US Large-Cap Stocks — Valuations, P/E & Dividends

This earnings season, the US corporate landscape is defined by a stark divergence: the AI infrastructure buildout is supercharging tech and power producers, while the energy patch, especially natural gas, is bleeding revenue. The median revenue growth across large-cap SEC filers hit +24.5%, but that headline masks a chasm—from Talen Energy's +189.5% surge to CNX Resources' -35.7% collapse. The winners are selling the picks and shovels of the AI era; the losers are stuck with a commodity that the market no longer prizes.

Revenue growth by industry (median YoY)

AI Infrastructure Is the Only Game in Town, and It's Winning Big

The technology sector's median revenue growth of +20.6% is powered by a handful of extraordinary performers. Micron Technology is the star: revenue exploded +345.7% year over year, with EBITDA turning positive and net profit swinging to a massive gain—a memory-chip supercycle driven by AI demand. Amphenol, a connector maker, grew revenue +55.0% and EBITDA +79.7%, while Broadcom's +47.9% revenue growth and +155.3% EBITDA expansion show the networking and custom-silicon boom is far from over. Even NVIDIA, the poster child, delivered +85.2% revenue growth with net profit up +210.6%, though its 3-year CAGR of +100.0% suggests the law of large numbers may eventually bite.

Power Generators Are the New Tech—Talen Energy Leads a Historic Re-rating

The surprise winner of the season is independent power producers, with Talen Energy's revenue up +189.5% year over year—the highest in the entire dataset—and net profit +146.7%. This is not a one-off: Talen's prior quarter grew +57.4%, so the acceleration is real, powered by data-center electricity demand. The market is paying up: Talen trades at 40.2x EV/EBITDA, a multiple that would make a tech stock blush, but with EBITDA growth of 'n/m' (not meaningful) due to a swing to profitability, investors are pricing in a structural shift. Utilities as a whole lag (+1.7% median revenue), but the independent power niche is where the AI electricity story lives.

Energy Patch Is a Bloodbath—Natural Gas and E&P Names Are Cratering

The energy sector is the clear loser this season. Natural gas and NGL producers saw median revenue collapse -35.7%, with CNX Resources down -35.7% and EBITDA -43.9%. Even diversified E&P names are hurting: California Resources revenue fell -87.0% with EBITDA -282.3%, while Matador Resources dropped -33.8% revenue and -49.4% EBITDA. The pain is broad: Devon Energy (-14.5% revenue, -75.7% net profit), EQT (-29.2% revenue), and Comstock Resources (-24.9% revenue, -93.0% net profit) all show the commodity downturn is not selective. The only bright spot is Northern Oil and Gas, which grew revenue +74.7%—but that's a consolidation story, not a price recovery.

The Plot Twist: Micron's Explosive Growth Masks a Choppy Semiconductor Cycle

The biggest surprise is not that Micron grew—it's the magnitude. Revenue accelerated from a prior period of 'n/a' to +345.7% year over year, a pace that dwarfs even NVIDIA's +85.2%. But this is not a uniform boom: Lam Research revenue fell -13.0% and net profit -41.9%, while Qualcomm slipped -4.0% revenue and -24.9% net profit. The semiconductor industry is bifurcating—memory and AI accelerators are in a supercycle, but mature segments like smartphones and PCs are weak. Micron's P/E of 18.3x and EV/EBITDA of 13.5x look cheap for this growth, but investors should question if the memory cycle is peaking.

Valuation Sanity Check: Growth Is Priced for Perfection in AI, but Some Names Are Bargains

The AI trade is expensive, but not uniformly. Palantir trades at 129.0x P/E despite growing revenue +84.7%—that's a bet on future acceleration, not current value. Conversely, Micron's 18.3x P/E with +345.7% growth is a rare bargain, though cyclicality explains the discount. In energy, APA Corporation trades at 8.6x P/E and 3.2x EV/EBITDA despite revenue -14.2%—cheap for a reason, but with net profit +28.5% it may be a contrarian play. The real value trap is in software: Datadog's P/E of 0.7x (likely a data artifact) and EV/EBITDA of 22.1x with +32.2% growth is reasonable, but Snowflake at 33.5% growth and no P/E given is hard to assess. For value investors, the oil patch offers deep discounts, but only if commodity prices stabilize.

Income Is Scarce, but Telecom and Energy Offer Yields That Reward Patience

For income investors, the standout is AT&T, trading at 7.7x P/E and 6.4x EV/EBITDA—likely offering a dividend yield above 5%, though not specified. Verizon at 12.2x P/E and 7.4x EV/EBITDA also provides a solid yield, with revenue -0.7% but EBITDA +7.0%. In energy, Antero Resources at 9.9x P/E and 5.9x EV/EBITDA, with revenue +20.2% and net profit +71.9%, may offer both growth and yield. But the real income play is in oil: Chord Energy at 5.4x EV/EBITDA and Devon at 5.4x EV/EBITDA, both with net profit declines, suggest high yields are a compensation for risk. Investors should demand a margin of safety.

The Long View: Compounding Machines Are Rare—Robinhood and Super Micro Stand Out

Looking at 3-year revenue CAGRs, the true compounding stories are Robinhood Markets (+48.8%), Palantir (+32.9%), and SentinelOne (+33.4%)—all growing at a pace that justifies premium multiples if sustained. But the most impressive is NVIDIA, with a 3-year CAGR of +100.0%, a feat rarely seen at its scale. The watch item is whether the AI capex cycle continues: Dell's +87.5% revenue growth and Super Micro's +122.7% suggest demand is still accelerating, but the energy sector's collapse is a reminder that cycles turn. Next quarter, watch if Micron's memory supercycle persists and if power producers like Talen can maintain their growth—those will be the bellwethers.

Players: growth & yield (no absolute levels)

A look at PepsiCo (Nasdaq: PEP) on the fresh Q2 2026 report. The prompt was 'a good report, but the stock is at six-year lows' - we checked both halves against primary data. Short version: the report is normal rather than good, the lows are real, and the heart of the case is that the market now pays almost half as much for the same earnings (multiple compression from 31x to 18x) while the business kept growing all along - and a large activist shareholder with a change agenda is already inside.

The price is back to 2019: seven years of business growth, none for the stock

$135 per share is summer-2019 territory; in seven years the price was lower only in the covid crash of March 2020 ($104). Formally 'six-year lows' is almost exact: excluding three weeks of 2020 panic, the current price is the bottom of the entire seven-year range. Over those same seven years revenue grew from ~$67bn to $94bn and the dividend was raised every single year.

The Q2 2026 report: 'it did not get worse' is what the market called good

The quarter's facts: revenue $24.18bn (+6.4%) - above expectations; organic growth +2.4%; adjusted EPS $2.20 - 1-3 cents below consensus; full-year guidance confirmed (organic +2-4%, core constant-currency EPS +4-6%). Geography is two-speed: North America is weak (snacks losing pricing power, sluggish volumes) while international - Asia, EMEA, the beverage franchise - grows volumes and pulls the group. In other words, after two years of downgrades a 'good report' literally means 'guidance confirmed, nothing broke' - the expectations bar is on the floor, which is information in itself.

Business vs stock: the entire decline is multiple compression

The decomposition is simple. Five-year revenue: $79.5bn to $94bn (+18%). Dividend per share: $4.25 to $5.75 (+35%), the 54th consecutive annual increase - Dividend King status. Meanwhile the multiple compressed from 31x earnings to 17.7x - against its own five-year norm of 24-28x. All of the stock's negative return is de-rating, not business degradation. The 4.3% dividend yield is the highest in the company's modern history.

What GLP-1 is and why it made the stock cheaper

GLP-1 is a class of weight-loss and diabetes drugs (best known: Ozempic, Wegovy, Mounjaro - the 'weight-loss injections'). They suppress appetite, and people taking them snack noticeably less - and snacking, Frito-Lay, is the most profitable part of PepsiCo's business. Hence the market's main fear: that falling US snack volumes are not the temporary belt-tightening of an inflation-weary consumer but an irreversible change of habits.

What independent data say. Consumer-purchase researcher Circana (checkout scanner data - actual sales, not surveys): someone in 23% of US households already takes a GLP-1 drug, and such households buy 10.1% less salty snacks and 8% less sweets; by 2030 households with a GLP-1 user may account for up to 35% of US food and beverage sales. Simple arithmetic: 23% of households x minus 10% on snacks is roughly minus 2-2.5% for total market volumes - noticeable, but not 'the end of snacks'. Meanwhile forecasts for the global salty-snacks market remain positive: +5.4% a year to 2030 (Wissen Research; other agencies put it at 4.7-6%) - driven by international markets, where PepsiCo is doing well. So the pessimists have two arguments - cyclical (the US consumer economising after inflation) and structural (GLP-1) - and so far the facts point more to a cyclical character of the decline: volumes are falling precisely in expensive US snacks while international markets and beverages grow. But only a turn in US volumes can prove that to the market conclusively - and that is what everyone is waiting for.

Elliott with $4bn inside: the catalyst is already working

Since late 2025 Elliott Management holds a ~$4bn position with a public plan: focus the beverage portfolio, optimise bottlers, review the North America supply chain, cut costs aggressively. Management has already announced an NA logistics review and a cost programme. Seven months in, the market wants proof - every next quarter becomes a turnaround checkpoint. An activist of this scale inside a Dividend King is a rare configuration: downside protected by the dividend, upside driven by the activist's plan.

Forecasts: consensus target $169 vs a $135 price

The median target of twelve analysts is $169 (+25%), but the spread is honestly wide: Citi cut to $145 after the report (and the rating to Neutral), Barclays raised to $144; the average rating is Hold. Translated from analyst-speak: 'the business is worth more, but buying before the volume turn is scary'. The company's own 2026 guidance is confirmed: organic +2-4%, core EPS +4-6% in constant currency.

Estimated upside: three scenarios over three years

From $135 with forward earnings of ~$8.3 per share (fwd P/E ~16.3). Base: EPS +5%/yr, exit P/E 19 - +48% total return, IRR ~14%/yr (of which ~13pp from dividends). Bull: a volume turn plus the Elliott programme, EPS +8%/yr, return to P/E 22 - +83%, IRR ~22%. Bear: GLP-1 proves structural, EPS flat, de-rating to 14x - roughly zero over three years: the dividend fully offsets the price decline. Probability-weighted (40/30/30) - roughly 12-14%/yr expected dollar return with rare asymmetry: the worst case is ~0, not a loss.

Risks

— GLP-1 may prove structural. Then the bear case is not a tail but the base: US snack volumes never return and 14-16x becomes the new normal forever.

— A long stretch of US consumer weakness. Even without GLP-1 the volume turn may take years; every quarter without it risks new downgrades.

— Rates. A 4.3% dividend against ~4%+ on Treasuries is a minimal premium; if the Fed goes higher, the dividend support for the price weakens.

— A 78% payout. Room to grow the dividend faster than earnings is nearly exhausted - payout growth will be slow until EPS turns.

— The activist may leave. If Elliott sees no progress and sells, part of the catalyst premium leaves with it.

Prepared by Enhanced Investments from PepsiCo filings (Q2 2026 press release, SEC 8-K), Nasdaq data and consensus forecasts; July 2026. Not individual investment advice.

Thirteen ideas from our coverage across frontier markets, commodities and US large caps. Each was produced by one of five live engines (EM bank value with monthly regulatory data, EM compounders, dividend+catalyst, commodity spot mean-reversion, US Leaders GARP), screened across ~500 issuers and verified against primary filings. Currency math is a first-class criterion: a 12% dividend behind a 13%/yr hedge is worth less than 9% behind a 2.8% one.

Live track records of the two systematic books contributing ideas: US Leaders +274% since Aug 2019 (CAGR 21%, Sharpe 1.0); Global Commodities +137% since May 2020 (CAGR 15%). The commodity engine's latest published call - PBF at $42 on June 25 - is +36%.

The numbers layer: valuation vs history and an indicative IRR sketch

Is each name cheap against ITS OWN history, and what annual return does the position imply if the multiple drifts back to the 5-year median while the business grows and pays out? Charts below; per-name assumptions in the cards. Growth assumptions are deliberately conservative vs recent prints; CPIN capped at 15x; VLO on forward earnings; MBRF and KEGOC excluded as not meaningful.

Part I. EM bank value

1. Bank Mandiri (IDX: BMRI; ADR PPERY) - the flagship

Indonesia's #2 bank: wholesale-led franchise, 71.6% CASA funding, ROE 20.4% in Q1'26, monthly regulatory data showing acceleration. Expected hedged USD return ~15-16%/yr before re-rating. The Q1 consolidated 'asset shrinkage' is the BSI deconsolidation, not the business. Risks: Danantara governance, directed lending, NIM pressure.

2. Bank Rakyat (IDX: BBRI; ADR BKRKY) - the higher-yield pair

The world's largest microlender (65m borrowers). 2025 was the clean-up year; the cycle is turning. With Mandiri - a sector bet at ~5x earnings with double-digit dividends. Risk: BBRI carries most of the $12bn village-cooperative programme.

3. Banorte (BMV: GFNORTEO) - the dividend machine with direct access

Mexico's #2 financial group: a durable ~10% peso cash yield from a franchise that survived every Mexican cycle since 1899; nearshoring optionality is free. Risks: Banxico cuts, US-Mexico politics.

Part II. EM compounders

4. Kaspi.kz (Nasdaq: KSPI) - quality growth at a value price

Kazakhstan's payments+marketplace+fintech monopoly funding a second act in Turkey (Hepsiburada) from its own cash flow. If Turkey works - a second engine; if not, the core justifies the price. Risks: tenge, fee regulation, Turkish execution.

5. Charoen Pokphand Indonesia (IDX: CPIN) - protein for 280 million people

Indonesia's poultry leader. The free-school-meals programme that worries bank investors adds 5-7% to broiler demand - a natural internal hedge in an Indonesian basket. Margins will normalize; the per-capita protein story is early. Risks: feed costs, normalization.

6. Indofood (IDX: INDF) - the Indomie monopoly at half book

Indomie is close to a de-facto noodle monopoly and one of Asia's most recognizable brands. The discount is structural (holding, country), not operational. Risks: holding discount persistence; wheat/palm oil.

Part III. Dividend + catalyst

7. Telkom Indonesia (NYSE: TLK) - value with a spin-off attached

Indonesia's incumbent telecom after a governance reset. Tower/data-center carve-outs have historically unlocked value across Asian telecoms; here one comes with a 9% yield attached. Risks: price war, timelines.

8. PLDT (NYSE: PHI) - dividends plus a free fintech option

The Philippine telecom leader whose #2 digital bank the market values at zero. Capex peak passed, FCF positive, deleveraging to 2.0x. Risk: leverage 2.56x EBITDA - the highest here.

9. KEGOC (KASE: KEGC) - the grid monopoly after a tariff reset

Half of book value for the national grid with regulated returns flowing through. Last despite best momentum: currency and access fail the USD framework; strong local-book idea. Risk: tariff politics.

Part IV. Commodities - what the model actually likes

10. Valero (NYSE: VLO) - the quality way to own the refining cycle

The same thesis that made PBF +36% in three weeks, in institutional form. PBF (+141% model upside) stays the aggressive pure-play; DK (+150%) and PARR (+132%) are the small-cap tail. Risks: crack normalization, product imports.

11. MBRF (B3: MBRF3, ex-BRF + Marfrig) - the new global protein giant

Cheap corn, a weak real and merger synergies compound Brazilian chicken's global cost advantage. Risks: US beef cattle cycle, controlling-shareholder governance, BRL.

12. EQT (NYSE: EQT) - the structural natural gas winner

Not mean-reversion but a multi-year demand story; the latest quarter already beat on volumes and prices. Risks: gas volatility, LNG contracting.

Part V. US Leaders

13. PepsiCo (Nasdaq: PEP) - a dividend king at a decade-high yield

GLP-1 fears pushed a pricing-power franchise to its highest yield in decades; the dividend streak is a hard floor, volume recovery is upside. Risks: structural GLP-1 drag, weak US consumer.

Summary

Three baskets. Buy-now via any US broker: BMRI (PPERY), BBRI (BKRKY), KSPI, TLK, PHI, GFNORTEO (IB direct Mexico), VLO, EQT, PEP. Local-access watchlist: CPIN, INDF (Jakarta), MBRF (B3), KEGOC (KASE). Published and running: PBF +36% since June 25.

Prepared by Enhanced Investments, July 2026, from primary filings, exchange and central-bank data. Live track records, not backtests. Not investment advice.