Why Stocks Fall After Earnings Beats: WDC, SNDK, APP, ELF

Last updated: August 6, 2026, 5:15 a.m. EDT

Four companies reported results after the U.S. market closed on August 5, and all four shares initially moved lower. Yet the earnings were not remotely alike. Western Digital and Sandisk delivered large revenue and profit gains tied to the artificial-intelligence infrastructure boom. AppLovin still grew at a rate most companies would envy, but its revenue and adjusted EBITDA landed below the demanding targets investors had built into the stock. E.l.f. Beauty beat consensus estimates and raised its full-year outlook, while the details showed that acquired brand rhode and a tariff-refund benefit did much of the heavy lifting.

The common question—why do stocks fall after an earnings beat?—has a simple answer and a complicated one. The simple answer is that a stock price reflects expectations about the future, not merely a comparison between reported results and a published analyst consensus. The complicated answer is that those expectations include revenue mix, margins, cash conversion, prior share-price gains, management credibility, competitive signals, one-time accounting items and assumptions that may never appear in the headline “beat” or “miss.” A company can exceed the average forecast and still disappoint the market’s more ambitious informal expectation. It can also post excellent operating results while reporting a statutory profit that is flattered by a non-operating gain.

That distinction mattered across all four reports. Western Digital’s hard-disk-drive business produced record profitability and strong free cash flow, but the stock had already risen sharply and its first-quarter outlook did not extend the upside narrative far beyond what investors had anticipated. Sandisk’s NAND flash economics were extraordinary—an 84.6% non-GAAP gross margin and a 372% year-over-year revenue increase—but two-thirds of its sequential revenue growth came from price, the consumer segment contracted, and the shares had climbed nearly 470% during 2026 before the report. AppLovin’s quarter was not a clean beat: revenue was near the low end of its prior range, adjusted EBITDA slipped below the company’s own guidance, and the next-quarter EBITDA range trailed the consensus cited by market reports. E.l.f. Beauty’s numbers were cleaner at first glance, but investors had to separate organic brand momentum from acquisition revenue and a temporary tariff-related windfall.

The immediate selloffs therefore did not show that earnings were “bad.” They showed that the market judged each report against a different hurdle. For Western Digital and Sandisk, the hurdle was set by exceptional demand, shortages, pricing power and enormous share-price appreciation. For AppLovin, the hurdle was a history of rapid AI-driven advertising gains and unusually high margins. For e.l.f. Beauty, the hurdle was whether its core namesake brand could regain momentum while the company integrated a $1 billion acquisition and carried more debt.

Key Takeaways

• Western Digital: Fiscal fourth-quarter revenue rose 44% year over year to $3.747 billion, non-GAAP EPS reached $3.56 and free cash flow was $1.281 billion. The shares still fell more than 9% in late trading as investors weighed a strong but largely anticipated outlook against a large prior rally.
• Sandisk: Fiscal fourth-quarter revenue climbed 372% year over year to $8.965 billion and non-GAAP gross margin reached 84.6%. The stock fell nearly 8% at one point because expectations had risen even faster, the outlook was interpreted differently across consensus providers, and pricing—not volume—drove most sequential growth.
• AppLovin: Second-quarter revenue increased 53% to $1.924 billion, but it was near the bottom of prior guidance. Adjusted EBITDA of $1.614 billion fell below the company’s May range, while third-quarter EBITDA guidance was softer than the consensus cited by financial media.
• e.l.f. Beauty: Fiscal first-quarter sales rose 36% to $479.4 million and adjusted EPS was $1.75. The company raised annual guidance, but rhode contributed roughly $160 million and tariff refunds added about 1,050 basis points to gross margin.
• The broader lesson: A published consensus is only one benchmark. Share prices react to the quality, durability and surprise value of earnings, as well as to what was already embedded in valuation.

The Earnings Scorecard: What Was Reported and What the Market Saw

Figures are in U.S. dollars. Market reactions refer to extended trading on August 5, 2026 and varied during the session. Company results are reported; consensus comparisons are from the cited financial-news sources and may differ by data provider.

Why Stocks Fall After an Earnings Beat

An earnings release is not a school examination with a fixed passing grade. The published consensus is an average assembled from participating analysts at a particular time. It may not capture last-minute estimate changes, buy-side models, options-market positioning or the assumptions held by investors who own the stock. Even when a company beats that average, the distribution of expectations can sit well above it.

The difference is often described as the “whisper number,” but that label can make the process sound more mysterious than it is. In practice, investors constantly update their estimates using peer-company reports, channel checks, commodity prices, management conferences, customer spending plans and movements in related securities. Seagate’s strong results one week before Western Digital, for example, raised the reference point for hard-drive demand. A published Western Digital consensus calculated before Seagate’s report could be less informative than the revised expectations investors carried into August 5.

Several dimensions determine whether an earnings beat is valuable:

• Magnitude: A one-percent revenue beat means less when the stock price assumes a five-percent surprise.
• Quality: Organic growth, repeatable pricing and operating cash flow usually carry more weight than acquisition revenue, tax benefits, fair-value gains or delayed spending.
• Mix: Growth from a lower-margin or less strategic business can be less valuable than slower growth in a high-retention segment.
• Guidance: Investors frequently look through the completed quarter and focus on the next one. A strong backward-looking result can be overshadowed by an outlook near consensus.
• Durability: Shortages can create extraordinary margins, but they also invite supply responses, customer substitution and eventual normalization.
• Valuation: The same financial result can produce different stock reactions depending on how much optimism is already priced in.
• Positioning: A heavily owned momentum stock can fall sharply when investors rush to protect gains, even if the long-term thesis remains intact.

These principles help explain why the four August 5 reactions should not be treated as one phenomenon. AppLovin missed important operating benchmarks. E.l.f. Beauty beat them, but the composition invited questions. Western Digital and Sandisk exceeded many forecasts, yet their stocks had already capitalized years of expected AI-storage demand into current prices. The market did not deny their growth; it demanded evidence that the growth could keep outrunning an already elevated curve.

The Market Backdrop Added Another Layer

The reports arrived after a divided U.S. trading session. The Dow Jones Industrial Average rose 263.18 points, or 0.49%, to a record 54,349.06 on August 5. The S&P 500 slipped 0.17% to 7,723.52, and the Nasdaq Composite fell 0.83% to 26,363.44. Healthcare and selected Dow components performed well, while technology lost momentum. The divergence mattered because three of the four reporting companies had benefited from powerful growth narratives linked to AI, digital advertising or consumer-brand expansion.

A softer Nasdaq can make an after-hours disappointment more severe. Investors who have already reduced exposure to expensive technology during the regular session may be less willing to defend a stock that delivers merely “good” results. That does not mean the index decline caused the company-specific moves. It means the reports were assessed in a market that had become more selective after several sessions of gains.

The Dow’s record also illustrated the rotation visible beneath the index level. A rising market does not imply that every growth stock receives a generous response. On August 5, earnings winners such as Eli Lilly and Disney helped the blue-chip average, while technology shares weakened. Western Digital and Sandisk therefore reported into a session in which investors were already differentiating between sectors and between individual earnings stories rather than buying risk indiscriminately.

A Timeline of the August 5 Earnings Event

1. July 28, 2026: Seagate reported fiscal fourth-quarter revenue of $3.629 billion, a 52.7% non-GAAP gross margin, $5.71 in non-GAAP EPS and $1.1 billion of free cash flow. Its fiscal first-quarter revenue guide was $4.1 billion, plus or minus $100 million. The report established a demanding peer benchmark for Western Digital.
2. August 5 regular session: The Dow closed at a record while the S&P 500 and Nasdaq Composite declined. Western Digital and Sandisk both fell during regular trading before releasing results.
3. Shortly after the close: Western Digital reported fiscal fourth-quarter and full-year 2026 results. The company beat commonly cited revenue and adjusted-EPS estimates and guided fiscal first-quarter revenue to approximately $4.1 billion.
4. After the close: Sandisk reported extraordinary NAND pricing, gross margin and data-center growth. It guided fiscal first-quarter revenue to $10.3 billion to $10.8 billion and non-GAAP EPS to $44 to $46.
5. After the close: AppLovin reported second-quarter revenue of $1.924 billion and adjusted EBITDA of $1.614 billion. Both fell short of the more demanding expectations surrounding the stock, and adjusted EBITDA was below the company’s prior guidance range.
6. After the close: E.l.f. Beauty reported fiscal first-quarter revenue of $479.4 million, raised annual sales and earnings guidance, and disclosed the large contribution from rhode and tariff refunds.
7. Late extended trading: All four stocks were lower, although the magnitude changed as investors parsed releases and awaited management commentary. The sharpest reaction was in AppLovin; the mildest was in e.l.f. Beauty.

Western Digital: A Strong Hard-Drive Quarter With a Complicated GAAP Profit

Western Digital’s report contained nearly everything investors had been waiting for from the newly focused company: faster revenue growth, expanding margins, large free cash flow and confidence in cloud demand. Fiscal fourth-quarter revenue reached $3.747 billion for the period ended July 3, up 12% sequentially and 44% from a year earlier. Non-GAAP gross margin rose to 54.4%, non-GAAP operating margin reached 44.2%, and non-GAAP diluted EPS was $3.56. Operating cash flow was $1.389 billion and free cash flow was $1.281 billion.

Those are exceptional figures for a manufacturer operating in a historically cyclical industry. They reflect tight supply, higher average selling prices, disciplined capacity and the growing storage requirements of hyperscale cloud customers. Western Digital said cloud represented 89% of its business in the quarter. That concentration is both the core of the earnings story and one of its principal risks.

The company’s fiscal first-quarter 2027 outlook also pointed higher. Management guided revenue to $4.1 billion, plus or minus $100 million, with a non-GAAP gross margin of 55% to 56% and non-GAAP EPS of $4.00, plus or minus $0.15. At the midpoint, the forecast implies another step up in sales and profitability. The board declared a quarterly dividend of $0.15 per share, payable September 17 to holders of record on September 8.

Why, then, did Western Digital shares fall more than 9% in late trading? The best explanation is not hidden deterioration. It is the interaction between a strong report and a stock that had already priced in a powerful cycle. Barron’s reported that the shares had gained roughly 201% during 2026 before the release. When a stock triples in little more than seven months, investors do not need a reason to reject the business. They need a reason to believe the next set of expectations will rise faster than the price already has.

Western Digital Is No Longer the Same Company Investors Followed Before 2025

A crucial point is frequently lost in automated summaries: Western Digital and Sandisk are now separate public companies. Western Digital completed the separation of its flash business in February 2025. The continuing Western Digital operation is primarily a hard-disk-drive company, while Sandisk owns the separated flash-memory business. Calling both companies generic “memory makers” obscures the different technologies, cost structures and demand cycles.

Hard disk drives store data magnetically on spinning platters. They remain essential for mass-capacity storage because their cost per terabyte can be attractive for large data sets that do not require the lowest possible latency. NAND flash stores data electronically and is used in solid-state drives, smartphones, PCs and a growing range of data-center applications. AI infrastructure needs both. High-performance accelerators require fast local and tiered storage, while the enormous data generated, retained and reused by AI systems also expands the need for economical capacity storage.

Western Digital’s post-separation strategy is therefore narrower and easier to evaluate. The company is no longer balancing two capital-intensive businesses with different technology cycles. Investors can assess its areal-density roadmap, cloud qualifications, unit economics and supply discipline directly against Seagate. The benefit is strategic clarity. The cost is concentration: the company has fewer unrelated businesses to offset a hard-drive downturn.

The Operating Result Was Better Than the Headline GAAP Comparison Suggests

Western Digital reported GAAP net income attributable to the company of $3.195 billion, or $8.21 per diluted share, compared with $284 million, or $0.85 per diluted share, a year earlier. Read without the income statement, that looks like an elevenfold surge produced by hard-drive operations. It was not.

The quarter included $1.684 billion of “interest and other income,” and Western Digital’s cash-flow reconciliation identified a $2.05 billion gain connected to the company’s retained interest in Sandisk. For the full fiscal year, the corresponding retained-interest gain was $6.498 billion. These mark-to-market effects increased GAAP earnings but did not represent revenue from selling more drives or cash generated by customers in the quarter.

That does not make GAAP earnings invalid. Accounting standards require companies to recognize relevant changes in the value of investments. It does mean that operating income, adjusted earnings and cash flow are more useful measures for judging the quarter’s hard-drive performance. On those measures, the result was still excellent. GAAP operating income was $1.563 billion, and free cash flow exceeded $1.28 billion. The analytical correction therefore does not weaken the operating story; it prevents an investment gain from being mistaken for manufacturing profit.

Cloud Concentration Is Producing Exceptional Economics

Western Digital’s cloud mix shows how dramatically the hard-drive market has shifted. Enterprise and hyperscale customers buy high-capacity drives in large volumes, qualify products through lengthy technical processes and care intensely about total cost of ownership. A drive that stores more data in the same physical footprint can reduce the number of racks, power supplies and supporting systems needed for a given amount of capacity. That creates room for manufacturers to capture part of the economic value through pricing.

The company highlighted next-generation energy-assisted perpendicular magnetic recording, or ePMR, products with capacity reaching as high as 40 terabytes per drive. The technology roadmap matters because the hard-drive industry competes not merely on units shipped but on exabytes and dollars per terabyte. A manufacturer can sell fewer physical drives while delivering more total capacity and potentially better margins.

Cloud customers also value supply assurance. Training and operating AI systems requires data pipelines that extend well beyond accelerators. Model checkpoints, training corpora, logs, synthetic data, media files and inference outputs all need storage. The result has been stronger demand for both high-performance flash and mass-capacity disk. In Western Digital’s case, the constraint has shifted from finding customers to allocating output and managing a long qualification cycle.

Yet concentration changes bargaining dynamics. When 89% of quarterly revenue is tied to cloud, a small number of large customers can influence product schedules, qualifications and purchasing patterns. Hyperscalers have substantial technical resources and can shift workloads across storage tiers. They may accept higher prices during scarcity, but they also have strong incentives to dual-source, redesign systems and negotiate lower cost when supply loosens.

Seagate Raised the Reference Point Before Western Digital Reported

Seagate’s July 28 results made Western Digital’s challenge harder. Seagate reported fiscal fourth-quarter revenue of $3.629 billion, a 52.7% non-GAAP gross margin, $5.71 in non-GAAP diluted EPS and $1.1 billion in free cash flow. It guided fiscal first-quarter revenue to $4.1 billion, plus or minus $100 million, and non-GAAP EPS to $7.30, plus or minus $0.20.

The companies’ per-share earnings cannot be compared directly because they have different share counts, capital structures and accounting adjustments. The operating comparison is nevertheless useful. Both reported sharply higher margins, strong cash generation and the same midpoint for next-quarter revenue. Seagate’s report confirmed that the demand and pricing environment was already strong across the industry. That reduced the novelty of Western Digital’s beat.

This is a classic example of a “read-through.” When a close competitor reports first, investors use its results to revise assumptions for the company that reports later. Western Digital may beat the formal consensus while only matching the expectation updated after Seagate. The market reaction can look irrational if the observer sees only the old consensus number. It becomes more coherent when the competitor’s report is incorporated.

Western Digital’s result also arrived after its shares had gained far more than the broad market. A large rally changes the burden of proof. The company was no longer being valued as a recovering cyclical manufacturer. It was increasingly being valued as a durable beneficiary of AI-related data growth. The distinction affects how investors treat guidance. A cyclical recovery can justify one or two strong quarters; a structural revaluation requires confidence that elevated margins and cash flows will persist across capacity expansions and technology transitions.

What the First-Quarter Guidance Did—and Did Not—Prove

At the midpoint, Western Digital’s $4.1 billion revenue guide represents approximately 9% sequential growth from the fourth quarter. The 55% to 56% non-GAAP gross-margin range is also above the 54.4% just reported. That is not weak guidance. It signals ongoing pricing power and favorable product mix.

The limitation is that it may have been close to what the market already expected after Seagate and after Western Digital’s share-price run. Guidance can be objectively strong and still lack incremental surprise. Investors also had to decide how much of the margin expansion reflected lasting product economics and how much reflected scarcity. If shortages ease sooner than expected, the high-end margin assumptions embedded in valuation can reverse quickly.

Management’s visibility is better than in a traditional spot-driven cycle because cloud qualifications and supply agreements extend forward. But visibility is not the same as certainty. Customer forecasts can change; capacity can arrive; product transitions can slip; and buyers can optimize workloads. The key question is not whether Western Digital will sell drives next quarter. The guidance answers that positively. The harder question is whether the company can sustain current returns when the industry is no longer capacity-constrained.

The Bull Case for Western Digital

The strongest supportive interpretation begins with industry structure. The hard-drive market has consolidated around a small number of suppliers, with Western Digital and Seagate dominating mass-capacity enterprise demand. Building advanced drive technology requires years of research, precision manufacturing and customer qualification. New entrants cannot quickly add competitive capacity merely because margins rise.

Second, data creation is not slowing. AI expands both the volume and value of stored data. Even when a workload runs primarily on flash, organizations often move older or less latency-sensitive information to hard drives. Higher-capacity drives can lower total infrastructure cost, giving suppliers room to innovate without making the product economically obsolete.

Third, Western Digital is converting profits into cash. The $1.281 billion of quarterly free cash flow is not an accounting revaluation. It gives the company flexibility to reduce debt, invest in product development, pay dividends and repurchase shares. The reinstated quarterly dividend adds evidence that the board views cash generation as sufficiently durable to support distributions, although future dividends remain discretionary.

Finally, the separation from Sandisk has removed a conglomerate discount and made management accountability clearer. Investors can evaluate Irving Tan’s execution against a focused set of operating targets. If the company maintains supply discipline and advances areal density without costly delays, current margins may prove more durable than historical hard-drive cycles suggest.

The Skeptical Case for Western Digital

The skeptical case starts with the same margin expansion that supports the bullish view. A 54% gross margin in a hardware business attracts scrutiny. Customers may tolerate elevated pricing when capacity is scarce, but high margins encourage every participant to improve yield, accelerate product ramps and seek additional output. The industry’s concentration limits a sudden flood of new competitors, yet existing suppliers can still misjudge demand and create excess inventory.

Cloud concentration is another concern. Hyperscale customers are financially strong, but their purchasing is lumpy. A delayed data-center build, a change in architecture or a period of inventory digestion can move quarterly demand materially. Western Digital’s reliance on a few sophisticated buyers means strong end-market growth does not eliminate customer power.

Technology execution also matters. Higher-capacity drives require complex transitions. Western Digital is pursuing ePMR while Seagate has emphasized heat-assisted magnetic recording, or HAMR, through its Mozaic platform. Different roadmaps can coexist, but delays or yield problems can affect cost, qualification timing and customer share. A product that performs well in a laboratory does not create revenue until it is manufactured at acceptable yield and approved by customers.

Valuation is the final concern. A company can execute well while its stock produces poor returns if the starting price assumes too much. The late-trading decline was a reminder that investors had already paid for substantial improvement. The report supported the business thesis; it did not necessarily support every price reached before the release.

What Western Digital Investors Need From the Earnings Call

The earnings call’s most useful information is unlikely to be another description of AI demand. Investors already know demand is strong. They need precision about contracted visibility, capacity commitments, pricing, customer concentration and the cost of upcoming technology ramps.

Management should clarify how much fiscal 2027 output is covered by firm agreements, how pricing mechanisms work, and what conditions could allow customers to alter purchases. It should explain whether gross-margin guidance depends more on price, mix, cost reductions or shipment volume. It should also separate near-term scarcity economics from the normalized return profile it expects after supply and demand rebalance.

Capital allocation deserves equal attention. Free cash flow has risen rapidly, but the company must balance shareholder returns with debt reduction and investment in future platforms. A generous buyback at an elevated valuation can create less value than a disciplined program tied to cash generation and technology needs. The dividend is modest relative to the quarter’s cash flow; its significance lies more in the signal of confidence than in its immediate yield.

Western Digital’s quarter was therefore not a failure disguised by headline beats. It was a strong operating result that confronted an even stronger set of expectations. The stock reaction says more about the price investors had assigned to future excellence than about the quality of the drives sold during the quarter.

Sandisk: Extraordinary NAND Economics Meet an Even Higher Bar

Sandisk’s report was the most dramatic of the four. Fiscal fourth-quarter revenue reached $8.965 billion, up 51% sequentially and 372% from a year earlier. GAAP gross profit was $7.582 billion, GAAP operating income was $7.037 billion and GAAP net income was $6.903 billion. Non-GAAP gross margin was 84.6%, and non-GAAP diluted EPS was $39.25.

Those figures resemble the economics of a highly scalable software company more than a semiconductor manufacturer. They were made possible by an acute NAND shortage, rapid price increases, AI-data-center demand and a new contracting model that gives Sandisk customers longer supply visibility. Yet the shares fell nearly 8% at one point in extended trading. The reaction becomes understandable only after considering the stock’s prior rise, the source of the growth and the uncertainty around how long such margins can last.

Sandisk said approximately one-third of its sequential revenue increase came from higher volume and two-thirds from higher pricing. That split is the center of the analysis. Price-led growth is highly profitable when manufacturing costs do not rise at the same pace. It can also be cyclical. A shortage permits suppliers to reset prices rapidly; additional supply or weaker demand can reverse the mechanism.

Sandisk’s End Markets Told Three Different Stories

Data-center revenue was $2.977 billion, up 103% sequentially and 1,298% year over year. Edge revenue—which includes mobile, client computing, automotive and other connected applications—was $5.432 billion, up 48% sequentially and 392% year over year. Consumer revenue was $556 million, down 32% sequentially and 5% from a year earlier.

The data-center result validates Sandisk’s exposure to enterprise solid-state drives and AI storage. AI clusters require more than accelerators and networking. They need high-throughput storage close to compute, rapid access to training data and reliable movement across memory and storage tiers. Sandisk’s enterprise products benefit as data-center operators seek capacity and performance at a time when supply is constrained.

Edge growth shows that the shortage is not confined to hyperscale customers. Higher NAND pricing is flowing through phones, PCs and other devices. Device manufacturers need to secure components before they can ship finished products, and the fear of being unable to obtain supply can reinforce forward purchasing.

Consumer weakness is the counterpoint. Retail products and removable storage do not have the same strategic priority or contractual protection as data-center demand. Consumers can delay purchases, choose lower capacity or switch brands. A 32% sequential decline in consumer revenue does not threaten the entire Sandisk story, because the segment represented only about 6% of quarterly sales. It does show that the boom is uneven and that high component prices can suppress demand in less essential channels.

The New Business Model Changes Visibility—but Not Every Risk

Sandisk has been moving customers from short-cycle purchasing toward what it calls its new business model, or NBM. The company disclosed eight agreements across data-center and edge customers with minimum contracted revenue of $93.9 billion at floor pricing. Reuters reported that the agreements involved six customers and had a median duration of four years. Sandisk said about half of fiscal 2027 production and two-thirds of fiscal 2028 production would be sold under these arrangements.

The scale is remarkable. Long-term agreements can reduce the classic semiconductor problem in which customers over-order during shortages, cancel when supply improves and force manufacturers to absorb inventory. Floor pricing can protect economics, while customer prepayments or guarantees help fund capital and demonstrate commitment.

But a headline contract value is not the same as recognized revenue. Investors need to understand timing, volume obligations, price-adjustment formulas, termination clauses, credit support and the treatment of customer deposits. Sandisk reported $59.8 billion of remaining performance obligations at quarter end and approximately $91.1 billion when including agreements signed afterward. It also referred to $16.5 billion of guarantees. Those figures measure related but different concepts and should not be added together casually.

The cash-flow statement illustrates the distinction. Sandisk generated $7.083 billion of free cash flow, but its adjusted free-cash-flow measure was $5.035 billion after excluding $1.938 billion of customer prepayments and deposits connected to the new agreements. The prepayments are genuine cash. They are not the same as cash earned from delivering products in the quarter. Excluding them helps readers evaluate underlying conversion, while including them shows the financing benefit of the contracting model.

Long-term contracts also shift risk rather than eliminate it. Customers gain supply certainty, but Sandisk may surrender part of the upside if market prices rise above contract formulas. Conversely, floor prices can cushion a downturn. The value depends on contract design and enforceability, details that public summaries cannot fully reveal.

An 84.6% Gross Margin Is Powerful—and Difficult to Extrapolate

Gross margin measures the portion of revenue left after the direct costs of producing goods. Sandisk’s 84.6% non-GAAP margin means that less than sixteen cents of every sales dollar was consumed by the adjusted cost of revenue during the quarter. That level reflects a rare combination of high prices, limited supply, favorable mix and manufacturing leverage.

Investors should resist two opposite errors. The first is dismissing the margin as unsustainable and therefore meaningless. Even if margins normalize, the cash generated during the current cycle strengthens the balance sheet and funds technology, buybacks and joint-venture commitments. A temporary period of exceptional profitability can create lasting financial value.

The second error is treating the margin as a permanent baseline. NAND has a long history of boom-and-bust cycles. Manufacturers invest in process transitions and new capacity; bit output rises; customers digest inventory; and prices can fall quickly. The current market differs because AI demand is large and Sandisk has adopted longer contracts. It does not follow that the economic cycle has disappeared.

Sandisk’s first-quarter gross-margin guidance of 83% to 85% indicates that management expects the extraordinary economics to continue at least in the near term. The revenue range of $10.3 billion to $10.8 billion and non-GAAP EPS range of $44 to $46 were above the LSEG averages cited by Reuters. Other data providers carried higher estimates, contributing to the negative stock reaction. This divergence is a concrete example of why “the consensus” is not a single universal number.

Why Different Estimate Providers Produced Different Narratives

Reuters reported an average first-quarter revenue estimate of $10.47 billion and adjusted EPS estimate of $43.12. Against those figures, Sandisk’s midpoint guidance was a beat. Other financial reports cited a revenue consensus near $10.8 billion, putting the high end of the company’s range around the forecast rather than above it. Both descriptions can be accurate if the services include different analysts, update at different times or classify estimates differently.

For a fast-moving company, estimate timing is especially important. Analysts may have raised forecasts after price data, competitor comments or investor meetings. A consensus pulled one day before earnings can differ substantially from one compiled earlier. The market trades on the freshest available information, not on whichever average produces the cleanest headline.

That helps explain why Sandisk could deliver a quarter far above older forecasts and still fall. The stock had risen nearly 470% during 2026, according to Reuters, and more than 3,000% over the preceding year according to another market report. Such performance implies that investors were not merely expecting a normal beat. They were capitalizing a shortage, sustained price increases, long contracts and expanding AI demand far into the future.

When a stock has moved that far, a guide that is “above consensus” by one measure may still be insufficient. The relevant question becomes how much the next twelve months of estimates can rise after the report. If current prices already assume 80%-plus gross margins and multiyear scarcity, the incremental upside must come from an even tighter market, longer duration or higher terminal profitability.

The Kioxia Joint Venture Is Central to Sandisk’s Supply Economics

Sandisk does not manufacture NAND in isolation. Its Flash Ventures arrangements with Kioxia are structured with Sandisk holding 49.9% and Kioxia 50.1%. The partnership operates fabrication capacity in Japan and supports the development and production of NAND technology.

The joint venture provides scale and shares capital requirements, but it also makes coordination important. Technology transitions, wafer allocation, equipment spending and output plans affect both partners. Sandisk’s capacity cannot be understood solely from its own capital-expenditure line because the joint-venture structure influences how manufacturing investments and costs appear in financial statements.

This matters during a shortage. Investors want more output, but a rapid supply response can weaken pricing. Management must balance customer commitments with the risk of adding too much capacity. The most valuable outcome is not necessarily maximum units. It is sufficient output to meet strategic agreements while preserving return on invested capital.

Sandisk said its BiCS8 technology represented a majority of bit production, while BiCS10 was ramping and its QLC “Stargate” product had begun revenue shipments. These transitions can lower cost per bit and improve product capability. They also carry execution risk. Yield, qualification and product-mix changes can affect margins even when market pricing remains favorable.

What Sandisk’s Buyback Says About Capital Allocation

The board added $14 billion to Sandisk’s repurchase authorization, bringing remaining authorization to $15.5 billion after approximately $4.5 billion of repurchases. The size reflects the extraordinary cash generated by the shortage and customer prepayments. It also raises a valuation question.

Buybacks create value when a company repurchases shares below their intrinsic value and does not compromise investment needs or balance-sheet resilience. They can destroy value when management buys aggressively after a speculative rise or when the cash would earn a better return in technology and capacity. An authorization is not a completed purchase; the board can change pace as conditions evolve.

Sandisk’s management can argue that long-term contracts and prepayments reduce financial risk, allowing more cash to be returned. A skeptic can counter that NAND remains cyclical and that the company may need flexibility when customers, governments and partners seek additional capacity. The appropriate pace depends on the durability of contract cash flows and on whether management’s estimate of normalized earnings is realistic.

The Bull Case for Sandisk

The strongest case rests on a structural change in demand. AI data centers need larger and faster storage pools, and enterprise SSDs are moving closer to the core of system architecture. Sandisk’s data-center revenue more than doubled sequentially, which suggests that the opportunity is already visible in recognized sales rather than existing only in forecasts.

Long-term agreements improve visibility. With half of fiscal 2027 output and two-thirds of fiscal 2028 production expected to be covered by the new model, Sandisk may be less exposed to abrupt order cancellations than in previous cycles. Floor prices and customer guarantees can protect returns, while prepayments finance working capital and investment.

The company also has a powerful technology and manufacturing position through its Kioxia partnership. Barriers to entry are high, and qualifying enterprise products takes time. Even if new supply is announced, converting it into reliable, customer-approved output can require years.

Finally, current earnings produce enormous optionality. Sandisk can invest, repurchase stock and strengthen its financial position simultaneously. If high margins persist longer than the market expects, the earnings base can outrun even aggressive estimates.

The Skeptical Case for Sandisk

The skeptical reading begins with the source of growth: two-thirds of the sequential increase came from price. Price is valuable, but it can reverse. Customers will redesign products, qualify alternatives and resist capacity increases if NAND costs make finished devices uneconomic. Consumer weakness may be an early example of demand responding to higher prices.

Contracts reduce volatility only to the extent their terms are firm. Minimum-revenue figures at floor pricing do not reveal every contingency. If technology changes, customers may seek different products; if market prices fall, disputes over specifications and commitments can become more important. The concentration of agreements among six customers also creates counterparty and negotiating risk.

The 84.6% gross margin may be near a cyclical extreme. A small change in price can have a large effect on profit when margins are this high. Investors who value the stock on peak earnings could suffer even if revenue remains above historical levels. The business does not need to collapse for the multiple and earnings base to normalize at the same time.

The prior share-price rise magnifies each concern. After a multi-hundred-percent advance, selling can reflect profit-taking, risk limits or a reassessment of position size rather than a new conclusion that the company is weak. That makes short-term price action an imperfect guide to long-term value, but it also makes the stock vulnerable when guidance does not decisively exceed the highest expectations.

What to Watch at Sandisk’s August 13 Investor Day

Sandisk scheduled an investor day for August 13, giving management an opportunity to explain the new contracting model in greater depth. Investors should look for a bridge between the $93.9 billion minimum-revenue figure, remaining performance obligations, guarantees, deposits and the company’s expected annual revenue recognition.

They should also examine the assumed NAND price path embedded in guidance. Floor prices can protect the downside, but the economic value depends on how contracts participate in market price increases and how cost reductions are shared. Management’s view of normalized gross margin will be more important than a claim that current demand is strong.

Technology and capital plans deserve scrutiny. The timing of BiCS10, the mix of QLC products, enterprise qualification schedules and investment through Flash Ventures will determine how much supply can be added without sacrificing returns. Investors also need a clear hierarchy for cash: required joint-venture investment first, balance-sheet resilience second, and repurchases based on valuation rather than a desire to exhaust the authorization.

Sandisk’s quarter was objectively extraordinary. The negative stock reaction did not disprove that. It showed that a company can produce historic margins and still disappoint a market that has already converted scarcity into a very large valuation.

AppLovin: The Sharpest Selloff Followed a Real Expectations Miss

AppLovin’s reaction differed from Western Digital’s and Sandisk’s because its report contained a clearer operating disappointment. Second-quarter revenue rose 53% year over year to $1.924 billion, net income increased 55% to $1.267 billion, and diluted EPS was $3.76. Adjusted EBITDA climbed 58% to $1.614 billion, equal to an 84% margin. Those figures would qualify as outstanding for almost any mature public company.

They were not enough for AppLovin. The stock fell about 17% in one late-trading snapshot and more than 20% at other points. The company had guided second-quarter revenue to $1.915 billion to $1.945 billion in May, so the actual result was within the range but below its midpoint. More importantly, prior adjusted EBITDA guidance was approximately $1.62 billion to $1.65 billion. The reported $1.614 billion was below the low end.

The next-quarter outlook did not restore confidence immediately. AppLovin guided third-quarter revenue to $2.055 billion to $2.085 billion and adjusted EBITDA to $1.710 billion to $1.740 billion, with an 83% margin. The revenue midpoint was close to commonly cited consensus estimates, while even the top of the EBITDA range was below the roughly $1.75 billion consensus reported by financial media.

AppLovin therefore was not an example of a clean beat rejected irrationally. EPS met or modestly exceeded some forecasts, but the metrics most closely tied to the advertising platform’s momentum—revenue and adjusted EBITDA—fell short of the bar. For a company valued on sustained rapid growth and extraordinary margins, a small miss can imply a large change in the expected trajectory.

Why AppLovin’s Revenue Quality Commands So Much Attention

AppLovin operates a marketing platform that uses data and machine-learning systems to match advertising demand with available inventory and optimize campaigns. Its value proposition depends on producing measurable returns for advertisers and better monetization for publishers. If the platform can identify valuable users more effectively, advertisers can spend more while maintaining their target economics.

This model can scale rapidly because the incremental cost of processing more auctions and serving more campaigns is low relative to revenue. That operating leverage helps explain an adjusted EBITDA margin above 80%. It also creates demanding expectations. Investors assume that model improvements will translate into additional advertiser spend, and that a larger pool of data will improve matching further.

The risk is that growth can be sensitive to the timing of product releases and model changes. AppLovin indicated that improvements arrived later than expected, after the quarter had largely ended. Chief Executive Adam Foroughi characterized the result as below the company’s standard and attributed the shortfall to timing rather than to a deterioration in advertiser demand. That explanation is plausible, but investors must see the improvement in subsequent revenue rather than treat it as completed progress.

A delayed model upgrade is different from lost demand. If advertisers remained willing to spend and the improved system began performing shortly after quarter end, part of the shortfall can move into the third quarter. If the delay reflects difficulty extracting incremental performance from a maturing system, the issue is more structural. The earnings call needed to help investors distinguish between those possibilities.

Adjusted EBITDA Was the Most Important Miss

Adjusted EBITDA is not the same as net income or cash flow. It removes interest, taxes, depreciation, amortization and specified adjustments. AppLovin emphasizes it because the measure shows the profitability of the operating platform before financing and several non-cash or non-core items. The company’s 84% adjusted EBITDA margin demonstrates exceptional operating leverage.

That makes the miss important. When a company has such a high margin, a modest revenue shortfall can pass through heavily to adjusted EBITDA. Investors had expected model improvements and revenue growth to support another quarter near the top of the company’s range. Instead, adjusted EBITDA came in below the previous guide.

The third-quarter margin outlook of 83% also represented a slight step down from 84% in the second quarter and 85% in the first. One percentage point may appear trivial, but on more than $2 billion of quarterly revenue it represents roughly $20 million of EBITDA. More importantly, the direction can affect valuation assumptions. If a stock is priced for growth and stable or expanding margins, a guide to lower margin challenges both sides of the equation.

None of this means an 83% adjusted EBITDA margin is weak. The analytical issue is change relative to expectation. AppLovin does not compete against an average company’s margin; it competes against the earnings path already embedded in its own share price.

Cash Flow Was Strong, but Lower Than Net Income

AppLovin generated $869 million of operating cash flow and $863.3 million of free cash flow in the quarter. Those are substantial amounts. They were below the $1.267 billion of net income, reminding investors that accounting profit and cash generation do not move identically in every period.

Cash taxes were an important factor. For the first six months of 2026, the company reported approximately $639.8 million of cash paid for income taxes, compared with about $100.6 million in the comparable prior-year period. Working-capital movements can also affect quarterly cash conversion. AppLovin ended the quarter with approximately $2.171 billion of accounts receivable, reflecting the scale and timing of collections across the platform.

Investors should avoid reading the difference as evidence that earnings are fictitious. A profitable company can pay taxes and build receivables as revenue grows. The useful question is whether free cash flow tracks adjusted profit over a longer period. If cash conversion repeatedly falls behind net income without a clear timing explanation, concern would rise. One quarter should be evaluated in the context of the full year and the company’s payment cycles.

The company used $551.3 million to repurchase and withhold 1.1 million shares during the quarter. It ended with 335 million Class A and Class B shares outstanding. Repurchases can offset equity compensation and reduce dilution, but their economic value depends on the purchase price. A volatile stock creates a wide range of possible outcomes from the same dollar authorization.

AppLovin’s Balance Sheet Supports Flexibility but Is Not Debt-Free

AppLovin ended the quarter with approximately $3.053 billion of cash and cash equivalents and $3.515 billion of long-term debt. The near balance between cash and debt gives the company flexibility, while the operating platform’s cash generation supports servicing obligations. Still, gross debt matters because cash may be needed for taxes, working capital, acquisitions and repurchases.

Investors often focus on net cash or net debt, but the maturity schedule, interest rates and restrictions are equally important. A company can have a modest net-debt position and still face refinancing costs if debt matures during an unfavorable credit market. AppLovin’s profitability makes that risk manageable on current figures, but it should not be ignored in a valuation based primarily on enterprise value and adjusted EBITDA.

Accounts receivable also deserves attention. Advertising platforms sit between advertisers, agencies, publishers and payment systems. Timing differences can make reported revenue, receivables and cash flow move unevenly. Strong revenue growth can absorb cash temporarily even when customers remain creditworthy. The quality of collections and exposure to concentrated counterparties are therefore more informative than the receivable balance alone.

The Expansion Beyond Mobile Gaming Is the Strategic Test

AppLovin built much of its expertise in mobile applications and gaming. Its larger ambition is to apply the same automated advertising technology to a broader range of commerce and performance-marketing categories. The opportunity is large because merchants want measurable customer acquisition, while digital ad budgets can move toward channels that demonstrate return on spending.

Expansion creates a compelling growth story, but it changes the competitive environment. Mobile-gaming advertisers and app publishers operate in ecosystems where AppLovin has deep data and workflow integration. E-commerce merchants may have different attribution windows, margins, repeat-purchase behavior and data sources. They also have established alternatives across search, social media, retail media and independent ad-technology providers.

A successful move into e-commerce would diversify revenue and increase the platform’s addressable market. It could also strengthen the data flywheel if more transactions improve targeting. The risk is that early growth reflects a limited set of advertisers or incentives that do not scale at the same economics. Investors should look for retention, repeat spending, contribution margins and advertiser diversification rather than relying only on gross spend or launch announcements.

The company’s self-service and AI platform initiatives are meant to lower barriers for advertisers. Automation can expand the customer base, but self-service systems also require fraud controls, measurement credibility and dependable customer support. A platform that optimizes aggressively without transparent attribution can face skepticism from marketers even when near-term reported growth is strong.

Why a Small Miss Produced a Large Stock Move

AppLovin’s after-hours decline exceeded the roughly 11% move reportedly implied by the options market before the release. That gap indicates that the result fell outside the range many traders had priced. It does not prove that the stock’s long-term value declined by the same percentage.

Several forces can amplify an after-hours move. Liquidity is thinner than during regular trading. Momentum investors may use automatic risk limits. Options dealers may hedge rapidly as the stock moves. Analysts and portfolio managers may update revenue and margin models at the same time. A company with a large market capitalization can still move sharply when the shareholder base shares similar expectations.

The stock had also fallen substantially during 2026 before the report, despite continued business growth. That background cuts in two directions. Bulls can argue that the valuation already incorporated skepticism and that a timing-related miss may be temporary. Bears can argue that the decline reflects recognition that extraordinary growth and margins are harder to sustain as the base expands.

The appropriate interpretation depends on what happens next. If third-quarter revenue accelerates, model improvements perform as management expects and EBITDA lands above the guided range, the second quarter may look like a timing anomaly. If guidance is merely met and margin drifts lower, investors may conclude that the platform is entering a more normal growth phase.

The Bull Case for AppLovin

The supportive case starts with the absolute numbers. Revenue grew 53%, net income exceeded $1.26 billion, and adjusted EBITDA margin was 84%. Few businesses combine that rate of growth with that level of profitability. A quarter at the low end of guidance does not erase the platform’s economic strength.

Management’s timing explanation is testable. A model improvement delivered shortly after quarter end can affect third-quarter performance quickly. If advertiser demand remained strong, AppLovin may recover delayed spending rather than rebuild demand from scratch. Early-quarter trends described by management would then have greater significance.

The platform also has room to expand beyond its original mobile-gaming base. Performance advertisers care about measurable returns, and a system that consistently delivers them can win budget across categories. Scale can reinforce the model by producing more data, more auctions and more opportunities to improve matching.

Finally, cash generation provides strategic freedom. AppLovin can invest in computing and engineering, repurchase shares and manage debt without relying on external financing. That resilience matters in a volatile advertising market.

The Skeptical Case for AppLovin

The skeptical case focuses on expectations and concentration of value in the model. When adjusted EBITDA margin exceeds 80%, investors may assume that incremental revenue remains extremely profitable. Any sign that new verticals require more sales support, incentives or infrastructure can lower long-term margin estimates materially.

Advertising technology is also competitive and exposed to platform rules, privacy changes and measurement disputes. AppLovin’s performance depends on access to data, the quality of its optimization and confidence among advertisers and publishers. Changes by mobile operating systems, app stores or regulators can alter available signals and economics.

Management’s explanation that timing caused the shortfall is reasonable but self-interested. Investors need evidence in reported results. A software update that arrives after quarter end may improve performance, yet it can also reveal that expected gains were harder to achieve than forecast. Repeated references to timing would become less persuasive.

Valuation magnifies these risks. A business priced for exceptional execution can decline even while growing quickly. The stock does not need a recession or platform failure to disappoint; slower growth, a lower margin or a reduced valuation multiple can be enough.

What AppLovin Must Clarify

The most important disclosure is a bridge from prior guidance to actual revenue and adjusted EBITDA. Investors need to know how much of the difference was caused by model timing, customer behavior, pricing, mix, expenses and taxes. A quantified explanation is more useful than a broad statement that the quarter fell short of internal standards.

Management should also distinguish demand from delivery. Did advertisers attempt to spend more but encounter performance limitations? Did campaigns launch later? Did certain verticals slow? How did retention and same-customer spending change? These operational indicators can show whether the issue was temporary.

For e-commerce, investors need cohort evidence: how advertisers behave after initial testing, whether spending expands, and whether returns remain attractive at larger budgets. The addressable market is not the same as revenue opportunity unless the platform can win and retain customers profitably.

AppLovin’s decline was the least surprising of the four after-hours reactions. The company remained highly profitable, but it failed to clear the specific growth and EBITDA hurdles that mattered most. The next report will determine whether August 5 marked a temporary timing gap or the beginning of a slower trajectory.

e.l.f. Beauty: A Beat-and-Raise Quarter With Two Large Qualifications

E.l.f. Beauty produced the cleanest headline beat of the group. Fiscal first-quarter net sales rose 36% year over year to $479.4 million, compared with an LSEG consensus of approximately $429.5 million cited by Reuters. Adjusted diluted EPS was $1.75, more than double the 71-cent consensus. The company raised its fiscal 2027 sales outlook to $1.938 billion to $1.968 billion from $1.835 billion to $1.865 billion, and lifted adjusted EPS guidance to $3.50 to $3.55 from $3.27 to $3.32.

Shares nevertheless traded about 3% lower after the close. The move was mild relative to AppLovin or the storage companies, but it reflected legitimate questions about the composition of growth. Rhode, the skincare and cosmetics brand founded by Hailey Bieber, contributed approximately $160 million of quarterly revenue. The namesake e.l.f. brand declined slightly, according to post-earnings reporting. Meanwhile, tariff refunds added roughly 1,050 basis points—10.5 percentage points—to the company’s 1,400-basis-point gross-margin expansion.

The result was therefore strong, but not all of the strength represented repeatable organic improvement. Acquisition revenue and a tariff benefit are economically real. They need to be separated from the performance of the legacy brand and from the margin the company can sustain once refunds are no longer available.

Rhode Has Already Changed the Shape of the Company

E.l.f. Beauty announced its agreement to acquire rhode in May 2025 for up to $1 billion. The terms included $800 million of consideration at closing—$600 million in cash and $200 million in newly issued e.l.f. shares—plus a potential $200 million earnout tied to the brand’s growth over three years. Rhode had generated $212 million of net sales during the twelve months ended March 31, 2025.

Against that base, approximately $160 million of revenue in a single quarter demonstrates rapid expansion. The contribution exceeded management’s expectations and supports the strategic logic of adding a premium-adjacent, digitally native skincare brand to e.l.f.’s portfolio. Rhode’s distribution through Sephora, planned expansion into nineteen European countries and strong founder visibility can extend growth beyond its original direct-to-consumer channel.

The acquisition also changed the balance sheet. E.l.f. ended the quarter with $344.2 million of cash and $834.2 million of debt, compared with $170 million of cash and $256.7 million of debt a year earlier. Debt increased partly because the transaction required substantial cash financing. The company can service the obligation if rhode continues growing and the core business generates cash, but integration performance now matters more to financial flexibility.

The quarter included a $16.1 million charge from remeasuring contingent consideration because rhode was outperforming earnout thresholds. That charge reduced GAAP income while signaling that the acquired business was ahead of the assumptions used to value the earnout. It is a useful reminder that successful acquisitions can create accounting expenses even as they strengthen operating results.

Why Acquisition Growth Is Not the Same as Organic Growth

Consolidated revenue includes acquired businesses after a transaction closes. That is the correct accounting treatment, but it can obscure how the pre-existing company is performing. If a buyer adds a fast-growing brand while its core brand slows, total sales can rise even before any integration benefit appears.

Organic growth attempts to compare the same business across periods by excluding acquisitions, divestitures and sometimes currency movements. E.l.f.’s reported 36% growth tells investors how large the consolidated company became. It does not, by itself, answer whether the namesake brand gained or lost momentum.

The distinction matters because the legacy e.l.f. brand remains the foundation of distribution, retailer relationships and operating scale. A small quarterly decline can reflect shipment timing, systems transitions or difficult comparisons rather than consumer rejection. Management said it expected the core brand to grow for the full year. Investors will need sell-through, unit and market-share evidence to determine whether that confidence is justified.

Rhode’s strength can offset temporary weakness, but it should not become a permanent substitute for core-brand health. Acquisitions are finite. A company cannot rely indefinitely on buying growth without increasing debt, dilution or integration complexity. The best outcome is that rhode expands the portfolio while the namesake brand resumes unit growth and both benefit from shared distribution capabilities.

The Tariff Refund Created a Large but Temporary Margin Benefit

E.l.f. Beauty reported a GAAP gross margin of 83%, an increase of 1,400 basis points from the prior year. Approximately 1,050 basis points of that improvement came from refunds connected with tariffs imposed under the International Emergency Economic Powers Act. In plain English, three-quarters of the margin expansion was linked to a refund rather than to a permanent improvement in product economics.

The benefit was substantial. Management indicated that tariff refunds contributed roughly 68 cents to quarterly earnings per share. The company also referred to an expected fiscal-year benefit of about $50 million. These amounts improve cash and profit, but an investor forecasting future years should not assume the refund recurs indefinitely.

The remaining margin improvement came from pricing and lower tariff costs, according to the company. Those drivers can be more durable, although pricing interacts with demand. E.l.f. has built its identity around accessible price points; Reuters reported that roughly 75% of products sell for $10 or less. Raising prices can protect dollars of gross profit but may weaken the value proposition if consumers trade down or reduce units.

Management took an unusual step by lowering prices on about 10% of the portfolio after testing consumer response. Chief Executive Tarang Amin said the action increased unit momentum and could grow gross-profit dollars over time. That strategy accepts a lower margin per unit in exchange for greater volume. It is economically sound if the extra units, customer acquisition and retailer productivity more than offset the price reduction.

The “Lipstick Effect” Is Useful but Incomplete

Beauty products are often described as resilient because consumers may continue buying small indulgences during economic stress even when they postpone expensive purchases. The pattern is sometimes called the lipstick effect. E.l.f.’s low price points strengthen that argument: a consumer can experiment with a new product without making a large commitment.

But affordability does not make demand immune to inflation. Consumers can delay replacement, buy fewer items or choose private-label alternatives. Retailers can also manage inventory cautiously if traffic slows. The relevant evidence is not the existence of a catchy economic theory but unit sales, repeat purchase, retailer sell-through and market share.

E.l.f. has an advantage when premium beauty prices rise faster than household income. Its products can attract consumers trading down from more expensive brands. Rhode complicates the positioning in a productive way by adding a brand with a different image and price architecture. The combined portfolio can serve more occasions, but management must avoid confusing customers or weakening the clarity that made e.l.f. successful.

International Expansion Offers Growth and Execution Risk

Only about 20% of e.l.f. Beauty’s sales come from outside the United States, according to Chief Financial Officer Mandy Fields, compared with more than 70% for some larger competitors. That gap creates a credible expansion opportunity. The company plans to launch rhode across nineteen European countries, introduce e.l.f. in Brazil through Sephora and broaden availability through Boots in the United Kingdom.

International growth can diversify revenue and extend the life of successful products. It also adds currency exposure, regulatory requirements, logistics costs and local competition. Beauty preferences differ across markets, and a digital campaign that works in the United States may not translate directly. Retail partners can accelerate reach but also take margin and influence merchandising.

Rhode’s strong social recognition may help initial launches, especially among consumers who already follow Hailey Bieber. The long-term test is repeat purchase after the launch excitement fades. Skincare brands depend on perceived efficacy and habit. Distribution can create trial; product performance and availability determine retention.

E.l.f.’s established retailer relationships and operational systems can help rhode scale more efficiently than it could as an independent brand. The same integration can create disruption if systems changes affect shipments or inventory. Management attributed part of the core brand’s softness to prior shipping peaks and systems transitions, making execution a near-term area to watch.

Inventory and Working Capital Need Attention

Inventory rose to $246.8 million at June 30 from $220.2 million at the end of March and $170.4 million a year earlier. Part of the increase is natural because the company is larger and includes rhode. Inventory growth still deserves monitoring because beauty products can become less valuable when trends change, packaging is refreshed or retailer orders slow.

The proper comparison is inventory relative to sales and expected launches, not the absolute balance alone. Building stock ahead of international distribution or retail expansion can prevent lost sales. Excess inventory, by contrast, can lead to discounts, retailer allowances or write-downs.

Cash generation also must support higher debt and acquisition obligations. E.l.f.’s adjusted earnings remove items such as stock compensation, amortization and the contingent-consideration charge. Those adjustments can clarify operating trends, but debt service is paid with cash. Investors should follow operating cash flow, capital needs and working-capital conversion alongside adjusted EBITDA.

The Bull Case for e.l.f. Beauty

The supportive case is that e.l.f. has acquired a rare brand that is scaling faster than expected while preserving its own value-led positioning. Rhode’s $160 million quarterly contribution compares favorably with its $212 million of sales in the twelve months before the acquisition announcement. If international distribution converts awareness into repeat demand, the purchase price can look increasingly reasonable.

The company also retains pricing flexibility. Lowering prices on selected products is not necessarily a sign of weakness; it can deepen the value advantage and expand units. E.l.f.’s ability to offer most products below $10 gives it a differentiated position when consumers face inflation.

International whitespace is substantial. A business with only one-fifth of sales abroad does not need to displace the largest global competitors to create meaningful growth. Successful launches through Sephora and Boots can leverage partners’ existing traffic and infrastructure.

Finally, management raised both sales and adjusted-profit guidance. The outlook incorporates more than the first-quarter beat and indicates confidence that portfolio growth will continue through the fiscal year.

The Skeptical Case for e.l.f. Beauty

The skeptical case is that consolidated growth has become too dependent on rhode while the core brand softens. A strong acquisition can mask weaker organic trends for several quarters. If the namesake brand does not resume growth, the company may be trading one concentrated brand risk for another.

The tariff refund makes the first-quarter margin difficult to annualize. Investors who apply the 83% gross margin to future years would overstate normalized profitability unless the company replaces the benefit with lasting pricing, sourcing or mix improvements.

Debt increased materially to finance the acquisition. That is manageable if rhode performs, but it reduces flexibility if demand slows or integration costs rise. The earnout can require additional consideration precisely when the brand succeeds, while underperformance would challenge the original strategic assumptions.

Beauty is also trend-sensitive. Founder visibility accelerates customer acquisition but creates reputational and concentration risk. The brand must become durable beyond any one celebrity association. Product efficacy, innovation and distribution execution will determine whether rhode develops into a multiyear platform or remains tied to a particular cultural moment.

What e.l.f. Beauty Must Prove Next

The next several quarters should separate three elements: the core e.l.f. brand, rhode and temporary tariff effects. Investors need organic growth, unit trends and market-share data for the namesake business. They need geographic and channel performance for rhode. They also need a normalized gross-margin bridge that removes refunds.

Management should explain how much inventory supports planned launches, how quickly international distribution will affect revenue and what incremental marketing is required. A brand can grow rapidly while consuming more cash if retailer expansion, inventory and advertising rise ahead of collections.

The August 5 decline was not a verdict that the acquisition had failed. The opposite is closer to the evidence: rhode performed so well that it contributed a contingent-consideration charge. The market’s concern was whether the strength could be repeated and whether the legacy brand was healthy underneath it.

Four Different Earnings Lessons Hidden Behind One Red Screen

It is tempting to group all four declines under the slogan “good news was priced in.” That phrase is partly right and mostly incomplete. It describes Western Digital and Sandisk better than AppLovin, and it describes e.l.f. Beauty only after the earnings quality is examined. Each company illustrates a separate analytical mistake that investors can make when reading results.

Lesson One: Start With the Company’s Own Guidance

Analyst consensus is useful, but management’s prior range is often the more direct accountability measure. AppLovin’s revenue remained inside its May guidance, yet adjusted EBITDA fell below the range. That is more informative than an EPS beat because adjusted EBITDA is one of the operating measures management emphasizes most heavily.

Western Digital and Sandisk, by contrast, delivered results that were strong against both prior operating trends and commonly cited consensus estimates. Their negative reactions require valuation and expectations analysis rather than a claim that the companies simply missed.

Lesson Two: Separate Operations From Accounting Revaluation

Western Digital’s $3.195 billion GAAP net income included a large gain connected to its retained Sandisk investment. The company’s operating income and free cash flow still showed excellent execution, but the statutory profit should not be treated as if it came entirely from selling drives. The correct response is not to discard GAAP or adjusted earnings. It is to reconcile them.

E.l.f. Beauty presents a different reconciliation. Its adjusted EPS excluded items such as amortization and a contingent-consideration remeasurement, while its GAAP gross margin benefited from tariff refunds. Readers need both measures and an explanation of what can recur.

Lesson Three: Ask Whether Growth Came From Price, Volume or Acquisition

Sandisk’s sequential growth came primarily from higher NAND prices. E.l.f. Beauty’s year-over-year increase included roughly $160 million from acquired brand rhode. Western Digital benefited from pricing, product mix and cloud demand. AppLovin’s growth came from the scaling of its advertising platform, but the rate was below what investors expected from model improvements.

These sources have different economic value. Volume growth can indicate expanding demand but may require capital and working capital. Price growth can produce exceptional margins but invite substitution and supply. Acquisition growth can accelerate scale but costs cash, debt or shares. Organic platform growth can be highly valuable but sensitive to technology performance and competition.

Lesson Four: Cash Flow Needs Its Own Analysis

Western Digital’s $1.281 billion of free cash flow confirmed that operating gains were translating into cash. Sandisk reported $7.083 billion of free cash flow, but $1.938 billion came from customer prepayments and deposits that it excluded from adjusted free cash flow. AppLovin’s $863.3 million of free cash flow was strong but below net income, partly amid much higher cash tax payments. E.l.f. Beauty must fund inventory, international launches, debt and integration.

No single cash-flow figure settles the question. Prepayments are valuable because they reduce financing needs, yet they may precede future product delivery. Working-capital investment can support growth, yet it can also signal slower collections or excessive inventory. The goal is to understand timing and durability.

Lesson Five: A Prior Rally Raises the Burden of Proof

Western Digital and Sandisk had risen by triple-digit percentages before their reports. That does not make them automatically overvalued, but it changes what a new buyer is paying for. The stock price already reflects some combination of shortages, AI demand, pricing power and long contracts. Earnings must provide incremental evidence beyond that package.

AppLovin demonstrates the opposite complication. Its shares had declined sharply during 2026, yet the stock still reacted violently because the remaining valuation depended on exceptional growth and margins. A falling stock is not necessarily cheap relative to revised expectations.

E.l.f. Beauty’s smaller reaction reflects a more balanced setup. The company raised guidance and the acquisition performed well, limiting the negative interpretation. The shares slipped because investors questioned composition, not because the quarter invalidated the business model.

How the AI Boom Connects Western Digital, Sandisk and AppLovin

Three of the companies are exposed to artificial intelligence, but in different parts of the economic chain. Western Digital supplies mass-capacity storage. Sandisk supplies NAND flash and enterprise SSDs. AppLovin applies machine learning to advertising optimization. The phrase “AI beneficiary” therefore conceals distinct revenue mechanisms.

Storage demand depends on data-center construction, customer architecture, capacity per system and the amount of data retained. It is capital-intensive and supply-constrained. AppLovin’s demand depends on advertisers’ return on spending, publisher inventory and model performance. It is asset-light relative to semiconductor manufacturing but exposed to digital-platform rules and measurement trust.

The difference matters when comparing valuations. A storage manufacturer can earn unusually high margins during scarcity, but it must invest continually in process technology and capacity. An advertising platform can scale at high incremental margin, but competitive models can improve quickly and customers can shift budgets. Neither business deserves a premium merely because AI appears in the description. The premium depends on barriers to entry, cash generation and the duration of advantage.

AI infrastructure also creates second-order risks. Rapid capital spending can pull demand forward. Customers may order aggressively to secure scarce components, then pause to digest inventory. Technology can improve efficiency, reducing the number of components required per unit of output even as total workloads grow. For AppLovin, better AI can increase advertiser returns, but rivals gain access to similar tools and computing resources.

The August 5 reports showed real AI-linked revenue rather than speculative announcements. Western Digital’s cloud mix, Sandisk’s data-center growth and AppLovin’s profitability demonstrate that each company has monetized the trend. The stock reactions showed that monetization alone is no longer sufficient. Investors want to know how much remains unpriced and how long the advantage lasts.

Valuation Without a Price Target

This analysis does not require a recommendation or a precise fair-value estimate. A useful valuation framework can be built from scenarios instead.

For Western Digital, the optimistic scenario assumes sustained cloud demand, disciplined hard-drive supply and gross margins remaining well above historical norms. The cautious scenario assumes pricing normalizes as output and customer inventory adjust, while technology spending remains high. The key variables are exabyte growth, dollars per terabyte, gross margin and free cash flow.

For Sandisk, the optimistic scenario assumes long-term contracts change the NAND cycle, AI drives persistent enterprise demand and floor pricing protects margins. The cautious scenario assumes current prices represent a cyclical peak, consumer elasticity spreads to other markets and contracts provide less protection than headline values imply. The key variables are NAND price, bit growth, contract recognition, joint-venture investment and normalized gross margin.

For AppLovin, the optimistic scenario assumes delayed model improvements convert into third-quarter acceleration and e-commerce expansion increases the addressable market without damaging margins. The cautious scenario assumes incremental model gains become harder, new verticals cost more to serve and the adjusted EBITDA margin trends lower. The key variables are advertiser retention, same-customer spend, revenue growth, adjusted EBITDA margin and free-cash-flow conversion.

For e.l.f. Beauty, the optimistic scenario assumes rhode scales internationally, the core brand resumes growth and value pricing wins share. The cautious scenario assumes rhode’s launch momentum fades, the namesake brand remains soft and debt limits flexibility after temporary tariff benefits disappear. The key variables are organic sales, unit volume, gross margin excluding refunds, inventory, debt and rhode’s repeat purchase.

A stock price can be tested against these scenarios without pretending to know a single correct future. The most dangerous valuation error is to use peak margins as a permanent base while applying a premium multiple. The second is to assume every temporary disappointment marks permanent decline. Scenario analysis forces both possibilities into view.

Risks That Cut Across the Four Companies

• Expectation risk: Strong operating results can produce losses for shareholders when the valuation assumes even stronger results.
• Concentration: Western Digital depends heavily on cloud customers; Sandisk’s long-term agreements involve a limited customer group; AppLovin depends on digital ecosystems; e.l.f. Beauty increasingly depends on rhode for growth.
• Normalization: Storage margins and pricing can fall when shortages ease. E.l.f.’s tariff benefit will not repeat indefinitely. AppLovin’s margin can decline as new verticals require investment.
• Execution: Product ramps, AI-model upgrades, international launches and systems integration can miss schedules.
• Capital allocation: Large buybacks, acquisition debt and technology investment create competing claims on cash.
• Measurement: GAAP, non-GAAP and cash-flow figures can tell different stories. Investors need reconciliation rather than allegiance to one metric.

What Happens Next

Western Digital’s fiscal first-quarter performance will test whether its guided margin expansion is supported by pricing, mix and cost improvements. Commentary on cloud commitments and product qualifications should help determine whether current visibility extends beyond a shortage.

Sandisk’s August 13 investor day is the closest scheduled catalyst among the four companies. The event should provide more detail on the $93.9 billion of minimum contracted NBM revenue, customer prepayments, technology transitions and capital allocation. Investors should focus on contract mechanics and normalized economics rather than the largest headline number.

AppLovin’s third quarter will test management’s timing explanation. The company does not need merely to stay inside its new guidance range; it needs evidence that model improvements restored the trajectory expected before the second-quarter shortfall. Revenue growth, margin and cash conversion will all matter.

E.l.f. Beauty’s next reports should distinguish rhode from the core brand and remove the tariff refund from the margin bridge. International launches will add growth opportunities and execution costs. Inventory and debt will show whether expansion is translating into healthy cash flow.

Frequently Asked Questions

Why did Western Digital stock fall after beating earnings estimates?

Western Digital reported strong revenue, adjusted earnings and cash flow, but the shares had already risen sharply and investors expected exceptional AI-storage demand to continue. Its guidance was strong without clearly exceeding the highest expectations, and GAAP net income included a large gain tied to the retained Sandisk investment.

Did Western Digital actually earn $3.2 billion from operations?

No. Western Digital reported about $3.2 billion of GAAP net income attributable to the company, but the quarter included a major non-operating gain connected to its retained Sandisk interest. GAAP operating income was $1.563 billion, and free cash flow was $1.281 billion.

Why did Sandisk stock fall despite an 84.6% gross margin?

The margin and revenue growth were exceptional, but the stock had risen nearly 470% during 2026. Investors also focused on guidance that was only around the highest consensus estimates, consumer weakness and the fact that two-thirds of sequential growth came from higher pricing.

Are Western Digital and Sandisk still the same company?

No. Western Digital separated the flash business as Sandisk in February 2025. Western Digital now focuses primarily on hard disk drives, while Sandisk focuses on NAND flash and solid-state storage.

What are Sandisk’s long-term customer agreements?

Sandisk calls them part of its new business model. The company disclosed eight agreements with six customers representing at least $93.9 billion of minimum contracted revenue at floor pricing. The contracts improve visibility, but their timing, pricing formulas and conditions remain important.

Did AppLovin beat earnings?

AppLovin’s diluted EPS of $3.76 met or exceeded some forecasts, but the report was not a clean operating beat. Revenue was near the low end of prior guidance, and adjusted EBITDA of $1.614 billion was below the company’s previous range.

Why was AppLovin’s stock reaction so large?

The company is valued for rapid growth and unusually high margins. A small revenue shortfall and softer adjusted-EBITDA outlook can lead investors to reduce multiple and earnings assumptions at the same time. Thin after-hours liquidity and crowded positioning can amplify the move.

Did e.l.f. Beauty raise its guidance?

Yes. E.l.f. raised fiscal 2027 net-sales guidance to $1.938 billion to $1.968 billion and adjusted EPS guidance to $3.50 to $3.55. The stock still slipped as investors assessed core-brand softness, rhode’s contribution and temporary tariff benefits.

How much did rhode contribute to e.l.f. Beauty?

Post-earnings reporting put rhode’s fiscal first-quarter contribution at approximately $160 million of revenue. That was a major part of e.l.f. Beauty’s consolidated growth and exceeded management’s expectations.

How much did tariff refunds help e.l.f. Beauty?

E.l.f. said tariff refunds contributed about 1,050 basis points of the quarter’s 1,400-basis-point gross-margin increase. The benefit added roughly 68 cents to quarterly EPS and should not be treated as a permanent annual margin improvement.

Does a falling stock mean the earnings were bad?

No. A stock move measures the difference between new information and what investors had already expected. Western Digital and Sandisk reported objectively strong quarters. AppLovin missed key operating expectations. E.l.f. Beauty beat estimates but had a more complicated revenue and margin mix.

What is the most important metric to watch next?

It differs by company: Western Digital’s normalized hard-drive margin and cloud commitments; Sandisk’s NAND pricing and contract economics; AppLovin’s revenue acceleration and adjusted EBITDA margin; and e.l.f. Beauty’s organic core-brand growth excluding rhode and tariff refunds.

Final Assessment

The August 5 earnings reactions demonstrate why a headline beat cannot substitute for financial analysis. Western Digital delivered a powerful hard-drive quarter, but its GAAP profit was inflated by an investment gain and its stock price already reflected a transformed industry. Sandisk produced one of the most profitable quarters imaginable for a NAND supplier, yet the result depended heavily on scarcity pricing and arrived after a historic rally. AppLovin remained extraordinarily profitable but missed the operating targets that mattered. E.l.f. Beauty raised guidance while revealing that acquired growth and tariff refunds made the quarter look stronger than the legacy business alone.

The strongest shared evidence is cash generation and real revenue growth. These were not companies promoting distant concepts without customers. Western Digital and Sandisk generated billions in cash from storage demand. AppLovin generated more than $860 million of quarterly free cash flow. E.l.f. Beauty converted a large acquisition into immediate consolidated growth and raised its annual forecast.

The strongest shared concern is extrapolation. Scarcity margins, tariff refunds, acquisition growth and model-driven advertising gains all have limits. The relevant question is not whether the latest quarter was impressive. It is what level of earnings remains when prices normalize, temporary benefits expire, acquired growth enters the comparison base and technology improvements become harder.

Investors should therefore treat the after-hours declines as information, not verdicts. The moves reveal where expectations were highest and which details the market considered fragile. They do not establish that every seller was correct or that each company’s long-term outlook deteriorated by the same percentage as its stock.

The next evidence will be company-specific: contract detail from Sandisk, margin and cloud visibility from Western Digital, proof of post-quarter acceleration from AppLovin, and organic brand data from e.l.f. Beauty. Those disclosures will determine whether August 5 was a temporary reset after excessive optimism or an early warning that the earnings curves had reached their most demanding point.

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

Sources

• Western Digital: Fiscal Fourth Quarter and Fiscal Year 2026 Financial Results
• Seagate Technology: Fiscal Fourth Quarter and Fiscal Year 2026 Financial Results
• Sandisk: Fiscal Fourth Quarter 2026 Earnings Presentation
• Reuters: Sandisk Forecasts Upbeat Quarterly Revenue on AI-Driven Demand
• AppLovin: Second Quarter 2026 Financial Results
• AppLovin: Second Quarter 2026 Financial Update
• The Wall Street Journal: AppLovin Says Second-Quarter Results Fell Short of Its Standards
• e.l.f. Beauty: First Quarter Fiscal 2027 Results
• e.l.f. Beauty: Definitive Agreement to Acquire rhode
• Reuters: e.l.f. Beauty Raises Annual Forecasts
• Reuters: U.S. Market Close on August 5, 2026
• Barron’s: Western Digital Stock Falls After an Earnings Beat

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.