The Lyceum: Macro & Markets Weekly — Apr 23, 2026

The S&P 500 closed at a fresh all-time high of 7,137 on Wednesday's session while Iranian gunboats seized two ships in the Strait of Hormuz that morning. That's not a contradiction the market is going to resolve; it's a feature of the current regime — where an AI capex supercycle and a geopolitical oil shock are running on parallel tracks, each ignoring the other until something forces reconciliation. This week, the forcing events start lining up: a Fed chair nominee who told the Senate he won't be Trump's "sock puppet" but also couldn't say what he'd do about rates, an ECB meeting on April 30 that's priced for a hold while the IMF urges a hike, and a Q1 GDP print that will tell us whether the economy was already slowing before any of this began.

Kevin Warsh sat for a committee-level confirmation hearing before the Senate Committee on Banking, Housing, and Urban Affairs on April 21, 2026 and delivered one of the more consequential performances of the year — not for what he promised, but for what he signaled. He pledged not to be Trump's "sock puppet," then proceeded to outline what can only be described as regime change at the Federal Reserve: fewer FOMC meetings, fewer press conferences, a smaller balance sheet (down from the current $6.7 trillion, which Warsh noted has grown from $800 billion when he arrived at the Fed in 2006), and — critically — less forward guidance, the practice of telegraphing future rate moves so markets can price them in advance.

CNN's framing captured the discomfort well: get used to being confused about where Warsh stands, because his answers were deliberately non-committal on inflation, rates, and framework. For markets that have spent a decade pricing off Fed forward guidance, a quieter Fed is a structurally more volatile Fed — wider credit spreads, fatter risk premia, and a short end of the curve that starts behaving like it did in the 1990s.

The confirmation itself is stuck. Republicans hold a 12-10 advantage on the Senate Committee on Banking, Housing, and Urban Affairs, meaning one dissent blocks the nomination, and Senator Thom Tillis has signaled he will vote "no" until the DOJ investigation into Jerome Powell is dropped. A near-term FOMC meeting would be Powell's second-to-last unless Warsh is not confirmed by May 15, in which case Powell could remain as FOMC chair until a replacement is named.

What to watch: whether the White House drops the Powell probe. That's the unlock. Until it happens, the Fed walks into its next meeting with a lame-duck chair, a nominee in limbo, and fed funds futures that have already pushed 2026 rate-cut odds close to zero.

Six months ago the ECB was cutting rates toward 2% to rescue a sluggish Eurozone. This week, the IMF's Alfred Kammer told Reuters the ECB should raise rates by about 50 basis points in 2026 "in order to maintain a neutral monetary stance." That's a full reversal, amid an energy shock that's pushed the ECB's 2026 headline inflation projection to 2.6%.

The Governing Council is genuinely split. Bundesbank President Joachim Nagel has publicly said an April hike is "an option." Prediction markets are pricing a 73.5% probability of a hold at the April 29-30 meeting, while most economists surveyed by Reuters still expect the ECB to hold through all of 2026. That gap between the consensus view and the hawkish minority is exactly where policy surprises are born.

Success for the ECB looks like a credible hold-with-hawkish-bias that anchors inflation expectations without tipping the economy into recession. Failure looks like either a surprise hike that breaks the euro carry trade or a dovish hold that lets inflation get away. The observable signal: Lagarde's press conference language on April 30. If she drops "meeting-by-meeting" for anything more committal, European banks and the euro/dollar cross move first, hardest.

March retail sales came in at 1.7% month-over-month, more than double the 0.8% consensus and the strongest print in three years. A victory lap for consumer resilience — until you read what's inside it.

A significant chunk of the surge was gasoline station receipts. When gas prices spike, the dollar value of fuel purchases rises even if Americans are buying the same number of gallons. That's not strength; that's an energy tax showing up as spending. The same dynamic inflated March CPI to 0.9% month-over-month — triple February's pace and the hottest print since mid-2022, driven almost entirely by energy.

The more honest read is mixed: core retail (ex-gas and autos) still posted a solid gain, and discretionary categories like electronics held up. But furniture and sporting goods declined, consistent with households reallocating dollars from wants to needs. That's a classic regressive-shock pattern. The April PCE print on April 30 will be the first clean read on whether this is bleeding into core services — the number the Fed actually cares about. If it accelerates, the rate-cut clock that briefly restarted after the Strait appeared to reopen gets pushed into 2027.

Forget the EPS beat. On the conference call, CEO Scott Strazik said GE Vernova added $13 billion to its backlog in 90 days and now expects to hit $200 billion in 2027 — a year ahead of prior guidance. The stock jumped 12% on the session and is up more than 60% year-to-date.

The read-through is direct: AI data centers need power, and the power-infrastructure buildout is running faster than anyone modeled a year ago. GE Vernova booked roughly $2.4 billion in data-center equipment orders in a single quarter — more than all of 2025 combined, per the AP's reporting. The constraint on AI is no longer compute or capital. It's grid capacity and transformer lead times.

If this holds, the winners are heavy electrical, copper, and the utilities that can actually connect hyperscale loads. Failure looks like hyperscaler capex guidance pulling back next week when Microsoft, Alphabet, Amazon, and Meta report. Watch those calls on April 29 — they are the single most important forward indicator for the entire AI industrial complex right now.

Wednesday morning: two container ships in Iranian custody, a third with rocket-propelled-grenade damage. Wednesday close: S&P 500 at 7,137, a fresh all-time high. The index has developed impressive compartmentalization skills over the past 54 days.

Look underneath. Per Bank of America, investors pulled $15.4 billion from U.S. stocks in March, and total U.S. outflows over the past year have reached $284 billion — a sustained, year-long reallocation away from American assets. The Russell 2000 is up roughly 12% year-to-date while the Nasdaq trails at about 5% year-to-date. Healthcare and international funds are leading. That is not a risk-on melt-up; that is a market hedging its bets on U.S. exceptionalism while still showing up to the party.

Earnings are genuinely excellent — 85% of S&P 500 companies have beaten estimates. The question is whether that strength survives a second quarter of $100 oil and a Fed that can't cut. And when the Strait briefly flickered open earlier this month, oil crashed 7% on the day and small caps ripped higher — until seizures reversed most of it. The headline index can ignore physical-world shocks; its constituents cannot.

Tesla posted Q1 adjusted EPS of $0.41 versus a $0.37 consensus and revenue of $22.39 billion against a $22.64 billion expectation. The beat is cosmetic; the revenue miss is the signal. Tesla's margin story depends on volume, and volume is under pressure from competition and a macro environment where auto loans sit near 7% and EV demand has been disproportionately financed.

The broader read: the market is bifurcating between companies with direct AI infrastructure exposure — GE Vernova, the hyperscalers, semiconductor toolmakers — and companies that are merely adjacent to AI. Tesla is the cleanest example of the latter. Narrative has carried the stock for fifteen years; fundamentals are increasingly telling a different story.

The signal to watch: prime auto ABS delinquency rates. If they tick up, the financed-demand ceiling is real, and every consumer-durables company with a credit arm gets repriced. Microsoft, Alphabet, Amazon, and Meta report next Wednesday. Their capex language — raising, holding, or cutting — is the macro event of the week.

President Trump extended the 14-day Iran ceasefire ahead of its expiration earlier this week, with no specified end date — it runs until Iran submits a new peace proposal. Markets read it as risk-off; the physical reality is more complicated. Three ships were attacked in Hormuz within hours of each other even as the ceasefire held on paper. Britain is hosting military planners from more than 30 countries this week to discuss reopening the strait.

WTI sits at roughly $101 as of April 23, 2026, after spiking to nearly $115 intraday on April 7. The ceasefire is a political construct; the blockade is a physical one. Until tankers actually move, every inflation print, every central bank meeting, and every earnings guidance call is being made against structurally elevated energy costs. The IMF's adverse scenario — 2.5% global growth, 5.4% inflation — becomes the base case if the Strait stays closed through Q3.

The short-term signal to watch: tanker transits per day via Hormuz. They've been down materially week-on-week during closure episodes. That's the number that matters, not the ceasefire headlines.

A Fed nominee who wouldn't say what he thinks, two container ships held hostage in the world's most important oil corridor, and a record-high S&P 500 closing six miles from the headlines that should have broken it. The consensus view is that nothing matters until it suddenly does — which is also, conveniently, the entire investment thesis of the people who just stripped out their downside protection.

Stay liquid.

Forward this to the CFO who's still modeling 2026 on $80 oil and a dovish Fed — they'll want the warning now rather than in the earnings call.