Haymaker (Substack) · co-founder / ex-CIO Evergreen Gavekal · macro newsletter writer — running synthesis of his issues, with per-post breakdowns. Mostly macro, but he does give sector and single-name calls (e.g. his June-5 David Lin interview).
Hay writes mostly macro (deficits, rates, AI productivity, the cycle), but he does offer sector and single-name ideas — usually as ETFs or rotation themes (energy, gas, coal, financials, EM, mid-caps) — so this index fills in when an interview gets specific. The substance lives in each post's talking points and feeds the master macro viewpoints.
Nuclear-microreactor milestone (2026-JUL-08; Positive, private) — Aalo Atomics, the microreactor startup Haymaker positively profiled in March; over the July-4 weekend its test reactor achieved criticality, joining just three other companies to reach that bar. A true microreactor (vs an SMR) whose small reactors "lend themselves to mass production, as well as delivery on the same trucks that transport standard containers" — an edge in the "exciting race" to power data centers as nuclear supplements/replaces natural gas. Private, not investable; a thesis to watch. (Reiterated positive.)
Index-removal buy (2026-JUL-27; Positive) — Atlas Energy Solutions, "one of the most mercurial companies we've brought to your attention," moved back up to the Buy List sub-$12 (a partial position bought, "though not as a Strong Buy") after the full round trip: Buy at 18 (Apr 2025) → DCA'd down to a ~8¾ bottom → Hold near $12 in January → Trim on May 15 near $19 → a crash back sub-$12 driven by removal from S&P Small Cap Growth indexes — the exact inverse of the late-2024 index inclusion that lifted it to $24 against deteriorating fundamentals and gave him "a great time to ring the cash register." This time the mechanical selling lands as the oil-service cycle turns up. The leading Permian frac-sand supplier ("you simply can't produce shale crude without frac sand"), volumes recovering with pricing still soft, into "the unprecedented oil shortage which has developed — and is worsening, despite the various peace treaty head fakes"; the Dune Express conveyor is a moat "none of its competitors have," the market is "almost a duopoly" (pricing power in the next up-cycle), and the growing power-supply business is free optionality. Caveat: peer Liberty Energy's (LBRT) weak quarter is assumed to repeat here, but "well-discounted" after the collapse since May.
Silver-miner trade → gradual accumulation (2026-JUN-25; Positive) — First Majestic, written up bullishly last summer and played as a high-beta silver trade: the Jan-26 Trading Alert gain-harvested after ~+170% ("could keep running, but is likely to get pummeled… on any correction in 'poor man's gold'"), and they concede they should have trimmed again at its ~$30 spike. After silver crashed from ~$120 to ~$58 (AG fell with it), the metal's fundamentals are "extremely encouraging" — very constrained supply vs. robust munitions / AI-data-center / grid demand pushing the market from surplus into a pronounced deficit, with silver-futures open interest at a multi-year low (extreme bearishness). They'd now suggest gradual accumulation of AG (and silver the metal) despite the poor technical set-up.
Back to the Buy List at $29 (2026-JUL-27; Positive) — Alamos Gold, "a long-time favorite of both this newsletter and… Fred Hickey," cut in half from its ~$55 February peak while the senior gold-miner ETF fell 35% from end-February. Haymaker had dollar-cost-averaged out of the miners through the 2025-26 "hockey stick phase" and stayed silent on the group since — "however, at $29… an attractive buy, or buy up, point": 10.4× trailing earnings, under 10× 2027's $3.27 consensus ($2.32 for 2026), reserves "almost tripled" in ten years and grades ~35% better. The Young-Davidson earthquake (−12% Q2 output, lower full-year guidance) is treated as a fading one-off — "natural disasters… tend to fade into insignificance over time." Bullion's drop is "a correction within an ongoing bull market" (central banks "aggressive buyers," rising government-bond-crisis odds, fiat debasement), and retail's "truly remarkable" negativity after a +150% year is read as a contrarian positive. "Before long we may bump it up to a Strong Buy." On the 2022 buy list around $8.
Reference (2026-APR-24; Positive) — Allstate, another Sy Jacobs win: a "qualified plug" that had run 40% but was breaking out and "has continued to surge." "Another 'W' for Sy" — Haymaker "should have been all, or ALL, over it" given its love of upside range expansions, "but that one got away." A breakout endorsement in passing, not the week's pick.
"A good illustration" (2026-MAY-27) of his energy-accumulation method — a 2-year breakout that has settled back; dollar-cost-average energy on ceasefire-hope weakness (lost-supply + SPR-replenishment thesis means oil doesn't snap back to $60–70).
Gas-producer winner recap (2026-JAN-20; Positive) — Antero Resources, recapped from a Sept-2024 bullish write-up (with CTRA & EXE) on the natural-gas thesis; has "performed considerably better" than CTRA since. Cited as confirmation of the gas-producer leadership behind the DVN pick.
"Second chance" buy (2025-FEB-03) — the Global X Argentina ETF, an admitted Haymaker "error of omission." Milei vanquished hyperinflation (200%+ 2023 → ~40% and falling) and the economy is back to growth; "the Argentinian ETF" broke out above 3-yr resistance (~40) in early 2023 ahead of his win — "a resoundingly accurate bullish signal."
Haymaker POW! Buy (2026-SEP-18) — "We recommend a BUY on AZN" at ~$164, 23% off its high after "three specific setbacks in three months": the most productive large-cap pharma pipeline bought on its hit rate (20+ Phase 3 readouts in 18 months; 16 positive pivotals in 2025) — "two wins and one loss… that ratio is the investment thesis." ~14.9× forward P/E (vs market 21) and P/S 4.16×, both near decade lows last seen in 2016–17; oncology 48% of product revenue +16%; 11% core EPS growth; tozorakimab (COPD, $5B+ peak, Priority Review, Q1 2027 decision) the unpriced asset; $80B 2030 ambition. Risks: more trial misses, US drug pricing (~40% of revenue), a cut to the 2030 target; no tight-range breakout, but prior corrections were buying opportunities.
Buy maintained on a 20% loss — the evidence improved, the price didn't (2026-AUG-24; Positive) — recommended "May 14, 2026, at $162.11" and down 20%: "the 20% loss is real and it's been a painful few months to be a holder." The Q1 FY2027 print beat on revenue (RMB 268.95bn, +9%) and missed badly on adjusted EPS (RMB 8.52 vs RMB 10.72), a shortfall "largely due to heavy AI infrastructure capital expenditure compressing near-term profitability" — reclassified as "the accounting cost of building what this powerful competitive advantage requires." What advanced the thesis: cloud revenue "grew 45% year-over-year, its fastest pace in 22 quarters," triple-digit AI product growth, and Qwen's emergence as the dominant open-source model ecosystem — "3 billion downloads… more than 50% of all open-source model downloads worldwide," outpacing Llama "more than 13 to 1," 300,000 derivative models, Hugging Face calling it "the default workflow" — with the commercial link made explicit: "developers building on Qwen deploy on Alibaba Cloud." At $148bn trailing revenue, $15bn net income and a PEG of 0.59 (46% below its own historical average P/E), "the discount is simply a fact and the key question is whether it is deserved… we think the discount has overshot." Buy maintained, with a suitability caveat that is the real risk disclosure: only for those who "can stomach a hefty dose of volatility" — others "should trim or exit." Addendum: a $10.2bn issuance at an 8.4% discount, all earmarked for AI infrastructure, "nearly three times oversubscribed." (Prior: 2026-JUL-20 back on the Buys list rated SB; 2026-JUN-01 accumulate on the "China's Amazon and Google" call.)
FXI-constituent evidence (2026-JUL-17; Positive) — Baidu ("the Google of China"), cited in the FXI reiteration for AI now comprising 52% of its total revenue, a platform far along in converting China's AI-usage surge into earnings. (Also an Uber AV / Uber-Autonomous-Solutions white-label partner in the 14-partner "toll booth across all AV winners" thesis.) A constituent data point behind the reiterated FXI buy, not a standalone Baidu call.
POW pick (2026-MAY-15) — Brasil Bolsa Balcão (B3 ADR), Brazil's monopoly exchange operator; a "HALO" (Hard Assets/Low Obsolescence) toll-booth with network effects/high switching costs (Trader Ferg's idea). Zero net debt, EPS $0.75 (2022) → $0.97 → $1.20E 2026 (~10%/yr, no down years), ~6% dividend; ~14× (forward ~9–11×) vs peers 20–30×, PT $21.60 at 18×; two chart breakouts. "Anti-fragile" (volumes rise in growth and volatility); falling Brazilian rates a tailwind. Risks: BZ politics/currency, market already up a lot.
POW pick (2026-JUL-10; pick & core analysis by contributor Daniel Bustamante) — Copart, the leading global online salvage/used-vehicle auction & remarketing platform; an asset-light intermediary (no vehicle ownership) earning 80%+ high-margin service fees via its VB3 platform (275+ locations, 11 countries, since 1982). Down ~45% from highs / −55% since last April to ~17× fwd P/E; fortress balance sheet ($4.2B cash+HTM, $5.5B liquidity, minimal debt), $1.6B YTD buybacks (43M+ shares) vs ~$27B cap, gross margin 46.3%. Two catalysts: a rising auto-loan-delinquency tailwind (90-day+ rate 5.6% in Q1 2026, highest since 2010; next data Aug 4) feeding non-insurance auction volume, and star CEO Jay Adair back as CEO Jul 31 (1,100%+ return over his 2010–24 tenure). Technicals ugly — wants a mid-$20s base for ~a month; high-$20s an ideal long-term entry; decisive break of $25 concerning, ~$24 a stop/sell-down threshold. "Genuinely a long-term compounder," not a trade.
Back onto the Buy List (2026-AUG-03; Positive) — CRH plc, the global aggregates / cement / asphalt / ready-mix supplier with a dominant North-American road-and-infrastructure franchise. A long-held Haymaker position that moves off the Trims/Holds table back to the Buy List, rated B on both lots (01/29/2024 @ $73.23, +33.27%; 04/28/2025 @ $98.51, −0.93%; marked $97.59) as one of that day's blue-highlighted rating changes. No fresh prose accompanies the move — read it as the building-materials name being re-promoted into the hard-asset / reindustrialization leg of the book Hay keeps favouring over the capital-hungry AI complex.
Rumored DVN acquirer (2026-JAN-20; Positive) — Coterra Energy, reportedly pursuing a merger with Devon (combined entity ~two-thirds gas / one-third oil) — a combination Haymaker thinks "would be highly attractive," and which could prompt a Chevron counter-bid. Given a bullish write-up in Sept 2024 (with AR & EXE); "still fans," up ~10% from that plug. Its pursuit of DVN "validates our view" of Devon's value.
POW pick (2026-MAR-06; featuring Daniel Bustamante) — Deckers Outdoor, parent of UGG and HOKA, at ~$105 (off ~$83 lows) after a ~50% 2025 fall on tariff fears. The crux: FY26 net tariff hit slashed to ~$25M (from $75–110M, a ~77% cut) via pricing + supply-chain moves, re-rating the multiple by raising the "E." Record Q3 FY26 (rev $1.96B, GAAP EPS $3.33 well above consensus); FY26 guide ≤$5.425B rev / $6.80–6.85 EPS. ~14.7× P/E (~13.25 ex-$1.74B net cash), zero debt, ~$1B FCF ≈ after-tax income, $1.8B buyback remaining (bought into a depressed price). HOKA low-teens growth; a third brand could lift long-run EPS CAGR ~12%. Broke the 2024 downtrend, back above the 200-day; case for the mid-$150s into next earnings. A 2–4-quarter opportunity; the Iran-correction dip is "a better entry."
Previously-endorsed energy name (2026-MAR-20; Positive) — Devon Energy, generating a multi-year upside range expansion; still well below its ATH and trading at a thrifty P/E, cited alongside Range Resources as confirmation of the energy-producer leadership behind the EQT pick.
POW pick (2026-MAY-22) — Estée Lauder, "a turnaround morphing into a growth story" in the 80-year prestige-beauty brand. New CEO's PRGP ("Beauty Reimagined") is working: adj. operating margin 15.0% vs 11.4% (+360bps), FCF rebounded to $891M (from $276M), China flipped from ~30%-of-sales headwind to tailwind (+11%). At decade-low P/S & P/E; since collapsed margins inflate the P/E, P/S is the turnaround lens — a return to 3× sales ≈ 60% upside (analyst PTs $90–100). Wait for a pullback after the ~12% Puig-merger-cancelled pop; a 2–3-yr hold for a possible double, "an investment, not a trade."
The predicted breakout arrived — let it run (2026-AUG-24; Positive) — the multi-year ceiling broke: "this month, it broke above overhead resistance extending back to the powerful energy sector rally during the early stages of Russia's attack on Ukraine," though "this is not yet decisive" until confirmed by "a continuing surge above the upper $140s where it had previously hit the wall." From the June 17 "vehement endorsement" it has run "from $133 at that point to $151 today, over six times greater than the return on the S&P," and the valuation is checked after the move rather than before: "irrespective of this surge, the valuation metrics for EOG remain extremely undemanding… the Price/Sales ratio is particularly modest." Untrimmed against the house's own trim rule because all three conditions still pass — "the chart patterns, the valuation, and the fundamental stories all remain highly encouraging" — so "sit tight and, hopefully, let them run"; Super-Major takeout named and discounted as "always longshots." (Prior, 2026-AUG-11: the Permian takeaway beneficiary — upgraded from a Jan-20 valuation benchmark to the Daily's named pick: "EOG is likely to be one of the main beneficiaries of this extraordinary delivery surge." The Permian, "America's most productive oil basin," has become the country's second-biggest and fastest-growing gas region, and for want of pipe its associated gas "has often traded at negative prices, causing much of this valuable resource to be flared into the atmosphere" — "another source of profit suppression." Midstream is fixing it: "multiple new pipelines under construction" totalling "some 15 billion cubic feet per day, about 12½% of aggregate U.S. marketed gas output," delivering long-stranded gas to "a marketplace that will soon be in desperate need of it" — "a notable earnings kicker for companies with significant Permian gas production." The payoff is basis, not price: flared/negative-priced molecules start clearing at a real price with no incremental drilling and no Henry Hub rally required, leaving decade-low gas bullishness and an 18th-percentile real price as free optionality. Trigger supplied by the chart — "on the verge of breaking out of what is nearly a five-year trading range." No rating stamp, price or Buy-list action.)
Performance recap (2026-JAN-30; Positive) — Equinor, up ~13% in the month since recommendation ("now at 26.90, up from $22.90 when we first recommended it"). The Norwegian energy major, a recent Haymaker winner. A winner update, not a fresh entry.
POW pick (2026-MAR-20) — EQT Corporation, America's largest, only large-scale vertically-integrated Appalachian gas producer, entirely unhedged for 2026 so every dollar of gas price falls straight to FCF: "a pure, unlevered call option on the current gas price environment," just as the Hormuz crisis + Qatar LNG force majeure made demand "urgent in 13 days." Clean Q4 (adj EPS $0.90 vs $0.73, +22.7%; FY net income $2.04B from $231M; FY FCF $2.5B), 2026 FCF guide $3.3B (~8% yield, set at lower strip), net debt $9.1→$7.7B → ~$4.7B YE26 (deleveraging unlocks the multiple). ~12–13× fwd FCF; consensus median $65 / to $80 on $5 gas; 34-analyst Strong Buy; broke out above all prior resistance (eclipsed the 2014 ATH). Buy a smaller initial position post run-up, add on weakness. Follow-up 2026-SEP-21: rated a Strong BUY at ~$50 despite ~−23% from entry ("Call us obstinate") — 12.4× P/E, ~40% below the S&P, "silly-cheap"; "among the best AI plays" as ~50 GW of gas plants are planned and Japan funds ~18 GW of US gas generation (a 9.2 GW Portsmouth, Ohio plant) near EQT's footprint; US gas ~$3 vs >$20 in Europe/Asia.
New name via Sy Jacobs (2026-JUL-20; Positive) — EverQuote, a QNST-peer online insurance-comparison marketplace and a Sy Jacobs "Super Investor" pick; up 59.7% since his Mar-3 Haymaker Webinar (as of Jul 17) and "on the verge of a five-year breakout." Bull case = abundant free cash flow funding substantial, accelerating share buybacks. Cited as a winning peer corroborating the QNST insurance-marketing thesis, not a stand-alone deep dive.
Let it run — up ~75%, still cheap on the right metric (2026-AUG-24; Positive) — recommended "April 7th, 2025, at $121" and added to the Buy List then, an entry made into the "Liberation/Obliteration Day period of extreme broad market weakness" that "pretty much caught the three-year low"; now "up almost 75%" against "the S&P's roughly 50% rally since the Tariff Tantrum." Still "at the low end on a Price/Sales basis" — the house's "preferred valuation metric for a cyclical company like FANG," justified with the 2022 worked example in which $100 oil more than doubled EPS and "deflat[ed] the P/E to an unsustainably low point" while P/S fell "to a significantly lesser extent." No trim despite the gain: "the chart patterns, the valuation, and the fundamental stories all remain highly encouraging" so "sit tight and, hopefully, let them run," with Super-Major takeout named and discounted as "always longshots." (Prior: 2026-JUN-22 energy Buy-list pick on chart and undemanding P/E and P/S.)
Gold royalty/streamer (2025-FEB-03) — flagged off a Rick Rule podcast (Dec-9 MHM): "capital-lite business model," corrected hard on Panama news, "a more than respectable rally." A quality standout in the rebounding, long-despised gold miners.
POW reiterated Buy (2026-JUL-17) — iShares China Large-Cap ETF, first bought $38.42 (2026-FEB-25), now ~$34 (~11.5% drawdown; 52-wk $31.19–$42.00). The four original bull pillars hold (Dec-2025 policy pivot, the DeepSeek moment, post-Xi/Jack-Ma regulatory normalization, 10–11× earnings); the loss is macro/sentiment (US tariffs to ~17% effective, a Hormuz-driven Beijing slowdown, a weak consumer — May retail −0.6%), not fundamental. AI monetization is accelerating (China daily AI-token usage +1,400× in 26 months; Alibaba Cloud $1.3B/qtr triple-digit, Tencent net profit +21%, Baidu AI 52% of revenue) while FXI sits at 10.78× earnings ("the cheapest major stock market in the world") with light positioning (~$895M outflows) and a controlled correction after a four-year breakout. Reiterating a Buy at $34, 12–18-mo target $42–48 (re-rate to 12–13×); eventual break of $45 resistance / new ATH >$60. Size for binary VIE/tariff risk (holdable through a 25% drawdown). "We are staying in."
The junior leg of the same apathy trade (2026-JUL-26; Positive) — the VanEck Junior Gold Miners ETF, shown on the same fund-flow chart as GDX: record outflows in a year the miners delivered "tremendous returns" and real profits, which Hay reads as contrarian apathy rather than information. Same conclusion — "the gold miners and silver miners look pretty interesting" — and the same fundamental kicker of widening margins plus incoming energy-style capital discipline. Higher beta than the seniors, so it swings harder both ways.
Haymaker POW! Buy (2026-AUG-07) — the $10.5B Q2 GAAP net loss is "an accounting event, not a business event": acquisition write-downs GAAP forces to be expensed immediately, in the same quarter that RAISED FY product-sales guidance to $30.1B. Underneath, a 46.9% adjusted operating margin on ~$30B revenue = ~$14B adjusted operating income, >$12.5B estimated FCF next year on a $163B cap (a 7.6% FCF yield) and net debt of only ~$13.4B (repayable in ~1 year of FCF). The engine is a ~70%-share HIV franchise — Biktarvy protected to 2036, PrEP past $1B/qtr, a Merck-partnered once-weekly oral pill positive in ISLEND-1/2 — funding a Trodelvy first-line oncology build-out (US+EU metastatic TNBC; +Keytruda in lung) that is "not incremental label expansion." At 12-13× adjusted earnings vs the S&P's 20× with a 2.5% dividend, resting on the 200-day after the spring-2024 two-year breakout — "the pause that refreshes." Bears must be right on BOTH IRA pricing and acquisition integration; expects a re-rate to 15-16×.
Haymaker POW! Positive (2026-AUG-28; written with contributor Daniel Bustamante) — "Generac is no longer simply a residential generator company dependent on hurricanes and power outages." The pick is the equipment supplier paid by the hyperscaler capex others fear: "while others are worrying about CapEx spend on the hyperscalers Generac has been, and will continue to be, a net benefactor of said spend," because "every large-scale data center requires multi-megawatt backup generation capable of carrying the full facility load through grid outages" at a "scale (and service network) that new entrants cannot replicate quickly." Two prints, not one: Q2 C&I sales +29% to $556M after +28% in Q1 — "no longer based on a single quarter or a hypothetical pipeline." Visibility is the product: a ~$1.6B data-center backlog (+~$1B since the prior update, $1.35B scheduled for 2027, and excluding the second hyperscaler's committed volume), ~$1B of new orders in the 90 days to Jun 30 (~$11M/day), FY data-center revenue guidance raised to $450M, ~$700M of committed 2027 volume with one hyperscaler plus a global supply agreement with a second — with Enercon and an Illinois large-megawatt packaging plant bought "against contracted demand, not merely building capacity in anticipation of demand that may never arrive." The re-rating is a mix shift: C&I ~47% of Q2 external sales vs ~41%, so it "increasingly resembles a diversified power-infrastructure company," and such businesses "generally receive higher and more durable valuation multiples." Residential is now optionality (Q2 −2%, 34.7% EBITDA margin of which ~9pp was tariff refunds; a probable "Super" El Niño the free call) — "the current thesis does not require a major hurricane season." Consensus: revenue +17.4%/+21.3% to ~$4.94B then ~$6.0B, adj. EBITDA $690M → $1.01B → $1.19B, adj. EPS $5.14 → $9.75 → $11.82, FCF ~$353M/~$478M, $1.33B debt vs $265M cash. At ~18.2× forward earnings and 2.7× trailing sales "well below its 2021 peak" — "earnings expectations, not valuation expansion, are now doing most of the work" — with a $500M buyback authorization (Feb 9, 2026) entirely unused. Entry: after a +90% run and a one-third correction it sits on the 200-day and the 50% Fibonacci retracement that "already found buyers back in July" — "an intriguing play near the $200 area." Follow-up 2026-SEP-21: BUY reiterated at ~$205 (+~10%) after an 8-K Amazon supply agreement for up to $8B of data-center backup generators through 2033 ($2.4B in 2027–28), with Amazon able to buy up to 1.7M shares at $200.93 only as it orders — "That alignment is the moat"; 17.9× forward, ~57% below the 2021 peak; pre-deal targets ($284–$295) "almost certainly stale"; next catalyst Oct 27 Q3.
Buy List — STRONG BUY, added 2026-AUG-31 at $36.79 ("One 'HAL' of an Opportunity"). The oil-service pick for "the greatest oil-supply shock in history": Mid-East facility repairs (and, a decade out, Venezuela) are years of work for the few firms with "the unique set of capabilities to rebuild global oil stocks." Cheap on trough earnings, not peak — "rarely traded as inexpensively as it is now on a P/E basis," with the P/S cross-check disclosed ("cheap, but not as decidedly") and resolved by the industry check: conditions "far from boom times," so "neither revenues nor profits are maxed out." Stuck below ~$43 for years and knocked to the low-$30s in late July on a guided-down Q2; preferred over SLB near-term and funded by trimming it. (First added 2026-JUN-22 as an oil-service proxy tracking oil & gas producers.)
Haymaker POW! Buy (2026-SEP-04) — "We recommend a Buy on INCY" at ~$129: a profitable, debt-free biopharma whose base case already clears the price, with a potentially practice-changing cancer drug taken for free. Valuation: ~16.5× the 2026 EPS consensus of ~$7.50; EV ~$22.4B (cap less $4B net cash) on $5.2B of guided FY net sales = "approximately 4.3x EV/revenue, which is modest for a business growing net sales 40% year-over-year"; on the house Price/Sales yardstick the headline ~4.4× becomes ~3.6× EV/sales net of the ~$4.5B cash pile — the discount "largely reflects the fear that its newer products won't replace its legacy ones." Base business: Jakafi $817M in Q2 (+7%), ~$3.2B annualized and shared with Novartis ex-US, the cash cow; Opzelura $450M, "the growth driver that has consistently outrun consensus," the only approved treatment that restores skin colour in vitiligo, where "each approved indication adds a new patient population at minimal marginal cost." Q2 (Jul 28): revenue $1.67B +38%, EPS $3.09 vs $1.84 — a 68% beat deflated by its own author to ~55% operational after a $246M non-cash Medicaid settlement — FY net sales raised to $5.13-$5.26B. Depth: 10 Phase 3 studies, four approvals expected within 12 months, $4.0B+ net cash and no meaningful debt. The free option: KRAS G12D in first-line pancreatic cancer (~40% of PDAC; five-year survival below 12%, below 3% metastatic; the target "considered undruggable for three decades," with G12C in lung cancer having "proved the mechanism could work in humans") — DAWN-303 EU registry shows 62.5% response / 95.8% disease control with GemNabP and 75% / 100% with mFOLFIRINOX, against 20-30% for chemotherapy alone: "less of a lottery ticket and more of a practice-changing drug." The market prices none of it — the $127.13 consensus target "is essentially at the current price… a Jakafi-plus-Opzelura base case with limited KRAS credit," while Leerink $155 / Canaccord $152 / H.C. Wainwright $150 "are the analysts adding in KRAS optionality." Bear case stated, not softened: the Jakafi patent cliff (US compound protection expires 2028, formulation patents only "potentially" to 2033) on ~55% of net sales; a "small interim cut" of KRAS data where "Phase 1 combination data has misled markets before," with no durability or long-term safety yet; and at $129 after +49% in 12 months this "is not a margin-of-safety purchase." Entry deliberately unhurried — "we are a bit late to this one" (the breakout fired into the $90s late last year), but the advance since has been "earnings-driven"; stretched over the 200-day, so "a retracement back to $110 or so would not be at all surprising. More conservative investors might want to hold off." Catalysts: the ESMO oral presentation in October, then earnings October 27. The fourth name in the healthcare sleeve after MDT, DGX and GILD.
Equity Highlight (2026-JAN-05; first flagged Apr 2024) — Jacobs Solutions, a "picks and shovels" play on US reindustrialization + the hyperscaler/data-center capex boom. Record >$23B backlog (+14%, 1.2x book-to-bill), FY26 EPS guide $6.90–7.60, fwd P/E ~17.7–19.6 (~25% below peers) vs a trailing P/E distorted by nonrecurring charges; Value Line 90/100, ~$10 EPS by 2029 → ~$230 3-yr target vs ~$135. Cost-plus discipline (no fixed-price turnkey work) vs peers that took write-downs; ~$1.1B returned (~3% of shares), ~1.1% yield. A prime "run-it-hot" beneficiary.
The bank call already scored (2026-JUL-31; Positive) — JPMorganChase, "glowingly highlighted" with CEO Jamie Dimon in January of 2024 and "since then it's been an outstanding performer, more than doubling from ~$170 to $353." Named first among the mega-banks "that broke out a year or two ago (2024, in the case of JPM and C)" — the multi-year range expansion Haymaker now finds across regional-bank charts after reviewing "literally hundreds" of them. A scorecard entry and the technical template behind the KRE pick rather than a fresh entry point, but the stance on the name is unambiguously positive.
Rotation-not-correction tell (2026-JUL-07; Positive) — the SPDR S&P Regional Banking ETF has completed a multi-year breakout and sits just under five-year resistance; a push above ~80 would be a new all-time high, "an even more positive development." Its strength (with the cumulative A/D line at fresh peaks) is the anti-2008/anti-2022 signal — in those tops both KRE and breadth were rolling over under an index near its highs — so Haymaker reads a major rotation as more probable than a significant correction.
Trading Alert (2026-JUL-11, re-post of the Jul-9 Ringside alert; Positive) — Lockheed Martin, a bullish accumulation call at an "undemanding" 17× earnings vs the S&P's 21× (LMT's earnings not inflated by extraordinary items), after four flat-lined years as defense bears — like the oil bears before them — decided "the shooting war is in the rear-view mirror." Engine: the global Patriot MIM-104 shortage (Ukraine's inventories nearly exhausted; even the US short) locks in strong production "for years to come," atop surging worldwide defense spending few non-US (non-China) producers can serve at scale; the Ultra Maritime acquisition starts the drone/naval next-gen pivot. Another run to a new all-time high north of ~$700 "highly likely." Risk kill-switch: a decisive break above 4.6% on the 10Y / 5% on the 30Y. Accumulate weakness / trim strength (clearest trim Oct-2024 ~$600).
POW pick (2026-JUN-19) — the bond, not the stock: the M 6.7% senior note due 7/15/2034, BB+, at 95.474 for a ~7.45% YTM (~290bp over USTs); BB-rated is Haymaker's "favorite slice" (the <1%-default "BB anomaly"). Asset-covered (CoStar real estate $7.9–10.5B > $6.5B mkt cap, covers $5.1B debt), FCF up double-digits 2yrs (+11% 2026E, ~$1.4B), ~$1.3B cash / ~$3.8B net debt, Berkshire bought ~1% of shares, Change-of-Control LBO protection; return seen beating the S&P over the decade.
New name via Sy Jacobs (2026-JUL-20; Positive) — MediaAlpha, the third QNST/EVER insurance-marketplace peer flagged by Sy Jacobs; up 37.8% since the Mar-3 webinar. Same engine — heavy free-cash-flow generation funding accelerating buybacks. Supporting evidence for the insurance-marketing cohort, not a stand-alone call.
POW pick (2026-JUN-12) — strongest top line in 10 years (FY26 +8.4% rep / +5.8% org) with four platforms inflecting (Affera PFA, Hugo robotics, AiBLE spinal AI, Symplicity) at 13.6× fwd EPS vs peers 22–28×; 12-mo PT $103, 3-yr bull $130–145, bear ~flat + dividend. New Haymaker Strong Buy.
Personal stake, hedged conviction (2023-JUN-05 back-fill; Positive) — private molten-salt-reactor (MSR) start-up where Hay's guest Richard McPherson is a senior executive. "In full disclosure, I'm an investor in one of its subsidiaries" — but "I have no clue if I've bet on the right horse," so he is looking for other ways to play MSRs, which he is "more and more convinced will be the safe and pollution-free energy-step change." Encouraged by its bench of U.S. Navy nuclear-propulsion veterans; hopes the pending patents are "rock solid." Not investable publicly.
Gold miner (2025-FEB-03) — highlighted in the Dec-2 MHM alongside Wesdome; "recovered a much more modest amount," $2.70 → $2.82. An earlier-stage recovery name in the rebounding gold-miner basket.
Portfolio winner (2026-JUN-01) — bought ~$102 (Apr 2) as "one of the most ridiculously mispriced stocks," now ~$135 (+~30%). Q1 confirmed (sub rev +22%, RPO +25%, +14.56% on the print); AI Control Tower launched at Knowledge 2026 as the enterprise-AI governance layer. Thesis confirmed, but after a 65% rally "some gain-harvesting may be in order."
Haymaker POW! Buy, reiterated (2026-AUG-21) — "We are reiterating PALL as a Buy," the physically-backed palladium ETF first recommended April 2024 and now back "in out-of-favor status." The trigger is a three-condition alignment: a 1/3 correction from the January 2026 apex of just over $2,000/oz to ~$1,300 spot, "within what we believe is a long-term uptrend"; a strengthened supply/demand story; and a macro backdrop that "has moved further in our direction" — Bessent's "twist the yield curve" read as "a definitive first step toward YCC" that "ignited a roaring rally in the hard-asset space." Palladium is a "special case" — both a quasi-monetary PGM and an industrially-deficit metal. Demand: 85–90% of global demand is automotive catalytic converters, and the consensus wrongly assumed the ICE→EV transition would be "binary and swift" — instead PHEVs became the dominant growth segment, carry full converters and "often consume more" Pd than a petrol car (Chinese NEVs >60% of monthly sales; BYD's PHEV export mix 11.6%→37%; Chinese PHEV exports to Europe +700% YoY; 21% of Tesla trade-ins going to hybrids). Supply: deficit 14 consecutive years (0.9Moz 2023, 0.5Moz 2024), the forecast surplus repeatedly deferred, Lac des Iles closed mid-2026, recycling down ~700koz (~10% of total supply) as the US fleet aged to a record 12.6 years, and Nornickel ~40% + South Africa ~35% of primary supply with the sanctions exemption "not remotely priced" at $1,300. BofA Q4-2026 target ~$2,200/oz (~70% upside; base case flat to +23%). Risks argued: the 2022 $3,430→−70% burn, EU ICE bans in the 2030s (doubted — ICEs are ~95% of the global fleet), and platinum substitution, "limited" at a Pd/Pt ratio of ~0.70×. Technically below the 200-day but holding the prior breakout with the 2025 downtrend broken. Vehicle chosen for having no counterparty risk; "we're adding to the position." Reaffirmed by reference 2026-SEP-03 — palladium is cited as "a salient example" of the "lengthy list of critical commodities in extremely short supply" that decides the Bloomberg Spot Commodity Index's 2026 high is the next leg of the 2022 bull market rather than a double-top around 740; no new call, target or sizing is issued in that note.
POW pick (2026-APR-10; research lead Daniel Bustamante) — PepsiCo, a defensive "big staple" at a rarely-cheap P/E & P/S, with the Dec-2025 Elliott Investment Management activist campaign as catalyst: aggressive cost cuts (3 plants closed, ~20% US SKU reduction), automation, and potential Quaker-brand divestitures that could lift margins 200–300bps and re-rate the multiple. $10B buyback (~5% of cap) + 3.5% dividend ≈ ~9% cash return; record productivity savings drove +18% core op profit / +16% core EPS; ~$30B FCF, debt covered 11×, Value Line 100/100. Broke its 2023 downtrend, holds the 200-day MA; target +$210 (+33%) into 2027. A staples "harbor" for a volatile, mid-term-election market — buy the corrective pullback.
Upgraded to Strong Buy (2026-AUG-03; Positive) — Sprott Physical Silver Trust, the vaulted-bullion vehicle behind Haymaker's standing "gradually accumulate silver and the silver miners" stance. Bought 08/26/2024 at $9.82 and marked $18.82 (+91.65%), it moves from B to SB on the published Buy List — one of the day's blue-highlighted rating changes. The underwrite is unchanged from the Jun-25 silver post-mortem: very constrained supply against munitions / AI-data-center / grid demand swinging the market into a pronounced deficit, with silver-futures open interest at multi-year lows (capitulation) after the crash from ~$120 to ~$58.
POW pick (2026-FEB-13; idea: Sy Jacobs) — QuinStreet, an atypical small-cap (~$11, ~$600–650M cap) sold off as a presumed AI victim despite a profound turnaround. The mispricing: market treats it as a cyclical lead-gen middleman at ~0.76× sales, but it owns a proprietary insurance comparative-rating platform (QRP, 100+ carrier integrations) that became mission-critical in the 2024–25 auto-insurance crisis (FY-Q2-2025 auto-insurance revenue +615% YoY). FY2025: revenue +78% to $1.1B, Adjusted EBITDA +299% to $81M, ~$74M FCF, zero debt. The $115M HomeBuddy deal (closed Jan 2026) adds high-margin auction-exclusive leads + social-media reach. Management targets 10% EBITDA margins (from ~7.3%) → ~$23–28/share; analyst PTs $22–26 (~56–86%). Risks: Google/Meta disintermediation, customer concentration, cancelable contracts, cyclicality. Average into a small position; a turnaround bet, not a blue chip.
POW pick (2026-MAY-15) — Ryanair (ADR), Europe's ultra-low-cost airline, added back to the recommended list. Fell $73 → $56 on fuel fears "unfairly": hedged 70% of fuel through 2027 at $67/bbl, zero net debt, 15% net margins vs 4% industry; O'Leary's 30-yr cost discipline; 300 new fuel-efficient 737 MAXs next year as Boeing turns. Value cyclicals on P/S — at 2× sales → $70 PT (often hits 3×). "A superior money-maker over the next 12-24 months."
Hold/Trim — moved back off the Buy List 2026-AUG-31 (rated H/T, $59.47 vs the 09/23/2024 lot at $42.00, +41.60%): "harvesting some of the latest appreciation might make sense, and an obvious redeployment candidate is HAL" — but "we wouldn't, by any means, suggest exiting SLB in full. Breaking above $60 could, and should, happen." The business case is unchanged and strong — "the most international of all global oil service companies," a digital arm some call "a software company in drag," +52% in 2025 vs HAL's +30%, through short-term resistance and near a five-year high, and underpriced against its own history (a long-time S&P premium now a slight discount). The trim is relative value, not bearishness, and follows a published 14-year swing ledger ($36 buy 2022 → ~$50 trim → harvests 2023 → sub-10 P/E at $32½ Apr-2025 → $37 → $50 → $57 → $45 → ~$60).
Uranium bull reiterated (2026-SEP-09; Positive) — named again as “one of the few vehicles available for U.S. investors to participate in this nuclear renaissance, other than mostly profitless (for now) U-92 miners,” reached by elimination: the US faces a “mammoth deficit of electricity over the next decade,” China has 36 reactors under construction against “precisely none” in the US, Russia exports 21 designs totalling 40GW (vs the 97GW entire US fleet) and holds 14% of mining / 20% of conversion / 43% of world enrichment — but “publicly traded investment options in the conversion and enrichment cycle are extremely limited,” so the physical trust is the expression (no price, target or sizing given). Earlier (2026-JUL-16) — Sprott Physical Uranium Trust, "the leading uranium ETF," a physical-uranium vehicle held on the Haymaker Buy list (2 lots). Pulled back to ~$18–20 after twice hitting $25 even as spot uranium is ~$85 (long-term contract prices higher). A decade of utilities contracting below consumption (since the 2011 Fukushima overreaction shut 65 plants) has fully depleted excess inventories, just as 70+ new reactors are planned and 16+ mothballed plants restart — so Haymaker "believes prices are poised to surge again… once the utility industry wakes up to the implications of severely deficient supplies at a time of erupting demand."
FXI-constituent evidence (2026-JUL-17; Positive) — Tencent Holdings (ADR), cited in the FXI reiteration for growing net profit 21% YoY through AI integration across social, gaming and enterprise platforms; a dominant Chinese platform monetizing the AI-usage surge and a core FXI communication-services holding. Constituent evidence for the reiterated FXI buy, not a standalone call.
Maintaining Buy (entered ~$84 Jan-2026, now ~$70) — ~$10B FCF +42%, 18× trailing FCF, PEG 0.64, SOTP DCF $110–130; AVs are a tailwind not a threat (Uber's demand network gives AV operators ~30% higher utilization; removing the ~70%-of-fare driver cost lifts take rate, ~$20B incremental rev at 20% AV penetration) and Uber is the "toll booth" across 14+ AV partners via Uber Autonomous Solutions.
Valar (private) — small modular reactor (SMR) maker
SMR milestone (2026-JUL-08; Positive, private) — Valar, Aalo's rival, "notched an even more ambitious milestone" by demonstrating full temperature and power conversion (FTPC) and signing a contract to supply 30 MW of atomic power to an Nvidia (NVDA) data center — "likely a sneak preview" of nuclear entering the data-center power stack. Positioned in small modular reactors (~5× the size of Aalo's micros). Caveat: NRC licensing runs years, so its "laudable and impressive" lead "could well diminish or outright vanish over the course of months-long NRC approval testing." Private, not investable.
Haymaker POW! Buy (2026-AUG-14) — "We are buyers of VST at today's price." The frame: "at its core, the AI story is fundamentally an energy story." Vistra is ~40,000MW of nuclear/gas/solar/battery across ERCOT and PJM (~40 large nuclear plants' worth) plus ~5m retail customers and the nation's second-largest competitive nuclear fleet, at ~$145 — 36% off the $219.82 high on a ~$58B cap. The Q2 revenue "miss" ($4.02B vs $5.46B) is unrealized mark-to-market on forward hedges — "real in the GAAP sense and irrelevant in the economic sense" — and the program causing it is what makes ~100% of 2026 generation hedged; underneath, adjusted EBITDA $1.767B (+30%, a beat) and generation EBITDA +68% at 97%+ availability through record heat. The valuation engine is the exclusions: FY26 guidance of $6.8-7.6B EBITDA / $3.925-4.725B FCF leaves out Cogentrix ($4.7B, ~5,500MW), the Meta nuclear PPAs (~2,600MW PJM), the Helix Platform (KKR + NVIDIA + Kuwait Investment Authority; VST commits up to $1B as preferred power partner) and nuclear PTCs — a 2028 base >$9B annualized, ~$5.4B normalized FCF = an 11.4% yield, at 10.6× EV/EBITDA vs Constellation's 13.6×. Demand is evidenced not projected: record July peak loads of 168GW (PJM) / 91GW (ERCOT). Against the bearish WSJ piece the thesis is "evolving, not breaking" — from passive scarcity rents to who can deliver incremental dispatchable power. Bear case = ERCOT curve softness + Helix execution + PJM IRAS, justified only if all three hit. Caveat: below the 200-day for over a year, stabilizing ~150 — suggested stop at 130. 20 analysts Buy, avg target ~$217.
Fred Hickey's gold pick (2025-FEB-03) — Wesdome: Q4 production +37% YoY (~$130M rev, +77%), modeled ~$0.26 EPS, FY25E $1.02 → "<10× earnings... a ridiculous discount" at $9.90 US. Up ~25% from early Dec, so "be judicious with any new purchases."
Structurally bullish, tactically extended (2026-AUG-19) — the leading energy-producer ETF's "breakout to an all-time high" is read as "confirmation of the long-term uptrend" on a still-disdained sector, but it is "currently quite extended on a near-term basis" and WTI backwardation is back at the ~$10 level that has historically marked oil-price corrections. Caution is "of a modest nature based on the exceptionally supportive fundamental set-up": buy a dip, don't chase the breakout.
Coal add (2026-JUL-20; Positive) — Yancoal Australia, a recommended thermal-coal miner down 23% ("particularly pummeled"). Haymaker blames misguided Strait-of-Hormuz-reopening optimism (thermal coal competes directly with LNG); with LNG "trapped in the Persian Gulf" and bullish Indonesia (top thermal-coal exporter) supply news, the fundamentals are strengthening — urges adding now, "time is of the essence," for those who bought higher. Full story in Friday's POW!
Portfolio Update #288 (2026-JUN-29; Neutral/Hold) — Alcoa, an integrated aluminum major (bauxite → alumina → smelting) up ~50% since the 10/31/25 pick (~$54). A physical-layer electrification play ("like Hudbay in copper"): aluminum at a four-year high on the Strait-of-Hormuz supply disruption (~3.5M tons of 2026 output at risk) plus structural EV/grid/data-center demand. The core tension is "long aluminum, short power" — power-intensive smelting means the AI buildout that lifts aluminum demand lifts its biggest input cost. Fortress balance sheet, but cyclically-elevated/partly-geopolitical peak prices, weak alumina segment, soft cash flow (Q1 FCF −$298M); the cheap P/E is the cyclical mirage — fair ~$52-55 now vs ~$36-42 mid-cycle, the ~$80 Street target pure leverage to aluminum near peak. HOLD sized for cyclicality; add on weakness toward the low-$40s (next print July 15).
Downgraded Buy → HOLD (2026-SEP-14) — moved to Holds/Trims after a -24.6% post-Q2 drop: comps -0.5% (from +3.5%) on a fragile DIY customer, while adj. op margin rose to 5.6% (3.0%), YTD FCF turned positive and leverage fell to 2.1× — but $0.31 of the $1.03 EPS was tariff refunds and the guide raise came from interest income. Overreaction "in magnitude" not direction; turnaround intact, ~12.4× 2026 EPS / ~7× normalized at a 7% margin. Needs DIY to stabilize and margins to hold ex-refunds; a weekly close above $48 the technical trigger. Original case: POW pick (2026-MAY-01) — Advance Auto Parts, a multi-year retail turnaround (CEO Shane O'Kelly, ex-HD Supply) entering the "let's see if it's working" phase. Phase 1: the $1.5B WorldPac sale fixed solvency; Phase 2: 500+ stores + ~200 independents closed, 38→~16 DCs, Market Hubs, first positive full-year comps in 3 years (+1.1% Q4), adj. op margin +200bps to 2.5%. Upcoming earnings the catalyst — wants comp acceleration, margin toward 3.8–4.5% FY26 guide (vs 10–12% history), $160M FCF. Deep discount to AZO/ORLY on still-low margins; on P/S the stock "could easily surge 100%, or more" over 12–24 mo. Run-up then eased $70→$59 — partial position / DCA; an investment, not a trade.
The capital-light era ending (2026-JUL-26; Neutral) — named in passing with Alphabet and Microsoft as one of the "very low capex, very high free cash flow… capital light" businesses with "great network effects" that is "why they were such great stocks for so long." The AI spending era erodes exactly that quality, and he closes the thought pointedly: "it wasn't that long ago they were trading at 10 11 times earnings but maybe someday they'll be back there given what they're doing." A comment on the group's business model, not a rating on Apple.
Held, trim date corrected (2026-AUG-03; Neutral) — Agnico Eagle Mines, the senior gold miner held since 02/10/2025 at $97.04 and marked $146.76 (+51.24%), rated H on the Holds/Trims table. The only change this week is a record correction Haymaker flags explicitly: the position's trim date moves from 07/20/2026 to 07/13/2026 — the sale happened a week earlier than the published table had shown. Housekeeping on the newsletter's public scorecard (the trim date sets the price the sale is credited at), not a fresh stance on the shares.
The super-app benchmark/competitor — Uber's Expedia tie-up is framed as the first move toward something that "competes structurally with Amazon Prime."
Cited as data (2026-SEP-02; Neutral) — Cathie Wood's ARK Innovation ETF, the growth-side line in the relayed manager table: +8.0% YTD 2026, −4.6% vs SPY as of Aug 28, 2026. Paired deliberately with the value managers so the underperformance reads as style-agnostic, which is the note's evidence that the index's edge is a megacap AI concentration no discretionary portfolio could match. No stance on the fund is taken.
Named only as the competitor (2026-SEP-18; Neutral) — Arvinas' rival drug already approved and leading in the hormone-driven metastatic breast-cancer indication where AstraZeneca's camizestrant failed Phase 3; no view on the stock.
Peer benchmark (2026-MAY-01; Neutral) — AutoZone, cited with O'Reilly as the high-margin auto-parts comp: AZO/ORLY run 14–18%+ operating margins vs AAP's 2–4%, the gap AAP must close for its re-rating. Not a call on AZO itself.
POW pick (2026-MAR-13; idea: Dan Bustamante) — Boeing at ~$210, a contrarian turnaround the bearish headlines mask. Thesis is financial: $682B record backlog (all 3 segments at records; 1,173 net commercial orders), Q4 2025 revenue $23.9B (highest since 2018, +57% YoY), and FCF inflected positive at $375M after years of burn — on track for $1–3B (2026), $6.7B (2027 consensus), $10B long-term. New CEO Ortberg cleaned up the balance sheet, restored deliveries (Feb the highest since 2017), fixed FAA relations. Half of an Airbus/Boeing duopoly with no third option at scale, so cancelled orders redistribute (China's came back as a ~500-jet deal). Consensus PT ~$246 (~15%); break of $260 opens $300. Caveat: the Iran war could pressure airline orders — start small or wait.
Breakout-precedent mention (2026-JUL-31; Neutral) — Bank of America, cited only as one of "the mega-banks like JPMorganChase (JPM), Bank of America (BAC) and Citigroup (C) [that] broke out a year or two ago," evidence that the multi-year range expansion now appearing across regional-bank charts already played out in the money-centers. No view taken on the shares.
Passing mention (2026-JUN-19) — Berkshire Hathaway (now run by Greg Abel) bought ~1% of Macy's shares (~$55M); "walking around money" but a "nice vote of confidence" in M's turnaround, cited to support the M bond credit (not a Berkshire call).
Qualified winner, gains partly harvested (2026-JUL-24; Neutral) — Peabody Energy, given "a heavily qualified tout" on 2025-MAR-24 near $13 (equivocal because BTU primarily produces the more environmentally-unfriendly thermal coal). It "went on to triple from there," but "since March it has come down very hard, to the $22 vicinity" — "still a nice gain," with partial gain-harvesting suggested just under $24. Named as the sharpest illustration of the coal group's slump (why YACAF's slide was "not surprising"); no fresh buy call.
The demand evidence, not a call (2026-AUG-21; Neutral) — BYD, "the world's largest auto exporter," is the single company carrying the PHEV argument behind the PALL palladium reiteration: PHEVs grew from 11.6% to over 37% of its total export mix, Chinese NEVs "led heavily by PHEVs" passed 60% of monthly auto sales in 2026 "with BYD at the center of it," and Chinese PHEV exports to Europe grew >700% YoY (Spain/UK/Belgium each up multi-hundred to over 1,000%). Two of the three named drivers are BYD-specific — the EU anti-subsidy tariffs that hit Chinese BEVs but exempt PHEVs "while BYD builds factories in Hungary and Turkey," and the DM 5.0 platform's "over 2,000 km (about 1,200 miles) of combined range." Every exported PHEV carries a full catalytic converter, so the fastest-growing slice of the "EV market" increases palladium demand. No view is taken on BYD's shares — "the China data, exemplified by BYD, continues to accumulate."
Breakout-precedent mention (2026-JUL-31; Neutral) — Citigroup, the third mega-bank in the regional-bank POW!'s breakout retrospective; with JPM it broke out in 2024 ("2024, in the case of JPM and C"), a year or two ahead of the range expansions Haymaker is now finding across regional banks. Context for the technical argument behind KRE, not a call.
Valuation comp (2026-JUN-26; Neutral) — Chubb, cited at ~12.3× NTM / ~1.64x P/B as the peer TRV trades below despite TRV's higher 22.7% core ROE; its international diversification and high-net-worth brand partly explain the premium. A reference for the TRV discount, not an argued call on CB.
Valuation comp (2026-AUG-14; Neutral) — Constellation Energy, the other large competitive US nuclear operator, cited at ~14.5× earnings (restated as 13.6× trailing EV/EBITDA) as the yardstick VST's 10.6× forward EV/EBITDA sits below — and VST's multiple is struck on guidance that excludes Cogentrix, the Meta PPAs and Helix. Used to show the discount is name-specific rather than a sector de-rating; no view is taken on CEG's shares.
Asset reference (2026-APR-10; Neutral) — Celsius Holdings; PEP owns ~11% of Celsius, "a big name in the energy-drink space," cited in PEP's "consumer staple = win" verdict. Not a standalone call on CELH.
Second trim, core kept (2026-SEP-11; Neutral) — Global X Copper Miners ETF, ~doubled from the $46.67 Aug-25-2025 highlight (first trim May-15-2026). "The thesis… was correct" (AI data centers, electrification, reshoring, a decade-long mine lead time, Hormuz-driven grid spending) but "what has changed, in a big way, is the price," and $331M of one-month outflows show institutions "selling into the strength" — trim another 25–33%, keep the core for a breakout above $99.
Update (2026-MAY-15; Neutral) — Coupang, a recent "loser": sales +8% but cash flow disappointed; near 13.5–14 longer-term support, which "it would be best to hold." Still fans of the Amazon-like logistics moat (+ a big insider buy); expects strong profits once it stops over-spending on growth. "Sticking with it but not advising buying more."
The March-2000 tell (2026-JUL-26; Neutral) — cited as the historical analogue, not a rating: at the NASDAQ's March-2000 peak "it was companies like Cisco reporting great earnings and the stocks went down" — precisely what Alphabet did the day of this interview — so "maybe this is the big shakeout we've been looking for with tech." The pattern (good numbers, falling stock) is the signature he watches for a leadership top.
Valuation benchmark (2026-APR-03; Neutral) — Deere & Co at ~30× earnings, paired with Caterpillar as the slower-growing industrial comp NOW now trades below on P/E. Not a call on DE.
DeepSeek (private) — Chinese open-weight AI models
Cheap-model competition (2026-JUL-26; Neutral, private) — Hay agrees ("No, that's true") with the point that Chinese open-weight models are genuine competition: Microsoft has sought government permission to use DeepSeek because it is "so much more affordable than Open AI and Anthropic" now that "token costs have gone through the roof" — the Chinese products "can't quite do as much, but they can do a lot for an awful lot cheaper." Feeds his view that "the whole OpenAI business model… is coming under tremendous suspicion." Private, not investable.
Delivery Hero — Uber's 24.99% stake (explored full takeover) read as a Gulf super-app, not "dumb money": Talabat dominates food delivery across nine GCC countries where Uber has Vision 2030 AV ops.
Illustrative mention (2026-JUL-01; Neutral) — D.R. Horton, the largest US homebuilder, named (with Lennar) only as the case study for how buyer subsidies have "slashed profit margins for the leading homebuilders by one-third to one-half" as new- and existing-home prices converge in a market where a median home costs ~5× median income. Haymaker notes the homebuilder equities/ETF still sit within 10% of an all-time high despite that margin squeeze. Not a rated buy/sell call.
Passing mention (2023-JUN-05 back-fill; Neutral) — Dow Chemical cited only as one of the "major companies" "jumping on the SMR bandwagon"; evidence for the small-modular-reactor theme, no view on the stock.
Core long (2025-FEB-03) — the iShares Brazil ETF as the easy U.S.-investor route into a market "deeply undervalued": Bovespa P/E ~8 (CAPE <8), ~9% dividend yield (> its P/E, "exceedingly rare"), Real ~40% below parity on REER, de-equitizing supply shrinkage; sentiment at GFC intensity. +10% to ~25.30 so expect a pullback, but "a lot more gas left in its tank" — "a once in a one- or two-generation opportunity."
Valuation peer (2026-MAR-27; Neutral) — FEMSA (incl. OXXO), "Walmex's biggest competitor," trades ~22× earnings, the comp that makes Walmex's ~18× look cheap. Not a call on FMX.
Miner apathy is the setup (2026-JUL-26; Positive) — the VanEck Gold Miners ETF, presented off Hay's own fund-flow chart: the largest outflow year on record was last year, after the miners' "tremendous returns" and "really good profits." Big inflows at that point would have been "a big warning sign like it was back in 2016"; record outflows instead "indicate a lot of investor apathy" — his best explanation being that precious-metals investors bailed on the equities early in the year when the metals worked and the miners didn't. Conclusion: "the gold miners and silver miners look pretty interesting," reinforced by a fundamental shift — mining, "very poorly run for a long time," is "starting to emulate the energy industry" on capital discipline. Sits under a structurally bid gold price (central banks buying on tonnage, monetary debasement, no Volcker this time).
Cited as data (2026-SEP-02; Neutral) — David Einhorn's Greenlight Capital Re, the value-side line in the @AskLivermore manager-performance table relayed by Trader Ferg: +8.5% YTD 2026, −4.1% vs SPY as of Aug 28, 2026. Its only role is to show the shortfall spans styles — "both growth- (Cathie Wood) and value-oriented investors (David Einhorn and Bill Ackman) are lagging" — which points the argument at index concentration rather than a style rotation. No view, valuation or recommendation on the shares.
Trim (2026-FEB-20; Neutral) — Alphabet (Google), a past Haymaker favorite written up at ~$140 (Feb-2024) alongside TSM, now ~$315 and "on the very spendy side" on its 5-yr P/S and P/E charts. Already gain-harvested on the way up; cited (with TSM/XOM) as why pervasive over-valuation warrants raising 20–25% cash. A take-some-off, not a sell-out.
Trim/sell after the win (2026-JUN-15) — twice highlighted in 2024 (~$460 in Aug), now ~$1,080 (+130% ex-div); the team now sees "a lot more downside over the next year or so than upside." (Rationale paywalled.)
Portfolio update (2026-JUN-22; Hold → Hold/Trim) — Hudbay Minerals, recommended ~$10.46 last July, now ~$29 (+175%). Three peak-cycle drivers (record copper ~$6.50/lb, a record-gold by-product credit that drove Q1 cash cost negative, and a full re-rating to a tier-one copper growth story) make it a concentrated double-cyclical bet; Price/Sales near a 5-yr high, consensus target only ~$30. Trim to bank the gain, hold a core for the genuine 2029+ deficit, re-add on a pullback (copper $5–5.50/lb or a deeper gold drop).
Met-coal winner recap (2026-JUL-24; Neutral) — Warrior Met Coal, the metallurgical (steelmaking) coal name Haymaker followed Trader Ferg into in January 2025, "a most profitable example" of the coal variety it will own, since met coal avoids the thermal/power-generation pollution objection. It "subsequently soared by ~51%, despite having corrected by 20%." Cited in the coal POW! as evidence of the group's recent weakness (with BTU) rather than as a fresh call — a carried winner and part of the track record behind backing Ferg's Australian coal names.
Profit-taking trim (2026-JAN-05) — Growth-portfolio note: H&P has ~doubled ($15.20 → $30.94) since the June-30 SLB→HP swap, "regaining far more than the hit on SLB"; sell one-third "to capitalize on the massive move." A trim, not a sell-out.
Cited as data, with a jab (2026-SEP-02; Neutral) — Carl Icahn's Icahn Enterprises in the relayed manager table: −9.4% YTD 2026, −22.0% vs SPY as of Aug 28, 2026. The one editorial line is a joke about the holdings, not the units: "the late-in-life dog rescuer, Carl Icahn, apparently has a number of canines in his portfolio." No valuation, target or recommendation follows, so it is recorded Neutral rather than Negative.
The Hugo analogy and the Hugo risk — Hugo sits where Intuitive stood in 2005 pre-$130B, but da Vinci's installed-base/evidence/ecosystem moat is substantial and actively defended.
The bear case's winner (2026-JUL-31; Neutral) — the Invesco KBW Bank ETF, the money-center basket named against Haymaker's own regional-bank pick: "the money-center banks (via funds like KBWB) have crushed the regionals on deregulation and capital-markets strength. Put another way, the regionals are the laggards, and sometimes you're considered a laggard for a reason." The honest counterweight to KRE — the large banks captured the deregulation and trading-revenue benefit first — not a recommendation to own it.
Helix partner (2026-AUG-14; Neutral) — KKR "provides capital and deal sourcing" in the Helix Platform, the strategic data-center infrastructure partnership with Vistra, NVIDIA and the Kuwait Investment Authority that Haymaker calls VST's "long-term game-changer," structured to generate additional contracted capacity commitments as it deploys capital. Cited as evidence of private-equity, chipmaker and sovereign-fund capital assembling around scarce dispatchable generation; no view is taken on KKR's shares.
Valuation benchmark (2026-APR-10; Neutral) — Coca-Cola, PEP's "chief rival," trades ~24× earnings; PEP "often traded at parity with Coke's multiple" historically, so the gap is PEP's re-rating opportunity. Not a call on KO itself.
The peer-miss caveat (2026-JUL-27; Neutral) — Liberty Energy, named only as the near-term risk to the AESI call: the "quasi-competitor… recently reported a disappointing quarter," and "the market seems to be assuming AESI will, too. That's certainly possible but such an outcome now appears to be well-discounted after the price collapse since May." No stance taken on Liberty itself — it is the read-across, not a recommendation.
The Lucid–Nuro partnership is how Uber deploys its own AV supply in San Francisco — evidence Uber owns AV capacity, not just dispatch, rebutting "Waymo kills Uber."
Illustrative mention (2026-JUL-01; Neutral) — Lennar, a leading US homebuilder named alongside D.R. Horton purely to illustrate the housing-affordability thesis: the subsidies builders use to move product (rate buy-downs, incentives) have cut builder margins by one-third to one-half. A passing example, not a stance on the shares.
Valuation/playbook benchmark (2026-JAN-30; Neutral) — MercadoLibre, named with Amazon & Sea as the platform that went through the same capex/margin-inflection a decade after Amazon; trades ~3× sales vs Coupang's ~1×, illustrating how cheap CPNG is at a comparable stage. A peer comp, not a call on MELI.
AI-capex-cut optionality (2026-JUL-26; Neutral) — a hypothetical rather than a call: "perhaps it's going to be Meta is going to go, guys, we're actually going to cut way back on our AI capex — that stock might go straight up," which he thinks might in turn give the other hyperscalers "a reason to, or at least a pause about" continuing to throw money at "something that's highly questionable." Sits inside his broader lament that the megacaps traded away a capital-light, high-free-cash-flow model for industrial-scale capex.
Income winner, qualified trim (2026-SEP-08; Neutral) — an "inflation-hedged income name" (a payout raised at least in line with CPI, so the real return isn't eroded the way a bond's is), flagged in Feb-2024 on the two-legged test: an upside range expansion threatening an eight-year high, plus a "Magnificent Double Nine" — a bit over 9× earnings and a 9% distribution yield. Total return since, with distributions: 93.7% vs 61% for the S&P 500 and 95.7% for the Magnificent Seven, on far higher cash flow and lower risk; units still sit below the 2015 peak of $78 though EPS/unit has more than tripled. Forward: "a lot less," so "trimming a bit might now make sense" — but the decision is framed on tax, not valuation (MLP-in-an-IRA taxable income; recapture of the 60–70% of distributions that were sheltered; the stepped-up basis at death). Still yields over 7% with the payout +9%/yr. Deliberately NOT on the Buy/Hold/Trim/Sell tables, which cover growth rather than income equities.
Partner / context (2026-AUG-07; Neutral) — Merck is Gilead's co-developer on the once-weekly oral HIV pill whose ISLEND-1/ISLEND-2 Phase 3s read out positively at AIDS 2026 in Rio, and the owner of Keytruda, with which Trodelvy won first-line metastatic lung approval. Named only as GILD's counterparty; no view is taken on Merck's shares.
Accounting-quality reference (2026-JUL-15; Neutral) — Microsoft, named as the earnings-augmentation exemplar via Gerard Minack's "earnings are an opinion; cash is a fact" frame, not a rated call: the "massive non-cash gains" hyperscalers book on stakes in loss-makers like OpenAI are "even more aggressive" because those start-ups are "often major customers of companies such as Microsoft" — a circular loop (mark up the stake and book the revenue) that inflates reported profit. A caution on the quality of big-tech earnings, not a view on MSFT shares.
Cited as data (2026-SEP-02; Neutral) — Michael Saylor's MicroStrategy/Strategy, the worst line in the relayed @AskLivermore manager table: −19.0% YTD 2026, −31.6% vs SPY as of Aug 28, 2026. Appears in the table image only and is not mentioned in Hay's prose; it extends the sample to the levered-proxy end, reinforcing that every celebrated approach trailed a broad index year. No view on the shares, the treasury strategy, or Bitcoin is expressed.
Off the Buy List to Hold (2026-AUG-03; Neutral) — M&T Bank, a large US regional lender bought 09/23/2025 at $187.34 and marked $248.58 (+32.69%), moved to the Holds/Trims table rated H as one of the day's blue-highlighted rating changes. The timing is the tell: it lands three days after the Jul-31 POW! made KRE — the equal-weighted ~150-name regional-bank basket — a rated Buy, so the sector call is now carried by the diversified vehicle while the single-name winner that has already run ~33% is demoted to a hold. Consistent with the same post's instruction to raise cash "particularly in grossly inflated highly valued securities."
Reference (2026-MAY-15) — Nasdaq, Inc., cited (via Ferg/Horizon Kinetics) as the proof case for the exchange-operator thesis: NDAQ beat the NASDAQ 100 since its July-2002 listing (index presently ahead, "a matter of timing"). Backs the BOLSY pick, not a fresh call.
"It's trim time" — bank gains, keep the position (2026-AUG-24; Neutral) — the thesis worked and is explicitly intact: the out-of-consensus call that the "EPS estimate of $5 for 2025 was low" landed at "almost $7," and the structural catalyst — Newmont is "the only gold miner in the S&P 500" — drew the institutional bid that "has undoubtedly happened and been a contributing factor in its moonshot." What changed is price. The relative check is measurable: "NEM is back at its early 2025 peak whereas the senior gold miner ETF, GDX, remains about 10% below its high" — a name outrunning its own sector index. Add the shape of the advance — "the straight up nature of the recent recovery run" makes "a non-trivial retracement… highly probable" — and a Price/Sales ratio that "looks a bit stretched" on the same yardstick that says EOG and FANG are cheap. Verdict: "cashing in some of your gains on this one is a prudent move" — a trim, not a sale, with the gold-miner sector view unchanged. History disclosed rather than airbrushed: an "embarrassment for us early on" after the "37% dividend cut in early 2024" cost ~25% of the position's value. (Prior: 2025-FEB-03, "a chicken way to play" the miner rally on a tarnished blue-chip and rock-bottom expectations.)
Aluminum sibling (2026-JUN-29; Neutral, held) — Norsk Hydro (ADR), "essentially the Alcoa of Europe" on low-cost Norwegian hydro power; suggested Dec '25 as a place to recycle AA trim proceeds, up ~23% since (+60% at one point). A held position carried along in the same electrification-aluminum theme as AA.
Tax-loss-package cautionary (2026-FEB-27; Neutral) — Novo Nordisk (ADR); Haymaker "whiffed" by not suggesting a sale after NVO ran ~30% in <60 days; it then "flipped from a 30% gain into a 23% loss," the package's one self-criticized miss and a lesson in taking partial profits. Not a fresh view on NVO.
Structural reference, not a call (2026-SEP-04; Neutral) — Novartis is Incyte's ex-US partner on ruxolitinib (Jakafi/Jakavi): the drug's ~$3.2B annualized run-rate is "shared with Novartis outside the U.S." That split is why the INCY case is underwritten on US net sales, and why the 2028 US compound-patent expiry is the cliff that matters. No view, valuation or recommendation on Novartis shares is expressed.
Peer benchmark (2026-MAY-01; Neutral) — O'Reilly Automotive, the other 14–18%+-margin auto-parts comp AAP is measured against; the margin gap is the re-rating opportunity, not a call on ORLY.
Referenced only (2026-SEP-18; Neutral) — private Pathos AI, AstraZeneca's second partner in its oncology foundation-model effort. Not investable publicly.
Deep-value Brazil resource behemoth (2025-FEB-03) — Petrobras at 4–5× earnings, excluded from the already-cheap Bovespa multiple; David's "soft spot" — "an extremely lucrative holding for clients, and personally, a few years back."
Valuation comp (2026-JUN-26; Neutral) — Progressive, cited at ~12.8× NTM / ~3.92x P/B on an exceptional ~38% ROE and a more stable personal-auto book; the gap to PGR's multiple is what Haymaker argues the market hasn't yet credited TRV for. A reference, not a call on PGR.
Cited as data (2026-SEP-02; Neutral) — Bill Ackman's Pershing Square Holdings (Euronext Amsterdam), the second value-side line in the relayed manager table: −0.8% YTD 2026, −13.4% vs SPY as of Aug 28, 2026. The table's own footnote supplies the caveat Hay lets stand — "Pershing Square NAV return was −7.1% as of Aug 18 per official weekly report" — i.e. the closed-end discount narrowed while the portfolio lost money, so the quoted unit return understates the underlying loss. No stance is taken.
Watch, don't chase (2026-AUG-21; Neutral) — Sibanye Stillwater, "the world's largest palladium producer," is the leveraged equity expression of the same PALL thesis and the endorsement that was drafted and then pulled by the tape: "we were intending to include an endorsement… which we previously brought to your attention as a high-risk play on this critical metal in May 2024. (We suggested profit-taking on it last December.) However, as we were preparing to write this note, it began a ripping rally and is now up 15% this week alone. We will keep an eye on it and advise if it settles back." No Buy is issued and the revisit trigger is a pullback, not further strength; the rated call went to the unleveraged physically-backed ETF instead, on the "no counterparty risk" argument.
Valuation benchmark (2026-JAN-30; Neutral) — Sea Limited, named with Amazon & MercadoLibre as the comparable-stage platforms whose EV/sales multiples dwarf Coupang's, underscoring the CPNG discount. A peer comp, not a call on SE.
The index-graduation derating (2026-AUG-17; Neutral) — Super Micro Computer, Haymaker's own 2024 example of the mid-cap style "catch[ing] companies as their growth accelerates." The prediction was that its explosive market-cap increase would get it "pulled up to the S&P 500 out of the Mid-Cap index" — which happened "almost as soon as we hit 'Send'" — and from that "graduation" it has derated "like by nearly 70%" (10:1 split; adjusted peak ~$114). Negligible for any one name, but "multiplied by dozens, if not hundreds, of these events" it is the structural drag Hay cites behind mid-caps' 21st-century outperformance of the S&P. A worked example, not a call; no view taken on the shares today.
The curve mechanism, proved offshore (2026-JUL-31; Neutral) — Sumitomo Mitsui Financial Group (ADR), whose five-year chart is the first exhibit in the regional-bank POW!: "bank stocks love steep yield curves; the extraordinary performance of Japanese banks since 2022 is a vivid example of that phenomenon. (Japan has the steepest yield curve in the developed world.) Who said you can't make a lot of money on banks?" It doubles as the chart precedent — "the Japanese banks generated a terrific breakout signal in early 2023," the same signal Haymaker reads on KRE now. Out-of-sample evidence for the KRE thesis, not a call on the shares.
Referenced only (2026-SEP-18; Neutral) — Tempus AI named as AstraZeneca's partner (with Pathos) building "the largest multimodal foundation model in oncology"; evidence for the AI-accelerates-drug-discovery view, no view on the stock.
Reference / catalyst (2026-APR-24; Neutral, private) — "the old nemesis of this newsletter" (still no audit), but its $150M GOLD share purchase (~10% stake) is "perhaps the most bullish recent development" for GOLD: Tether seeks hard-asset backing and plans a US gold-backed stablecoin, making it a future big customer of GOLD's storage/logistics. The world's largest non-central-bank holder of gold bullion. Gold's volatility still strikes Haymaker as "improper collateral for a stablecoin… but that's Tether's problem, not GOLD's."
Haymaker POW! Buy (2026-JUN-26) — P&C compounder repriced to rate adequacy (22.7% trailing core ROE, 85.3% underlying combined ratio); a $106B fixed-income float earns a rising tail in a higher-for-longer world ($3.3B after-tax NII 2026E), ~20% annualized capital-return yield (14th straight dividend hike), at ~10.8× NTM vs peers Chubb 12.3× / Progressive 12.8× despite a higher ROE — re-rate toward ~13× the upside. Barclays Underweight (Jun-12) the bear case; a long-term hold, not a trade.
Profit-taking case study (2026-FEB-20; Neutral/trim) — Taiwan Semiconductor, featured not as a fresh buy but to teach disciplined incremental trimming: first flagged ~$84 (Aug 2022), bought the $70 dip, trimmed 25%-at-a-time up to ~$303 (a "four-bagger"; DCA-out sim 215% on the original $12,500, ~46% IRR). Verdict today: "no longer the bargain it once was" — P/E ~20 is okay but 13× sales "leaves scant room for disappointment." Dominant in precision wafer-etching for Nvidia's AI chips, but richly valued: exit the rest or do another partial sale (the "elderly Haymaker" keeps a modest position). A trim/hold, not an add.
Bullish-oil positioning signal (2026-JUN-30; Positive) — the United States Oil Fund, the main oil ETF, cited for carrying the "largest short position ever" — read as contrarian-bullish (per John Kemp, those shorts "will eventually have to be repurchased, creating a considerable reservoir of potential future buying and upward pressure on prices"). Layered on WTI open interest back at its COVID-nadir low into the worst-ever supply shortage, crude stuck ~$70 despite a mammoth inventory drawdown, record futures bearishness, and SPR-replenishment demand (large nations + Pakistan) — Haymaker expects an oil-price spike in the months ahead. A positioning read, not an explicit "buy USO" trade call. (USO is also carried as a Haymaker Trims/Holds portfolio lot.)
Peer-set mention (2025-FEB-03) — named with Petrobras as one of Brazil's two natural-resource behemoths at 4–5× earnings (via PauloMacro), cheap but not singled out as a hard endorsement.
The "Waymo kills Uber" bear case rebutted — 500k weekly rides vs Uber's ~24M daily trips (>300:1; ~170:1 even at Waymo's 1M target), and 30% higher utilization dispatched via Uber vs standalone; a lead AV partner, not a killer.
Parent / valuation reference (2026-MAR-27; Neutral) — Walmart Inc, the "slower growing parent (at least longer term)," trades near 1.4× sales vs Walmex's ~1×, framing the relative cheapness of the Mexican franchise. Not a standalone call.
WeRide — Uber's robotaxi-ops partner: launched Spain's first commercial robotaxi pilot in Madrid and runs the live Riyadh/Abu Dhabi ops; a flagship AV-partner relationship.
The other half of the "dynamic duo" (2026-JUL-29; Negative) — SK Hynix. With Samsung it effectively is the KOSPI's semiconductor sector, and it is down 40% in five weeks "despite ballistic earnings, leading to modest P/Es." Haymaker's transmission channel from the Korean blow-off — which by June 18th "had generated more return in the past eight and half months than… the prior 45 years" — into the US semiconductor complex: "this is also when the U.S. Semiconductor Index, the SOXX, began its descent."
Half the two-stock engine of the KOSPI blow-off (2026-JUL-29; Negative) — Samsung Electronics. "The main driver of this KOSPI blow off was its semiconductor sector. In reality, that is made up of essentially two stocks: Samsung and Hynix" — and "despite ballistic earnings, leading to modest P/Es, this dynamic duo has plunged by 40% over the last five weeks." Cited as proof that a low multiple on peak-cycle earnings is no floor once a mania unwinds; the same five weeks mark the start of the SOXX's descent, making Korea the leading indicator for the US semis rather than a separate story.
SOLD IN FULL (2026-SEP-08; Negative) — the May-8 POW pick at ~$178 is closed at $178.26 for a −0.03% round trip, "one of our rare full sells." It fell to $125 within six weeks, and the post-mortem names three errors: treating the prior fall from $385 as a floor, ignoring the decisive break below $250 ("violations of support levels are as meaningful as breaking above multi-year resistance"), and not averaging in at nine times earnings. The exit is a business call made against the valuation — the stock "remains a very cheap stock" but "our confidence level in it retaining that hallowed status is eroding" after an FT report (cross-checked with Gemini) that customers are "demanding price concessions to stick with ACN": pricing-power erosion, not defections. Doubles as pain-free cash-raising into hurricane season and a "messy in the extreme" mid-term election.
Valuation benchmark (2026-APR-03; Neutral) — Caterpillar at ~30× earnings, cited (with Deere) as a slower-growing industrial that NOW now trades below on P/E despite NOW's far higher growth profile. Not a call on CAT.
Second trim on rising deal risk (2026-SEP-11; Negative) — CK Hutchison, bought $6.78 (Nov-2025) for the sum-of-parts unlock from selling ~43 ports to a BlackRock-led consortium at $22.8B EV; trimmed at $8.95 (Jul-20). Flat since, while Panama voided the Balboa/Cristobal concessions and Cosco + MSC joined the buyers — complexity "increased rather than decreased," a 259% premium to some fair-value estimates may be "overpricing deal certainty"; the $1.5B arbitration is "multi-year, not a near-term catalyst." Trim and monitor the deal.
Recap / trim (2026-FEB-27; Negative) — iShares MSCI Japan ETF; "another trim is in order" — the yen is extremely undervalued but "the Japanese market itself appears equally extended to the upside." Still bullish Asia, just taking some off.
Moved to a Sell (2026-JUN-08) and off the Hold/Trim list — up ~50% since the Jan-29-2024 highlight, now its highest valuation in a decade (P/S & P/E loftiest in 25 years); raising cash on "ominous" market signals + concentrating on best ideas; ~22× 2026E vs UBER's 18.6×.
Margins from scarcity, not innovation (2026-JUL-26; Negative) — Micron looks cheap on the wrong metric: its P/E is "very low, single digits" while price-to-sales is "extremely high because their margins are just off the charts" — so the earnings are a peak, not a run-rate. Crucially the margins come from shortage rather than product: "they haven't changed the product really at all." With "a lot of overordering going on in the semiconductor world — that's always been the case in that industry," a merely slower AI-data-center buildout (his base case: still growing, "but at a much reduced rate") turns "those boom earnings… to bust" and the stocks "just hit air pockets." Semis are "about as cyclical as an industry can get… especially RAMs."
Loss-maker exemplar (2026-JUL-15; Neutral, private) — OpenAI, cited by Haymaker (via Gerard Minack) as the poster child for inflated US profits: companies book "massive non-cash gains on investments in loss-generating entities like Open AI" even though its "red ink… is astronomical," and the treatment is "even more aggressive" because OpenAI and peers are "often major customers of companies such as Microsoft." Several such loss-makers are "vectoring to go public" (future stock supply). Private, not investable; an illustration of the earnings-quality problem.
The AI-credit canary (2026-JUL-26; Negative) — "Oracle bonds are trading almost like junk," which Hay calls hard to believe for a company that "at one point got up close to a trillion dollar market cap" — but it "really leveraged up… trashed balance sheet. The market's kind of unforgiving." He treats the credit market as the early-warning system for the AI buildout's financing leg (IG spreads are tight and give "no warning at all," so the signal has to come from single names, private credit and the BB-vs-CCC spread). Read alongside his symmetry rule, drawn from Oracle adding ~$200B of market cap in a single day on its 2025 AI-spend announcement: "when you see that kind of stuff on the upside, almost invariably you're going to get the payback on the downside."
Update / full exit (2026-MAY-15; Negative) — Sprott, Inc., a "monster winner" (Sy Jacobs' idea), $32 (Dec-2023) → $138, a four-bagger in <2½ yrs; now "quite spendy" (9× P/S, 29× P/E even after a 14% correction). Sy himself worries about a give-back — "one of our rare suggestions to exit it entirely" (mind the tax bill). (First referenced 2026-APR-24 as the foil to AMRK/GOLD.)
Take a big knife to the semis; sell the bounce (2026-JUL-29; Negative) — the iShares Semiconductor ETF began its descent in the same five weeks Samsung and SK Hynix fell 40%, "one that is becoming increasingly sickening for its formerly jubilant holders." "No slice of the S&P 500/NASDAQ was more adored than the semis were this year," and the Korean pair shows cheapness on real earnings is no defence. Tactical instruction: the sell-off is intense enough that "a bounce before long is possible, even probable" — and that bounce "will likely be an opportunity to once again reduce exposure for disciplined investors," because "the bigger the bubble, the bigger the bust."
Negative, restated on flows (2026-SEP-02) — the earlier "Bubble 4.0" narrow-breadth case is explicitly not the argument any more: the S&P is "up roughly 12 ½% this year on a total return basis," the advance is "broader than in the past" and "the equally weighted S&P index is actually slightly ahead." The bear case is now relative performance plus flow. Per Gerard Minack, "the U.S. market itself has been lagging overseas indexes since the start of 2025" — called "surprising" given the "mammoth inflows into U.S. equities due to the AI boom/mania" and Hartnett/BofA tech-fund flows on pace for a record ~$216bn in 2026. Underperforming while receiving the largest inflow on record means the inflow is the only prop, so "already underperforming U.S. shares will trail overseas markets to an even greater extent should — we'd argue, when — these epic inflows shift into outflows," with the damage landing on "the fully invested U.S. index funds which are also highly concentrated in AI-related shares" (no cash buffer, cap-weighted forced selling into the most crowded names). Expressed as a rotation, not a short: Americans are "dramatically underweight foreign markets" and "the long-overdue rotation into international equities is well underway." (Earlier framing: ~50% tech, dangerously narrow breadth, vicious-unwind risk.) Reinforced 2026-SEP-03 from the other side of the same trade — on the FactSet/Jefferies ratio of the GSCI Commodity Index to the S&P 500 (1970–2026, relayed by ZeroHedge), commodities sit near ~0.7× the index: below the 1999 Tech Bubble trough and the low of the entire 56-year series, against peaks of ~9.5 (Gulf War), ~8.3 (2008) and ~7.6 (oil embargo). Stated as an identity rather than a forecast: "it's highly probable that either stocks need to fall precipitously or commodities need to continue their ascent, or some combination of both." The direction is deliberately left open there; the Sep-02 flow argument supplies it.
Tactical only — long-bond rallies are "to be sold"; structurally negative, prefers shorter treasuries. Sep-14: TLT's 10-yr chart "clearly illustrates the price breakdown" to new lows as long yields break out globally (30-yr UST 5.33%; France near a sovereign debt crisis) — one of four reasons to raise cash.
Losing owners to hybrids (2026-AUG-21; Negative) — cited in the PALL palladium reiteration as the demand datapoint that inverts the pure-BEV case: "PHEVs are taking market share from Tesla. As The Financial Times reported last week, 21% TSLA owners are trading in their 'Musk-mobiles' for hybrids" — i.e. defecting to vehicles that carry a full catalytic converter and "often consume more" palladium than a conventional ICE. No target, valuation or recommendation is given on the shares; the reference is to share loss in the segment Tesla defines. (Earlier framing, 2026-JUN-05: the counter-example — its IPO looked overvalued too, yet became one of tech's biggest gains.)
Cautionary (2026-JUN-02) — a former Nifty-Fifty "invulnerable" great that "now appears to be vectoring toward Chapter 11," the live example of the dangerous one-decision "never sell" myth.
In one line: the call remains the great rotation, not the great meltdown — out of the ~45%-of-market-cap AI complex and into energy (oil equities, cheap US gas, coal), uranium/nuclear, precious metals & miners, regional banks (the Jul-31 POW! pick, KRE, on a confirmed curve-driven NIM inflection), deep value and international markets — but as of Aug-3 two macro shocks have landed at once and the stance hardens from rotation to "raise cash, particularly in grossly inflated highly valued securities": the 30-year Treasury yield broke out to 5.27%, an 18-year high, with only the 2003-era ~5.5% wall before open air (the 1999 template: extreme valuation is the vulnerability, the rate spike the coup de grâce), and Friday's BOJ yen intervention was joined by the U.S. Treasury, sparking a 4%+1.4% rally in a currency whose cheapness has financed "trillions of overseas investments" — a carry-unwind/repatriation channel "off almost everyone's radar." On Price/Sales — the house's preferred yardstick, because sales are less volatile and much harder to fake — the S&P is more overvalued than at the 2000 peak, with earnings flattered by "accounting… bordering on circular financing" and Mag-7 margins now carrying $700B→$1T of AI capex. The tech trade's "greatly overdue" reality check is underway with the semis the epicentre (the KOSPI's two-stock engine −40% in five weeks despite ballistic earnings; SOXX and MU following; sell the probable bounce). The energy leg is unchanged and still the buy side of the rotation: as of Aug-4, ceasefire-driven dips in crude are to be bought — the Hormuz chokehold is held by the IRGC, not by the negotiators; inventories are drawing down even in a self-sufficient U.S. ("diminishing at an alarming rate," per John Kemp); and the shortfall nobody is watching is in refined products — jet fuel and gasoline stocks drawing just as hard, invisible only because the data is hard to get, "at least until airports and gas stations… start running short on fuel." Aug-5 hardens that into the downstream leg: John Kemp's U.S. "Big Three" composite is significantly drawn down because refiners prioritised distillates "at the expense of gasoline" with no spare capacity, record refined exports falsify the demand-destruction story on volumes, California (LA + Benicia gone, ~20% of gasoline imported, Chevron vectoring to close its last two in-state refineries) faces Asian suppliers short themselves — "the possibility of 1970s-like gas lines" — and the complex running flat out has put refinery margins essentially at their 21st-century high, exploding CVX's profits "despite that oil prices have barely risen over the last 20 years (and, adjusted for inflation, are down by roughly 50%)." Aug-6 puts the number on the earnings-quality charge — S&P earnings are "the strongest since 1955," but ~75% of Alphabet/GOOG's first-half 2026 profits came from marking up its stakes in Anthropic and SpaceX, and per David Rosenberg the ex-energy, ex-AI profit increase would be "a much more pedestrian 4%"; earnings spikes have always been followed by "equally dramatic declines," so "at least some mean reversion is to be realistically expected" (Dimon's "this time is different" and Fisher's Oct-15-1929 "permanently higher plateau" as the two warnings). And the buy side of the rotation gets a healthcare destination on Aug-7: GILD an explicit Buy at ~$133, where a $10.5B GAAP loss is "an accounting event, not a business event" masking ~$14B of adjusted operating income, a 7.6% forward FCF yield and a patent-protected HIV cash engine funding a first-line oncology build-out — at 12-13× against an S&P at 20×, resting on its 200-day: "the pause that refreshes." And on Aug-11 the energy leg gets its own single-name expression, in the commodity he has called his highest-conviction long since Jun-11: natural gas is "America's #1 source of electricity generation (at 43% of the total… nearly double all renewables combined)" and the only thing keeping a data center from being an "inert shell" — yet bullishness is "close to the lowest it's been this decade, outside of Covid" and the real price is in the 18th percentile back to 2010, "as though natural gas is being valued as little more than a scarcely used bridge to nowhere." The trade is the plumbing: ~15 Bcf/d of new Permian takeaway under construction (~12½% of U.S. marketed output) turns flared, negative-priced associated gas into sold gas — "a notable earnings kicker" — with EOG "likely to be one of the main beneficiaries," on the verge of breaking a five-year range. Then on Aug-12 the precious-metals leg gets the same treatment, borrowed rather than written — Haymaker republishes and endorses Kevin Muir's gold piece ("we agree strongly enough to broadcast the main thesis here"): the late-2025 mania has broken, "China has quietly begun buying again" into the pessimism, and, decisively, gold "refused to break below 4000" all summer while the dollar and real interest rates were both spiking higher — "gold was telling you that it was being accumulated" — before a break of the 50-day/downtrend forced shorts to cover and took it "350 dollars higher." Apathy is measured in real money (the 1-yr 25-delta gold call skew at its lowest since before COVID — "no one is paying up for right-tail risk") and the miners at "11x" carry a decline that is "already in the earnings," so stabilisation alone bottoms EPS. Haymaker's own marks: the Jul-27 AGI Buy alert "popped about 15%," and GDX is an accumulation candidate on any pull-back — a reinforcement of the standing gold stance with an explicit chase warning ("both bullion and the miners have had a snappy rally of late"). As of Aug-17 the overvaluation charge gets its cleanest formulation and its relative-value destination: the S&P's 50% total return since March 2024 came entirely from earnings — its forward P/E actually fell21× → 20× — and those earnings lean on "companies such as GOOG reporting massive gains on their holdings in AI entities like Anthropic," which "represented more than 50% of its profits in the first two quarters of this year," an "immense, though unsustainable, earnings booster rocket." Hence "a looming mean reversion in the S&P 500's profit margins" and the summary line: "on a normalized profit-margin basis, the S&P 500 is far more overpriced than it superficially appears" — with margin debt "having gone vertical" named as the most serious of "an unusually large number of threats." The beneficiary is mid-caps (IJH), which never had the booster and which structurally sell their parabolic winners into S&P "graduation" (SMCI −70% post-promotion, Tesla the bigger case) — "over the next 10 years, we are adamant that Mid-Caps will dramatically outperform," though "for risk wary types, we wouldn't quibble with booking some gains" near term. Aug-19 puts a timing flag — not a rating change — on the energy leg, and does it by deliberately publishing evidence against the house book ("because we have a decidedly pro-energy outlook… we have to guard against only relaying data which backs up our bullish views"): persistent WTI backwardation has stopped being informative as a state ("for years now… the prevailing condition"), but its degree still marks tops — the 2022 Ukraine-invasion spike to a $30+ spot premium over the 12-month contract and March's Iran-War spike both preceded a "cliff dive," and the ordinary, less extreme peaks show that "whenever the backwardation hit $10 the oil market corrected, at least to a degree." It sits "right around $10" now, and XLE's breakout to an all-time high — read as "confirmation of the long-term uptrend" — is simultaneously "quite extended on a near-term basis." The conclusion is a buy plan rather than a trim: caution "of a modest nature based on the exceptionally supportive fundamental set-up… don't be surprised if oil dips a bit but, if it does, prepare to be on the buy side." And Aug-21 adds the hard-asset leg's macro trigger and its least-crowded expression: "we are entering what we believe is a new era of YCC" — governments "suppressing interest rates to manage sovereign debt costs (explicitly as Japan has done, implicitly as other OECD sovereigns are doing)" — with Bessent's announcement to "twist the yield curve," selling short-term Treasuries to buy longer-term issues, having "ignited a roaring rally in the hard-asset space" and read as "a definitive first step toward YCC." The allocation conclusion reaches past the obvious: "precious metals are the obvious beneficiary, but industrial metals also benefit as hard assets, often with highly favorable supply-demand characteristics of their own." The expression is a re-entry rather than a new idea — PALL reiterated as a Buy (originally April 2024) after "a 1/3 correction from the January 2026 apex of just over $2,000/ounce" to ~$1,300, on the three-condition test that defines the house's re-entry discipline: corrected price, strengthened fundamentals, and a macro backdrop moved further in its favour. Palladium is "a special case" for being both a quasi-monetary PGM and an industrially-deficit metal: 85-90% of demand is catalytic converters, and the consensus "assumed the internal combustion engine-to-EV transition would be binary and swift" when in fact PHEVs — full converters, "often consume more" Pd than a petrol car — became the growth segment (Chinese NEVs >60% of monthly sales, BYD's PHEV export mix 11.6% → 37%, Chinese PHEV exports to Europe +700% YoY, and 21% of Tesla trade-ins going to hybrids), an error "exacerbated by the confusing manner in which many sources report EV sales by including plug-in hybrids in the total." Supply is in deficit "for 14 consecutive years" (0.9Moz 2023, 0.5Moz 2024) with the forecast surplus "repeatedly pushed back," Lac des Iles closed mid-2026, recycling down ~700koz (~10% of total supply) as the US fleet aged to a record 12.6 years, and Nornickel ~40% + South Africa ~35% of primary supply behind a sanctions exemption "not remotely priced into the current $1,300 spot price." The discipline is visible in what was withheld: the drafted endorsement of SBSW, the world's largest producer, was pulled because "as we were preparing to write this note, it began a ripping rally and is now up 15% this week alone" — watch, don't chase. And Aug-28 extends the power leg to the equipment layer with GNRC, written with contributor Daniel Bustamante: "Generac is no longer simply a residential generator company dependent on hurricanes and power outages" but a C&I/data-center supplier sitting on the receiving side of the capex the market frets about — "while others are worrying about CapEx spend on the hyperscalers Generac has been, and will continue to be, a net benefactor of said spend," since "every large-scale data center requires multi-megawatt backup generation capable of carrying the full facility load through grid outages." Q2 C&I sales +29% to $556M after +28% in Q1, an approximately $1.6B data-center backlog ($1.35B scheduled for 2027, excluding a second hyperscaler's committed volume) growing at "roughly $11 million per day," and C&I now ~47% of external sales vs ~41% — a mix shift that earns the multiple, because "infrastructure businesses with visible backlogs and secular growth generally receive higher and more durable valuation multiples." Residential is demoted to optionality (a 34.7% margin of which ~9pp was tariff refunds; a probable "Super" El Niño the free call), and after a +90% run and a one-third correction the stock rests on the 200-day and the 50% Fibonacci retracement at ~18.2× forward earnings — "an intriguing play near the $200 area." And Sep-8 puts the cash-raising stance into practice on the sell side rather than in the abstract: ACN is closed in full at a flat price — "one of our rare full sells" — with the reason a change in the business, not the drawdown ("it remains a very cheap stock… but our confidence level in it retaining that hallowed status is eroding," after an FT report that "a large number [of customers] apparently are demanding price concessions to stick with ACN" — pricing power, not defections), and the timing explicitly calendar-driven: "we continue to worry the market overall is facing intensifying head-winds… we're entering hurricane season. Market history is rife with examples that this is not just a meteorological phenomenon," plus "an upcoming mid-term election that has the very real potential to be messy in the extreme… raise some cash while the raising's good." The same issue restates the entry test the whole book is built on — an upside range expansion above a multi-year ceiling plus a cash valuation, "almost always a winning combination" — via MPLX's "Magnificent Double Nine" (9× earnings, 9% yield, Feb-2024) and its 93.7% total return against 61% for the S&P 500, and its symmetric half, the support-break rule the ACN post-mortem admits was ignored: "violations of support levels are as meaningful as breaking above multi-year resistance." Sep-11 applies the same stance to two winners instead of a loser, with no new name: "a big part of our value-add… is ensuring that you take gains." COPX gets a second trim of 25–33% with the core kept — the copper-deficit thesis "was correct," but "what has changed, in a big way, is the price" after a double, with $331M of one-month outflows read as institutions "selling into the strength" — and CKHUY is trimmed again because its price is flat since the July trim while the Panama/consortium deal risk "increased rather than decreased." Sep-14 turns it into a general instruction with a sourcing rule — all but the most risk-tolerant "should consider raising cash. This is NOT a call to exit the stock market," drawing first on "weaker names (like those rated Hold/Trim, Trim of Sell)" and selling inside IRAs/401(k)s — on a "confluence of negatives": diesel over $200/bbl (double crude), the Houthi strike on Saudi Arabia's East-West Pipeline leaving most of the pre-war 20 mb/d Gulf flow off-line with stocks near a record, an AI backlash, and a global long-bond breakdown (TLT new lows, 30-yr UST 5.33%, France near a sovereign debt crisis); AAP is cut to Hold in the same issue. Sep-15 adds the historical warning: the 10-year up tenfold (0.5% → 5%) while the S&P returned 16%/yr is the widest stock-over-bond gap since 1926 — the 1986–87 and late-1990s analogues both ended badly — so "cash is definitely not trash."
Crosscurrents. Depressed sentiment, sluggish big-ticket sales, surging energy prices, decelerating tax receipts and pessimistic CEOs — set against AI data-center spending and a >$2T deficit (+$365B) plus planned military spending.
The stimulants fade. ~Half of planned data centers are delayed/cancelled on power shortages, and rising interest outlays + a 10-yr Treasury threatening >5% start to bite.
Productivity is the durable story. Per JPM's Cembalest, Real GDP & Gross Domestic Output have accelerated since ChatGPT (late 2022) — vital with US labor-force growth absent, and "in its early days."
The June-5 view. A severe, under-priced energy supply shock (Hormuz; $120–180 oil risk) → bullish energy, especially cheap US natural gas and coal; an over-concentrated "bubble 4.0" market facing a ~$4.5T mega-IPO supply wave → rotate toward energy, financials, fertilizers, EM and small/mid-caps.
Healthcare rotation (Jun-12). Back-to-back healthcare POW picks (DGX, then MDT) — a deliberate signal that "healthcare as a whole looks undervalued and we're seeing a broad rotation into the space"; the MDT case is the de-rated-giant template (13.6× fwd EPS vs peers 22–28× just as organic growth inflects). The sleeve has since grown to five POW picks — GILD (Aug-7), INCY (Sep-4, a base case plus a free KRAS option) and AZN (Sep-18, a de-rated mega-pipeline at ~14.9× forward bought on its trial hit rate, with AI pitched as an underappreciated drug-discovery accelerant for big pharma).
Into BB-rated credit (Jun-19). The picks now extend past equities into corporate bonds: the Jun-19 POW! is the Macy's (M) 6.7% senior note due 7/15/2034 — BB+, at 95.474 for a ~7.45% YTM (~290bp over USTs). BB-rated bonds are Haymaker's "favorite slice" of credit (sub-1% historic default rate yet high-yield coupons — the "BB anomaly"); the M credit is asset-covered (CoStar real estate $7.9–10.5B > $6.5B market cap, covers $5.1B debt), FCF up double-digits two years (+11% 2026E), with Berkshire's ~1% share stake and a Change-of-Control LBO put. Haymaker believes the bond's return can beat the S&P 500 over the rest of the decade — a low-risk way to lock ~7½% for years.
Market-top alarms sharpen (Jun-23/24). Two back-to-back Dailies hardened the "book gains, hold cash" stance with valuation/leverage gauges: margin debt relative to nominal GDP has made a new all-time high and "looks extremely extended" (Jun-23, via The Felder Report) — prior peaks closely tracked market tops — and the NASDAQ's market-cap-to-GDP ratio now sits far above its 2000 dot-com peak (Jun-24, via David Stockman). Combined with trillion-dollar-IPO insider selling and narrowing leadership, it "further justifies" raising cash. The one hopeful note: the Fed's new chairman seems "determined to break the cycle" of enabling asset inflation (a regime change from the asset-supporting Fed).
Silver — buy the washout (Jun-25). Post-mortem on the silver round-trip: the Jan-29 Daily called the vertical move a blow-off top, and silver crashed from ~$120 to ~$58 (below the 200-day, "one sick chart"), with Hay gain-harvesting along the way via futures (bought 2,000 oz ~$69, sold half ~$79) and trimming First Majestic (AG) (the silver miner, +~170% at peak) on the spikes. But "the fundamentals for silver are extremely encouraging" — constrained supply vs. munitions/AI-data-center/grid demand swinging the market into a pronounced deficit, and silver-futures open interest at a multi-year low (capitulation) — so he'd now gradually accumulate the metal and AG despite the poor chart. Consistent with the structurally bullish precious-metals view (SLV/GLD).
Oil — buy the positioning washout (Jun-30). A bullish-oil Daily built on three sentiment/positioning extremes all pointing up: WTI futures open interest is back at its COVID-nadir low "given this is the worst supply shortage ever seen," with crude "plummeted to $70 despite the mammoth inventory drawdown that has occurred this year"; futures speculators are near record bearish; and there's the largest short position ever on the main oil ETF, USO. Every prior open-interest washout preceded a big rally (~$18→$90 after 2020, $70→$90 after 2022), and (per analyst John Kemp) the record shorts "will eventually have to be repurchased, creating a considerable reservoir of potential future buying" — colliding with large nations (and smaller ones like Pakistan) needing to replenish/establish strategic petroleum reserves. Net: expect an oil-price spike in the months ahead. Reinforces the standing energy-bullish stance (Jun-11/18 crude-rally calls, DCA-energy on ceasefire weakness).
The "AI casualties" are AI beneficiaries — hold NOW + ACN as a pair (Jul-6). A Portfolio Update reframes two beaten-down held names the market treated as AI roadkill. ServiceNow (NOW) and Accenture (ACN) launched a joint agentic-AI security offering on June 26 (ACN's managed security built on the NOW AI Platform + a migration tool onto agentic AI) — "a blatant rejection of the bear case on both names at once," with the division of labor (NOW = the governed platform agents run on, ACN = the deployment partner) exactly the split the original bull case rested on. NOW (~$105, ~51% off its high) stays Buy / Accumulate: fundamentals never fit the "SaaSpocalypse" story (~20% subscription growth, ~32% op margin, $4.6B FCF +34%, $30B-revenue-by-2030 path) and a thickening integrator ecosystem (IBM, Cognizant, Google/HCLTech, Accenture) is making it the agentic-AI "control tower"; consensus Strong Buy (~$141 target). ACN moves from "conditional starter" to Hold / Accumulate after a solid June-18 "prove-it" Q3 (op margin +20bps to 17.0%, EPS +9% to $3.80, FY FCF $10.8-11.5B, ~13% FCF yield at ~12-13× fwd; OpenAI's first "AI Transformation Partner of the Year," Anthropic's most-certified partner) — accumulate patiently since organic growth is still low-single-digit and bookings slightly soft. Also a Petrobras (PBR) quick-hitter: recommended ~$12.70 (Jun 2025), +75% then trimmed near $20 in March, now ~$16.45 — a modest buy… or buyback as a play on a snap-back in grossly oversold oil (Q1 EPS beat on record Búzios/Tupi output, ~10% yield, ~$23 target), not yet a table-pounding buy.
Salvage-auction compounder POW (Jul-10). The latest Friday POW! is Copart (CPRT) — pick and core analysis by contributor Daniel Bustamante — a Positive/Buy for long-term compounders. The leading global online salvage-vehicle auction platform, an asset-light intermediary earning 80%+ high-margin service fees via VB3 (275+ locations, 11 countries), is down ~45–55% from its highs to ~17× forward P/E on soft volumes, with a fortress balance sheet ($4.2B cash+HTM, $5.5B liquidity, minimal debt) and $1.6B of YTD buybacks against a ~$27B cap. Two catalysts: a rising auto-loan-delinquency tailwind (90-day+ rate 5.6% in Q1 2026, highest since 2010; next data Aug 4) feeding non-insurance auction volume, and star CEO Jay Adair returning Jul 31 (1,100%+ return over 2010–24). Ugly chart — wants a mid-$20s base, high-$20s the long-term entry, ~$24 the stop/sell-down. A long-term compounder, not a trade. (The Buys/Trims/Holds/Sales tables were deferred to Monday's Portfolio Update.)
Nuclear microreactors enter the data-center stack (Jul-8). Extending the Jul-2 "Watts happening" data-center-power thread, two milestones frame nuclear as beginning to supplement/replace natural gas as the primary data-center fuel: microreactor startup Aalo Atomics (positively profiled in March) hit test-reactor criticality over the July-4 weekend (one of only four companies), and rival SMR maker Valar demonstrated full temperature and power conversion (FTPC) and signed a 30MW contract to power an Nvidia (NVDA) data center. Hay's caveat: NRC licensing runs years, so Valar's lead "could well diminish or outright vanish" in approval testing. Aalo's micros (~⅕ the size of Valar's SMRs) favor mass production and truck delivery; the market is big enough for multiple designs — "a significant positive for America" given China's enormous lead. Both private.
Yen bull reiterated (Jul-9). A macro/FX Daily restates the long-standing (and admittedly unrewarding) bullish yen call at 162/USD: per Vincent Deluard the yen's ~50% discount to the USD is "the greatest on record," and it looks "dirt-cheap" even after adjusting for the US-vs-Japan yield gap that fuels the carry trade. The BOJ is intervening around this level and under pressure to keep hiking, with "the very real potential for coordinated support that includes the U.S." Caveat: the yen "functions best as a hedge against market turmoil or an alternative to US cash/Treasuries" — a big near-term move is only 10–15%, with no stock-like upside and no income. No ticker named (macro/FX view; distinct from the equity-market call behind the earlier EWJ trims).
Boeing to Hold; Brazil bull reiterated (Jul-13). A Portfolio Update downgrades Boeing (BA) from Buy to Hold for the 2027-28 cash ramp — the March-13 turnaround is genuinely working (737 MAX production 38→toward 47/mo, guided 52 by early-2027; Q2 deliveries ~167 beat; FCF inflected positive at $375M in Q4-2025, guided $1-3B in 2026 toward ~$6.8B/2027 and ~$10.5B/2028; >5,000-jet backlog + Airbus duopoly), and the position is ~+5% — but the recovery timeline slipped on every front (777X to 2027+, Commercial profitability to 2027, a costly Spirit AeroSystems reacquisition, $47-51B debt, dividend/buybacks off indefinitely), the stock is valued entirely on future FCF (trailing P/E >100 is a turnaround "math quirk"; a 20-25× multiple on 2028 FCF pencils to ~$275-345 only if the ramp executes), and China/litigation headline risk keeps resetting the mood — so with "very little valuation cushion for a stumble" he'd hold, not add, and wait for better entries on a strike/FAA/delivery setback (catalysts: the Jul-28 Q2 print, SPEEA labor talks into October, the FAA directive; thesis-breakers: a prolonged engineers' strike, a quality glitch, or the FCF ramp slipping). The same update reiterates and "amps up" Brasil Bolsa Balcão (BOLSY) — Brazil's exchange operator, down ~8.5% (its own swoon ~25%) despite robust Q1 earnings on Brazilian-market weakness, at trough P/S & P/E with a ~5% dividend, a "massive discount" to global exchange peers (20×+) tied to Brazil's single-digit-P/E market and record-high real rates Hay expects to fall "in a big way" (→ "a massive bull market in Brazilian equities"); risk = a Lula re-election (age 80, dead-heat polls). The Brazil ETF EWZ (profit-taken earlier this year, off ~15% from its April-14 apex) is referenced for contrast — "a correction within a structural bull market for real assets." (Consistent with the standing EWZ/PBR/ARGT Brazil-&-EM-value thread.)
China reiterated — FXI POW (Jul-17). The latest Friday POW! reiterates the February buy of the iShares China Large-Cap ETF (FXI), now ~$34 vs a $38.42 entry (~11.5% down). Haymaker re-underwrites it pillar-by-pillar and concludes the drawdown is macro/sentiment, not fundamental: US tariffs to ~17% effective, a deliberate Hormuz-era Beijing slowdown, and a weak "K-shaped" consumer (May retail −0.6%) — while the four original bull pillars hold and AI monetization has accelerated (China daily AI-token usage +1,400× in 26 months; Alibaba Cloud $1.3B/qtr triple-digit, Tencent net profit +21%, Baidu AI 52% of revenue). At 10.78× earnings — "the cheapest major stock market in the world" — with light positioning (~$895M outflows) and a controlled correction after a four-year breakout, likened to oil/energy just before their July bounce. Bear case respected (tariffs, the never-stress-tested VIE structure, the consumer), so size for the binary risk (holdable through a 25% drawdown). Reiterating a Buy at $34, 12–18-mo target $42–48, eventual new ATH >$60. "We are staying in." (Constituents BABA, TCEHY, BIDU cited as strengthening evidence.)
Coal — the lonely trade, in full (Jul-24). The Friday POW! delivers the write-up promised in Monday's Portfolio Update: an explicit buy / re-buy and dollar-cost-average into the two Australian coal producers — Yancoal (YACAF) and New Hope (NHPEF, carried as NHC). The frame: coal, "humanity's first high-density energy source… once known as black diamond," was supposed to be extinct, yet global usage is up ~65% this century and has risen even over the past decade. Haymaker has avoided the group on pollution grounds — favouring met coal (steel) over thermal (power) — which is how Warrior Met (HCC) became a Jan-2025 Trader Ferg win (+~51% despite a 20% correction) and why the Mar-2025 Peabody (BTU) tout near $13 was "heavily qualified" (it tripled, then fell hard to ~$22; partial gain-harvesting had been suggested just under $24). The mechanism is substitution: escalating Mideast hostilities threaten LNG shipments to Asia, and LNG "competes directly with coal" for power generation in China, India and Japan (~3bn people) — with thermal coal about as cheap as it has ever been relative to LNG and LNG "once again ripping." The peace narrative is refuted by the physical data: after the "Memo of (Mis)Understanding" briefly freed trapped Persian Gulf carriers, "all product shipments have once again fallen off a cliff" (tankermap.com). Supply is tightening too — Indonesia, the world's largest thermal-coal exporter, is materially cutting exports — into hot weather in Europe/China/India and drawn-down emergency reserves. YACAF ("a dud," now ~$4.25 after a ~10% weekly pop) is the cheaper, more beaten-up vehicle; NHPEF "hung in there admirably well" (three-year high last month, ~12% off) and is Ferg's preference — "we like both of them… a lot." Haymaker explicitly prizes the loneliness: "similar to our extremely against-the-grain aggressive buy on oil at the end of June." (Extends the energy-shortage / Hormuz thread into the coal leg.)
The great rotation, not the great meltdown (Jul-26, Thoughtful Money). A two-hour interview that ties the year's threads together and marks the first sell-side action on the oil call. The June-30 washout trade worked — WTI 69 → ~93, Brent over 100, "one of the biggest rallies I've ever seen" — so having been long 10,000 barrels of futures he is now gradually liquidating and explicitly blesses viewer profit-taking ("like with USO, I don't disagree"), while insisting "any dip is to be bought," keeping "a lot of contracts left" for a possible blow-off top, and framing oil as rangebound ~$70–100 with China as "the new OPEC" (imports had fallen 4–6M bbl/d; global SPR refills — including countries that never had one — plus Jeff Currie's "day zero" ahead). The upside now sits in the equities that haven't caught up: EOG is named as the June-17 Haymaker recommendation (PE 10, a five-year multi-year breakout pending, takeover kicker), XLE is +33% YTD vs the NASDAQ's ~8% with energy tied with tech for the best five-year sector returns yet still 3% of the S&P, and US natural gas at $3 vs >$20 abroad is "ridiculously cheap" with Europe's storage at 15-year lows ("they're going to pay a severe price this winter"). Uranium "looks like oil in late June" and is the better supply story — SRUUF at a ~10% discount to an $85 spot that itself sits below the ~$95 utility contract market = "a discount upon a discount" (~20%) — with Aalo Atomics now DOE-approved for criticality (one of four) though "three to four years away from producing in size," and the grid, not the reactor, the binding constraint. The Monday coal alert (YACAF / NHC) is reiterated on Indonesian export cuts into an Asian LNG shortage. Against that: the tech reality check "greatly overdue" may be here — GOOGL whacked on good earnings where "almost two-thirds" came from marking up stakes in its own customers ("circular financing") with depreciation still to hit the P&L, ORCL's bonds "trading almost like junk" on a "trashed balance sheet," IBM −25% in a day (sold months earlier), MU/semis with single-digit P/Es but scarcity-driven margins heading for "air pockets," CAT at ~32× earnings / 6× sales ("take some partial profits"), SPCX ~$116 vs a ~$170 average secondary cost, and OpenAI's IPO likely pulled or "quite a fiasco." The constructive side of the rotation: KRE's regional-bank breakout "very bullish" at 10–12× earnings (bought on the pullback, not the breakout), deep value "unbelievably cheap" (GMO) with software "AI victims" at ≤10× earnings, international markets still outperforming, and precious metals turning — SLV "looks pretty interesting for a recovery" after his blow-off-top call, GLD underpinned by central banks buying on tonnage, and GDX/GDXJ's record 2025 outflows in a record-profit year read as contrarian apathy plus incoming energy-style capital discipline. Tail risks he is watching: the BB-vs-CCC spread "on the verge of a multi-year breakout" (+~300bp) and soaring private-credit defaults at ~9% vs junk's 7.3% ("at least some fire") while IG spreads give "no warning at all"; the 30-year at 5.18% as a breakout, with 6% "really bad news"; a yen rally (~50% PPP undervalued, Japan repatriating on the Korea playbook, a ~2%-of-GDP deficit vs America's 6–7%) that would be "a real nasty hit to the AI trade"; and the next-crisis endgame — Fed games, then the Treasury buying equity index funds (Bagehot's dictum, the Hong Kong Monetary Authority precedent, "project Zimbabwe") which is why there's "a persistent bid under precious metals." Closing plea: "don't go to a bomb shelter, but be on alert for all these opportunities" — the danger is souring on gold miners, energy, financials and overseas all at once and going to cash at the wrong time.
Buying the two washouts — miners and sand (Jul-27). The first of a new, shorter Monday portfolio-update series ("two or three… each Monday") is all buying, and both moves are back onto the Buy List. Alamos Gold (AGI) at $29 — "cut in half" from its ~$55 February peak while "the primary ETF for senior gold-mining stocks has plunged 35%" since end-February — is "an attractive buy, or buy up, point": 10.4× trailing and under 10× 2027's $3.27 consensus ($2.32 for 2026), reserves "almost tripled… over the last 10 years" with grades ~35% better, and the Young-Davidson earthquake (a 12% cut to Q2 output plus lower full-year guidance) dismissed as the kind of one-off that "tend[s] to fade into insignificance over time." Crucially he re-states the exit discipline that got them here — profit-taking by dollar-cost-averaging out as the group went vertical, "our default tactic when prices enter the hockey stick phase" — and the macro underwrite is unchanged: bullion's drop is "a correction within an ongoing bull market" because "central banks around the world remain aggressive buyers," "the odds of government bond crises becoming more acute are rising by the day," and the trend stays one of "fiat currency debasement" — while retail's "truly remarkable" negativity after a +150% year is a contrarian buy signal ("their recurring tendency to be on the wrong side"). "Before long we may bump it up to a Strong Buy." Atlas Energy Solutions (AESI) gets a partial position bought sub-$12 and moves back up to the Buy List ("though not as a Strong Buy") on a pure index-flow arbitrage: its removal from S&P Small Cap Growth indexes crushed it just as the oil-service outlook turned up — "almost the inverse situation" of the late-2024 inclusion that lifted it to $24 against deteriorating fundamentals and made it "a great time to ring the cash register." The business case is the Permian bottleneck ("you simply can't produce shale crude without frac sand"), the Dune Express conveyor "none of its competitors have," a near-duopoly implying "very strong pricing power in the next up-cycle," and a growing power-supply business — against a demand backdrop stated more forcefully than anywhere else this month: "the unprecedented oil shortage which has developed — and is worsening, despite the various peace treaty head fakes." (A direct rebuttal of the de-escalation narrative, and the buy-side counterpart to the Jul-26 decision to harvest futures gains while keeping the equity upside.) LBRT's disappointing quarter is the acknowledged near-term risk, but "well-discounted after the price collapse since May."
"That's not a bubble. That's a bubble" — the semi bust (Jul-29). The bearish counterpart to the rotation call, and the sharpest sell-side statement of the month. Yes, parts of the US market are bubbles and "some have already been punctured" — SpaceX (SPCX)'s "initial post-IPO rocket launch has come crashing down to earth," halved from the late-June peak — but the real extreme was offshore: by June 18th the KOSPI "had generated more return in the past eight and half months than it had in the prior 45 years of its existence" (Grant Williams' KOSPI TURVY). That statistic is the whole sell thesis, and Haymaker had already acted on it — "one of our very rare outright sell recommendations" on longtime favorite EWY on May 6th at $151; "we were a tad early, but the basic call was spot-on," −17% since, "and based on its spectacular appreciation, there could be much more to come on the downside." The US transmission is explicit: that blow-off was "essentially two stocks: Samsung and Hynix," and "despite ballistic earnings, leading to modest P/Es, this dynamic duo has plunged by 40% over the last five weeks" — "this is also when the U.S. Semiconductor Index, the SOXX, began its descent, one that is becoming increasingly sickening for its formerly jubilant holders." The cheapness defence is retired by example: "no slice of the S&P 500/NASDAQ was more adored than the semis," whose fans cited "how cheap stocks like Micron (MU) were even at their peaks" — yet MU "has lost one-third of its market cap since June 24th… despite reporting blow-out earnings." The rule it proves is the same exit discipline that got them out of the gold miners: "the prudence of systemically selling into these hockey stick-like moves, irrespective of how exciting the story sounds." And the tactical instruction pre-commits the next move — a bounce is "possible, even probable," and "will likely be an opportunity to once again reduce exposure for disciplined investors," because "history is alpine lake clear that the bigger the bubble, the bigger the bust." (Hardens the Jul-26 "tech reality check greatly overdue" leg: the rotation's source of funds is now named and dated.)
The crack in the oil "logic" (Jul-30). A macro-only Daily that supplies the missing evidence for the standing bullish-oil stance by attacking the market's stated reason for cheap crude rather than the price. Two anomalies are stacked. The first is the nominal-price illusion: crude in the upper $70s/$80s "were in this range almost 20 years ago, and considering that consumer prices have risen approximately 60% since 2007, oil is an inflation-adjusted bargain" — WTI "hit almost $98" in the fall of 2007 — which is extraordinary because "almost no one on Planet Earth would have believed oil would be anywhere close to this inexpensive with the Strait of Hormuz still essentially closed," while "egress out of the Red Sea, where about 12% of global crude supplies transit, is at-risk, as well." The second is the piece's real contribution: the crack spread — "the cost of converting oil into refined products, like gasoline" — has "nearly converged" with crude itself, at $61.80 against $84.19, a 73% ratio where "even last fall, when the crack spread spiked as oil was languishing around $60, it was barely above 50% of the crude price. This is unlike anything seen in the 21st Century (and, likely, in the 20th Century, as well)." The inference is a falsification, not an opinion: a fat crack is refiners bidding for barrels to satisfy end-product demand, so that ratio — "along with very strong demand for jet fuel" — shows "global demand for oil-based products remains extremely robust," and "these realities run counter to the view that a key reason oil prices have stayed subdued, relative to the severity of the supply shock, is because of plunging consumption of refined products like gasoline." Bottom line: "oil prices are far too low based on the continuing debacle in the Middle East as well as the high demand for oil-related products. Accordingly, sell-offs in crude, such as seen this week, are opportunities for accumulating oil and, particularly, oil-producer equities." (No ticker named. The producer-equity preference is the buy-side counterpart of the Jul-26 stance — gradually harvesting the 10,000-barrel futures position while "any dip is to be bought" and the remaining upside sits in the equities that haven't caught up — and of the Jul-27 line that the shortage is "worsening, despite the various peace treaty head fakes.")
The rotation gets its destination — regional banks POW (Jul-31). The Friday POW! puts a rated Buy behind the financials leg of the rotation: KRE, the SPDR S&P Regional Banking ETF ("an equal-weighted basket of roughly 150 U.S. regional and community lenders"), rated Buy / Accumulate — the sharper, dated version of the "very bullish" KRE breakout comment made to Thoughtful Money five days earlier. The setup is the house template — "while the market pays 30x + for everything that touches AI, this entire sector trades at 11x earnings because it broke three years ago and no one has forgiven it. That's where we often like to look" — applied to a group still carrying "the psychological scar tissue of March 2023" (First Republic, Silicon Valley, Signature), when it "got repriced for possible extinction." The diagnosis is what makes it buyable: those failures were "a correctable mistake," massive unrealized losses in long-duration bonds funded by "flighty, uninsured deposits," and "that imbalance has been substantially addressed across the group. However, the multiple never fully recovered" — so today's ~12× "embeds a permanent risk premium for a crisis that has already passed… exactly the kind of lingering odor that creates mispricing." The operating case is reported, not projected: "for a bank, everything flows from net interest margin," and Q1 2026 confirmed deposit costs rolling over while loan yields hold — "the single most important operating trend for banks, and it just turned positive." Critically it is curve-driven, not cut-dependent: it works "even without Fed rate cuts" because the curve "has steepened meaningfully" and banks "make money borrowing short and lending long," which "makes the thesis more robust than the simplistic 'rate cuts save the banks' version" — a deliberate stripping-out of the Fed as a single point of failure (and of the tailwind "the bulls counted on in February"). Proof is imported from offshore: "bank stocks love steep yield curves; the extraordinary performance of Japanese banks since 2022 is a vivid example" (a five-year Sumitomo/SMFG chart; "Japan has the steepest yield curve in the developed world"), and those banks "generated a terrific breakout signal in early 2023" — the same signature now on KRE, whose chart shows "multiple upside breakouts" after a bulk review of "literally hundreds of stock charts" turned up "a plethora of multi-year range expansions" among regionals. The domestic precedent is his own scorecard: JPM, BAC and C broke out in 2024, and "we glowingly highlighted JPM, and its CEO, Jamie Dimon, back in January of 2024… more than doubling from ~$170 to $353." Valuation is argued on the industry's own yardstick — earnings, tangible book (~1.4–1.5× vs a sub-1× 2023 trough and ~1.7–2.0× in rosier times), ROTCE and dividend yield — with ROTCE climbing "from the low-teens toward the mid-teens" as "the engine that closes that gap" ("a bank earning ~14-15% ROTCE justifies a materially higher multiple of book than returns stuck near 11%"), making the re-rating "earnings-driven, not sentiment-driven… durable rather than a momentum blip." Catalysts are ranked as free options: an M&A/consolidation wave across "~150 sub-scale regional banks" now that approval pathways are clearer (an equal-weight ETF "captures the takeout premiums broadly") and Basel III capital relief freeing capital for buybacks and lending — "either catalyst firing would force the market to re-rate the group; both firing would do it violently." What he concedes, and leads with: "the deep-discount entry is most likely behind us" (KRE ~$76, up ~25–28%, just under a $78.10 52-week high, the re-rating "visibly underway" on fund inflows and KBW's "banner year" call — "a cheap sector re-rating further, not a bottom-fishing trade"); a fresh inflation shock "would revive genuine 2023-style duration and deposit stress," though longer rates rising on a Fed seen as "behind the curve" actually helps (new chair Kevin Warsh was "roundly criticized this week for his failure to provide specifics on containing inflation"); KBWB's money-centers "have crushed the regionals on deregulation and capital-markets strength… sometimes you're considered a laggard for a reason"; and the thesis-breaker — ~$875B of CRE debt maturing in 2026 into a weak office market (one large holding flagged "a ~20% potential loss rate on its general-office book"), where the equal weighting means "the structure that diversifies deposit risk actually concentrates CRE risk." Verdict: "the risk/reward favors the owner… sized for volatility and understood as a catalyst-dependent, higher-beta sector play rather than a sleep-at-night holding. Accumulate on weakness, size for the beta, and treat the CRE data as a critical factor." (No Buys / Trims-Holds portfolio tables accompanied this issue.)
Two seismic events, and the order to raise cash (Aug-3). The Monday Portfolio Update argues last week produced two potentially regime-changing breaks, and converts them into the most explicit defensive instruction of the year. Part I — rates. The 30-year T-bond yield made a "clear breakout to an 18-year high" at 5.27%, raising "the very real possibility of taking out the resistance at just under 5.5% that runs all the way back to 2003"; above that "a bit of friction at 5.75%" and then nothing until 6.75%, set in late 1999 when the Fed was flooding the system for Y2K and a tech-bubble "wealth effect" was stimulating the economy. The 1999 analogy carries the argument, and its structure is the reusable part: the yield spike "undoubtedly played an important role in popping the dot-com bubble… however, outrageously high valuations had put the conditions in place for a waterfall decline. The rate surge simply applied the coup de grâce" — rates are the trigger, valuation the vulnerability, so a market "totally unperturbed by the upside range expansion in long-term Treasury yields might suddenly snap to attention." Are the 1999 preconditions present? "It's most fortuitous today that there aren't similar valuation extremes… If you sense a whiff of sarcasm, you detect well": SpaceX (SPCX) at $225 (+70% from the $135 IPO) was worth "nearly $2½ trillion… almost 70 times sales," and "showing how quickly bubble-like valuations can reverse, its market cap is down to $1.4 trillion." The measurement discipline is stated outright — "we consider the Price/Sales ratio to be superior to Price/Earnings… sales are considerably less volatile than earnings" and "much harder to inflate… with accounting gimmickry" — and on that basis "the S&P today looks significantly more overvalued… than it was even at the height of" the dot-com bubble. Both bullish rebuttals are dismantled: strong earnings are "the AI frenzy and… extremely flattering accounting, bordering on circular financing" (big tech booking "huge sales to companies in which they have received stock… then booking enormous gains on those shares, or warrants, as income — legal but… extremely unsustainable"), and rising margins came from "capital-light business models" that "have become highly capital intensive" in the data-center "arms race" — $700 billion of hyperscaler AI capex this year, an estimated $1 trillion next, with "serious questions… about the ultimate returns." The years-long avoid-long-Treasurys call (a deluge of issuance funding "$2 trillion-type deficits" while "the largest foreign buyers have been exiting stage left") is scored: "Alas, that is no longer a prediction but a reality." Verdict: "Raising cash right now, particularly in grossly inflated highly valued securities." Part II — Japan. The yen hit a 40-year low against "the highly flawed, and overvalued, U.S. dollar" despite negligible external debt, "a monster current account surplus" and status as "the world's largest creditor nation" — "a 180° rotation from America's present sorry fiscal and trade status." The culprit is a BOJ "pathetically behind the curve" whose "sloth-like tightening pace makes the Fed under Jay Powell look like the second coming of… Paul Volcker." Its repeated solo interventions were "an exercise in futility," but Friday's was different in the one way that counts: "the U.S. Treasury has stated it is also participating in the stabilization effort" — the coordinated support Haymaker "anticipated… on July 9th when we reiterated our positive — and highly contrarian — stance" ("anyone who took our advice is nicely in the black since then… and it's about time!"). The yen "spurt[ed] by 4% (and another 1.4% today)," and it ranks as the week's second event "because of how important borrowing in depreciating yen has been to financing trillions of overseas investments." Position: "we remain bullish on the yen as a hedge against global financial markets losing this important source of buying pressure, though we wouldn't chase this rally," with Japan's "new efforts to bring capital back to its shores" able to "turbocharge the liquidity exodus from offshore markets, a macro risk that appears to be off almost everyone's radar" — the 2024 "global selling tsunami" proving the channel is real. "The one-two punch of soaring long-term U.S. Treasury yields and a ripping yen could be a hard blow to the U.S. stock market"; the mitigant is that "the BOJ continues to raise rates at a glacial speed," and the escalation trigger is the BOJ "signaling a much more aggressive stance," which could make investors "remember something they have long forgotten: downside risk." The book, restated in the tables: nine blue-highlighted rating changes execute the same instruction — Strong Buy upgrades for AGI, BOLSY, EQT (both lots), PSLV and YACAF (four of the five below cost: conviction rises as price falls against an intact thesis), CRH back onto the Buy List (both lots, rated B), and winners downgraded to fund the cash: ACN, MTB and TRV off the Buy List to Hold plus a PBR trim dated 08/03/26. MTB's demotion lands three days after KRE became the rated way to own regionals — basket over single name. Housekeeping: the AEM trim date is corrected 07/20/2026 → 07/13/2026; GDX/GDXJ now show on the Buy List and CPRT/KRE appear for the first time; HAL is absent from all three tables. (No ticker is attached to the yen — as on Jul-9 it stays a macro/FX viewpoint.)
A tale of two inventories — buy the ceasefire dip (Aug-4). The next morning's Daily keeps the energy leg intact even as the equity stance turns defensive: oil is "once again retreating on the latest Iran War ceasefire news" — dismissed on the base rate ("repeated breakdowns of prior negotiations") and on who actually holds the chokepoint ("an obvious unwillingness on the part of Iran's Islamic Revolutionary Guard Corps to relinquish their Strait of Hormuz chokehold"), i.e. a deal signed by governments does not reopen a strait held by the IRGC. Underneath the headline, "inventories continue to be drawn down globally, even in the U.S." — the decisive case, since America is "largely self-reliant when it comes to crude output vs consumption," so a draw there cannot be blamed on blocked imports; per John Kemp, "America's petroleum supply is diminishing at an alarming rate." The new leg is the second, unwatched inventory: per Cornerstone Analytics, "it's not just oil stocks that are being severely depleted. Refined-products inventories, like jet fuel and gasoline, are also experiencing dramatic drawdowns" — missed because "the data on these stocks are much less easily obtained… prone to being ignored… at least until airports and gas stations around the world start running short on fuel." That is the direct sequel to Jul-30's crack-spread argument (margins proved product demand; stocks now show it being drained). Verdict: the retreat to $76 is "a compelling opportunity for investors who appreciate the severity of the current energy crisis," with two technical confirmations — crude has "retreated to its 200-day moving average" (a retest, not a break, and Hay's preferred entry) and "the largest ETF of oil and gas producers made an all-time high earlier this year and has held very near that critical breakout point," i.e. the producer equities refuse to confirm the crude fade. That ETF is not named and no prior issue in this archive identifies it, so no ticker is inferred — macro viewpoint only.
The downstream leg, with the data (Aug-5). The direct sequel supplies what the previous day called un-instrumented, and it changes the shape of the energy call: the shortage is now specifically refining. John Kemp's chart of U.S. "The Big Three" — a composite of jet fuel, distillate and gasoline, in the country with "some of the planet's most accurate inventory data" — shows those stocks "significantly drawn down," and the cause is a fixed product slate with no slack: "refineries have prioritised production of jet fuel and diesel at the expense of gasoline to relieve the worldwide shortage of middle distillates. But the lack of spare capacity has meant gasoline stocks have come under intense pressure." The demand-destruction narrative is falsified a second time — Jul-30 did it on margins (the crack spread at 73% of crude), this does it on volumes: "American refined exports to the rest of the world have risen to a record rate," which "calls into question the prevalent belief that there has been significant demand destruction for oil-related fuels." The shortage is then given a geography and a date. California lost "two large facilities… Los Angeles and Benecia, shuttered since last fall," now imports "almost 20% of its gasoline consumption… likely to rise as more in-state refineries are taken off-line" — with the conditional that carries the risk, "(this assumes that offshore suppliers have the ability to deliver)" — and "ominously, Chevron (CVX) is vectoring to close its last two California refineries." The counterparty can't fill it either: the Asian suppliers of those cargoes "are facing a serious shortage of refining capacity themselves and are highly likely to restrict shipments across the Pacific," raising "the possibility of 1970s-like gas lines before long unless the U.S. limits overseas shipments and redirects those to California" — a fix immediately priced ("doing so, however, would damage America's reputation as a reliable supplier"). The investment consequence is a margin fact: the complex "has been running flat out," which "accounts for the spike in refinery profit margins to what are essentially tied for the highest of the 21st century," and that is "a key reason why profits at companies like Chevron have exploded, incurring the wrath of President Trump." The closing frame is the standing grievance plus the standing measurement discipline: "no matter which party is in the White House, the oil industry continues to be a recurring scapegoat, despite that oil prices have barely risen over the last 20 years (and, adjusted for inflation, are down by roughly 50%)" — the wider-lens restatement of Jul-30's nominal-price illusion. CVX is the only ticker named; no refiner is.
Mean reversion, and the 75% number (Aug-6). The earnings-quality charge that has run all summer finally gets a company-specific figure, and it is the other half of the Aug-3 valuation argument: if the index is already more overvalued than 2000 on Price/Sales, this attacks the "E" that keeps the P/E respectable. The bullish fact is conceded first — per ZeroHedge ("which some deride as the financial equivalent of the National Enquirer, but we find a frequent source of interesting factoids") S&P 500 earnings are "the strongest since 1955," and the year is chosen deliberately: the post-WWII interstate-highway peak, when "civilian atomic energy was also just lifting off, promising an era of cheap and nearly limitless power." Then decomposed: the explosion "has enjoyed a tremendous tailwind from the AI spending frenzy," and the flattering is non-operating — hyperscaler profits are "greatly flattered by gains on share ownership in their largest AI customers and partners," with the headline number "in GOOG's case, approximately 75% of its first-half 2026 profits have come from the mark-up of its stakes in, primarily, Anthropic and SpaceX." That is an escalation from "almost two-thirds" on Jul-26 and the generic "booking enormous gains on those shares, or warrants, as income… legal but… extremely unsustainable" of Aug-3 — and it makes Alphabet's reported profit largely a mark-to-market on two private valuations, one of which (SpaceX) Haymaker documented falling from ~$2.5T to $1.4T three days earlier. A mark-up is a mark-down waiting for the next round. The independent cross-check comes from outside the accounting: "Canada's highest profile economist, David Rosenberg, calculates that the S&P profits increase this year exclusive of energy and AI-related sectors would be a much more pedestrian 4%" — a sector strip landing where the accounting decomposition landed. The conclusion is a base rate rather than a forecast: "past times of earnings spikes have been followed by equally dramatic declines," so "at least some mean reversion is to be realistically expected." And the only rebuttal is pre-empted twice — Jamie Dimon's "this time is different" as "the most dangerous words in the English language," rivalled by Irving Fisher, who "publicly, and infamously, declared on October 15th, 1929, that 'Stock prices have reached what looks like a permanently higher plateau'" (nine days before the crash). Close: "nothing is permanent… other than constant change and extreme volatility. As the above charts clearly illustrate, that applies to corporate profits as well."
Healthcare gets its destination — the GILD POW (Aug-7). The Friday POW! puts an explicit Buy on Gilead Sciences (GILD) at ~$133 (16% below the $157.29 52-week high) and is the cleanest statement in this archive of Haymaker's favourite setup: a quality franchise mispriced by an accounting artifact rather than by business deterioration. The frame is stated at the top — "one of the more quietly compelling setups in large-cap biopharma… the reason it is under the radar for most investors is the same reason we find it attractive." The artifact: the $10.5 billion Q2 GAAP net loss reported August 4 is "almost entirely driven by acquisition-related write-downs and upfront payments… required to be expensed immediately under GAAP… rather than any deterioration in the company's operations" — "an accounting event, not a business event." Why it persists is mechanical: "that number will appear in every screener, valuation tool, and news headline about Gilead for the next several months," making it "look broken to investors who look at GAAP earnings without reading the footnotes." The operating reality comes from the same release — FY product-sales guidance raised to $30.1B, a 46.9% adjusted operating margin (Q1) that "produces approximately $14 billion in annual adjusted operating income," FCF estimated "north of $12.5 billion next year" = "a most alluring 7.6% FCF yield" on the $163B cap, and net debt of just ~$13.4B ($8.6B cash vs >$22B debt) that "could [be repaid] nearly all… should it choose to," i.e. roughly one year of free cash. The engine is HIV: ~70% U.S. share anchored by Biktarvy, "the world's most prescribed HIV regimen," protected through 2036 — "in pharmaceutical terms, that is a license to print the cash that funds R&D, acquisitions, dividends, and buybacks for the better part of the next decade" — with the lenacapavir estate beyond it, PrEP sales past $1B in Q2 (Yes2Go), an FDA-accepted Yeztugo once-weekly oral PrEP filing, and the Merck-partnered once-weekly oral pill positive in ISLEND-1/ISLEND-2 at AIDS 2026 in Rio. The transformation is Trodelvy, from the criticised $21B Immunomedics purchase: first-line approvals in metastatic triple-negative breast cancer (US and EU) and, with Keytruda, first-line metastatic lung — "these are not incremental label expansions," because first-line indications "are among the largest oncology markets by patient volume and revenue potential" and "uptake in first-line indications historically bears little resemblance to the niche second-line volumes the bears are modeling." Valuation: 12-13× adjusted earnings against peers and an S&P at 20×, a 2.5% dividend, targets $120-$165 (avg ~$150) with the dispersion itself informative (Leerink Market Perform $127 vs Morgan Stanley $165). Technically it "achieved a two-year breakout in the spring of 2024" yet "on a forward P/E basis it's just as cheap now as it was when it broke out," and after a nine-month pullback is "resting on its 200-day moving average… the pause that refreshes" — the same buy-the-retest discipline applied to crude on Aug-4. Risks are conceded rather than dismissed — IRA Medicare price negotiation on Biktarvy before 2036 ("the most credible bear case… we'd be lying if we said it's completely bogus") and integration risk across Arcellx, Tubulis and Oral Medicines ("if any acquired pipeline asset fails in development, the write-downs compound") — then bounded on joint probability: "the bears need to be right about both… simultaneously to justify the current discount… a higher burden of proof than the current price implies." Target outcome: a re-rating to ~15-16× adjusted earnings "as the GAAP loss normalization removes the… selling overhang and the oncology revenue ramp becomes visible in consecutive quarterly results." Extends the standing healthcare rotation leg (DGX Jun-5, MDT Jun-12) with the first name whose mispricing is argued from accounting rather than from sector flows. (No Buys / Trims-Holds / Sells tables accompanied this issue.)
Gas — the sentiment-versus-physics trade, and an oil name to play it (Aug-11). The cleanest statement of the house screen: measure a commodity's physical standing, measure its financial standing, and buy the gap. Physically, natural gas is "America's #1 source of electricity generation (at 43% of the total, it is nearly double all renewables combined)," and it is now being asked to be the destination rather than the bridge — "there is now a growing awareness that wind and solar are unequal to the task of providing reliable baseload power at scale," reinforced by data centers that require "uninterrupted power" and "can't operate in accordance with daylight hours and wind patterns." The demand is sized, not asserted: the U.S. has "10 times as many" installed data centers as China, "beforeMeta, Microsoft, Amazon, Oracle, and Google invest another $5.8 trillion by 2030… without electricity these facilities will be nothing but inert shells" — note that Haymaker's quarrel with the AI buildout is with its financing and accounting (Jul-26, Aug-6), never with its electricity bill. Financially the opposite: per two John Kemp charts, "bullishness on the blue fuel is close to the lowest it's been this decade, outside of Covid" (most depressed in two years) and the real, inflation-adjusted price sits in the 18th percentile going back to 2010 — "investor apathy (bordering on antipathy)," as if gas were "a scarcely used bridge to nowhere." That apathy has "obviously inhibited the market performance of the largest U.S. natural gas producers, like Expand Energy (EXE), Range Resources (RRC), and EOG." The second, mechanical injury is the Permian, now "the country's second-biggest gas-producing region, as well as the fastest growing," whose associated gas has "often traded at negative prices" and been "flared into the atmosphere" for want of pipe — "another source of profit suppression." The fix is under construction and funded by someone else: "multiple new pipelines" out of the basin totalling "some 15 billion cubic feet per day, about 12½% of aggregate U.S. marketed gas output," reaching "a marketplace that will soon be in desperate need of it" — "a notable earnings kicker for companies with significant Permian gas production." The named beneficiary is EOG ("likely to be one of the main beneficiaries of this extraordinary delivery surge"), with the usual multi-year range-expansion trigger — "on the verge of breaking out of what is nearly a five-year trading range." The structural elegance: EOG's payoff comes from basis, not price, so it needs no Henry Hub rally and no incremental drilling, leaving the sentiment and real-price extremes as free optionality. Third instance this month of "the selling is structural, the asset is not" (after GDX/GDXJ and AESI). Caveat: EXE and RRC are Haynesville/Appalachian, so they get the demand and sentiment legs but not the Permian basis kicker — and 15 Bcf/d of liberated Permian gas is arguably incremental competition for their molecules.
Gold & the miners — the same screen, run by a guest and endorsed (Aug-12). A wholly guest-sourced Daily: "an abbreviated re-posting" of Kevin "The Macro Tourist" Muir's Aug-8 edition, published because "Kevin is making a convincing case for gold, and the miners thereof… one with which we agree strongly enough to broadcast the main thesis here" — the reasoning is Muir's, the stance and the two securities are Haymaker's. Muir's structure is the house method stated more precisely than the house usually states it. First, he models the strategic buyer's incentives rather than its size: China is "interested in buying as much gold as they can over the next decade… for as little as possible," with "zero interest in letting hedge funds and other speculators move up the price… and then sell it back to China for a profit," so western euphoria kept it out and the broken "speculative fervor" has brought it back — "China has quietly begun buying again" (Poland actually led last quarter; official figures are "just a fraction of their actual buying"). Second, he disqualifies that bid as the trigger — "these whales have been softening the decline… yet gold has been steadily falling" — and locates the trigger in a price refusing its own drivers: real rates and the dollar "were spiking higher, yet the price of gold refused to break below 4000. Gold was telling you that it was being accumulated," after which the break of the 50-day/downtrend forced covering and gold ran "350 dollars higher. Sometimes trading is that simple." Third, apathy is measured with money at risk rather than a survey: gold uniquely pays up for calls, and the "1-yr 25-delta call skew," normally 3-4 vol points, "dropped to the lowest since before COVID… no one is paying up for right-tail risk," while "few care about gold. They are more interested in semiconductor and memory stocks" — his filter being to own what sits "on Page 17… on its way to Page 1," and to have no view at all on a page-1 mania. Fourth, the cheap-cyclical trap is defused on timing: miners at "11x," cheapest outside the GFC and 2011-13, where "all of that gold decline is already in the earnings" and "analysts were previously slow to raise the price of gold in their models," so mere stabilisation lifts estimates and bottoms EPS. He is "buying gold, platinum, and a bunch of different gold mining stocks" with a one-line stop — "losers average losers" — which is notably the opposite discipline to Haymaker's own average-into-weakness habit (AESI, APA), published without comment. Haymaker's contributions are the two names and the hedge: the Jul-27AGI Buy alert has "popped about 15% since then," an Adam Taggart chart makes GDX an "accumulation candidate" on any pull-back, and the whole endorsement carries the caveat that "both bullion and the miners have had a snappy rally of late." This reinforces rather than shifts the standing precious-metals stance (Jul-26's record-outflow apathy read on GDX/GDXJ, Jul-27's Alamos buy at $29) by supplying the price-action confirmation and the valuation floor those two lacked — while explicitly discouraging a chase. No rating stamp, no new price or target, no portfolio tables.
"AI is fundamentally an energy story" — the VST POW (Aug-14). The Friday POW! states the power thesis as a first principle and then buys the generation: "at its core, the AI story is fundamentally an energy story. Who will win? How will it be (effectively) regulated? Will it massively displace labor? All these questions and handwringing are secondary to the fact that AI is going to require an enormous amount of reliable power." The vehicle is Vistra (VST) — ~40,000MW of nuclear, gas, solar and battery across ERCOT and PJM ("equivalent to around 40 large-scale nuclear power plants"), the nation's second-largest competitive nuclear fleet — at ~$145, 36% off the $219.82 high on a ~$58B cap, and the verdict is unqualified: "We are buyers of VST at today's price." Structurally it is the Aug-7 GILD template run on a different distortion: the $4.02B revenue "miss" (vs $5.46B) is unrealized mark-to-market on forward hedges — "real in the GAAP sense and irrelevant in the economic sense" — and the program causing it is what leaves ~100% of 2026 generation hedged, i.e. the optical cost and the earnings certainty are the same fact. Underneath: adjusted EBITDA $1.767B, +30% (a beat), generation EBITDA +68%, 97%+ availability through record heat. Two contributions are new to the archive. First, an evidence standard: "PJM and ERCOT set all-time peak loads of… 168 and 91 gigawatts in July. That is the structural AI demand proof the thesis requires, visible in real-time load data rather than forward projections" — a meter reading rather than a consultant's forecast, and the cleanest answer yet to his own long-running scepticism about data-center financing (he separates the power demand, which he underwrites, from the capex funding, which he doesn't). Second, a valuation method: FY26 guidance of $6.8-7.6B EBITDA "explicitly excludes" Cogentrix ($4.7B, ~5,500MW), the Meta nuclear PPAs (~2,600MW PJM), the Helix Platform (KKR + NVIDIA + the Kuwait Investment Authority, VST committing up to $1B as "preferred power partner") and nuclear PTCs — so "the 2026 guidance midpoint… is the starting point, not the destination… the 2028 trajectory almost certainly exceeds $9 billion annualized. At a $58 billion cap, the market is pricing none of it," implying ~$5.4B normalized FCF, an 11.4% yield, at 10.6× EV/EBITDA vs Constellation's 13.6×. Against a bearish WSJ "Heard on the Street" piece he concedes the scarcity-rent cap and relocates the moat to physics — "the bottleneck is more than just policy, it is the physical ability to deliver reliable megawatts" — so "the thesis is evolving, not breaking: less about passive scarcity rents and more about who can actually deliver incremental dispatchable power," which is the sharpest update to the standing power/uranium/gas leg since the Jul-8 microreactor note. Bear case is three-legged (ERCOT curve softness, Helix execution, PJM IRAS) and "justifiable only if all three occur simultaneously"; the build-out risk is answered by rate rather than denial ("could slow meaningfully, but… almost certain to continue at a rapid clip"). The honest caveat inverts the GILD entry: VST has been below its 200-day for over a year, stabilizing near 150, so the discipline offered is a stop at 130 rather than a confirmed retest.
Mid-caps as the destination for the margin call (Aug-17). The Monday Portfolio Update turns the summer's earnings-quality charge into a relative-value position, and reviews two held names. DGX, recommended Jun-5 at $200.29, is ~$236 — "roughly 18% in just over two months, plus the $0.86 quarterly dividend" vs the S&P's ~6%, on a Jul-23 Q2 print that moved it 11.15% in a day. The entry is disclosed as a repeatable tactic: Haymaker "missed the key upside range expansion" but caught the "retracement to near the breakout point not that far above $180… also smack dab on the 200-day moving average," because "buying relatively close to the resistance penetration spot and at the 200-day is typically a profitable tactic." Consensus $239-246; Argus/BofA $260-265 needs "another beat-and-raise on October 27," with the Corewell JV lab (2027) and double-digit advanced diagnostics "not yet fully in run-rate numbers." Valuation and chart are "undeniably extended," so the instruction splits: "hold the core position through October 27… trim only if the position has grown to an uncomfortable share of your portfolio's weight" — or "sell one-third" if raising cash. IJH is re-underwritten after lagging: +30% / +33% TR / ~12½% CAGR since Mar-18-2024 vs the S&P's 50%. The decomposition is the point — cap-weighting (top 10 = 38%), and the S&P's forward P/E falling21× → 20× across that return, i.e. all of it was earnings, and those earnings lean on "massive gains on their holdings in AI entities like Anthropic" (>50% of GOOG's 1H profits), "immense, though unsustainable." So: "a looming mean reversion in the S&P 500's profit margins… on a normalized profit-margin basis, the S&P 500 is far more overpriced than it superficially appears." Second structural leg, new to this archive — index graduation as a permanent drag: mid-cap indices hand their parabolic winners to the S&P at peak market cap (SMCI, Haymaker's own 2024 example, promoted "almost as soon as we hit 'Send'" and down "nearly 70%" since; Tesla "much more impactful — and controversial"), negligible per event but "multiplied by dozens, if not hundreds." Caveats volunteered: mid-caps at 17× forward are "close to the peak this decade," and "margin debt having gone vertical" heads "an unusually large number of threats," so "we'd be cautious even with Mid-Caps" and "wouldn't quibble with booking some gains on IJH." The conviction is placed on the horizon the argument actually works on: "over the next 10 years, we are adamant that Mid-Caps will dramatically outperform the world's most beloved equity index."
Let the winners run, trim the parabolic one (Aug-24). A four-stock Portfolio Update that states the house's valuation method for cyclicals and its trim rule more plainly than any prior issue, and volunteers the base rate first — "three are winners and one is a loser… about the hit vs miss rate our picks have had overall since 2022." EOG delivered the predicted break "above overhead resistance extending back to the powerful energy sector rally during the early stages of Russia's attack on Ukraine," though "not yet decisive" until it holds above the upper $140s; from the Jun-17 "vehement endorsement" it is $133 → $151, "over six times greater than the return on the S&P." FANG, bought Apr-7 2025 at $121 into the Liberation/Obliteration Day panic, is +~75% vs the S&P's ~50%. Both stay untrimmed because all three trim-test conditions still pass — "the chart patterns, the valuation, and the fundamental stories all remain highly encouraging" — so "we are inclined to sit tight and, hopefully, let them run" (Super-Major takeout named and discounted as "always longshots"). The P/S doctrine gets its worked example: in 2022, $100 oil "more than doubled" FANG's EPS, "deflating the P/E to an unsustainably low point" while P/S fell "to a significantly lesser extent" — hence P/S is "more reliable with companies where profits can fluctuate considerably." NEM is the control case: "it's trim time." The thesis is intact (the out-of-consensus "$5 EPS estimate is low" call landed at almost $7; the only-gold-miner-in-the-S&P catalyst "has undoubtedly happened"), but two of three conditions fail — the run is "straight up" so "a non-trivial retracement strikes us as highly probable," P/S "looks a bit stretched," and the relative check is measurable: NEM is "back at its early 2025 peak whereas the senior gold miner ETF, GDX, remains about 10% below its high." A trim, not a sale. BABA is the loser, defended on evidence rather than price: down 20% from the May-14 $162.11 recommendation, the EPS miss reclassified as "the accounting cost" of AI capex, with cloud +45% (fastest in 22 quarters), Qwen at 3bn downloads / >50% of all open-source model downloads (Llama beaten 13:1) and a PEG of 0.59 — "the discount has overshot," buy maintained but "for investors who… can stomach a hefty dose of volatility. For those who don't fit this description, we'd suggest trimming or exiting." Post-publication: a $10.2bn share issuance at an 8.4% discount, all for AI infrastructure, "nearly three times oversubscribed." Housekeeping only on the tables — "PALL Cost/Date corrected."
The YCC week — Bessent vs the long bond, and the profits that are really fiscal (Aug-25/26/27). Three consecutive Dailies build one argument, and it is the macro spine under the whole hard-asset book. Aug-25 — "a titanic struggle… between U.S. Treasury Secretary Scott Bessent and the long-term government bond market," with Hartnett's 5% 30-year as "the Maginot Line" (a line bypassed, not breached) and Druckenmiller's WSJ op-ed supplying the principle — "governments defending prices against fundamentals always lose… let the bond market speak." The anchor is empirical: the 30-year yield historically tracks nominal GDP; post-GFC suppression made bondholders "the bag holders for America's fiscal and monetary recklessness," delivering "one of the worst return phases for U.S. Treasury investors on record… despite (or because of) long bond yields consistently running below inflation." The fork has no benign branch: YCC "will almost certainly push real assets even higher," while letting yields clear "materially higher" is "extremely adverse" for everything — "if Druckenmiller's advice is followed, the pain will be immediate. If not… a much stiffer eventual price tag." Aug-26 scores Bessent against his own Three 3s (late Nov-2024): GDP ~2% vs 3% (credited as a near miss), oil +400k b/d vs three million — and the miss came "despite the incentive of higher prices caused by the war against Iran," which reads as evidence about the resource base and corroborates the standing shortage thesis — and the deficit "near 6%… basically double" target, with DOGE walked down $2T → $1T → ~$200bn while outlays rose ~$300bn. Per Jim Grant, publicly held Treasury supply is +8.4% Y/Y — implying "the government's accounting is overly flattering to the actual deficit" and making yield suppression "challenging in the extreme." Precedent: Japan's YCC left the yen −40% vs the dollar in a decade and −80%+ against gold — the gap between those two readings being the point. Aug-27 converts it into an earnings argument: 6%-of-GDP deficits vs a normal 2% generate "around 2% of artificial economic growth" even at a conservative 0.5 multiplier (i.e. roughly all of measured GDP growth), and under the Levy-Kalecki Profit Equation "a $1 trillion jump in the federal deficit inflates corporate profits by a similar amount" — so reported margins are partly fiscal, which is the mechanism beneath the Aug-17 claim that "on a normalized profit-margin basis, the S&P 500 is far more overpriced than it superficially appears." The second stimulant is $700bn+/yr of AI capex funded by "a deluge of hyperscaler debt raises" — a "reverse crowding out" pushing the government's own borrowing costs up (NVDA's Q2 confirming the spend "will continue to run white hot"). Per a Columbia Business School chart relayed by Luke Gromen, no prior efficiency-spending era has been this large a share of GDP. The verdict splits: the AI bull case "may well turn out to be the case," but excessive government spending "will turn out to be a big disappointment" — growth "dramatically decelerated over the last 25 years while deficit spending has exploded," and "the long-term bond market is clearly beginning to sniff out the unsustainability."
The AI-power trade, one step down the stack — the GNRC POW (Aug-28). The Friday POW!, written with frequent contributor Daniel Bustamante, buys the equipment supplier rather than the generator of the electrons: Generac (GNRC), positive, "an intriguing play near the $200 area." The thesis is a re-classification — "Generac is no longer simply a residential generator company dependent on hurricanes and power outages" but one building "a second earnings engine in Commercial & Industrial power equipment," such that "the company can produce meaningful earnings growth without requiring both businesses to peak simultaneously." Its position in the AI trade is deliberately the other side of the consensus worry: "while others are worrying about CapEx spend on the hyperscalers Generac has been, and will continue to be, a net benefactor of said spend," because "every large-scale data center requires multi-megawatt backup generation capable of carrying the full facility load through grid outages," at a "scale (and service network) that new entrants cannot replicate quickly." The evidence is order flow: Q2 C&I sales +29% to $556M after +28% in Q1 ("no longer based on a single quarter or a hypothetical pipeline"), an approximately $1.6B data-center backlog (+~$1B since the prior update, $1.35B scheduled for 2027, and excluding the second hyperscaler's committed volume), ~$1B of new orders in the 90 days to Jun 30 ("roughly $11 million per day"), guidance raised to $450M of FY data-center revenue, and capacity bought against contracts (Enercon for controls/packaging plus an Illinois large-megawatt plant) — "investing against contracted demand, not merely building capacity in anticipation of demand that may never arrive." The re-rating is a mix shift: C&I ~47% of Q2 external sales vs ~41%, so it "increasingly resembles a diversified power-infrastructure company," and such businesses "generally receive higher and more durable valuation multiples than highly cyclical consumer-equipment companies." Residential becomes optionality — sales −2%, a 34.7% EBITDA margin of which ~9pp was tariff refunds, with a probable "Super" El Niño as the free call ("the current thesis does not require a major hurricane season"). Consensus has revenue +17.4%/+21.3%, adj. EBITDA $690M → $1.01B → $1.19B and adj. EPS $5.14 → $9.75 → $11.82, at ~18.2× forward earnings / 2.7× sales where "the stock's multiple has recently compressed, suggesting earnings expectations, not valuation expansion, are now doing most of the work," and a $500M buyback authorization sits entirely unused. Entry is the confluence after a +90% run and a one-third correction: the 200-day plus the 50% Fibonacci retracement "which already found buyers back in July." Follow-up (Sep-21): the thesis was "validated" by an 8-K Amazon supply agreement for up to $8B of data-center backup generators through 2033, with Amazon able to buy up to 1.7M GNRC shares at $200.93 only as it places orders ("that alignment is the moat"). The pick is ~+10% at ~$205, and the Buy is reiterated. The same issue doubles down on EQT, a Strong Buy at 12.4× P/E despite ~−23% from entry, as the cheap gas-supply route into the same AI-power demand (~50 GW of gas plants planned; Japan funding ~18 GW).
Don't trade the headline, buy the rebuild — the HAL/SLB switch (Aug-31). The Monday Portfolio Update makes the energy leg's most concrete move since the coal and gas calls, and it starts by refusing a headline. Trump's "THE BIGGEST OIL DEAL IN HISTORY!" (Venezuela, "MORE THAN DOUBLES American Oil Reserves") is dismantled on three specifics — time ("a decade or so before they are ready for harvest"), law ("selling Venezuela's oil reserves to a foreign entity violates its constitution," JVs and production-sharing excepted) and counterparty memory (decades of expropriation; even Exxon's won judgements uncollectable — "they may not hurry back, despite the mammoth upside") — with the motive named: "a cynic could view this as the latest move… to depress oil prices in the run-up to the mid-term elections," of a piece with the Hormuz-is-open posts "front-run by not-so-mysterious insiders." So "meaningful production gains are unlikely to have any practical impact on current oil prices" — but the service companies get paid either way, and the far bigger job is "repairing the Middle East's war-ravaged production facilities." The standing depletion view is restated at maximum strength — supplies "enormously depleted" by the attacks plus "the loss of around half of the normal shipments through the Strait of Hormuz," "the greatest oil-supply shock in history" — so firms with "the unique set of capabilities to rebuild global oil stocks should have a number of robust operating years ahead of them." The trade is a pair: HAL added to the Buy List as a STRONG BUY at $36.79 (stuck under ~$43 for years, knocked to the low-$30s in late July on a guided-down Q2, "rarely traded as inexpensively as it is now on a P/E basis" — with the P/S cross-check volunteered, "cheap, but not as decidedly," and resolved as trough earnings: conditions "far from boom times," so "neither revenues nor profits are maxed out"), funded by trimming SLB back to Holds/Trims at H/T after $45 → ~$60 — "harvesting some of the latest appreciation might make sense, and an obvious redeployment candidate is HAL," though "we wouldn't, by any means, suggest exiting SLB in full." Both "poised to experience that happy combination of rising earnings and expanding P/E ratios." Full portfolio tables published (HAL new SB; SLB H/T; CRH, EOG, GNRC, DVN, EL, GDXJ also changed).
The Fed was dragged into hiking — expect more (Sep-17). The Jun-11 "one hike wouldn't be that deleterious" scenario arrives: new Chair Kevin Warsh, Trump's pick, delivers "the first tightening by the Fed in three years" — "with extreme reluctance," because with prediction-market odds at 85–90% skipping would have been "a disturbingly dovish signal." The evidence is the 2-year/fed-funds gap: funds 3.75% vs the 2-year at 4.67% (~90 bp), "highly unusual" on a 20-year Bloomberg chart where the 2-year leads. With Gundlach's view that the 2-year sets rates better than a politically exposed Fed, "brace… for several more hikes" — the only fast off-ramp being "a sudden and severe correction in stock prices" that Trump would pin on Warsh. Reinforces the Sep-14/15 raise-cash stance: higher short rates make cash pay while stocks carry the rate risk.
Raise cash now, from the weakest names first — "A Confluence of Negatives" (Sep-14/15). The most explicit defensive call of the year, bounded — "This is NOT a call to exit the stock market" — and given a sourcing rule: review "weaker names (like those rated Hold/Trim, Trim of Sell) as sources of cash," selling "in IRAs and/or 401(k)s" as the tax-efficient route. Four negatives: energy measured at the price users pay (diesel >$200/bbl, double crude — stagflation risk); the war in its eighth month with the Houthis striking Saudi Arabia's East-West Pipeline (4–5 mb/d), leaving most of the pre-war 20 mb/d Gulf flow off-line while stocks sit near a record; an AI backlash; and a global long-bond breakdown (TLT new lows, 30-yr UST 5.33%, France's debt-trap arithmetic — debt service +25% vs GDP +3%). The same issue cuts AAP from Buy to Hold (−24.6% post-Q2; turnaround intact, sales must stabilize; weekly close >$48 the trigger), routing it into that cash-source bucket. The Sep-15 Daily supplies the history: 10-yr 0.5% → 5% since 2020 while the S&P returned 16%/yr, a stock-over-bond gap above every peak since 1926 on Paulsen's chart — the 1986–87 analogue (yields 7.2% → 10%, stocks +40%, then the crash) and the late-1990s gap (then 2000–03) — so "cash is definitely not trash."
Taking gains is the product too — second trims on CKHUY and COPX (Sep-11). A Friday POW! with no new name: "a big part of our value-add… is ensuring that you take gains when certain positions have had a significant run up." COPX (highlighted at $46.67, first trim May-15, now ~$90) gets another 25–33% trim with the core kept — "the thesis… was correct… what has changed, in a big way, is the price," with $331M of one-month outflows read as institutions "selling into the strength" and a breakout above $99 the reason to hold the rest. CKHUY (bought $6.78, trimmed $8.95 in July) is trimmed again because the price is flat since then while "the geopolitical complexity has increased rather than decreased" — Panama voided the Balboa/Cristobal concessions, Cosco + MSC joined the BlackRock consortium, and a 259% premium to some fair-value estimates may be "overpricing deal certainty." The sell side of the Aug-3 raise-cash stance, applied one position at a time.
The healthcare sleeve gets its fourth name, and a free option — the INCY POW (Sep-04). The Friday POW! is Incyte (INCY) at ~$129 — "We recommend a Buy on INCY" — and it is the cleanest example in this archive of the house's free-option structure: buy a business whose base case already clears the price and take the contested asset for nothing. The valuation does the underwriting — ~16.5× the 2026 EPS consensus of ~$7.50, and an EV of ~$22.4B (cap less $4B net cash) on $5.2B of guided FY net sales = "approximately 4.3x EV/revenue, which is modest for a business growing net sales 40% year-over-year," with the Price/Sales yardstick re-run net of cash ("headline ~4.4x… reduces the more meaningful EV/sales multiple to ~3.6x"). The option is priced by the sell side at zero, and Haymaker reads the target list as the disclosure: "the consensus target of $127.13 is essentially at the current price… a Jakafi-plus-Opzelura base case with limited KRAS credit," while Leerink $155 / Canaccord $152 / H.C. Wainwright $150 "are the analysts adding in KRAS optionality" — so "the market is paying almost nothing for a program that is showing 62% to 75% response rates in one of the most lethal cancers in existence." That programme is KRAS G12D in first-line pancreatic cancer (~40% of PDAC; five-year survival "below 12% for all stages, below 3% for metastatic disease"; the target "considered undruggable for three decades," with G12C in lung cancer having "proved the mechanism could work in humans"): the DAWN-303 EU registry shows 62.5% response / 95.8% disease control with GemNabP and 75% / 100% with mFOLFIRINOX against a 20–30% chemotherapy base rate — "less of a lottery ticket and more of a practice-changing drug." What makes it affordable is the platform beneath it: Jakafi at $817M/qtr (~$3.2B annualized, shared with Novartis ex-US), Opzelura at $450M/qtr where "each approved indication adds a new patient population at minimal marginal cost," 10 Phase 3 studies, four approvals expected inside 12 months and $4B+ net cash with no meaningful debt — so "if it does not [hold], Incyte absorbs it from a diversified multi-franchise platform rather than a binary single-asset position." Two disciplines are on display and both cut against the author: the Q2 beat is deflated in its own sentence (EPS $3.09 vs $1.84, a 68% beat, "stripping that out" the $246M non-cash Medicaid settlement leaves ~55% operational), and the entry is conceded as late — "frankly, we are a bit late to this one," the breakout having fired into the $90s late last year — but excused because "the appreciation since then hasn't been excessive and… it's been earnings-driven," with a level attached: stretched over the 200-day, "a retracement back to $110 or so would not be at all surprising. More conservative investors might want to hold off." The bear case is stated in full — the Jakafi patent cliff (US compound protection expires 2028, formulation patents only "potentially" to 2033) on ~55% of net sales, KRAS data that is "a small interim cut" where "Phase 1 combination data has misled markets before," and a stock "not a margin-of-safety purchase" after +49% in 12 months. And it is explicitly sector construction rather than a one-off: "adding Incyte to the Haymaker portfolio is the continuation of the healthcare sleeve we started building with the inclusions of MDT and DGX in early June, as well as GILD a month ago. We believe healthcare and biotech are entering (or are in) a bullish recovery mode that will continue for the foreseeable future" — the Jun-12 "broad rotation into the space" now a four-name overweight, expressed through four uncorrelated drivers (devices, labs, large-cap pharma cash flow, and a clinical option). Catalysts: ESMO in October, then earnings Oct 27. No portfolio tables published with the issue.
The other side of the same trade — the commodity bull nobody owns (Sep-03). The mirror of the Sep-02 note, published the next day and best read as one allocation with it. Two questions, answered separately. Technical: is the 2026 high near 740 on the Bloomberg Spot Commodity Index a double-top with the 2022 spike? The tell is not the depth of the drawdown but its terminus — "the sharp correction from that spike ended right at the prior upside resistance level," the ~500 shelf that capped the index at both the 2008 and 2011 highs, i.e. a ceiling converted into a floor. 2026 repeats the sequence one level up (the "joint American/Israeli attack on Iran," an overshoot, then a correction "less severe than in 2022"). The bear case is conceded and then rejected on fundamental grounds, not chart grounds: "given the lengthy list of critical commodities in extremely short supply, and factoring in rising demand, we would argue for the latter interpretation" — the worked example being palladium and the Aug-21 POW!PALL reiteration. Integrity note: he grants that in 2022 "the feared shortages from that geopolitical shock never fully materialized." Valuation: the FactSet/Jefferies GSCI-to-S&P ratio back to 1970 sits near 0.7 — below the 1999 Tech Bubble trough and the low of the whole 56-year series, against peaks of ~9.5 (Gulf War), ~8.3 (2008) and ~7.6 (oil embargo). Which forces the note's only unconditional claim, stated as an identity: "either stocks need to fall precipitously or commodities need to continue their ascent, or some combination of both" — with the direction deliberately left open, since Sep-02's flow argument supplies it. Positioning is unchanged but explicitly two-sided: hard-asset exposure continues, "though we anticipate ongoing volatility," worked with "timely profit-taking suggestions on spikes, and buy-up guidance on weakness, with both the underlying commodities and their producers" (none named). Closes on Jeff Currie — "Scarcity is in the physical world. The illusion of abundance is behind us" — and nominates oil as where the reality check lands first, "as physical shortages become increasingly difficult to dismiss with social media posts."
The international leg, re-argued on flows — and the breadth complaint retired (Sep-02). A short allocation Daily that upgrades the long-standing "bullish on international markets" bias from preference to evidence, and does it by abandoning the house's own favourite bear argument. The narrow-market case is explicitly not available: the S&P is "up roughly 12 ½% this year on a total return basis," the advance is "broader than in the past," and "the equally weighted S&P index is actually slightly ahead." What replaces it is two anomalies. First, the professionals all trail a broad tape regardless of style — "both growth- (Cathie Wood) and value-oriented investors (David Einhorn and Bill Ackman) are lagging" (per an @AskLivermore table relayed by Trader Ferg: SPY +12.6% vs GLRE +8.5%, ARKK +8.0%, BRK.B +1.6%, PSH −0.8%, IEP −9.4%, MSTR −19.0% as of Aug 28) — which rules out style rotation and points at an index concentration no active book could hold. Second, and the real finding: per Gerard Minack, "the U.S. market itself has been lagging overseas indexes since the start of 2025," and Hay states plainly that this is "surprising" given the "mammoth inflows into U.S. equities due to the AI boom/mania" — sized by Michael Hartnett's BofA chart at a record ~$216bn of 2026 tech-fund inflows against a prior ceiling near $80bn. An asset underperforming while absorbing its largest inflow on record is being held up by the flow alone, so the conditional resolves itself: US shares trail overseas "to an even greater extent should — we'd argue, when — these epic inflows shift into outflows," with the damage concentrated in "the fully invested U.S. index funds which are also highly concentrated in AI-related shares" — no cash buffer, cap-weighted forced selling into the most crowded names. Positioning closes it: Americans are "dramatically underweight foreign markets" and "few seem to realize the long-overdue rotation into international equities is well underway," with "a long list of attractive overseas stocks" in the coverage universe but none named here.
Uranium — the ignored shortage (Jul-16). A short, contrarian Daily flags "inexplicable… complacency" in uranium even as oil/LNG shortages dominate headlines. Per a Trader Ferg chart, nuclear utilities have contracted for less uranium than they consume for over a decade (since the 2011 Fukushima overreaction shut 65 plants), so "excess inventories are now fully depleted" — just as 70+ new reactors are planned (mostly Chinese/Russian designs) and 16+ mothballed plants restart. The leading uranium ETF SRUUF (a 2-lot Buy-list holding) has pulled back to ~$18–20 after twice hitting $25, while spot uranium is ~$85 (term prices higher) — so Haymaker sees prices "poised to surge again" once utilities recognize the deficit. (Extends the nuclear/data-center-power thread; feeds the master Uranium/nuclear theme.)
Defense joins the contrarian book — LMT alert (Jul-9/11). The Ringside-Access Trading Alert (re-posted to the paid tier Jul-11) applies the oil mispricing template to defense: bears have decided "the shooting war is in the rear-view mirror," leaving Lockheed Martin (LMT) flat-lined four years at one of its lowest P/S & P/E of the decade — 17× clean earnings vs the S&P's 21× (index "E" inflated by heavyweights' non-recurring private-markup gains, e.g. Google). The engine is the global Patriot MIM-104 shortage (Ukraine nearly out; even the US short → production strong "for years to come") plus surging worldwide defense budgets few non-US, non-China producers can serve at scale, with the Ultra Maritime acquisition starting the drone/naval next-gen pivot. Bullish — a run to a new ATH north of ~$700 "highly likely"; the defined kill-switch is a decisive break above 4.6% on the 10Y / 5% on the 30Y (America unable to keep deficit-funding "defense").
Quality-compounder POW (Jun-26). The latest Friday POW! is Travelers (TRV), a Buy for long-term compounders — a "defensive" P&C insurer compounding like a growth stock (135% five-year return). It has repriced its whole book to "rate adequacy" (22.7% trailing core ROE, 85.3% underlying combined ratio), earns a rising tail on a $106B fixed-income float in a higher-for-longer world ($3.3B after-tax NII guided 2026), and returns capital at a ~20% annualized yield (14% dividend hike, 14th straight year). At ~10.8× NTM it trades below peers Chubb (12.3×) and Progressive (12.8×) despite a higher ROE, with a re-rate toward ~13× the upside; demand ties straight to the three-cycle regime (reshoring complexity, energy-security spend, debt-cycle liability). The Barclays Underweight (Jun-12) at the 52-week high is the bear case; Haymaker would use a likely ~5% pullback to buy more aggressively. A long-term hold, not a trade.
Red Sea reality check (Jun-18). The Hormuz-reopening euphoria over an Iran peace treaty is misplaced: Suez + Bab el-Mandeb transits remain ~50% below pre-2023 (pre-Houthi-attack) levels, the Mideast cease-fire has been an on/off affair, and Iran is likely to stay belligerent — so combined with Ukraine's drone strikes on Russian oil & gas facilities, Haymaker expects "another resounding rally by crude prices" (energy recovery plays teased for the next POW!). Reinforces the energy-bullish stance.
The Fed is a sideshow (Jun-11). Year-end hike odds went 0% → 30%, but one hike wouldn't be "that deleterious" — energy "way outweighs the Fed" (though a persistent energy crisis could force it to tighten far more than once). The shutdown is ~1B barrels lost heading to 1.5B with no demand destruction yet (flight data strong), so a major price spike has to do the rationing; natural gas stays his highest-conviction long — a ~90% US discount to global LNG, 37 GW of data centers under construction, 20% of global LNG offline, bearish positioning, gas "still trading in the threes."
The "run-it-hot" 2026 thesis (Jan-5). The year-opening Making Hay Monday frames 2026 as "the mother of all mid-term manipulations": a desperate Trump pushes a pliable Fed toward a deeply negative real rate (wants 1% funds vs ~3% inflation) and a fiscal flood (bonus depreciation, ~$100B refunds, ~$500B tariff rebates, cash-for-clunkers, an affordable-housing push), with Michael Howell's "liquidity dam" breaking out of markets into the real economy (Main Street over Wall Street). Reconciles record asset prices with 65-yr-low UMich consumer sentiment via the "K-shaped"/"Special K" economy (top 10% drive ~50% of spending). Playbook: own the "run-it-hot" beneficiaries — the Equity Highlight is Jacobs Solutions (J), a cheap "picks and shovels" reindustrialization/data-center play (record $23B backlog, fwd P/E ~18 vs trailing distorted by one-time charges, ~$230 3-yr target), plus the deep-value energy sector (XLE, <3% of the S&P). Portfolio housekeeping: trim ⅓ of HP after a ~double (funded earlier by halving SLB at a ~14% loss), and DCA PBR up to ~1.5%.
The product — Haymaker
What it is: a Substack newsletter (haymaker.substack.com — Haymaker Publications, including the written "Haymaker Daily" posts) by David Hay, retired co-CIO of Evergreen Gavekal. He's now only a passive owner of the SEC-registered adviser with no role in its research or investment management (the posts carry that disclosure), so Haymaker is a personal recommendation letter, not the firm's product.
From his own descriptions in the appearances and posts archived here (esp. 2026-JUN-05, 2026-JUN-09).
Trade type
What it is
How he runs it
Seen in the index
Newsletter recommendations
Explicit buy calls — and "rare sells" — on markets and sectors, published in the letter.
Recommended Korean stocks before they "went ballistic" — EWY was "a longtime favorite of this newsletter" — then issued "one of our very rare outright sell recommendations" on May 6th at $151; early, then vindicated (−17% by 2026-JUL-29, with "much more to come on the downside").
EWY
Tactical round trips
Trading calls around sentiment extremes and blow-off tops.
Silver: bullish 2024–25, called the >$100 move a blow-off top, issued a trading buy after it was crushed, then suggested selling the rally.
SLV (and TLT, held as "only a tactical trade")
Macro rotation themes
The structural book: out of the over-concentrated "bubble 4.0" index into energy (especially cheap US natural gas and coal), financials, fertilizers, EM and small/mid-caps.
Argued in the written posts (deficits, energy supply shock, mega-IPO supply wave).
Short, dated reviews of names already recommended — the ratings-change vehicle between the Friday POW! deep dives. As of 2026-JUL-27 the stated cadence is "two or three portfolio updates each Monday": "This allows us to provide overviews on a wider range of securities we've put before you over the last few years. As a result, these will be shorter and, in some cases, we'll follow-up with a more in-depth review, similar to what we did with Yancoal and New Hope last Friday."
Each update states the rating move in prose (Buy / Strong Buy / Hold / Trim / Sell) and the published Buy-List and Holds/Trims tables are reprinted at the end; the ratings history of a single name is often laid out in full (AESI: Buy 18 → DCA to 8¾ → Hold → Trim → Buy).
2026-JUL-27: AGI + AESI both back to the Buy List; Jul-13: BA → Hold; Jun-29: AA Hold, BABA → Hold
Friday POW! + the open book
A weekly "Pick of the Week" deep dive (thesis, valuation, technicals, an "Arguing the Other Side" section), with the full Haymaker portfolio published at the end of each POW — a Buys list and a Trims/Holds list showing every lot's cost basis, current price, % change and rating.
Ratings move B/SB → H → T — and now S (Sell): the Jun-12 post retitles the second list "Holds, Trims, Sells" and carries the first S (IBM). Dated changes per lot; the portfolio is captured in portfolio.json and drives the master table's ◆ filter.
A CIO's discipline at Substack prices. Decades of running an RIA's book, delivered as a consumer newsletter with named tickers and sectors.
Sell calls, not just buys. The "rare sales" (Korea, the silver blow-off) give subscribers exits — the half of the trade most newsletters skip.
Written and regular. Haymaker Daily posts are short written pieces (archived as "posts" below), so the reasoning is on the record and checkable after the fact.
Posts
One dated page per issue — each has its talking points and the saved article text. Newest first.
Friday POW! — Generac (GNRC), "Generating Power... And Profits": the AI-power thesis moved one step down the stack from Aug-14's VST — from the utility that generates the electricity to the manufacturer of the multi-megawatt backup generators every data center must install behind the grid. Written with frequent contributor Daniel Bustamante. The thesis is a re-classification: "Generac is no longer simply a residential generator company dependent on hurricanes and power outages," but one developing "a second earnings engine in Commercial & Industrial (C&I) power equipment, with data-center demand accelerating faster than the market previously expected" — so "the company can produce meaningful earnings growth without requiring both businesses to peak simultaneously." The positioning is contrarian by construction: "while others are worrying about CapEx spend on the hyperscalers Generac has been, and will continue to be, a net benefactor of said spend," because "every large-scale data center requires multi-megawatt backup generation capable of carrying the full facility load through grid outages," built at a "scale (and service network) that new entrants cannot replicate quickly." The acceleration is two prints, not one: "second-quarter C&I sales increased 29% to $556 million… this followed 28% C&I growth in the first quarter, meaning the acceleration is no longer based on a single quarter or a hypothetical pipeline which was the market concern just over a year ago." The visibility is the product: an approximately $1.6 billion data-center backlog, "an increase of roughly $1 billion since its previous update," which "does not include committed volume from the company's second hyperscale agreement," with $1.35 billion scheduled for 2027 — "a significant portion of next year's commercial and industrial revenue is already contracted" — fed by "approximately $1 billion in new data center orders in the 90 days ending June 30, or roughly $11 million per day," full-year data-center revenue guidance raised to $450 million, and CEO Aaron Jagdfeld calling the market "unlike anything we've seen before." Five capital actions are read as one signal — ~$700M of committed 2027 volume finalized with one hyperscaler, a global supply agreement with a second, 2027/2028 product terms under negotiation, Enercon acquired (generator controls and packaging) and an additional Illinois large-megawatt packaging facility — "these actions indicate Generac is investing against contracted demand, not merely building capacity in anticipation of demand that may never arrive." The re-rating mechanism is a mix shift, stated as a rule: data centers add visibility, scale and diversification, and "C&I represented approximately 47% of second-quarter external sales, compared with roughly 41% in the prior-year period," so Generac "increasingly resembles a diversified power-infrastructure company rather than a pure residential generator manufacturer… infrastructure businesses with visible backlogs and secular growth generally receive higher and more durable valuation multiples than highly cyclical consumer-equipment companies." Residential is demoted to optionality and its margin disclosed with the distortion attached: sales −2%, adjusted EBITDA margin 34.7% "although roughly nine percentage points came from tariff refunds," with "favorable mix and operating efficiencies" the durable half — "the current thesis does not require a major hurricane season," while a "Super" El Niño "this year and next seems probable" and retail flows chase weather events ("see the 2021 Texas freeze storm"). The numbers: revenue +17.4% (2026) / +21.3% (2027) to ~$4.94B then ~$6.0B; adjusted EBITDA $690M → $1.01B → $1.19B (margin 16.4% → 20.5%); adjusted EPS $5.14 → $9.75 → $11.82; FCF ~$353M then ~$478M; $1.33B debt vs $265M cash, EV ~$13.0B — "GNRC is entering a new earnings-expansion cycle." Valuation is doing none of the work: "roughly 18.2x forward earnings and 2.7x trailing sales, well below its 2021 peak… the stock's multiple has recently compressed, suggesting earnings expectations, not valuation expansion, are now doing most of the work," and the $500 million buyback authorization approved Feb 9, 2026 is entirely unused. The entry is technical and explicit: after "a +90% run up to start the year" and "a one-third correction of its previously overbought condition," the stock "is hovering right on the 200-day moving average" and is "testing the 50% retracement (from the December lows to the June 2026 highs) which already found buyers back in July." Verdict: "the story boils down to one question: 'How big can the C&I business become?'… now that it's pulled back, it provides an intriguing play near the $200 area." No portfolio tables with this issue.
Haymaker Daily — Steroids On Top Of Steroids: the week's payoff note, and the one that turns the fiscal argument into an earnings argument. Two stimulants are propping up growth and profits, and "there are substantial near-term benefits that make it hard to kick the habit." Dose one is the deficit: "when federal red ink is running at four percent above normal — say, 6% of GDP deficits vs. a more typical 2%… even assuming only a 0.5 'multiplier effect' this would still amount to around 2% of artificial economic growth" — set against Aug-26's "roughly 2%" measured GDP, essentially all of it. The under-appreciated half is profits: "in theory, there is a one-to-one relationship, per the widely accepted Levy-Kalecki Profit Equation. Accordingly, a $1 trillion jump in the federal deficit inflates corporate profits by a similar amount," qualified honestly by "negative offsets such as elevating trading deficits, as well as raising interest rates as surging government borrowing 'crowds out' the private sector" — which is the mechanism beneath the Aug-17 normalized-margin claim. Dose two is AI: "$700 billion plus annual expenditures on the great data center build-out," "highly unlikely to be maintained longer term," and financed in a way that creates the note's cleanest idea — the "deluge of hyperscaler debt raises" is "driving up the government's borrowing costs… in other words, this is a reverse 'crowding out'," a third force on the same long yield alongside the deficit and Aug-26's issuance surge. (NVDA's Q2, "reported yesterday," is cited in a single parenthesis as evidence the AI spend "will continue to run white hot" — no stance on the shares.) The historical check is borrowed twice over: "per the following graphic from the Columbia Business School, as relayed by Luke Gromen in his August 14th newsletter, prior eras of excessive spending to finance efficiency/technological breakthroughs have never represented such a large share of GDP" — raising "the reasonable question of what happens when these double doses of amphetamines inevitably fade." The verdict splits by evidence: the AI bull case "may well turn out to be the case," but "the odds are that much higher excessive government spending will turn out to be a big disappointment," because "America's economic growth dramatically decelerated over the last 25 years while deficit spending has exploded." Closing line: "the long-term bond market is clearly beginning to sniff out the unsustainability of this situation."
Haymaker Daily — O for 3 On The Three 3s: the sequel to Aug-25, and a lesson in the least glamorous method there is — score the policymaker against his own published targets. The Three 3s, unveiled "in late November 2024": "a 3% real GDP growth rate, three million barrels per day of increased oil production and… a 3% budget deficit. Unfortunately, the administration has achieved none of those objectives." The near miss is credited first, which is what makes the rest land — "to be fair… the GDP target hasn't been 'bigly' missed. That has averaged roughly 2%." Then the whiffs: "U.S. oil output has increased by just 400,000 barrels/day, despite the incentive of higher prices caused by the war against Iran" — a supply response that fails to appear at a high price is evidence about the resource base, not about policy — and "the federal deficit as a percentage of GDP is running near 6%, basically double Mr. Bessent's 3% target," with "no improvement in sight." DOGE is the anatomy of a walk-down: "$2 trillion" (fall 2024) → "$1 trillion" (February) → "around $200 billion" two months later, while "U.S. government expenditures have increased by approximately $300 billion" — "borderline comical considering the intent." The bridge back to the bond fight is Jim Grant, of Grant's Interest Rate Observer: "the new supply of Treasuries held by the public has surged by 8.4% on a year-over-year basis," which both implies "the government's accounting is overly flattering to the actual deficit" (issuance being the harder number) and makes holding yields down "challenging in the extreme." The precedent is Japan: "Japan's multi-decade experiment with YCC was a prime factor in the yen's 40% value shrinkage vs the dollar over the past decade," and "relative to gold, the yen's value loss has been truly breathtaking, exceeding 80%" — the gap between the two readings being the argument, since a currency measured against another currency understates a shared debasement. Current-cycle confirmation: stealth YCC "has already weakened the dollar and… reignited vigorous rallies in gold and Bitcoin." No security named.
Friday POW! — abrdn Physical Palladium Shares ETF (PALL): a POW! of a different shape from Aug-7's GILD and Aug-14's VST — not a new name found in an accounting artifact but a re-entry into an owned, already-proven thesis. "This is a reiteration of our original recommendation… of PALL, the palladium ETF, in April 2024. Our bullish thesis on it back then was contrarian and it has returned to out-of-favor status today." The trigger is stated as a three-condition alignment: "palladium has significantly corrected within what we believe is a long-term uptrend at the same time that the fundamental supply/demand story has strengthened. Additionally, the macro backdrop underpinning the hard-asset thesis has moved further in our direction, particularly the trend toward monetary debasement." Entry: a 1/3 correction from the January 2026 apex of just over $2,000/ounce to ~$1,300 spot, against a BofA Q4-2026 target of ~$2,200/oz, "implying ~70% upside from spot," with the base case quoted honestly as "flat to roughly a 23% gain." The macro leg: "we are entering what we believe is a new era of YCC. When central governments begin suppressing interest rates to manage sovereign debt costs (explicitly as Japan has done, implicitly as other OECD sovereigns are doing) the monetary debasement implications for real assets are profound" — with the dated catalyst being Treasury Secretary Scott Bessent's announcement "to 'twist the yield curve', by selling short-term Treasuries to buy longer term issues," which "ignited a roaring rally in the hard-asset space… the market reasonably believes this is a definitive first step toward YCC." Why this rather than gold: "precious metals are the obvious beneficiary, but industrial metals also benefit as hard assets… Palladium is a special case because it qualifies as both a quasi-monetary asset (part of the PGM complex) and it has industrial supply-demand dynamics that are separately compelling." The demand blindspot:85-90% of global palladium demand is automotive catalytic converters, and "the consensus assumed the internal combustion engine-to-EV transition would be binary and swift. What has actually happened is that hybrids (particularly plug-in hybrids) have become the dominant growth segment of the global 'EV market,' and these vehicles actually use catalytic converters, requiring Pd… at rates comparable to conventional gasoline vehicles. Actually, they often consume more." The evidence is Chinese and dated: NEVs surpassing 60% of monthly auto sales in 2026, BYD ("the world's largest auto exporter") seeing PHEVs go "from 11.6% to over 37% of its total export mix," and "Chinese PHEV exports to Europe grew more than 700% year-over-year," with Spain, the UK and Belgium up "multi-hundred to over 1,000%." Three drivers — range anxiety, EU anti-subsidy tariffs on Chinese BEVs from which PHEVs are exempt ("while BYD builds factories in Hungary and Turkey"), and DM 5.0-class platforms "over 2,000 km… of combined range" — plus two accelerants: the Iran War ("spiking oil prices and energy supply uncertainty have pushed consumers toward PHEVs… a hedge against grid failures") and the FT's finding that "21% TSLA owners are trading in their 'Musk-mobiles' for hybrids." The error persists because "the perception of the EV risk has proven durable, exacerbated by the confusing manner in which many sources report EV sales by including plug-in hybrids in the total." The supply leg: a deficit "for 14 consecutive years" (0.9Moz in 2023, 0.5Moz in 2024) with "an expected 2026 surplus… repeatedly pushed back as supply tightens"; the Lac des Iles mine in Ontario ceased commercial production mid-2026, "removing one of the few significant non-Russian, non-South African sources"; and the shock absorber is shrinking — "recycling supply has shrunk rather than grown: U.S. vehicles now average a record 12.6 years as owners delay scrappage, and recycling supply fell approximately 700,000 ounces between 2022 and 2024… around 10% of total palladium supply." Geography is the tail: "Russia's Nornickel controls approximately 40% of global primary supply, while South Africa represents approximately 35%," and "Western sanctions have thus far exempted palladium under automotive industry lobbying pressure, but Montana lawmakers are actively pushing the White House for harsher restrictions. Any escalation would be an acute supply shock not remotely priced into the current $1,300 spot price." Technically the scorecard runs both ways — Pd "essentially doubled in 2025" then "entered a deep corrective phase" that "shaken out most of those who simply chased the momentum" (with "Hormuz-related risk-off sentiment" pressuring the whole complex); it "has held around its initial breakout point," "remains well below its 200-day moving average," but "the steep downtrend it's been in for most of 2025 has been broken" — invoking the rule that "when a physically constrained commodity with a persistent structural deficit breaks a near-term downtrend, the prior trend has historically tended to regain its momentum." Arguing the other side is unusually candid: palladium bulls have been "burned, as in third-degree burned" ($3,430 in March 2022, then "collapsed approximately 70% as worst-case Russia sanctions failed to materialize"); European ICE bans in the 2030s are doubted "for practical reasons; ICEs still represent around 95% of the existing global vehicle fleet"; and platinum-for-palladium substitution is real but dormant — "at the current palladium/platinum ratio of approximately 0.70x the pressure is limited," and anyway "platinum prices could also soar." Verdict: "we're buying a physically scarce metal in persistent deficit since 2012… We're adding to the position at a price 1/3 below its February highs… the market has a double blindspot regarding palladium. PALL, the physically backed ETF, has no counterparty risk… We are reiterating PALL as a Buy." The footnote is the discipline: the drafted endorsement of SBSW ("the world's largest palladium producer," flagged May 2024 as a high-risk play, profit-taken last December) was pulled — "as we were preparing to write this note, it began a ripping rally and is now up 15% this week alone. We will keep an eye on it and advise if it settles back." No portfolio tables with this issue. Dating note: the apex is given as "January 2026" in the highlights and "February highs" in the close
Friday POW! — Vistra Corp (VST): "We are buyers of VST at today's price," on a thesis stated as a first principle — "at its core, the AI story is fundamentally an energy story. Who will win? How will it be (effectively) regulated? Will it massively displace labor? All these questions and handwringing are secondary to the fact that AI is going to require an enormous amount of reliable power." Vistra is "one of the largest competitive power generators in the United States" — ~40,000MW of nuclear, gas, solar and battery across ERCOT and PJM, "equivalent to around 40 large-scale nuclear power plants," ~5m retail customers, the nation's second-largest competitive nuclear fleet — at ~$145, 36% below the $219.82 high on a ~$58B cap. The structural twin of Aug-7's GILD: the $4.02B revenue "miss" (vs $5.46B consensus) is unrealized mark-to-market losses on forward hedges — "real in the GAAP sense and irrelevant in the economic sense" — and, in the reversal that makes the argument, "the hedge program creating the apparent miss is the same driver that gives the company near-complete 2026 earnings certainty" (~100% of 2026 generation hedged; 94% 2027, 72% 2028). The operational reality in the same print: adjusted EBITDA $1.767B, +30% YoY (beat $1.64B), generation EBITDA $994M, +68%, YTD $3.261B vs $2.589B, and 97%+ commercial availability through record heat. Demand is evidenced rather than forecast: "PJM and ERCOT set all-time peak loads of… 168 and 91 gigawatts in July. That is the structural AI demand proof the thesis requires, visible in real-time load data rather than forward projections." The valuation engine is the exclusions — FY26 guidance of $6.8-7.6B EBITDA / $3.925-4.725B adjusted FCF "explicitly excludes" Cogentrix ($4.7B, ~5,500MW, "+$470-590M annual EBITDA" at H2 close), the Meta nuclear PPAs (~2,600MW PJM), the Helix Platform and nuclear PTCs — so "the 2026 guidance midpoint of $7.2 billion… is the starting point, not the destination… the 2028 EBITDA trajectory… almost certainly exceeds $9 billion annualized. At a $58 billion cap, the market is pricing none of it." Helix — up to $1B alongside KKR (capital, deal sourcing), NVIDIA (compute expertise) and the Kuwait Investment Authority (long-duration mandate), with VST as "preferred power partner" supplying "the one thing none of them can build quickly: 24/7 dispatchable, carbon-friendly nuclear and gas generation" — is "the stock's long-term game-changer," and none of its revenue is in guidance; alongside an AWS PPA ~3,800MW at Comanche Peak and Perry Nuclear extended to 2046. Against the WSJ's bearish "Heard on the Street" piece he concedes the premise ("the easy version of the power thesis has weakened to a degree" as ERCOT/PJM regulators move to cap scarcity rents) and relocates the moat to physics — "the bottleneck is more than just policy, it is the physical ability to deliver reliable megawatts… Vistra already owns scarce generation, existing grid connections, nuclear and gas assets, and relationships with hyperscalers" — concluding that the market shifts from "own existing power and enjoy higher prices" to "own existing power and be one of the few companies capable of expanding reliable supply": "the thesis is evolving, not breaking." Valuation 10.6× forward EV/EBITDA vs Constellation (CEG) at 13.6× trailing (~14.5× P/E), with FCF yield "the most compelling entry-point metric" at ~9.1% on 2026 guidance and ~11.4% on ~$5.4B normalized 2028 FCF; 20 analysts Buy, avg target ~$217 (~54% upside). Bear case = ERCOT forward-curve softness + Helix execution + PJM IRAS uncertainty, "justifiable only if all three occur simultaneously… that threshold is too high"; the build-out risk is answered by rate ("could slow meaningfully, but… almost certain to continue at a rapid clip" — hyperscalers treat falling behind as "an existential risk"). The technical caveat inverts the GILD entry: below the 200-day for over a year, "stabilizing in the 150 range," so "you could put in a stop at 130." (Buys / Holds-Trims / Sells tables published as images, not transcribed — the only text deltas are "DVN added, along with earlier WDOFF recommendation, for visibility" and a Sells-table % Change formula revision)