Ten conversations this week, and taken together they map the fault line running under every portfolio: a US economy — and market — being remade from a neutral referee into an activist, interventionist state, even as the AI capex boom that powers the tape raises the question of whether the demand behind it is real. Cem Karsan opens with the grandest frame, an “America’s 250-year reset” in which positioning is everything and Washington’s answer to unsustainable debt, populism, and the contest with China is a sovereign-wealth-fund bid — $10–15 trillion of equities bought with printed money — that backstops a market too big to fail. Ed Cole gives the allocator’s response: the end of American exceptionalism, a stickier-inflation regime in which stocks and bonds move together and the 60/40 breaks, and an AI “earnings bubble” predicated on compute staying scarce. David Rosenberg warns that “there are no more bears left,” pairing an out-of-consensus disinflation call with a hollow-consumer, extreme-concentration, financialized-AI picture and the risk-management lesson of Charles Merrill in 1928. Kevin Muir presses the same defensive case — priced for perfection, a lost decade, a “token mirage” — and hides in cheap oil producers and gold miners. Then the counterweight: Warren Pies takes the bearish AI arguments apart with hour-by-hour GPU-availability and OpenRouter data, arguing the token panic is overblown even as a trapped Fed becomes the real threat. Scott Galloway and Ed Elson push the bear case to its edge, reading OpenAI’s proposed government stake and Meta’s cloud pivot as a 1999-style demand crisis and a bailout by design. Brian Belski stays bullish but warns to beware the earnings second derivative, leaning into the broadening-out trade — SMID, regional banks, and software as the AI hedge. Mike Green supplies the market-structure plumbing — SpaceX and the passive bid, a rate that’s simultaneously too high and too stimulative, the portfolio-rebalance channel, and sovereign gold flows. Josh Pristaw makes the halo trade concrete, explaining why a $73 billion real estate allocator avoids data centers and leans into senior housing, industrial, and the start of a new property cycle. And Peter Zeihan closes on the physical world: roughly 1.25 billion barrels of crude never delivered, a strategic reserve at a 1983 low, and an Iran that emerged stronger — a summer oil crunch the financial “glut” narrative is missing. Each summary is designed to be immediately actionable — whether you are allocating capital, running a business, or simply trying to understand the forces reshaping the world around you.
THIS WEEK’S LINEUP
EP 1 America’s 250-Year Reset: Positioning, the Activist State & the Coming Sovereign Bid — Cem Karsan — Kai Volatility Advisors Founder & CIO — Listen Full Episode
EP 2 The End of Exceptionalism: Diversifying Into a Stickier-Inflation Regime — Ed Cole — Man Group Head of Multi-Strategy Equities — Listen Full Episode
EP 3 There Are No More Bears Left: Disinflation, a Hollow Consumer & Merrill’s 1928 Lesson — David Rosenberg — Rosenberg Research President & Founder — Listen Full Episode
EP 4 Priced for Perfection: Playing Defense Into a Lost Decade — Kevin Muir — The Macro Tourist Founder & Author — Watch Full Video
EP 5 The $400 Billion Gap: Token Panic, a Trapped Fed & What Murders Bull Markets — Warren Pies — 314 Research Co-Founder — Watch Full Video
EP 6 Too Big to Fail by Design: The OpenAI Bailout & the 1999 Demand Crisis — Scott Galloway & Ed Elson — Prof G Markets Host & Co-Host — Watch Full Video
EP 7 Beware the Second Derivative: The Broadening-Out Trade — Brian Belski — Humilis Investment Strategies Founder, CEO & CIO — Watch Full Video
EP 8 SpaceX, Passive & a Rate That’s Too High: The Market-Structure Reckoning — Mike Green — Simplify Asset Management Chief Strategist & Portfolio Manager — Watch Full Video
EP 9 Why a $73 Billion Allocator Is Avoiding Data Centers: The Halo Trade Made Concrete — Josh Pristaw — Clarion Partners Managing Director & President — Watch Full Video
EP 10 Get Ready for the Summer Oil Crunch: The Missing 1.25 Billion Barrels — Peter Zeihan — Zeihan on Geopolitics Founder & Author — Watch Full Video
Full summaries with actionable insights and investment focus for each podcast follow on the pages below.
Cem Karsan — Kai Volatility Advisors Founder & CIO
Cem Karsan of Kai Volatility Advisors joins Niels Kaastrup-Larsen’s Systematic Investor series in the week of America’s 250th anniversary to argue the country sits at a civilizational inflection point — the roughly 250-year mark where empires and regimes tend to change fast once they finally move. His organizing insight is that nothing matters more than positioning: markets have become so large that a 20% two-month move swings some $50 trillion of collateral, so reflexivity, not fundamentals, drives outcomes — which is why the “inevitable” $200 oil never came and the crowd is now short at record levels. From there he lays out the regime: an activist, interventionist Washington answering unsustainable debt, accelerating populism, structural inflation, a market too big to fail, and the contest with China with a single elegant move — a state-backed equity bid financed by the dollar’s exorbitant privilege.
Actionable Bullet Points
Positioning Is Everything — Fade the Obvious: Karsan’s first principle is that nothing has a more reliable effect on outcomes than positioning; when the world is long the odds favor a fall, and when it is short a rise, because markets are now so large that reflexivity dominates sentiment and fundamentals (4:40). The “inevitable” $200 oil after the Iran shock never arrived precisely because everyone was positioned for it, and now that the crowd has thrown in the towel with oil shorts at record highs, the setup has flipped (6:30). This is George Soros’s reflexivity applied at a $50-trillion-collateral scale. Bet against the consensus that feels obvious and inevitable.
The K-Shaped Economy Runs on a Reflexive Capex Loop: Karsan reads the soft labor data as the signature of a K-shaped economy where capex — not labor — drives everything, and that capex is the non-labor-heavy buildout of AI infrastructure concentrated in a handful of names (7:36). Earnings growth, he argues, is now essentially 100% coming from that capex build in a reflexive loop — he cites Microsoft booking a large share of a recent quarter’s earnings from the rising value of its OpenAI stake, not end demand (8:22). The market itself, not the economy, has become the biggest driver of outcomes. Watch the capex loop, not the macro data, for what actually moves earnings.
The “Chinaification of America” — a State-Backed Equity Bid: Karsan’s central call is that Washington is adopting “socialism/communism with American characteristics,” using the dollar’s exorbitant privilege to print money and stand up a sovereign wealth fund that buys $10–15 trillion of equities over the next decade while backstopping the Treasury market (15:08, 19:57). He reads the Intel and OpenAI stakes as the opening moves of a plan that simultaneously competes with China, supports a too-big-to-fail market, short-circuits populism by making the public an owner, and monetizes the debt through inflation (23:39). This is Norway’s and Japan’s playbook at US scale. Expect the sovereign bid to arrive aggressively — he thinks after the midterms and before next June.
Warsh and Bessent as Good Cop, Bad Cop: Karsan sees it as no coincidence that two hedge-fund managers now sit atop the Treasury and the Fed, with the central bank’s “independence” shrinking and the two institutions working ever more closely (45:46). His sequencing: Bessent plays bad cop — buying stocks, demanding money printing, and having the Treasury buy back its own debt first — while Warsh plays the good cop whose apparent independence defends the dollar’s credibility, before the Fed ultimately backstops the long end of the curve into a crisis (47:56). Read Warsh’s hawkish posture as a credibility device, not a genuine constraint on the plan.
Iran Isn’t Ending, and November–June Is the Danger Window: Karsan argues Iran is not really about nuclear weapons but about controlling trade and keeping it denominated in dollars — which is why the Strait of Hormuz fight and the “deal” are temporary, held together only by the coming midterm (31:42). His three-legged stool — structural flows, macro, and an activist administration — is supportive into September–November but turns dangerous from November to June, when flows fade, macro degrades, and an administration that may want a controlled crisis loses its tailwinds (52:40). He even games out the sequence: wait until the midterms, then re-engage Iran (57:16). Treat late 2026 into mid-2027 as the volatile, high-risk window.
Investment Focus
Karsan’s is the week’s grand unifying frame, and it is near-term constructive but long-term uneasy. The investment template: (1) trade positioning over narrative — fade the obvious, which right now means respecting the record short in oil and the crowded consensus (4:40, 6:30); (2) don’t fight the activist administration while flows and policy are supportive into the fall, because the state-backed bid gives AI capex an implicit guarantee even as it inflates a bubble (19:57); (3) play the “summer of George” — vol compression, correlation breakdowns, and dispersion reward doing the opposite and owning the right tail (50:10); (4) mark November–June as the danger window, when the three-legged stool weakens and a controlled crisis becomes tempting (52:40); (5) position for an inflationary regime — the whole point of the plan is to monetize the debt, which argues for real assets and the sovereign bid, not disinflation trades (25:16). Karsan’s sticky-inflation, monetize-the-debt regime is the direct foil to Rosenberg’s disinflation call (EP 3), and his reflexive-capex-loop echoes the AI circularity that Rosenberg, Cole (EP 2), and Pies (EP 5) each dissect from a different angle.
▶ Listen to the full conversation
Ed Cole — Man Group Head of Multi-Strategy Equities
Ed Cole, head of multi-strategy equities within Solutions at Man Group, returns to Merryn Somerset Webb’s Merryn Talks Money to mark his own homework on the “end of American exceptionalism” — and finds the rotation more real than the headlines suggest. Under the surface of another 20% year for the S&P and the Mag 7, the Russell 2000 is up 38%, emerging markets nearly 50% (though that is really three Asian semiconductor giants), and onshore China roughly 30%. Cole’s central argument on AI is that this is an earnings bubble more than a valuation bubble — the numbers only work if compute stays scarce — and his broader thesis is a regime change into stickier inflation that breaks the 60/40 and demands that investors “diversify their diversifiers.”
Actionable Bullet Points
The Rotation Is Real — Look Under the US Exceptionalism Narrative: Marking his own scorecard, Cole notes that while the S&P and Mag 7 are up around 20%, the more cyclical, value-tilted Russell 2000 is up 38%, and the rotation into smaller-cap value has shown up everywhere, not just the US (3:29). Emerging markets are up nearly 50%, but he cautions that is not an EM story at all — it is three giant semiconductor names in Taiwan and Korea, so buying the index just buys the AI-momentum trade (4:16). Onshore Chinese equities, a deep and inefficient market, are up roughly 30% in dollars with many stories below the semiconductor surface (5:27). Diversify where your equities are, and don’t mistake the EM index for diversification.
AI Is an Earnings Bubble More Than a Valuation Bubble: Cole’s sharpest framing is that the market does not look expensive today only because earnings estimates have been revised up so aggressively — but those estimates are predicated on the assumption that compute is scarce and that Asian memory and logic makers can keep both price and volume rising (8:03, 8:57). If compute turns out not to be scarce — via open-source models, small language models for discrete tasks, cheaper Chinese models, or engineering breakthroughs like a CPU-powered Chinese supercomputer — the earnings collapse and the market re-rates sharply higher on valuation (11:07, 12:59). Test the compute-scarcity assumption, because that is the load-bearing wall.
Circularity Multiplies the Revisions — but Most Bubbles Are Productive: Cole revisits the vendor-financing circularity in which customers invest in the very companies whose chips they order, so a single order book ripples through the model-builders, hyperscalers, memory makers, and logic makers all at once (14:41). Yet he urges open-mindedness: most bubbles are productive, sucking capital into genuinely transformational infrastructure — railways, the dot-com build-out, even the tulip mania, which left the Netherlands its world-leading flower industry (16:39, 17:53). The technology is valid; the only question is whether the valuations are right. Judge the price, not the premise.
A Stickier-Inflation Regime Breaks the 60/40: Cole argues the stable presets of the past 30 years — globalization, frictionless capital and supply chains, low inflation — are reversing into a higher-friction, national-interest world of stickier producer-price inflation (20:41). The consequence that matters for portfolios is that stocks and bonds now move together: in the inflationary drawdown of 1974 a 60/40 lost far more than in 2008, when bonds bailed you out (23:54). Once the inflation genie is out, central banks may be unable to provide the reliable “put” investors have leaned on for three decades (26:07). Stop assuming bonds will diversify your equities.
Where to Hide: the Halo Trade, Liquid Alternatives & Gold: Cole’s prescription is to tilt toward shorter-duration cash flow with a real fixed-capital base rather than the intangible assets of the last cycle — the “halo” trade of hard assets and low obsolescence — much of it found outside large caps in smaller-cap value (28:19). He calls this a golden age for liquid alternatives: market-neutral and equity long-short strategies actually improve as higher inflation and costlier debt force more idiosyncratic corporate distress and greater dispersion (30:24, 32:09). Gold he likes on a weakening dollar and intact central-bank demand, while conceding it can flip to a risk asset as it did in March (35:26). Diversify your diversifiers — halo equities, true liquid alternatives, and gold.
Investment Focus
Cole is the institutional allocator’s translation of the regime Karsan describes. The investment template: (1) rotate where your equities sit — smaller-cap value globally and onshore China, while avoiding the EM index that is really an AI-momentum trade (3:29, 4:16); (2) treat AI as an earnings bubble hostage to compute scarcity, and watch open-source, small-language-model, and cheaper-Chinese-model adoption as the re-rating trigger (8:57); (3) rebuild for positive stock-bond correlation — the 60/40 is breaking and bonds are no longer reliable diversifiers, so broaden the toolkit (23:54); (4) tilt to the halo trade — short-duration cash flow, real fixed capital, smaller caps — and lean on liquid alternatives that benefit from inflation-driven dispersion (28:19, 32:09); (5) hold gold as a weak-dollar and central-bank-demand story while respecting that it can become a risk asset in a scramble (35:26). Cole’s compute-scarcity skepticism is the mirror image of Pies’s data-driven “demand is real” case (EP 5), and his stickier-inflation regime aligns with Karsan (EP 1) against Rosenberg’s disinflation call (EP 3).
▶ Listen to the full conversation
David Rosenberg — Rosenberg Research President & Founder
David Rosenberg of Rosenberg Research joins Dan Nathan on the Risk Reversal podcast on the last day of Q2 to press his out-of-consensus disinflation call and to warn that the more dangerous fact about this market is that “there are no more bears left.” His inflation argument is structural: productivity is doing 90% of the work of US growth, unit labor costs have collapsed to 0.5%, and market break-evens sit below their pre-war level — so the run-up in yields reflects a shift in the Fed’s reaction function, not the fundamentals. Beneath the strong price momentum he sees a hollow consumer, extreme concentration, a negative equity risk premium, and a financialized AI build-out — and he reaches for the lesson of Charles Merrill, who raised cash in 1928 and was told to see a psychiatrist.
Actionable Bullet Points
Disinflation, Not Inflation — Productivity Is Doing the Work: Rosenberg argues that with 90% of US growth over the past year coming from productivity — versus a normal 50/50 split with labor input — the structural setup is the antithesis of an inflation cycle, since productivity is “kryptonite for secular inflation” (12:40). Unit labor costs, the mother’s milk of future inflation, have fallen to 0.5% year-over-year from 3% a year ago, with no meaningful spillover of the oil shock into core or wages (14:16). Ten-year break-evens sit at 2.2%, lower than before the war, so the yield back-up reflects the Fed’s reaction function, not real inflation (15:32). Fade the inflation scare and the hikes the market has priced.
The Consumer Is Hollow — the Income Ledger Reads Zero: Rosenberg’s most important point is that the savings rate has fallen from a pre-COVID 8% to 3%, the single biggest under-discussed source of US stimulus, even as real disposable income growth runs at 0% for the year (25:09, 27:54). The high end is spending its equity-wealth effect while the low end taps credit cards at 20% rates on a $1.3 trillion balance, with delinquencies at a two-decade high of 15% (26:35). The savings rate is a classic mean-reverting series, and its reversion would deliver a consumer recession that catches most by surprise (29:56). Position for the income ledger, not the spending headline.
The Labor Market Is Flat, Not Stabilizing: Rosenberg warns against reading one or two prints as stabilization — the labor market moves in glacial trends, and on a year-over-year basis non-farm payrolls are running flat while the household survey is at minus 0.3% (33:56). The 4.3% unemployment rate flatters the picture: hold labor-force participation constant and it would be 5.1%, meaning more slack than meets the eye (34:40). With nominal wage growth cooling faster than inflation, real wages stay constrained just as the savings-rate cushion runs out (35:32). Trust the trend and the benchmark revisions, not the monthly gyrations.
“There Are No More Bears Left”: Rosenberg’s contrarian antenna is up because everyone is all-in — portfolio-manager cash ratios are at 1%, and 73% of US household financial assets sit in equities against barely 7% in bonds (39:08, 39:12). A negative equity risk premium is the market telling you it no longer regards equities as a risky asset class, exactly as in the late 1990s (40:32). And credit always leads: triple-C/double-B spreads have widened sharply and one private-credit fund after another is gating and capping redemptions — the canary in the coal mine (41:31). Watch credit, not equities, for the turn.
The AI Financialization & Depreciation Mirage: Rosenberg flags the circular financing behind the build-out — a $35 billion Blackstone/Apollo vehicle to fund Google TPUs used by Anthropic, plus Nvidia funding its own customers — as resembling the chaebol-style intertwining of the late 1990s (1:03:16). The depreciation accounting also skews earnings: the seller books pure revenue while the buyer expenses it, and adjusting for the depreciation allowance puts the S&P’s true P/E near 30, with a Shiller CAPE of 41 (1:06:41). His answer is risk management, not timing — Charles Merrill raised cash in 1928, was told to see a psychiatrist, and saved his firm (1:00:09). Being early is risk management, because bear markets reverse three-quarters of the prior bull, fast.
Investment Focus
Rosenberg runs roughly 50% equities — Asia-heavy, low-beta and sector-specific — with about half in bonds and near 10% in gold and the miners. The investment template: (1) lean into disinflation — expect a string of soft CPI prints and a Fed that does not hike as the market fears (15:32); (2) respect the hollow-consumer setup and position for savings-rate mean reversion and a possible consumer recession (29:56); (3) treat concentration, a negative ERP, and a CAPE of 41 as the risk, and watch credit — private-credit gating and CCC/BB spreads — as the lead indicator (40:32, 41:31); (4) risk-manage like Charles Merrill in 1928, accepting that being early is prudent because three-quarters of bull-market gains reverse quickly (1:00:09); (5) diversify by geography and asset class — Asian equities, bonds, and gold/miners, which he sees as only temporarily undercut by rates and the dollar (43:06, 57:29). Rosenberg’s disinflation call is the direct counterpoint to Karsan (EP 1) and Cole (EP 2); his concentration, circularity, and defensive-positioning warnings rhyme closely with Muir (EP 4) and with Pies’s own circularity concerns (EP 5).
▶ Listen to the full conversation
Kevin Muir — The Macro Tourist Founder & Author
Kevin Muir — the Macro Tourist — sits down with Thoughtful Money’s Adam Taggart to argue that investors have convinced themselves the market is safer than it was at the COVID lows when in fact it has become “immensely more risky.” His case is built on price, not story: an equity risk premium that has completely inverted since 2020, a Buffett indicator and Shiller CAPE at or near record highs, and an S&P whose returns for three-plus years have trained investors to expect an endless summer. Muir’s punchline is defense — reduce equity exposure into strength rather than add to it — and his hiding places are the cheap, dividend-paying oil producers and gold miners that almost nobody owns. The New Harbor Financial team of Mike Preston and John Llodra close the hour with the practical companion: financial-plan modeling that shows how a lost decade, sequenced like 2000–2025, can bankrupt a retiree even without a permanent bear market.
Actionable Bullet Points
The Market Is Priced for Perfection — the Equity Risk Premium Has Completely Reversed: In March 2020 the S&P’s earnings yield was 6% against a 1% ten-year, an equity risk premium screaming that stocks would outperform for a decade; today the ten-year sits near 4.5% and the earnings yield in the threes-to-fours, so equities are “priced for a lot of perfection” (2:51). The Buffett indicator (market cap to GDP) and the Shiller CAPE are at or near record highs — “everywhere you look, the market is expensive” (3:59). None of these are timing tools, but over five-to-ten-year horizons they work, which is why Muir would reduce equity exposure here rather than increase it like most investors (4:23). Position for lower forward returns, not another year of 20%.
Treat a Lost Decade as the Base Case — Concentration Makes It Worse: Muir says a lost decade is “probably what you should assume,” and index concentration makes this one more dangerous than prior bad periods: if the top-10 names are roughly 40% of the S&P and take a 50% drawdown, the index is down 20% on those alone (9:08). He points to 1968–1982, when the Dow struggled to hold 1,000 and investors were ground sideways for fourteen years while inflation ate real returns (14:21). The one comfort is that lost decades are punctuated by violent rallies and crashes, so avoiding the down legs — not sitting in cash — is the game. Plan for a sideways-to-violent decade and protect against the big drawdowns.
The “Token Mirage” and the Return of the Semiconductor Cycle: The memory and semiconductor complex added roughly $4 trillion of market cap in 2025 and about $8 trillion year-to-date in 2026 (19:25), a pace Muir argues rests on a three-month “token mirage” — companies mandating AI use, Meta running engineer leaderboards, and coders looping agents to top them, generating demand that is not real (33:10). He reminds investors the world has “forgotten that semiconductors are cyclical” (9:28), and that AI compute depreciates in a couple of years with replacement running about two-thirds of build cost — unlike dark fiber, this overbuild will not sit useful and idle (24:04). Do not extrapolate three months of subsidized token usage into a permanent demand curve.
The Warsh Fed Is More Hawkish Than the Market Believes — Don’t Position for Cuts: Muir read Warsh’s first major address as “about as hawkish as he could possibly have been” (50:00), hammering price stability and a hard 2% target while barely nodding at employment. With PCE up to 3.3% and three decent jobs prints, the Taylor rule mechanically pushes the whole committee — not just Warsh — hawkish, and December SOFR futures have already repriced from no hikes to roughly one-and-a-half (57:22). His line for the equity crowd: “he’s just not into lowering rates,” and Warsh only turns dovish after an accident (1:00:25). Stop pricing a rescue and trade the curve that is actually in front of you.
Where to Hide: Oil, Gold Miners & the Next Rolling Mini-Bubble: With the Iran deal treated as capitulation, Muir flips bullish oil on an SPR-refill thesis — the surprise for the next two months is that crude does not fall to $40 as governments worldwide bid to rebuild strategic reserves (1:04:55) — and prefers the producers, cheap near 16x earnings and paying you to wait (1:05:28). On gold he stays with the People’s Bank of China as the only buyer that matters and would own the dirt-cheap miners over metal (1:09:22, 1:11:26). Finally, he urges investors off page one and onto the next “rolling mini-bubble,” naming agriculture, refiners, and biotech as candidates (1:14:28). Buy the cheap, unowned, income-producing assets — and hunt the next rotation rather than the last.
Investment Focus
Muir delivers the week’s cleanest cut-through on valuation and defense, and the New Harbor segment supplies the arithmetic behind it. The investment template: (1) reduce passive equity into strength — New Harbor would cap a pure indexer near 30–40%, a level at which you could lose half and still not touch your plan (1:29:27); (2) respect the sequence-of-returns risk a lost decade creates — John Llodra’s plan modeling shows a retiree who begins drawing in 2000 and lives through the actual 2000–2025 return sequence runs out of money, while the same plan on average-return assumptions leaves $1.2 million (1:40:05); (3) own the cheap, dividend-paying real-asset trades Muir favors — oil producers, especially the long-life Canadian names, and gold miners — as the assets that behave well in a concentrated-index correction; (4) stay with the People’s Bank of China gold thesis and treat a pause in its buying, not the dollar or real yields, as the signal to get less bullish (1:09:22); (5) hunt the next rolling mini-bubble — agriculture, refiners, biotech — rather than chasing AI and semis on page one (1:14:28). On the Fed, cross-reference Rosenberg (EP 3), Pies (EP 5), and Green (EP 8): each reads a Fed that will not be the rescue the bulls expect, removing the multiple-expansion tailwind — and on oil, Muir’s SPR-refill bull case runs alongside Zeihan’s physical-crunch thesis (EP 10).
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.