SK Hynix just committed $720 billion to build what it calls the largest network of memory factories on Earth, and the bet is simple: the AI boom’s real bottleneck isn’t chips, it’s the memory that feeds them. The South Korean company controls 58% of the high-bandwidth memory market that powers every major AI processor, more than double the share held by rivals Samsung or Micron. That’s the kind of structural dominance long-term investors should notice, even as the stock has skidded 21% from its July highs.

The numbers behind the buildout are staggering. SK Hynix raised $26.5 billion in July through a Nasdaq listing, the largest capital raise ever by a foreign company on U.S. markets, and it’s plowing the proceeds into fabs so tall they’ll rival 50-story buildings. Nvidia has already signed a $500 billion deal for guaranteed supply and co-development of next-generation memory, and SK Group’s chairman says “if you name any company in Big Tech, they are all in Korea to sign a contract.” Demand from Microsoft, Google, Meta, and Amazon isn’t speculative chatter, it’s contracted, multi-year revenue.

What makes this more than a cyclical chip story is the shift toward custom HBM, memory co-designed directly with AI processors rather than sold as an interchangeable commodity. That’s a moat-widening development: pricing power increases when your product is engineered into a customer’s chip roadmap years in advance, not swapped out when the next price war hits. It’s also why South Korea’s president is pushing to double the country’s memory output within five years, and why Micron is racing to match SK Hynix with $150 billion in new U.S. fabs of its own.

The recent pullback reflects broader AI-trade jitters, not a crack in SK Hynix’s underlying position. A company that just tripled its market cap to over $1 trillion, locked in a decade of demand from the biggest AI spenders in the world, and is transitioning from commodity supplier to indispensable co-engineer isn’t a business getting weaker because the stock dipped 21%. So what for long-term investors: when a dominant, moat-building supplier to an entire industry goes on sale during a sector-wide wobble, that’s usually the setup, not the warning sign.

Aviva shares have climbed roughly 150% since Amanda Blanc took the helm in July 2020 — a rare feat for a UK insurer that burned through four CEOs in the previous 13 years. The turnaround is done. What matters now for long-term investors is the next chapter: a pivot from balance-sheet repair to a genuine growth engine in UK wealth management, a market Aviva pegs at £2.7 trillion today and projects will top £4 trillion by 2030.

Blanc’s playbook was disciplined capital allocation, not financial engineering. She raised roughly £8 billion selling eight non-core businesses — exiting France, Italy, and Poland — and returned about £10 billion to shareholders via dividends and buybacks while narrowing Aviva to three core markets: UK, Canada, and Ireland. She then redeployed capital into scale, buying rival Direct Line for £3.7 billion in December 2024 (now a fifth of UK motor insurance) and Succession Wealth for £385 million in 2022, the seed of a wealth unit expected to reach 10% of group earnings. Half-year results due this Friday are expected to show operating profit up 17.5% year-over-year to roughly £1.3 billion.

The risks are real: some Aviva Investors funds show patchy performance, autonomous vehicles pose a long-run structural threat to motor insurance premiums, and the broadened business mix invites comparison to lower-multiple European composites like Allianz and Axa rather than a premium UK life insurer. But the underlying moat is hard to replicate — nearly 22 million UK customers, the second-largest customer base of any bank or insurer in the country, and 500,000 retail shareholders who’ve stuck around since the Norwich Union days. So what for long-term investors: a management team with a proven record of turning divestiture proceeds into disciplined, accretive acquisitions is now aiming that same capital discipline at a structurally growing wealth market — the kind of compounding setup patient dividend investors should be watching closely into Friday’s numbers.

Baker Hughes just posted its best quarter in years, and the headline numbers tell a story that oil bulls and skeptics will both want to claim. Shares jumped more than 6% after second-quarter earnings per share came in at 64 cents, crushing the 50-cent consensus estimate, according to LSEG data. But the real signal for patient investors isn’t the beat — it’s where the growth is actually coming from.

Orders surged 49% year-over-year to a record $10.5 billion, and $7.1 billion of that came from the company’s Industrial and Energy Technology segment — the division serving LNG export terminals, power grids, and data centers, not oil rigs. Backlog climbed 19% to an all-time high, with $40.1 billion in contracted work now on the books. CEO Lorenzo Simonelli has been reframing Baker Hughes around what he calls a “demand decade for energy,” pushing the 100-plus-year-old oilfield services company deeper into the infrastructure that powers AI data centers and electrified grids rather than just wells and pipelines.

That pivot matters because the old business is softening. Baker Hughes itself said global oil and gas producer spending will decline modestly this year, as weakness in Europe and the Middle East — exacerbated by the ongoing U.S.-Iran standoff — offsets gains in Latin America and North America. Third-quarter revenue guidance for the IET segment, $3.17 billion to $3.47 billion, actually falls short of the $3.79 billion Wall Street wanted, and management flagged 1-2 points of Middle East-related drag plus rising logistics costs. This isn’t a company pretending everything is fine; it’s one candidly showing two divergent stories in the same earnings report.

Hedge funds appear to be underwriting the pivot ahead of the retail crowd. Insider Monkey’s database shows Baker Hughes had 72 hedge fund holders as of Q1 2026, up sharply from 59 the prior quarter, with dollar exposure roughly doubling from $797 million to $1.62 billion — before this earnings beat even hit the tape. That puts Baker Hughes on par with oilfield peers Halliburton (72 holders, up from 53) and SLB (74, up from 73), suggesting smart money sees the whole sector’s infrastructure angle re-rating, not just one name.

So what for long-term investors: a $40.1 billion backlog and a 19% jump in contracted work is the kind of visibility that lets a company weather a cyclical energy downturn without cutting the dividend or capex that builds the next decade’s moat. Baker Hughes trades as an oilfield services stock, but an increasing share of its economics now looks more like an industrial infrastructure company riding the same LNG and power buildout tailwinds as utilities and grid-equipment makers. The near-term guidance miss is real, but for investors who care about where backlog and order growth are heading over three to five years rather than the next quarter, this looks less like an oil stock in decline and more like an energy-transition compounder that hasn’t been repriced yet.

Oracle shares are down 26% this year, yet the company’s forward-looking numbers tell a starkly different story than the stock price suggests. Its Remaining Performance Obligations — contracted revenue not yet recognized — hit a record $638 billion in the most recent quarter, while cloud revenue grew 47% year-over-year and Infrastructure-as-a-Service nearly doubled, up 93%. For patient investors, that gap between sentiment and substance is exactly the kind of setup worth studying.

The market’s skepticism centers on Oracle’s aggressive AI capital spending, with capex running near $35 billion this year to build out data center capacity. Wall Street has punished the stock for the spend, treating it as a drag rather than an investment. But Mizuho analyst Siti Panigrahi, who added Oracle to the firm’s August top-picks list, argues the buildout is the fuel for a structural inflection: he projects 34% operating income compound annual growth through fiscal 2030, more than ten times the 3% CAGR Oracle delivered between 2010 and 2020. That’s not incremental improvement — it’s a business model shift, powered by database, infrastructure, and application layers all monetizing the same AI wave simultaneously.

What makes this interesting for value-oriented investors isn’t just the growth math — it’s the valuation mismatch it creates. A company compounding operating income at more than 10x its historical rate, with backlog visibility stretching years into the future, is trading at a discount most investors reserve for businesses in decline, not businesses accelerating. Oracle’s transformation from a legacy database vendor into critical AI infrastructure plumbing — sitting underneath enterprise workloads that are expensive and disruptive to migrate away from — is exactly the kind of durable, moat-widening shift long-term holders should want to own before consensus catches up.

So what for long-term investors: backlog conversion, not quarterly headlines, is the metric to track here. If Oracle converts even a fraction of that $638 billion RPO into recognized revenue at improving margins over the next several years, today’s 26% drawdown may look, in hindsight, like the entry point patient capital was waiting for.

Berkshire Hathaway just gave long-term investors the clearest signal yet that its capital-allocation drought is over. In its first full quarter under new CEO Greg Abel, the conglomerate repurchased roughly $4.5 billion of its own stock — a sharp jump from just $235 million in the first quarter — and became a net buyer of equities for the first time in 14 consecutive quarters, adding nearly $20 billion in net stock purchases. For a company that had been quietly hoarding cash while Warren Buffett said he couldn’t find value, that reversal matters more than the headline earnings print.

The underlying businesses did their part too. Operating earnings — the metric Buffett and Abel have long called the truer read on Berkshire’s health — rose 16% to $12.98 billion from $11.16 billion a year earlier. Manufacturing, service and retailing earnings jumped 24% to $4.47 billion, Berkshire Hathaway Energy’s profit surged 27% to $891 million, and BNSF’s railroad income climbed 6% to $1.56 billion. Insurance was the soft spot, with underwriting income down 13% to $1.73 billion and investment income off 9% to $3.06 billion — a reminder that even Berkshire’s steadiest moat has cyclical dents.

Berkshire’s fabled cash pile, which peaked near $397 billion last quarter, fell to about $365.5 billion as Abel funded buybacks, the $8.5 billion Taylor Morrison acquisition, and fresh equity stakes. Alphabet has now climbed into Berkshire’s top five holdings by market value, joining long-standing anchors American Express, Apple, Bank of America and Coca-Cola — a notable shift for a portfolio that avoided Big Tech for decades.

Berkshire shares are up just 3% this year, trailing the S&P 500’s 13% gain, even after a 9% rally over the past three months. That gap is the real story: a company sitting on diversified, cash-generating businesses and a newly aggressive buyback program is being priced like it’s standing still. So what for long-term investors — Berkshire’s discount to the broader market, paired with a management team finally willing to deploy capital rather than just stockpile it, is exactly the kind of quiet mismatch patient shareholders have historically been rewarded for holding through.

Otis Worldwide’s stock is down roughly 15% year-to-date, but the business generating over 90% of its profits — servicing the 2.5 million elevators it already has in the field — just got structurally more valuable. In April, Finland’s Kone agreed to buy Germany’s TK Elevator for nearly $35 billion, a deal that would shrink the industry from four major players to three. If it clears regulatory scrutiny, Otis inherits a more rational, less price-competitive market for the recurring service contracts that actually make it money.

That’s the part Wall Street’s AI-chasing capital seems to be missing. Otis’s new-equipment business — installing elevators — runs at a thin 4.8% operating margin; nobody gets rich building elevators. The real engine is the 20-year-plus service tail that follows each installation: maintenance, repairs, and eventual modernization, which together produced a 25.5% margin in 2025. The company services elevators in more than 200 countries and grew service sales 11% year-over-year in its most recent quarter, even as retention — the renewal rate on those contracts — hasn’t yet fully recovered from a slump that started in 2025. Otis is plowing an incremental $50 million into fixing that in 2026, betting that fewer outages mean happier, stickier customers. Renewals are somewhat automatic if no one is unhappy, as one analyst put it.

None of this is glamorous. There is no AI angle, no chip shortage narrative, no 40% quarterly revenue growth. What Otis offers instead is what patient capital has always prized: a toll-booth business with decadeslong tailwinds from urbanization, an aging population needing mobility solutions, and infrastructure modernization that is not going away regardless of what happens with data center capex. Management describes this as a decadelong opportunity, not a quarterly trade.

So what for long-term investors: a 15% pullback in a company where roughly 90 cents of every profit dollar comes from recurring, multi-decade service contracts — with a looming industry consolidation that could reduce competitive intensity — is exactly the kind of boring-is-beautiful mispricing patient capital should be watching. The retention numbers over the next two quarters will tell you whether the moat is widening or just holding steady.

Advanced Micro Devices delivered its strongest quarter on record in Q2 2026, reporting revenue of $11.54 billion — a 50% increase year over year — yet the stock fell nearly 9% in after-hours trading. For long-term investors who understand how durable competitive advantages compound, that post-earnings selloff may turn out to be the most important data point in the entire report.

The headline number tells only part of the story. AMD’s Data Center segment — which now includes both its EPYC server processors and Instinct AI accelerators — generated $6.72 billion in Q2, up 107% year over year and representing 58% of total company revenue, compared to just 42% a year earlier. Data center operating income reached $2.1 billion, implying a segment operating margin above 31%. CEO Lisa Su guided Q3 revenue to approximately $13 billion and, on the earnings call, stated that AMD expects data center revenue to more than double again in 2027 as Helios rack-scale AI infrastructure systems ramp at volume. Server CPU total addressable market is now projected to exceed $120 billion by 2030, according to AMD’s own modeling — a number that would have seemed absurd three years ago.

The Helios platform is worth watching carefully. Rather than competing with Nvidia only on individual GPU benchmarks, AMD is now shipping an integrated rack-scale solution that bundles EPYC CPUs, Instinct MI450 accelerators, and high-speed networking into a single deployable unit — precisely the kind of system-level architecture that hyperscalers like Microsoft, Google, and Oracle increasingly prefer. AMD’s growing AI partnership roster, combined with the 6th generation EPYC CPU launch, suggests this is not a company fighting for scraps at Nvidia’s table but one methodically building a parallel AI infrastructure stack. Unit sales growth outpaced average selling price growth last quarter, which means volume expansion — not just price inflation — is driving the revenue curve upward.

The stock’s decline after a beat comes down to expectations: AMD had rallied 21% in the five trading sessions leading into earnings, embedding a perfection premium that even a 107% data center growth rate couldn’t satisfy. That dynamic is not a business problem; it is a sentiment problem. For patient long-term investors, the distinction matters enormously. The fundamentals — accelerating data center share, a ramp in rack-scale systems, a server CPU TAM expanding toward nine figures, and operating margins expanding structurally — remain intact and are, if anything, reinforced by Q2’s results. When the market sells off a compounding business because the stock had run too far into earnings, the business does not get cheaper. But the entry point does.

When investors look at Amazon’s Q2 2026 earnings, most fixate on the headline risk: $220 billion in capital expenditures committed for the year, negative free cash flow of $7.6 billion in the quarter, and long-term debt that has climbed to $119 billion. It looks, on the surface, like a company spending recklessly into an uncertain AI future. But the number that reframes the entire story sits in a footnote most analysts walk past — a contracted AWS backlog of $496 billion, growing triple digits year over year. That is not speculative demand. That is revenue already sold.

AWS delivered $42.2 billion in Q2 2026 revenue — up 37% year over year, the fastest growth in 18 quarters — running at a $169 billion annualized pace. More importantly, AWS generated $16.6 billion in operating income at a 39.4% margin, accounting for roughly 61% of Amazon’s total operating income of $27.5 billion (itself up 43% year over year). This is the architecture of a compounding business: a cloud infrastructure division with near-monopoly switching costs, expanding margins at scale, and a $496 billion order book that dwarfs its current annual revenue run rate roughly three times over. The capex Amazon is deploying isn’t a bet — it’s the fulfillment leg of contracts already signed by enterprises and governments that have no practical alternative.

The broader Amazon flywheel reinforces the picture. Advertising, often overlooked, continues to grow at a double-digit clip, leveraging Amazon’s unmatched purchase-intent data. Total net sales crossed $200 billion in a single quarter for the first time. Operating cash flow for the trailing twelve months rose 33% to $161 billion. UBS projects Amazon’s net income could reach $509 billion by 2030, driven by AWS margin expansion and advertising leverage. The $220 billion capex program will weigh on near-term free cash flow — but patient investors who understand that pre-committed infrastructure spending against a $496 billion backlog is categorically different from speculative construction are seeing a company systematically converting AI demand into durable, high-margin annuity revenue. For long-term investors, the story isn’t whether Amazon is spending too much. It’s that the demand already exists at a scale that makes the spending look conservative.

When a $550 billion pharmaceutical company grows revenue 48% in a single quarter, it deserves more than a passing glance from long-term investors. Eli Lilly’s second-quarter 2026 results — posted before the opening bell on August 5 — were not a blip. They were structural proof that the company has built one of the deepest, most durable competitive moats in modern medicine.

The numbers are staggering. Total Q2 revenue came in at $23.0 billion, up 48% year-over-year, handily beating Wall Street’s $20.4 billion estimate. The engine behind that growth: Mounjaro (tirzepatide for diabetes) generated $9.9 billion in the quarter — up 91% from a year ago — while Zepbound (tirzepatide for obesity) added another $4.9 billion. Together, a single molecule is producing nearly $15 billion in quarterly revenue, with global penetration still in early innings. Non-GAAP EPS came in at $8.38, well above the $8.84 consensus. Management responded by raising full-year 2026 revenue guidance to $85–$87 billion, up from the prior $82–$85 billion range.

What long-term investors often miss about Lilly is how this dominance compounds. The GLP-1/GIP drug class is not a fad — it is being embedded into standard care pathways for Type 2 diabetes, obesity, cardiovascular disease, and sleep apnea. Retatrutide, Lilly’s next-generation triple-receptor agonist in late-stage trials, produced weight loss equivalent to bariatric surgery in Phase 2 data. Foundayo (orforglipron), recently FDA-approved as the first oral GLP-1, captured 8,000 prescribers in its first three weeks on market — 80% of whom were treating patients who had never used an incretin drug before. That last point matters enormously: it signals market expansion, not cannibalization. Lilly is not merely defending territory — it is continuously enlarging the battlefield it controls.

The moat here is multi-layered. Intellectual property on tirzepatide runs well into the 2030s. Manufacturing scale — Lilly has committed over $20 billion in U.S. facility expansions since 2023 — creates a structural barrier that would take a competitor nearly a decade to replicate. And physician habit-formation with branded drugs in chronic conditions is notoriously sticky. The switching costs in metabolic medicine are real.

So what does this mean for long-term investors? Lilly is not cheap — shares trade at a significant premium to the broader market. But premium multiples compress over time when the underlying earnings engine is this powerful. A company on track for $85+ billion in annual revenue in 2026, with pipeline visibility to retatrutide and potentially a dozen more assets, is one where the question is not “if” it compounds wealth — it’s “for how long.” Patient investors who focus on earnings power rather than near-term price noise may find that Lilly’s story is only in the middle chapters.

Most investors think of Mastercard as a credit card company. That framing undersells the business by a wide margin — sometimes literally. When Mastercard reported second-quarter 2026 results on July 30, the numbers underneath the headline told a more powerful story: a business running at 61.1% adjusted operating margins, growing profit 19% year-over-year to $4.4 billion, while collecting a quiet tax on virtually every digital dollar that moves across borders.

The quarter’s fundamentals were broad-based and durable. Revenue rose 14% to $9.3 billion, with adjusted earnings per share of $5.04 beating the consensus estimate of $4.77 by 6%. Gross Dollar Volume — the total value of transactions flowing through the Mastercard network — climbed 8% to $2.9 trillion in a single quarter. Switched transactions grew 9% globally, with international markets outside the U.S. expanding at 12%, nearly double the domestic rate. Cross-border volumes, the highest-margin revenue stream in the business, benefited from both a resurgent global travel environment and a World Cup 2026 spending surge that moved billions through the network. Value-added services — fraud prevention, analytics, identity verification, currency conversion — grew 20%, now functioning as a structural layer on top of the core payment rails and commanding premium pricing that pure network fees alone cannot.

What makes Mastercard’s long-term thesis particularly compelling is that the company earns a fraction of a cent on every swipe, tap, or tap-to-pay transaction — and that fraction scales effortlessly. There are no inventory costs, no manufacturing waste, no raw material exposure. As global commerce migrates further toward digital — from in-store NFC payments to e-commerce to cross-border B2B transfers — Mastercard’s two-sided network only becomes harder to displace. The company is now extending that moat into the next era of commerce. CFO Sachin Mehra confirmed that “Agent Pay,” Mastercard’s agentic AI commerce platform, is processing live transactions globally — meaning the same trusted network rails that settled $2.9 trillion in GDV last quarter are now being embedded into AI agent-driven purchases. Stablecoin settlement integration is also underway, ensuring Mastercard captures digital asset flows as that market matures rather than watching from the sidelines.

For long-term investors, the key insight is compounding at scale. Mastercard doesn’t need global GDP to boom — it only needs global commerce to continue digitizing, which has decades of runway remaining. Cash still accounts for roughly 80% of consumer transactions worldwide, particularly in emerging markets where Mastercard’s international volume growth is already running at twice the U.S. rate. Each percentage point of cash-to-digital conversion represents hundreds of billions in new GDV flowing through a network that exists, is trusted, and charges a fee. That combination — 61% operating margins, 19% profit growth, an unmatched global network, and a front-row seat to AI-powered commerce — is a compounding machine that patient investors have historically been rewarded handsomely for holding.