not recurring operations

Apple reported a gross margin of 50.1%, up from 46.5% a year earlier. Management said approximately two percentage points of the quarter’s gross margin came from favorable tariff refunds. Earnings per share included an approximately $0.11 benefit from the same factor.

That does not make the reported result incorrect. Refunds are real economic benefits when the company is entitled to receive them. It does mean that analysts should separate the benefit from the repeatable economics of selling devices and services. Without the refund, gross margin would have been closer to 48.1%, still strong but materially below the headline figure.

The distinction resembles the treatment of Amazon’s Anthropic gain, although the scale and accounting category are different. In both cases, reported net income included a meaningful item that was not a clean measure of the ordinary operating trajectory. Apple’s underlying operating performance was strong even after removing the tariff effect, but the one-time benefit raised the comparison point for anyone looking only at the published margin and EPS.

The September-quarter gross-margin outlook of 47% to 48% included an expected benefit of roughly one percentage point from tariff-related items, according to management commentary. That means the implied underlying margin could be lower than the headline range. Rising memory costs, product mix, component availability, foreign exchange and the timing of product launches can all influence the final result.

Supply constraints changed the meaning of strong demand

Demand is valuable only when a company can fulfill it on economically attractive terms. Apple indicated that advanced chip capacity was constrained and that memory prices were increasing. These are not identical problems. Foundry capacity can limit the number of devices Apple can produce, while higher memory prices can compress gross margin unless the company offsets them through pricing, product mix, supplier negotiations or efficiency.

The constraint is especially important during a product transition. Apple’s September quarter typically includes the early period of a new iPhone generation. If supply is limited, revenue can shift between quarters rather than disappear permanently. Yet a delayed sale is not always neutral. Some customers postpone purchases, some choose a competing product, and carriers or retailers may adjust promotions. The impact depends on the duration and severity of the constraint.

Apple has advantages in this environment. Its scale gives it purchasing power, its product margins provide room to absorb some cost increases, and its brand can preserve demand during short shortages. It also has a tightly coordinated supply chain and long-standing relationships with manufacturing partners. Those advantages reduce risk; they do not eliminate dependence on a concentrated set of leading-edge semiconductor and memory suppliers.

The broader AI infrastructure boom makes that dependence more visible. Cloud providers are competing for accelerators, networking equipment, memory and advanced packaging. Consumer-device makers require different chips, but they draw from overlapping manufacturing ecosystems. A company that has historically been praised for supply-chain discipline is now operating in a market where several of the world’s largest technology companies are simultaneously increasing demand for scarce components.

Apple’s AI strategy is intentionally different from the hyperscalers

Apple is not trying to build the same business as Amazon Web Services, Microsoft Azure or Google Cloud. It sells integrated consumer devices and software, and it can combine on-device processing with private cloud infrastructure and partnerships. That approach can reduce the amount of capital Apple must commit compared with a hyperscaler serving thousands of enterprise customers.

The strategy also creates a different test. Amazon can point to AWS revenue, operating income, backlog and customer demand. Apple must show that AI features improve product demand, retention, usage or Services economics without weakening privacy, reliability or margins. The financial benefit may appear indirectly through a replacement cycle rather than as a separately reported AI revenue line.

This makes Apple’s AI investment harder to evaluate from quarterly accounts. Research and development expense increased 32% to $11.7 billion, much faster than revenue. The increase supports management’s claim that the company is investing aggressively, but R&D does not identify how much was allocated to AI, which projects are nearing commercialization or what return they may produce.

Apple’s capital expenditure remained modest relative to the largest cloud providers. Cash spent on property, plant and equipment totaled $6.8 billion during the first nine months of fiscal 2026, down from $9.5 billion in the comparable prior-year period. That difference does not imply that Apple is underinvesting because the business model and accounting arrangements differ. Apple can purchase services from suppliers, use contract manufacturers and rely on external cloud infrastructure rather than own every asset itself.

The risk is that a capital-light approach can become strategically dependent if external partners control critical models, data-center capacity or platform capabilities. The benefit is that Apple avoids committing hundreds of billions of dollars before the economics of consumer AI are fully established. The correct judgment will depend on product quality, customer adoption and whether Apple can preserve differentiation while using technology supplied partly by others.

The all-new Siri is a product test, not an earnings line

Management emphasized the introduction of a redesigned Siri and described early feedback positively. Corporate commentary about early feedback should be treated as management’s assessment, not as independent evidence of mass adoption or long-term satisfaction. The more useful questions are measurable: Does the feature increase device upgrades? Does it improve retention? Does it raise engagement with Apple services? Does it reduce support costs or create new revenue?

Consumer AI products also face a stricter reliability standard than experimental chatbots. A voice assistant integrated into messages, calendars, photos, payments and device controls must understand context without exposing private information or taking the wrong action. The value of integration can be high, but so can the reputational cost of errors.

Apple’s installed base gives it distribution that most AI companies cannot match. A successful feature can reach hundreds of millions of users through operating-system updates. The same scale magnifies product defects and regulatory questions. Apple must therefore balance speed with the quality controls expected from a premium consumer platform.

The June-quarter report offered evidence that the core hardware and Services engine remained healthy while this transition was underway. It did not provide enough evidence to determine whether the new Siri will produce a material upgrade cycle or a new Services business. Investors were therefore left with a strong present and a less quantifiable future.

Geographic breadth was a genuine strength

Apple reported revenue growth in all five geographic segments. Americas revenue increased 11.1% to $45.8 billion. Europe rose 22.4% to $29.4 billion. Greater China increased 22.4% to $18.8 billion. Japan grew 13.4% to $6.6 billion, and the rest of Asia Pacific increased 15.6% to $8.9 billion.

The Greater China result was especially notable because the region has been a recurring source of concern. Apple competes with strong local smartphone manufacturers, operates amid U.S.-China trade and technology tensions, and must navigate consumer preferences that can shift quickly. One strong quarter does not establish a permanent reversal, but broad regional growth reduces the risk that the company’s performance was created by one market or one currency movement.

Regional revenue is reported based on customer location and includes multiple product and service categories. It does not reveal unit volumes, pricing, channel inventory or profitability by geography. A region can grow because of higher-priced models, favorable product timing or easier comparisons even when unit share is unchanged. The figures are therefore encouraging but not a complete competitive scorecard.

Cash generation and capital returns remained formidable

Apple generated $117.0 billion of operating cash flow during the first nine months of fiscal 2026. Over the same period, it spent $62.1 billion repurchasing shares and $11.8 billion paying dividends. The ability to fund product development, supply commitments and shareholder distributions from internal cash generation remains one of Apple’s central financial strengths.

Buybacks reduce the share count and can increase earnings per share even when net income grows more slowly. They are most valuable when shares are repurchased below a reasonable estimate of intrinsic value and when the company has no higher-return use for the cash. They can destroy value when conducted at excessive valuations or when they substitute for necessary investment.

Apple’s quarter did not suggest that capital returns were impairing day-to-day liquidity. The more relevant debate is strategic: whether the company should continue returning such a large share of cash while AI competition is increasing the need for research, infrastructure partnerships, acquisitions and component commitments. A lower-capital model can justify continued distributions, but only if the model delivers competitive products.

Tim Cook’s final earnings call raised the stakes of the outlook

The release also carried an unusual governance dimension. Apple had announced that Tim Cook would become executive chairman and that John Ternus, the company’s hardware engineering chief, would become chief executive effective September 1, 2026. The July earnings call was therefore Cook’s final scheduled quarterly call as CEO.

Leadership transitions can be overinterpreted. Apple is a large institution with established operating processes, a deep executive team and a product roadmap developed years in advance. The CEO change does not reset the company on September 1. It does, however, change who will be accountable for executing the AI strategy, managing the supply chain and deciding how much of Apple’s economic model should evolve.

Ternus brings a hardware background at a moment when Apple’s central strategic problem crosses hardware, software, services and AI infrastructure. His effectiveness will depend partly on coordination with software leadership, Services executives, finance chief Kevan Parekh and the wider operations organization. Cook’s role as executive chairman may provide continuity, but it will also require clear governance boundaries so that authority is not ambiguous.

The stock decline should not be attributed solely to succession. The market had known about the transition since April. The relevant connection is that cautious guidance and component constraints become more consequential when a new CEO is about to take responsibility. The September quarter will be judged both as an operating period and as an early test of how smoothly the leadership handoff proceeds.

What Apple’s fiscal fourth-quarter guidance actually implies

Apple expected fiscal fourth-quarter revenue to increase approximately 9% to 11% year over year. Reuters reported that analysts tracked by LSEG had expected growth of roughly 12%. The difference between the top of management’s range and consensus was therefore about one percentage point—not evidence of a collapse, but enough to disappoint a market that had rewarded the shares before the release.

Gross margin guidance of 47% to 48% was below the 50.1% reported in the June quarter, partly because the earlier quarter contained a larger tariff-refund benefit. Operating expenses were expected to be between $19.1 billion and $19.4 billion. The combination points to continued growth with more pressure from product costs, mix and investment.

Guidance is a range based on management’s assumptions, not a guarantee. Product availability, launch timing, foreign exchange, macroeconomic demand and component prices can all move the result. Apple also has a history of managing a complex launch quarter in which channel inventory and supply allocation can affect reported revenue.

The market reaction appears to reflect a reduction in the probability of an upside surprise rather than a belief that Apple’s franchise had suddenly deteriorated. The distinction matters. A lower stock price can be consistent with a still-profitable, still-growing business if investors revise the multiple they are willing to pay for that growth.

The strongest case for Apple

The constructive interpretation begins with the breadth of current performance. Revenue grew in every region. The two largest hardware franchises produced strong growth. Services reached a record. Operating income and net income rose by more than one-quarter. Cash generation remained enormous. The installed base expanded, giving the company an unmatched distribution channel for new software and AI features.

Under this view, supply constraints are a timing issue created by unusually strong demand and industry-wide component tightness. Apple’s scale should help it secure capacity, while a high-margin ecosystem can absorb some cost pressure. AI features can strengthen the upgrade cycle without requiring Apple to replicate the capital spending of the cloud providers.

The succession can also be interpreted positively. Ternus has deep product experience, Cook remains involved as executive chairman, and the transition was announced months in advance rather than following a crisis. Continuity may allow Apple to change leadership without abandoning the disciplined operating model that has produced strong returns.

The strongest skeptical case for Apple

The skeptical interpretation is that the quarter benefited from an unusually favorable product cycle and a tariff refund just as forward growth was slowing. Services did not exceed expectations, advanced chip capacity was constrained, memory costs were rising, and the gross-margin outlook declined. Those pressures could persist beyond one quarter if AI infrastructure demand keeps component markets tight.

The larger strategic concern is that Apple has not yet demonstrated that its AI products can create a durable new growth engine. A large installed base is an advantage only if the features are compelling enough to change behavior. If customers view them as incremental, the company may spend more on research and partnerships without producing a commensurate increase in device upgrades or Services revenue.

Leadership transition adds execution risk. Ternus must preserve hardware excellence while accelerating software and AI delivery. Cook’s continuing role can provide support, but investors will want evidence that decision rights are clear and that the company can move faster without compromising privacy, reliability or supply discipline.

The June quarter weakens any claim that Apple’s existing business is in decline. It does not answer whether the next leg of growth will be large enough to justify the expectations embedded in the stock before the report. That is why a record quarter and a falling share price were not contradictory.

Reddit’s Earnings Beat Was Overshadowed by a Distribution Warning

Reddit’s second-quarter report contained the fastest revenue growth and the sharpest stock decline of the three companies. Revenue increased 61% to $805 million. Net income rose 183% to $253 million. Adjusted EBITDA more than doubled to $343 million, and free cash flow reached $261 million. Management’s third-quarter revenue guidance exceeded the consensus estimates reported before the release.

Those results establish that Reddit’s advertising business was working. The company was attracting more users, generating more revenue per user and converting revenue into cash at an impressive rate. The selloff therefore requires a narrower explanation than “bad earnings.” Investors focused on the durability of user acquisition in the United States, where daily active-user growth slowed to 6% and the average audience declined slightly from the first quarter.

Management said search referrals became choppy and more volatile later in the quarter. That phrase mattered because many people reach Reddit through search engines rather than by opening the app or typing the site’s address. Search referrals can introduce new users at low acquisition cost, but they are controlled partly by algorithms and product decisions made by other companies. A change in search-result presentation can therefore affect Reddit’s traffic without any deterioration in the underlying communities.

Advertising was the primary engine

Advertising revenue rose 64% to $762 million and represented almost 95% of total revenue. Other revenue, which includes data licensing and related activities, increased 24% to $43 million. The mix shows that Reddit remains principally an advertising platform rather than a diversified software or licensing business.

That concentration has advantages. Advertising can scale rapidly when user growth, engagement, measurement and targeting improve. Reddit’s communities produce conversations around products, hobbies, financial decisions, entertainment and local interests that can be commercially valuable. The platform also contains high-intent discussions in which users seek recommendations or compare alternatives.

The disadvantage is cyclicality and platform risk. Advertising budgets respond to economic conditions, brand-safety concerns, measurement quality and the performance of competing platforms. A business that receives nearly all of its revenue from advertising must keep users engaged while convincing marketers that community discussion can be translated into measurable results.

Reddit’s quarter provided strong evidence on the second half of that equation. U.S. revenue increased 56%, international revenue rose 84%, and global average revenue per user increased 36% to $6.18. U.S. ARPU reached $11.85, while international monetization remained far lower. The gap is both a weakness and an opportunity: Reddit has considerable room to increase revenue per international user, but it must localize advertising tools, sales operations and content discovery without damaging the product.

User growth was broad but uneven

Daily active uniques, Reddit’s average measure of unique users who visited on a given day, increased 18% to 130.3 million. Weekly active uniques rose 24% to 514.6 million. International daily active uniques increased 28% to 77.1 million, while U.S. daily active uniques increased 6% to 53.2 million.

The international result demonstrates that the platform was still expanding rapidly at a global level. The U.S. result explains the market’s concern. U.S. users are worth more financially because U.S. ARPU is substantially higher. A slowdown in the most heavily monetized market can matter more than faster growth in markets where advertising revenue per user remains lower.

Sequential performance reinforced the concern. Reddit reported 53.5 million U.S. daily active uniques in the first quarter, slightly above the 53.2 million reported in the second. Seasonal patterns can influence quarter-to-quarter comparisons, and the difference is small. Yet a sequential decline is difficult to ignore when a high-growth valuation depends on continued expansion.

Logged-in daily active uniques increased 7%, while logged-out daily active uniques increased 27%. Logged-out traffic is often associated with visitors who arrive through search or external links. It can broaden reach, but logged-in users generally provide richer personalization, stronger retention signals and more direct product relationships. A business increasingly dependent on logged-out visitors may be more exposed to changes in external distribution.

Why search referrals became the decisive issue

Search engines have historically benefited from Reddit’s inventory of specific, first-person discussions. A query about a product, travel destination, health experience or technical problem often leads to a Reddit thread because users value direct accounts from other people. That behavior can generate enormous free traffic for Reddit.

The relationship is changing as search companies place AI-generated answers, summaries and other rich results directly on the results page. A user may receive a synthesized answer without visiting the source. Search engines can still cite Reddit discussions or rank Reddit links, but the number and quality of outbound clicks can become more volatile.

Reddit faces a complicated position in this transition. Its content is valuable input for AI systems because it captures informal language, niche expertise, disagreement and lived experience. The company can license access to data, use AI to improve its own search and translation, and help advertisers understand communities. At the same time, AI summaries can reduce the need for users to click through to the original discussion.

Management’s warning did not establish that AI summaries caused all of the choppiness. Search traffic changes for many reasons, including ranking updates, query mix, seasonality and changes in Reddit pages. The responsible conclusion is narrower: external search referrals were volatile, and that volatility matters because Reddit does not control the search platforms.

This is a distribution-risk problem rather than an immediate monetization failure. Revenue and ARPU were strong. The risk is that weaker top-of-funnel traffic eventually slows user growth, reduces advertising inventory or raises the cost of acquiring users through paid channels and product investment.

Reddit’s answer is to build a direct discovery product

The company has been investing in search, machine translation, onboarding and personalized recommendations to make Reddit less dependent on external search. A stronger internal search product can turn a one-time visitor into a repeat user by helping people find relevant communities and current discussions without leaving the platform.

That is easier to describe than to execute. Reddit’s archive is immense, inconsistent and highly contextual. Threads become outdated. Jokes and sarcasm can be mistaken for factual advice. The best answer can be buried in replies, and community norms differ. An AI-assisted search product must identify useful material without flattening disagreement or presenting unsafe content as authoritative.

Reddit also has to preserve the incentive to create content. Communities work because users contribute without being paid for every post. If AI interfaces extract the answer while reducing visits, comments and recognition, contributors may perceive less value in participating. The company must therefore use AI in a way that increases discovery and participation rather than merely summarizing the work of its communities.

Direct traffic and logged-in usage would reduce exposure to external algorithms. They could also improve advertising performance because Reddit would have more persistent signals about interests and engagement. The second-quarter report showed progress in monetization but did not yet demonstrate that the direct relationship was growing quickly enough to neutralize search volatility.

Profitability was not an adjusted-only story

Reddit reported GAAP operating income of approximately $236 million and net income of $253 million. The net margin was 31.4%. Operating cash flow reached $262 million, and free cash flow was $261 million. These results distinguish the quarter from many high-growth platform reports in which adjusted profit expands while statutory losses and cash burn remain substantial.

Adjusted EBITDA of $343 million produced a 42.6% margin. The difference between adjusted EBITDA and GAAP operating income included stock-based compensation, depreciation and amortization, and other excluded items. Investors should not treat adjusted EBITDA as equivalent to net income, but the strong cash result indicates that profitability was not created solely by excluding non-cash expenses.

Gross margin reached 91.3%, reflecting the economics of a digital advertising platform whose direct cost of revenue is relatively low. High gross margin does not guarantee high operating margin because research, sales, infrastructure, content moderation and general administrative costs remain substantial. In the quarter, revenue growth was fast enough to create significant operating leverage.

The durability of that leverage depends partly on investment needs. Reddit must improve search, moderation, translation, advertiser tools and infrastructure. It also operates under scrutiny over content safety, privacy, copyright and the use of user-generated material in AI systems. Maintaining a very high adjusted EBITDA margin while funding those areas will be an important test.

The share repurchase sends a mixed signal

Reddit repurchased approximately 1.5 million shares for $235 million during the quarter at an average price of $157.57. Repurchases can offset dilution from stock-based compensation and return capital when management believes shares are undervalued. They can also expose the company to criticism when the market price subsequently falls below the repurchase price.

The July 31 intraday price of roughly $136 was below the reported average repurchase price. That does not prove the decision was wrong because value should be assessed over a longer horizon and the quarter’s repurchases occurred at different times. It does illustrate the risk of using corporate cash to buy a volatile stock during a period when the business still requires substantial investment.

Reddit’s balance between repurchases and growth spending will matter more as the company generates cash. Offsetting employee equity dilution can be sensible, but shareholders should examine whether the net share count is declining, whether repurchases are opportunistic and whether management has better uses for the capital.

Third-quarter guidance was strong enough to raise the bar

Reddit guided third-quarter revenue to $860 million to $870 million. Reuters cited an analyst consensus of approximately $828 million before the release. At the midpoint, the guidance was roughly $37 million above that figure. Adjusted EBITDA was expected to be $385 million to $395 million.

The guidance demonstrates that management did not expect search volatility to produce an immediate revenue slowdown severe enough to offset advertising momentum. It also implies an adjusted EBITDA margin near 45% at the midpoint, which would represent continued operating leverage.

Guidance does not eliminate the user-growth issue because advertising results can remain strong for a period even if traffic growth weakens. Better pricing, ad relevance, measurement and sales execution can increase revenue per user. Eventually, however, a platform needs either more users, more engagement or more monetization to sustain rapid growth. Reddit delivered all three in the second quarter on a global basis, but the U.S. audience trend raised doubt about the first component in its most valuable market.

The selloff therefore appears to represent a reassessment of duration. Investors did not deny the quarter’s growth; they reduced confidence in how long the growth can continue without a stronger direct traffic channel. That is a much harsher standard than the one applied to a mature company, but it is common for stocks whose valuations depend on years of high growth.

International monetization is Reddit’s largest visible opportunity

International users accounted for approximately 59% of daily active uniques but generated a much smaller portion of revenue. Reuters reported international ARPU of about $2.26, compared with $11.85 in the United States. Closing even part of that gap could support growth without requiring U.S. user numbers to accelerate immediately.

Monetization gaps do not close automatically. Advertiser demand, purchasing power, language, regulatory rules and the maturity of local sales operations vary by market. Some communities are global and English-language; others require localized content and moderation. Machine translation can improve access, but literal translation is not enough when humor, cultural references and safety standards differ.

Reddit’s international growth therefore offers a real but execution-dependent opportunity. The company must build advertiser tools and measurement that work across markets while preserving the community character that attracts users. Aggressive ad loading could raise short-term ARPU and weaken engagement. A slower approach can protect the product but delay revenue.

Data licensing is useful but not yet a second pillar

Other revenue increased 24% to $43 million. That category includes licensing arrangements under which technology companies can access Reddit data for model development and other purposes. The revenue is strategically relevant because it recognizes that the platform’s archive has economic value beyond advertising.

At less than 6% of quarterly revenue, however, other revenue was not large enough to change Reddit’s fundamental business mix. Data-licensing contracts can also be concentrated among a small number of counterparties, subject to renewal negotiations and affected by legal or regulatory questions over user-generated content.

Licensing creates a tension that Reddit must manage carefully. Selling access can fund the platform and establish commercial terms for AI companies. Users and moderators may still question how their contributions are used, whether attribution is preserved and whether the resulting AI products compete with Reddit for attention.

The strongest version of the licensing strategy would create revenue while reinforcing Reddit’s direct product—for example, through technology partnerships, better search or improved moderation. The weakest version would collect temporary fees while helping external systems reduce future referral traffic. The second-quarter numbers were too early and too small to determine which outcome will dominate.

Content quality and moderation are financial issues

Reddit’s commercial value depends on the credibility and relevance of its communities. Low-quality automated posts, undisclosed promotion, harassment or manipulation can reduce user trust and advertiser demand. Moderation is therefore not merely a social-policy function; it is part of the economic infrastructure of the platform.

AI can help detect spam, translate content and assist volunteer moderators. It can also make spam, impersonation and coordinated influence cheaper to produce. Reddit must invest on both sides of that contest while avoiding systems that remove legitimate speech or overwhelm moderators with false positives.

The company’s decentralized structure gives communities flexibility but creates inconsistent rules and experiences. That diversity is part of Reddit’s appeal and a challenge for large advertisers seeking predictable environments. Better contextual targeting and community-level controls may improve monetization without requiring the platform to standardize every discussion.

The strongest case for Reddit

The constructive interpretation is that Reddit has moved from a popular but weakly monetized website to a high-growth, profitable advertising platform. Revenue increased 61%, advertising revenue rose 64%, free cash flow exceeded $260 million and third-quarter guidance was well above the pre-report consensus. International user growth and monetization provide years of potential expansion.

Search volatility can also be viewed as a manageable product problem rather than an existential threat. Reddit’s content remains differentiated because it contains current human discussion rather than only publisher pages or polished brand material. The company can improve internal search, translation and recommendations to convert external visitors into direct users.

Under this reading, the stock decline reflected a market that had set an unrealistically high standard. A single quarter of choppy referrals did not reverse the platform’s advertising progress or cash generation. Continued execution could make the selloff look disproportionate to the operating evidence.

The strongest skeptical case for Reddit

The skeptical interpretation begins with dependence. Reddit does not own the search engines that supply many logged-out visitors, and it does not own the advertisers’ budgets. AI search products can summarize community knowledge without sending equivalent traffic back. U.S. daily active-user growth slowed and declined slightly from the previous quarter despite excellent revenue growth.

Monetization can temporarily mask slower audience growth. Higher ARPU has limits, especially if increased ad load damages engagement. International users are less valuable today and require investment to monetize. Data licensing is too small to diversify the business meaningfully and may strengthen products that compete with Reddit for attention.

The company must also spend to preserve content quality, support moderators and defend against AI-generated spam. Its high adjusted EBITDA margin may prove difficult to maintain if those investments rise. A valuation based on prolonged rapid growth can fall sharply when investors perceive even a small reduction in the durability of the user funnel.

Reddit’s quarter answered the question of whether the platform can make money. It did not answer whether it can control enough of its own distribution to keep compounding at the pace the market had priced before the report.

Three Bottlenecks, Three Different Market Judgments

The cleanest comparison across Amazon, Apple and Reddit is not revenue growth, company size or valuation. It is control over the next unit of growth. Amazon can spend to add cloud capacity. Apple can negotiate for components and adjust product allocation, but it cannot instantly create leading-edge fabrication capacity. Reddit can improve its own product, but it cannot dictate how search engines rank links or present AI-generated answers.

That difference shaped the market response. Amazon’s constraint invited more capital because the quarter provided evidence of demand and high segment profitability. Apple’s constraint threatened to defer revenue and reduce margin during an important launch period. Reddit’s constraint raised a question about whether future users would arrive at all, particularly in the high-value U.S. market.

This does not mean Amazon has the safest business or Reddit the weakest. Amazon is making the largest financial commitment and faces the greatest risk if demand or pricing changes. Apple has extraordinary customer loyalty and cash generation. Reddit’s small base allows it to grow rapidly and improve monetization for years. The point is that the market evaluates risk in relation to expectations and the company’s ability to respond.

The AI Capital-Spending Divide Is Becoming More Precise

The reports arrived during a week in which investors were comparing the spending plans and cloud results of the largest technology companies. The debate is often framed as a binary choice between confidence in AI and fear of excessive capital expenditure. The actual distinction is more useful: how much current revenue, profit, backlog and customer adoption supports each additional dollar of spending?

Microsoft reported continued strength in Azure and a large commercial backlog. Alphabet reported rapid Google Cloud growth and raised its capital-spending outlook. Meta continued to invest heavily in data centers and computing capacity even as free cash flow came under pressure. Amazon’s result joined that pattern but provided unusually strong evidence that capacity was being absorbed.

Cloud demand can remain strong while shareholder returns disappoint. The providers must buy accelerators, memory, networking equipment, land and power; build data centers; develop software; and compete on price. Customers may optimize usage, switch providers or build more efficient models. Hardware can become obsolete quickly, and depreciation schedules may not perfectly capture economic life.

Backlog also requires careful interpretation. It is not cash, recognized revenue or guaranteed profit. Contracts can extend over years, contain cancellation provisions or depend on customer usage. Still, when backlog, current revenue growth and operating profit all increase together, the evidence is stronger than a capital-spending plan supported only by management forecasts.

Amazon benefited because AWS produced that combination. Apple occupies another position in the value chain: it can benefit from AI-driven device demand but also faces higher component costs. Reddit occupies a third: it can sell advertising and data access around human conversation, while AI interfaces may change how that conversation is discovered.

Free cash flow has become the discipline mechanism

As capital programs expand, free cash flow is the most direct measure of the trade-off between current operations and investment. It is not a perfect metric because companies define it differently and because necessary growth investment can reduce it temporarily. It does, however, show how much cash remains after capital expenditures under the chosen definition.

Amazon’s trailing free cash flow turned negative despite a large increase in operating cash flow. Alphabet reported negative quarterly free cash flow as capital expenditure accelerated. Meta’s free cash flow narrowed sharply. Those figures do not prove that the investments are uneconomic. They show that shareholders are funding a large construction cycle whose returns will be realized later, if at all.

Apple’s model produced strong cash while requiring much less reported property-and-equipment spending. Reddit converted a substantial share of its revenue into free cash without owning a comparable physical infrastructure base. Those differences help explain why capital intensity cannot be compared using revenue growth alone.

The discipline for Amazon will be AWS growth and margin. The discipline for Apple will be product adoption, gross margin and Services expansion. The discipline for Reddit will be direct user growth and advertising retention. Each company needs a financial outcome that connects AI investment or exposure to shareholder value.

Accounting Quality Matters More in a Volatile Earnings Season

The three reports also demonstrate why earnings per share should not be read without the income statement and cash-flow statement. Amazon’s EPS was dominated by a non-operating investment gain. Apple’s EPS and gross margin benefited from tariff refunds. Reddit’s adjusted EBITDA excluded items that do not disappear economically, even though its GAAP profit and cash generation were also strong.

None of those facts invalidates the reported numbers. They change what the numbers can tell readers. Amazon’s EPS says much about the accounting value of its investment holdings during the quarter and less about recurring retail or cloud profitability. Apple’s EPS includes a real benefit that is not a normal quarterly sales margin. Reddit’s adjusted EBITDA helps compare operating trends but should be reconciled with stock-based compensation and other excluded expenses.

A useful earnings review therefore asks four questions:

• What came from ordinary operations? Revenue, gross profit, operating income and cash generated by the core business deserve primary attention.
• What came from valuation or timing? Investment gains, refunds, tax items and other one-time effects can change net income without changing customer demand.
• What required cash? Capital expenditure, working capital, acquisitions and repurchases determine whether reported profit translates into financial flexibility.
• What is management asking investors to assume next? Guidance, capacity plans and product claims are forward-looking and should be tested against later results.

Applying that framework produces a more coherent result than a simple beat-or-miss table. Amazon’s ordinary operations were strong, but cash investment was larger. Apple’s ordinary operations were strong, but the next-quarter margin and supply outlook were less favorable. Reddit’s ordinary operations and cash flow were strong, but the audience-acquisition risk was external and difficult to measure.

What the Intraday Market Reaction Did—and Did Not—Prove

At approximately 11:26 a.m. Eastern Time on July 31, Amazon shares were up about 14.6%, Apple was down about 9.3%, and Reddit was down about 23.6%. Those moves were substantial, but they were intraday observations rather than closing returns. They could change before the end of regular trading.

A stock move reveals that the balance of buying and selling changed at a particular price. It does not reveal one unanimous investor belief. Traders can react to earnings, guidance, positioning, options hedging, analyst revisions, sector moves and macroeconomic data at the same time. A heavily shorted stock may rise more than fundamentals alone would imply; a crowded long can fall even after good results.

Amazon’s rally is best described as following the AWS acceleration and improved confidence in AI demand, not as being caused by one figure with certainty. Apple’s decline followed guidance, margin and supply concerns, but the stock’s prior performance also influenced the response. Reddit’s fall followed the search-referral warning and U.S. user trend, even though guidance was strong.

The reaction can also affect future fundamentals. A higher share price makes equity compensation and acquisitions less dilutive. A lower price can increase the cost of equity and affect employee retention. For companies with ample cash and established operations, these effects are secondary in the short term, but they become more relevant if volatility persists.

Valuation Changes the Standard of Proof

Market expectations are embedded in price before a report. A company expected to grow slowly can rise after modest improvement. A company priced for years of rapid expansion can fall after an excellent quarter if one leading indicator weakens. This is why the percentage growth rates alone did not predict the July 31 moves.

Amazon entered the report with investors debating whether capital expenditure had outrun near-term cloud growth. AWS acceleration answered that concern directly. Apple entered with confidence in its product cycle and AI roadmap; guidance below the cited consensus reduced the probability of continued upside surprises. Reddit entered as a high-growth advertising platform; a warning about search referrals challenged the duration of that growth.

Valuation should not be reduced to a single price-to-earnings ratio. Amazon’s reported net income was distorted by the Anthropic gain, making a headline multiple less useful. Apple’s earnings are more stable, but the appropriate multiple depends on growth, Services mix, capital returns and product risk. Reddit is earlier in its profitability cycle, so investors may emphasize revenue growth, free cash flow and long-term margins.

The common principle is that a high valuation requires evidence that the company can sustain the assumptions behind it. The required evidence differs by business. For Amazon, it is cloud growth and return on capital. For Apple, it is ecosystem durability and a successful AI-enabled product cycle. For Reddit, it is direct audience growth and international monetization.

What Investors and Business Readers Should Watch Next

Amazon: conversion of infrastructure demand into cash

The next Amazon reports should be examined for AWS growth, AWS operating margin, backlog, capital expenditure and trailing free cash flow. If AWS remains near its current growth rate while operating margin stays high, the argument for the capital program strengthens. If growth decelerates before free cash flow recovers, concern about overbuilding will return.

Readers should also watch the relationship between proprietary chips and external accelerators. Greater use of Trainium and Graviton can improve AWS economics if customers adopt them at scale. It can also require pricing incentives and software support. Management’s annualized run-rate claims will become more credible if later disclosures show sustained revenue and broad customer use.

Retail margins and advertising growth remain important counterweights. Amazon has more flexibility to fund cloud construction when North America, International and advertising are profitable. A deterioration in consumer demand or fulfillment economics would make the same capital plan harder to support.

Apple: availability, margin and proof of AI demand

Apple’s September-quarter report should be read for whether advanced-chip constraints limited shipments, whether memory costs reduced product margin and whether the final revenue result landed within the 9% to 11% growth range. The product mix will matter because higher-priced devices and Services can offset some component pressure.

The most important strategic evidence will not be a management description of positive feedback. It will be adoption, usage and purchasing behavior. Device upgrade rates, installed-base growth, Services engagement and developer support will show whether new AI features are changing the economics of the ecosystem.

Governance will also move from planning to execution when John Ternus becomes CEO on September 1. Investors should look for continuity in operations and clearer explanation of how Apple will divide AI workloads among devices, private infrastructure and external partners.

Reddit: direct traffic and U.S. user stabilization

Reddit’s next report should be assessed first on U.S. daily active uniques, logged-in user growth and management’s description of search referrals. A recovery in direct or logged-in usage would reduce concern even if external search remains volatile. Continued sequential weakness would increase the burden on ARPU and international growth.

Advertising demand should be separated into pricing, impressions and customer breadth where disclosures permit. Rapid ARPU growth is valuable, but relying indefinitely on higher monetization per user can eventually create product tension. The healthiest outcome would combine user growth, engagement and better advertiser tools.

International ARPU is another key indicator. A gradual rise would support the long-term opportunity. A rapid increase achieved through aggressive ad loading could weaken retention. Reddit must balance monetization with the community experience that creates the inventory advertisers want.

Competitive Context: The Results Cannot Be Read in Isolation

Each company’s report arrived in a market where competitors were also changing the standard of performance. AWS was measured against Microsoft Azure and Google Cloud, not only against its own prior quarter. Apple was measured against other device ecosystems and against the AI experiences available through competing platforms. Reddit was measured against digital advertising leaders, social networks, search engines and a growing group of AI discovery products.

Competitive comparison is difficult because the companies disclose different metrics. Microsoft reports Azure growth but not Azure revenue as a standalone figure. Alphabet reports Google Cloud revenue, which includes more than one product category. Amazon reports AWS as a separate segment with revenue and operating income. Apple does not disclose device unit sales or AI-related revenue. Reddit’s user metric is designed around unique visitors and is not directly comparable with the daily-active-person measures used by every other platform.

The correct approach is therefore to compare direction, scale and economics while acknowledging the definitions. A false precision—such as declaring one cloud platform definitively “ahead” from one percentage-growth number—would hide differences in base size, currency effects, product mix and accounting.

AWS rejoined the fastest part of the cloud growth race

Amazon’s 37% AWS growth represented a substantial acceleration from 28% in the first quarter of 2026, 24% in the fourth quarter of 2025 and 20% in the third quarter of 2025. That sequence matters more than a one-quarter comparison with a competitor. It shows that AWS was gaining speed as new capacity came online and AI demand expanded.

Microsoft reported continued rapid Azure growth in its fiscal fourth quarter, while Alphabet reported an 82% increase in Google Cloud revenue to $24.8 billion. The Alphabet comparison is imperfect because the prior-year base, product mix and accounting are different, but it demonstrates the intensity of cloud demand. Alphabet also reported a cloud backlog of $514 billion, while Microsoft disclosed a large commercial remaining-performance-obligation balance. Customers are making long-duration commitments across more than one provider.

This environment supports a multicloud interpretation rather than a winner-takes-all conclusion. Large enterprises often distribute workloads for resilience, regulatory requirements, bargaining power and access to specialized services. A customer can use AWS for one application, Azure for another and Google Cloud for data or AI services. Market-share changes can occur even when every major provider grows rapidly.

AWS retains several structural advantages. It has a broad service catalog, a large installed customer base, a mature partner ecosystem and a segment that already produces significant operating profit. Its custom silicon can offer alternatives to scarce third-party accelerators. Amazon’s retail, logistics and advertising operations also provide internal workloads through which it can test infrastructure at scale.

The competitive risks are equally clear. Microsoft can integrate cloud and AI tools into enterprise software relationships. Alphabet can combine models, data infrastructure, search expertise and its own accelerators. Smaller specialized providers may compete on price or access to particular chips. Open-source software can make it easier for customers to move workloads or avoid proprietary services.

As a result, the durable question is not whether AWS can grow during an industry boom. It is whether it can preserve pricing, margin and customer relevance after capacity becomes more available. The second-quarter margin of 39.4% was strong, but competition can shift from securing supply to winning workloads. The current shortage supports providers; a future surplus would test differentiation.

Power, memory and advanced packaging are now strategic inputs

The AI infrastructure cycle has changed the list of assets that technology executives must manage. Data centers require access to electricity, transmission, cooling, land, construction labor and networking. Accelerators require advanced manufacturing and packaging. Large models require high-bandwidth memory and fast interconnects. A shortage in any one layer can delay the entire system.

Amazon’s capacity problem and Apple’s component warning are therefore connected even though the companies sell different products. Both depend on a semiconductor supply chain operating near its leading-edge limits. Cloud providers can pay more for scarce components because the assets generate enterprise revenue. Consumer-device companies must protect retail price points and margins. The competition is not always for the identical chip, but it can affect foundry schedules, packaging capacity, memory pricing and supplier priorities.

This creates a second-order risk. A company can correctly forecast customer demand and still miss financial targets because power connections are delayed, factories cannot expand quickly enough or component prices rise. Capital commitments can also become stranded if technology changes before facilities are complete. Management quality in this cycle therefore includes procurement, engineering and project execution—not only product design.

Amazon’s response is greater vertical integration through custom chips and infrastructure design. Apple has long used custom silicon to differentiate devices and control performance per watt. Both strategies reduce dependence on off-the-shelf processors but do not remove dependence on contract manufacturing, memory and equipment suppliers. Vertical integration moves the boundary of control; it does not make the supply chain disappear.

Apple competes through ecosystem economics rather than cloud scale

Apple’s strongest competitive asset is the interaction among hardware, operating systems, services, retail distribution and a large installed base. A competitor may offer a superior individual AI feature without immediately breaking that system. Customers weigh camera quality, battery life, security, applications, device interoperability, accessories, support and resale value as well as the newest software demonstration.

This gives Apple time, but not unlimited time. Platform advantages erode when a missing capability changes how users work or communicate. The company must make AI useful inside ordinary tasks rather than treat it as a separate destination. Integration with messages, photos, calendars, health data and device controls can be more valuable than a general chatbot, but it also requires permission systems and reliability that are difficult to build.

Competing mobile ecosystems can use AI to improve search, translation, photography, productivity and assistance. Independent applications can reach Apple users without Apple owning the underlying model. That reduces the risk that the iPhone becomes immediately obsolete, but it can shift value away from the operating system if users form their primary relationship with an external AI service.

Apple’s strategic task is therefore to keep the operating system and device at the center of the experience. On-device processing can improve latency and privacy. Private cloud systems can support larger workloads. Partnerships can provide capabilities that would take longer to build internally. The trade-off is control: every external dependency can affect cost, product differentiation and negotiating power.

The company’s Services business strengthens the ecosystem but also creates regulatory exposure. App distribution, payment rules and default services can be interpreted as integrated product design or as the use of platform control to disadvantage rivals, depending on the conduct and jurisdiction. Changes forced by regulators could reduce commission revenue or alter the user experience. They could also encourage innovation by allowing more competition inside the platform.

Because Apple does not disclose the profitability of each Services component, investors cannot calculate the precise exposure. They can monitor total Services growth, gross margin, developer activity and regulatory outcomes. The June quarter showed continued expansion, but the result below expectations reminded the market that a record level does not guarantee accelerating growth.

The CEO transition is also a competitive choice

Selecting John Ternus signals confidence in Apple’s product-engineering culture at a time when the company must coordinate hardware and AI. A hardware executive can bring discipline to device roadmaps, performance and manufacturing. The role also requires decisions about software openness, cloud partnerships, Services economics and regulatory engagement.

No CEO manages those areas alone. The transition’s success will depend on the executive team and the systems through which product, software, operations and finance resolve trade-offs. Tim Cook’s move to executive chairman can preserve institutional knowledge and external relationships. It can become a weakness only if responsibilities are unclear or if the new CEO lacks freedom to make necessary changes.

Competitors will not pause during the handoff. The first year will include product launches and investment decisions whose development began under Cook. That makes attribution difficult. Investors should judge the transition over several cycles, while still holding the new leadership accountable for communication, capital allocation and the pace of execution.

Reddit competes for intent, not only social attention

Reddit is often compared with social platforms because users post, comment and vote. Its commercial value is also close to search and commerce because many discussions begin with a question or purchase decision. A person researching a laptop, brokerage account, vacation or health experience may arrive with more specific intent than someone scrolling a general entertainment feed.

That intent can support effective advertising even at a smaller user scale. Advertisers care about the probability that an impression changes behavior, not only the number of people on the platform. Community context can indicate interests that are difficult to infer from broad demographic data.

The weakness is that search engines recognize the same value. Reddit pages often answer long-tail queries that would be expensive for a centralized publisher to cover. Search platforms can direct users to those pages, quote them in snippets or use them as inputs for AI-generated summaries. Reddit benefits from discovery but risks becoming an upstream content supplier whose brand is less visible than the answer.

Other social platforms compete through scale, identity graphs, video inventory and sophisticated advertising systems. Reddit competes through topic depth, anonymity or pseudonymity, and the perceived authenticity of peer discussion. Those qualities are difficult to preserve while increasing ad load and introducing automated content.

The company’s investment in internal search is therefore a competitive necessity, not a peripheral feature. A successful direct search product would let Reddit capture more of the intent generated by its archive. It could also create new advertising formats around questions and recommendations. The risk is that summarization reduces the incentive to read and contribute to the underlying threads.

Human-generated discussion is valuable because it is imperfect

Reddit’s content contains disagreement, corrections, anecdotes and context. Those imperfections can make it more useful than a single polished answer because readers can evaluate multiple experiences. They also make the content difficult to summarize safely. Advice may be outdated, biased or wrong, and popularity is not evidence.

AI systems can organize the archive, but they should preserve uncertainty and point users toward the discussion rather than present a synthetic answer as unquestioned fact. This is particularly important for medical, legal, financial and safety-related topics. Reddit is not an authoritative source merely because many people discuss an issue there.

Commercially, the company must avoid confusing authenticity with accuracy. Advertisers may value authentic conversation, while regulators and users expect protections against manipulation and harmful content. Better provenance, community signals and moderation tools can raise the value of the platform without converting it into a conventional publisher.

The advertising overlap creates an unusual triangle

Amazon and Reddit compete for advertising budgets even though their businesses look very different. Amazon offers proximity to transactions and purchase data. Reddit offers context and community intent. Apple influences both through privacy rules, device distribution and advertising services within its own ecosystem.

Changes made by Apple to tracking and permissions can alter the effectiveness of advertising across other platforms. Amazon’s first-party shopping data can reduce reliance on cross-app tracking. Reddit can use community context and on-platform behavior, but its smaller logged-in base limits the signals available for some visitors.

AI advertising tools can lower the cost of creating campaigns and improve targeting, benefiting all three in different ways. They can also make advertising inventory more interchangeable if every platform offers automated creative and optimization. Durable advantage will then depend on unique data, user intent, measurement and brand safety rather than the mere availability of AI tools.

Amazon’s 26% advertising growth and Reddit’s 64% growth show that the digital advertising market can support more than one winner. The rates are not directly comparable because Reddit is growing from a much smaller base. The strategic overlap matters because both companies are trying to prove that their user context produces better commercial outcomes than broad audience reach alone.

Scale protects each company in a different way

Amazon’s scale spreads infrastructure research and procurement across a large cloud and commerce business. Apple’s scale supports supplier negotiations, software distribution and customer retention. Reddit’s scale is smaller financially, but the breadth of its communities creates a content archive that would be difficult to reproduce through centralized production.

Scale can also create rigidity. Amazon must deploy enormous amounts of capital to move its growth rate. Apple’s installed base raises compatibility and reliability requirements. Reddit’s community structure makes product changes sensitive because volunteer moderators and long-time users can resist decisions that appear to prioritize monetization over culture.

The second-quarter reports showed all three sides of scale. Amazon used scale to convert AI demand into $42.2 billion of quarterly cloud revenue. Apple used scale to produce broad regional growth despite a complex supply environment. Reddit used community breadth to grow advertising rapidly, but remained vulnerable to an external platform that can redirect traffic at global scale.

Competitive advantage is therefore not a fixed label. It is the ability to use an existing asset—capital, ecosystem or community—to solve the next constraint more effectively than rivals. The July reports offered strong evidence that Amazon was doing so. Apple and Reddit retained valuable assets but had more work to demonstrate that those assets could neutralize the particular constraints disclosed in their outlooks.

The comparison also cautions against labeling every technology company an AI stock in the same sense. AI changes Amazon’s capital requirements, Apple’s product proposition and Reddit’s distribution environment through different mechanisms. The relevant metrics, time horizons and failure modes must therefore remain specific to each business.

A Practical Framework for Reading the Next Earnings Cycle

The next set of reports will be easier to interpret if readers avoid treating every favorable development as confirmation and every unfavorable development as failure. The relevant signals should be grouped into demand, capacity, monetization and cash. Those categories reveal whether a company is solving the bottleneck identified in this quarter or merely shifting it elsewhere.

Demand must be measured by behavior, not enthusiasm

For Amazon, demand evidence includes AWS revenue growth, new customer commitments, usage expansion and the conversion of backlog into recognized sales. For Apple, it includes device purchases, upgrades, installed-base activity and engagement with new AI features. For Reddit, it includes visits, logged-in usage, community participation and advertiser spending. Management descriptions of strong interest are useful context, but behavior is the stronger proof.

Demand quality also matters. A customer signing a long cloud contract is different from one experimenting with promotional credits. An iPhone buyer attracted by a durable feature is different from a purchaser brought forward by launch timing. An advertiser increasing spend because returns improved is different from one testing a new placement for a quarter. Later disclosures should help determine whether the growth is repeatable.

Capacity should improve economics, not only scale

Amazon’s additional capacity should eventually produce more revenue and operating cash than the cost of constructing and refreshing it. Utilization, pricing, chip efficiency and energy costs will determine that outcome. A data center can be full and still earn an inadequate return if customers negotiate lower prices or if equipment becomes obsolete too quickly.

Apple’s supply capacity should be judged by availability and margin. Securing more components at any price would protect unit volume but could damage profitability. The better outcome is enough capacity to meet high-value demand while preserving product quality and an attractive cost structure. Product mix can help, but it should not obscure persistent increases in component cost.

Reddit’s relevant capacity is partly digital: the ability to surface useful discussions, moderate content, translate communities and serve ads without degrading the experience. Adding more advertising inventory is not automatically productive if it reduces engagement. The company needs discovery and moderation systems that expand usable attention rather than simply increasing the number of impressions.

Monetization should not weaken the underlying asset

Amazon can monetize retail traffic through advertising, but excessive sponsored placement could reduce shopping relevance. Apple can monetize its installed base through Services, but aggressive fees or restrictions can create regulatory pressure and developer dissatisfaction. Reddit can raise ARPU, but a heavier ad load or intrusive targeting can weaken community trust.

The strongest businesses increase monetization because the product becomes more useful, not because they extract more from a static audience. Faster delivery can make Amazon customers shop more frequently. Better device integration can make Apple services more valuable. Improved search and translation can help Reddit users find communities they would otherwise miss. Revenue built on those improvements is more durable than revenue created only by higher prices or more ads.

Cash should confirm the strategic narrative over time

One quarter is rarely enough to judge a multiyear investment cycle. Amazon’s negative trailing free cash flow can be consistent with a valuable expansion if later cash generation rises. Apple’s large free cash generation can coexist with strategic weakness if the company underinvests in products customers eventually demand. Reddit’s strong cash conversion can be temporary if traffic acquisition and moderation costs rise sharply.

The connection between strategy and cash should become clearer over several reporting periods. Amazon should show that capacity produces revenue before another spending wave is required. Apple should show that research and partnerships support a new product cycle without structurally eroding margin. Reddit should show that it can fund direct discovery, international expansion and safety while retaining operating leverage.

Three Scenarios That Would Change the Current Interpretation

What would weaken the Amazon case?

The constructive Amazon interpretation would weaken if AWS growth decelerated materially while capital expenditure remained near the new plan, if backlog stopped expanding, or if AWS operating margin fell without a clear temporary reason. A continued free-cash-flow deficit would become more concerning if operating cash flow also slowed. Evidence that customers were optimizing workloads faster than they were adding AI capacity would challenge the assumption that supply was the principal limitation.

Conversely, sustained AWS growth above the rates reported in late 2025, stable segment margins and an eventual improvement in free cash flow would strengthen the conclusion that Amazon was building against real demand. Greater disclosure about custom-chip adoption and the economics of AI services would further reduce uncertainty.

What would improve the Apple case?

Apple’s current interpretation would improve if the company met or exceeded September-quarter guidance despite component constraints, maintained an underlying gross margin near recent levels and showed measurable adoption of its new AI features. Strong upgrade activity, continued growth in Greater China and renewed Services acceleration would demonstrate that the June quarter was not simply a favorable comparison.

The case would weaken if supply problems persisted into later product cycles, memory costs forced repeated margin reductions or AI features failed to influence customer behavior. A disorderly leadership transition or unclear division of authority between the CEO and executive chairman would add governance risk, although that should be assessed through actions rather than speculation.

What would reverse the Reddit concern?

Reddit could reverse the distribution concern by stabilizing U.S. daily active uniques, accelerating logged-in growth and showing that internal search or recommendations were bringing users back directly. Continued revenue growth accompanied by healthy impressions—not only higher pricing—would demonstrate that the advertising business had a broad foundation.

The concern would intensify if search referrals remained volatile, U.S. users declined for several quarters or ad growth depended increasingly on extracting more revenue from a stagnant audience. Rising content-quality costs or weaker community participation would also matter because Reddit’s inventory is produced by users rather than by a centralized editorial organization.

These scenarios are not price forecasts. They are tests of the operating explanations supplied by the July reports. A stock can rise while its underlying evidence weakens, or fall while the evidence improves, because valuation and positioning change. The objective is to identify which facts would confirm or challenge the business case rather than infer business quality from the ticker alone.

Frequently Asked Questions

Why did Amazon stock rise after earnings?

Amazon shares rose sharply after AWS revenue increased 37% to $42.2 billion, the segment’s fastest growth in 18 quarters, while AWS operating income rose 64%. The result made Amazon’s large AI infrastructure program look more closely connected to current demand and profit, although free cash flow remained negative on a trailing basis.

Did Amazon really earn $5.75 per share from its core business?

No. The $5.75 diluted EPS figure was reported under GAAP, but net income included $53.4 billion of pre-tax non-operating income, primarily related to Amazon’s Anthropic investment. Operating income of $27.5 billion is a better starting point for evaluating the recurring quarter.

Why was Amazon’s free cash flow negative?

Amazon’s operating cash flow grew, but spending on property and equipment increased even faster as the company expanded data-center, chip, network and fulfillment capacity. Under Amazon’s definition, trailing-12-month free cash flow was negative $7.6 billion.

Why did Apple stock fall after a record quarter?

The market focused on the next quarter. Apple guided to approximately 9% to 11% revenue growth, below the roughly 12% consensus cited by Reuters, and warned about advanced chip capacity and higher memory costs. The June-quarter result was strong, but the outlook implied less upside and lower gross margin.

How much of Apple’s quarter came from tariff refunds?

Management said tariff refunds added approximately two percentage points to fiscal third-quarter gross margin and about $0.11 to diluted EPS. The refunds were real benefits, but they should be separated from the recurring margin generated by product and Services sales.

Who is replacing Tim Cook as Apple CEO?

Apple announced that John Ternus would become chief executive on September 1, 2026. Tim Cook is set to become executive chairman, providing continuity while Ternus assumes operating responsibility.

Why did Reddit stock fall despite strong revenue and guidance?

Investors focused on choppy search referrals, slower U.S. daily active-user growth and a slight sequential decline in U.S. daily active uniques. Those issues raised concern about the durability of Reddit’s traffic funnel even though advertising revenue, profit, free cash flow and third-quarter guidance were strong.

Is Reddit profitable?

Reddit reported $253 million of GAAP net income in the second quarter of 2026, a 31.4% net margin. It also generated $262 million of operating cash flow and $261 million of free cash flow. Adjusted EBITDA was $343 million.

How dependent is Reddit on advertising?

Advertising generated $762 million of Reddit’s $805 million in quarterly revenue, or almost 95%. Other revenue, including data licensing, was $43 million. Advertising therefore remains the company’s dominant business.

What do search referrals mean for Reddit?

Search referrals are visits that arrive from external search engines. They can bring large amounts of low-cost traffic, particularly from users seeking specific community discussions. Reddit does not control how search engines rank links or display AI summaries, making this traffic more volatile than direct or logged-in usage.

Which company reported the highest revenue growth?

Reddit had the highest percentage growth at 61%, followed by Amazon at 20% and Apple at approximately 16%. The comparison is limited because the companies differ enormously in size, maturity, business model and capital intensity.

What is the main lesson from the Amazon, Apple and Reddit earnings?

The market rewarded evidence that a company could control and monetize its next growth constraint. Amazon showed accelerated cloud demand against a capacity shortage. Apple warned that supply and costs could limit the next quarter. Reddit disclosed volatility in a distribution channel controlled by external search platforms.

Final Assessment

The July 31 divergence was not a referendum on technology earnings as a whole. It was a test of evidence. Amazon supplied evidence that its most expensive strategic commitment was producing faster cloud revenue and higher segment profit. Apple supplied evidence of outstanding current demand but less certainty about near-term supply, margin and the economic impact of its AI strategy. Reddit supplied evidence of a highly profitable advertising engine while exposing a weakness in the path through which many users reach it.

Amazon therefore received the most favorable interpretation, but its burden remains the largest. A $220 billion capital plan requires years of utilization, pricing power and operational discipline. The quarter moved that debate in Amazon’s favor without resolving it.

Apple’s decline should not obscure the strength of a $109.4 billion quarter or a business capable of producing nearly $30 billion of net income in three months. The concern is forward-looking: component constraints, lower underlying margin, a Services result below expectations and an AI transition that has not yet produced a separately measurable growth engine. The leadership handoff increases the importance of execution rather than changing the quarter’s accounting.

Reddit’s selloff was the most severe because high-growth platforms are valued on the durability of future traffic. The company showed that it can monetize attention and convert revenue into cash. It now must show that it can deepen direct usage, stabilize U.S. audience growth and benefit from AI without allowing AI search products to absorb the value of its communities while reducing referrals.

The strongest common conclusion is that investors are becoming more selective about the AI trade. Spending is not automatically punished, and a headline earnings beat is not automatically rewarded. The decisive question is whether management can connect investment, distribution and product constraints to repeatable revenue and cash. Amazon made that connection most convincingly in this reporting cycle. Apple and Reddit left more of it to be proven in the quarters ahead.

Sources

• Amazon second-quarter 2026 results
• Amazon first-quarter 2026 results
• Reuters reporting on Amazon’s AWS growth and capital-spending outlook
• Reuters reporting on Amazon’s post-earnings market reaction
• Apple fiscal third-quarter 2026 results
• Apple fiscal third-quarter 2026 consolidated financial statements
• Apple fiscal third-quarter 2026 SEC earnings exhibit
• Reuters reporting on Apple’s results, guidance and supply constraints
• Apple announcement of the Tim Cook–John Ternus leadership transition
• Reddit second-quarter 2026 results
• Reddit second-quarter 2026 shareholder letter
• Reddit second-quarter 2026 SEC filing index
• Reuters reporting on Reddit’s guidance, advertising growth and user trends
• Reddit first-quarter 2026 results
• Microsoft fiscal fourth-quarter 2026 results
• Alphabet second-quarter 2026 results
• Meta second-quarter 2026 results
• Reuters analysis of Big Tech capital spending and free-cash-flow pressure

This article is provided for general informational purposes and does not constitute financial, investment, tax, or legal advice.

Affiliate disclosure: Businessfinance.news may earn compensation from qualifying actions completed through selected links on this website, at no additional cost to the reader. Affiliate relationships do not influence our editorial reporting, analysis, or conclusions.