TO: Portfolio Managers, Risk Officers, Concerned Subscribers
FROM: The Markets Correspondent
DATE: Friday, 6:47 PM EST
RE: What it means when financials crater while energy rises
Let me be direct: something structural shifted this week, and it wasn't the geopolitical theater everyone's talking about.
The headline read like a rerun—Greenland threats, tariff posturing, the whole Davos kabuki on repeat. Tuesday we got the expected capitulation of equities (S&P 500 down 2.06%, Nasdaq down 2.39%, VIX spiking to 20.99). Wednesday and Thursday brought the reflexive recovery. By Friday, we had erased most of the losses.
But the composition of gains and losses tells a different story. And that story is about who believes the setup anymore.
Goldman Sachs dropped nearly 4% Friday. Caterpillar fell 3.34%. These aren't tech stocks spooked by consumer sentiment—these are supposed to be the beneficiaries of the current regime. Higher rates, tighter credit spreads, robust industrial demand, a stable labor market. Everything should be flowing to the dividend-paying, economically-sensitive blue chips.
Instead they're getting sold. Hard.
The Russell 2000, which had staged a 14-session winning streak built on "small caps are immune to tariffs," fell 1.7% Friday and posted its worst day since November. The trade was predicated on the theory that small-cap revenues aren't exposed to international headwinds. That thesis had legs for exactly three weeks. Now the reality is asserting itself: when the Fed isn't cutting rates, and tariffs might actually bite into growth, being domestic-only isn't a moat. It's a liability.
Meanwhile, Microsoft rose 3.45%. Amazon up 2.06%. Nvidia up 1.5-1.6% daily. The mega-cap, globally-diversified, high-margin software and semiconductor names are consolidating gains. That's not a rotation toward growth—that's a recognition that if the economy narrows, you want to own the names that can survive the narrowing and keep expanding margins anyway.
Gold hit a fresh record high this week. The 10-year Treasury yield stayed pinned to 4.24% despite the "recovery" in risk sentiment. That's not risk-on behavior. That's what you see when investors are saying: "I'll take equities, but only the ones I trust."
The University of Michigan sentiment index came in at 56.4, up from December's 52.9 and beating the 54.0 consensus estimate. Inflation expectations for the next year dropped to 4%, the lowest since January 2025. Real consumer spending grew 0.3% monthly across both discretionary and necessities.
On the surface, this is good news. The American consumer isn't panicking. They're still spending.
But here's what matters: sentiment revisions upward paired with PMI readings showing mild cooling in both services and manufacturing is the textbook signal of a late-cycle consumer. The data says people feel okay right now. It doesn't say they expect to feel okay in six months.
An Arctic storm is about to disrupt 10%+ of U.S. natural gas production. Goldman Sachs flagged this Friday morning. Food costs remain elevated despite easing core inflation. The "K-shaped recovery"—where the wealthy spend while lower-income Americans pull back—continues to widen, according to Stifel strategist Barry Bannister, who predicts falling aggregate labor income in 2026 will slow consumption.
Resilient today is not the same as resilient forward-looking.
Intel dropped 17% on weakened guidance and manufacturing snags. Broadcom fell 1.7%. The rest of the semiconductor complex held because Nvidia got the China card played again—Jensen Huang visiting Beijing, Chinese officials signaling that state-backed tech companies can prepare H200 orders.
What you need to see here: the win for Nvidia is real, but it's narrow. One company, one supply relationship, one geopolitical angle. The broader semiconductor narrative—that AI-driven capex is a tidal wave lifting all boats—is showing its limits. Intel can't deliver on nodes. The foundry advantage is consolidating around TSMC and Samsung. Broadcom's networking strength isn't enough to offset weakness elsewhere.
When a pillar of a sector fails this comprehensively, it's because the fundamentals were suspect for others too, they were just better hidden. Intel became the truth-teller.
The Federal Reserve meets January 27-28. Everyone knows it'll hold rates. The market is priced for a 95%+ probability of no cut. The three-month forward rate expectations show maybe one quarter-point cut all year—and probably not until late 2026.
But here's what matters: when Powell holds, will he signal any confidence that cuts could come later in the year? Or will the dot plot show the same one-cut projection it showed last time?
Because the market's entire equity recovery this week was built on the assumption that the tariff risks are negotiable and that the Fed will eventually capitulate to slower growth. If Powell comes out of his presser and the committee signals rates are just staying at 5.25-5.5% with no easing in sight, the calculus changes entirely.
Apple is down 8% year-to-date. Meta is down 8%. Microsoft is down 6%. The Magnificent Seven haven't broken, but they're bending. That's the canary. When the names that were supposed to drive 2026 returns start stumbling on higher-for-longer rates, the whole structure becomes a question mark.
And then there's the Fed chair decision. Treasury Secretary Scott Bessent told CNBC Tuesday that a choice is "maybe as soon as next week." The field is down to four candidates. If it's a dove, markets calm. If it's a hawk, watch the 10-year spike past 4.3%.
Monday morning: Any statement from the Treasury on tariff timing or scope. The "framework" conversation with NATO is too vague to trust.
Wednesday morning: S&P Global flash PMI for January. If that cooling in services accelerates, the job market pressure becomes real.
Thursday afternoon: The Fed hold and the dot plot. This is the match.
The week's moves look routine on the surface—whipsaw, recovery, consolidation. But underneath, you've got a market where the defensive trades are working (gold, mega-cap tech), the cyclical trades are failing (financials, small caps), and the yield curve is saying "I don't believe the growth story anymore."
That's not a narrative. That's a warning.
What's your read on the consumption data holding up through an Arctic storm and higher heating costs? I'm looking to understand if the January sentiment gain is real or borrowed. Reply to this email—I read the market views that matter.