Warsh Fails the First Test Warsh has failed his first test: hiking into a negative growth shock. His mistake exposes a contentious thesis that will slowly become consensus: the bond market is no longer the smartest money in the room. Hedge funds’ growing influence challenges the old conviction that credit markets and bond vigilantes deliver sober judgments about economic fundamentals. When short-term, leveraged players dominate, prices can reflect crowded trades and groupthink rather than sound economic analysis. A Fed driven by those prices compounds the problem. Traders predict the Fed; the Fed follows their predictions. The hall of mirrors turns positioning into a policy signal and then mistakes that signal for independent judgment. Hiking into a negative growth shock is what happens when the central bank follows the crowd instead of assessing the economy. This is not your grandfather’s bond market. The supposed smart money may now be the dumb money at the table and a Fed that follows it risks becoming the dumbest player of all.

Most people say the AI capital cycle is over. The financing says otherwise. Recent deals: • SpaceX: ~$40B GPU-equipment financing • IREN: $2.4B at 9% for 30 months • Nebius: $3.5B secured facility SpaceX’s CFO says infrastructure paybacks are under one year. Anthropic’s revenue is up 14x, cash flow turned positive, and it holds, $132B in cash. Banks publish skeptical research while their financing teams structure the deals. Higher yields pull capital in, not push it out. The market is getting more selective. It is not closed. Follow the transactions.

Trump’s green shoots are right here, you just have to ignore the doomers. Latest AAR rail data: 540,135 total carloads + intermodal units last week, up 8.5% year-over-year. Intermodal jumped 12.6%, carloads +3.5%. YTD volume is up 3.7%. Freight is moving. The economy is working.

AI’s Financing Window Is Open The claim that AI’s capital-spending cycle is over makes a compelling podcast title. Doom attracts views. The financing tells a different story. SpaceX announced a commitment for approximately $40 billion of GPU-equipment financing. IREN secured $2.4 billion at 9% for 30 months. Nebius obtained a $3.5 billion secured bank facility running roughly one year. Large pools of capital are financing the buildout. The market is imposing discipline, not closing its doors. SpaceX CFO Bret Johnsen says its infrastructure investments are seeing payback periods of less than one year. He says the lease diversifies funding sources and finances a substantial portion of its expansion. That is a different picture from infrastructure built without demand or a path to repayment. The institutional contrast is striking. Financial institutions arranging enormous AI transactions also publish research questioning AI’s productivity benefits and the boom’s economics. Their analysts debate the opportunity while their financing teams structure the deals. Follow the transactions, not just the commentary. Higher yields help explain why capital keeps arriving. A borrower’s interest expense is a lender’s income. Attractive rates, collateral and repayment protections draw institutions into infrastructure financing. Distinctions matter. A new loan supplies capital. A distressed yield on an existing bond reflects its falling price. Confusing the two turns borrower-specific stress into a verdict on the entire cycle. Anthropic’s numbers cut through the gloom. First-half revenue rose fourteenfold to $16.3 billion, operating losses narrowed, and operating cash flow turned positive. With roughly $132 billion in cash and securities and an IPO planned, it is hardly approaching a funding cliff. The buildout does not require every company to succeed. It requires viable projects to attract sufficient capital. The deals show that capital is arriving. The doomer narrative mistakes a more demanding market for a closed one. The AI boom is being underwritten, not written off.

The ECB’s Latest Blunder The European Central Bank raised interest rates into an energy shock, a fragile economy and a mountain of debt. It was a massive policy blunder. France’s debt crisis is exposing the consequences. The ECB hiked twice this year to combat energy-driven inflation. Now French borrowing costs have surged, tightening the squeeze on a heavily indebted state. Monetary restraint has collided with fiscal fragility—exactly the danger policymakers should have anticipated. The echo chamber of elite economic technocrats is at it again. Ignore the data. Ignore the theory. Fear the sword of Damocles: the second- and third-order inflation effects that keep failing to materialize. Actual economic weakness takes second place to hypothetical inflation persistence. To be sure, ECB acolytes will say all is fine., An oil spike makes an importing economy poorer. Tariffs raise costs and undermine efficiency. Neither creates excess purchasing power. Households and businesses are already absorbing a loss of real income. Raising their borrowing costs adds a second blow. Basic theory distinguishes a supply shock from a demand boom. The mechanical equation—higher prices require higher rates—is not analysis. It is doctrine. Extreme debt makes that doctrine dangerous. Higher financing costs squeeze budgets while weaker growth erodes the revenues needed to service debt. Governments retrench, investment falls and the slowdown deepens. The policy feeds the fragility. France illustrates the trap. Rising oil prices and expectations of tighter monetary policy have compounded fiscal and political weakness. By October 9, French ten-year yields had risen 150 basis points since the Hormuz crisis began. The ECB has rehearsed this disaster before. It raised rates in July 2008 as the global financial crisis unfolded—an overreaction to oil-driven inflation that its former vice-president subsequently acknowledged. It tightened twice again in 2011 during the eurozone debt emergency, then reversed both increases before year-end. With Greece already in crisis and contagion threatening the monetary union, the ECB made money more expensive. Now France faces the same brutal arithmetic: expensive debt, weak growth and tightening financial conditions. Central bankers call this defending credibility. Repeating an acknowledged mistake does not defend credibility. It exposes an institution more committed to its own orthodoxy than to understanding the economy. bloomberg.com/news/newslette…

Classic 👇

Ignore the noise. Ignore the Doomers. We are in a secular bull market. Have a nice day.

Small Cap Value Back to the upward sloping 200 day SMA Breadth is expanding. Ignore the noise $iwn

A normal healthy natural pullback. The SMH corrects right back to the upward sloping 200 Day SMA. Ignore the noise. $SMH

As Wall St pundits head into the bunker. S&P 500 Equal Weight 👇 What happened to there is no breadth complaints. $RSP

Wall Street’s AI Whiplash In 1998, Paul Krugman predicted that the internet’s economic impact would prove “no greater than the fax machine’s”. Wall Street should frame that quotation before declaring artificial intelligence a productivity disappointment. The latest bearish assessment points to flat revenue per worker and profit margins across the S&P 493, excluding the Magnificent Seven, and declares “zero financial payoff” outside technology. The observation deserves scrutiny. The verdict outruns the evidence. Revenue per employee is not real output per hour. Flat margins cannot establish what costs would have been without AI. Meanwhile, the narrative swings from “AI will take every job, hide your children” to “AI is useless”. Add a few charts, generate some noise, rinse and repeat. Manufacturing certainty from contradictory narratives is what Wall Street does best. Apparently, robots will never arrive, inference will never scale and businesses will never reorganise. Today’s deployment becomes tomorrow’s ceiling. Companies must prepare data, integrate systems and adapt CPU-based workflows to exploit GPU-accelerated AI where appropriate. Not every workload requires conversion. But useful deployment requires organisational change, not merely another subscription. Yes, adoption is a constraint. People must change habits and companies must change processes, as they did with computers and the internet. This is not profound. It describes technological diffusion, not technological failure. Physical constraints matter too. The International Energy Agency expects data-centre electricity consumption to more than double by 2030. Compute capacity and power connections cannot materialise on demand. None of this validates every investment. Projects will fail. Valuations will disappoint. Cheaper models may undermine suppliers while improving customers’ economics. But declaring economic futility during an unfinished buildout is ambulance-chasing analysis. The serious question is where AI creates value, at what cost and for whom. Wall Street’s oscillation between apocalypse and irrelevance answers none of it. Neither do its charts.

AI’s financing shakeout is not an earnings collapse This week’s earnings from Microsoft, Alphabet, Amazon, Vertiv, GE Vernova, ServiceNow and Intel will test whether AI spending fears are obscuring resilient demand and constrained capacity. Fears that artificial intelligence spending will overwhelm technology earnings confuse marginal developers’ vulnerabilities with the economics of established businesses. The distinction is financeability. An operator borrowing to finish a data centre faces a different challenge from a diversified platform with operating cash flow, powered facilities and paying customers. Tightening credit can punish one without undermining the other. If forecasts of another 800–1,000 US data centres materialise, competition for electricity, equipment and construction capacity will intensify. But buildings are not compute. Winners must finance, power and connect their infrastructure. This competition extends beyond commercial ambition. US-China rivalry makes computing capacity, advanced semiconductors and reliable electricity strategic assets. That guarantees neither profits nor sensible valuations. It does make the buildout more durable than quarterly spending anxieties suggest. Financing is evolving alongside demand. Corporate borrowing is being supplemented by lending against equipment, customer contracts and expected receipts. Securitisation could broaden funding further by pooling eligible loans or receivables for institutional investors. These structures cannot manufacture electricity or rescue worthless contracts. They can, however, finance viable projects beyond developers’ balance-sheet limits. Weak operators restructure; useful assets move to stronger owners; investment becomes more selective. That process could extend the cycle rather than end it. Staggered construction delays supply, while lower acquisition costs improve project economics. Established operators with powered capacity can serve customers while competitors struggle to deliver. Suppliers need funded buyers, not universal developer survival. A failed operator does not automatically mean a lost shipment. Nor do cancelled proposals eliminate electricity constraints. Investors should distinguish spending from returns. Capital expenditure reduces cash flow upfront; depreciation affects earnings over time. The test is whether assets earn above their costs. Nuanced exposure is the strategy: favour financial durability, productive capacity and enforceable revenues. The weakest borrowers are not a proxy for the entire AI investment cycle. bloomberg.com/news/articles/…

BC elects Conservative Government Another election where Capitalism wins. Socialism loses. Yes the West is moving back to the Right.

This is well worth your time.👇