Killer Bees — Market Brief January 15, 2026

The Sting

When TSMC reports a 35% profit jump and jacks its capex by 40%, the stock rips 5%. When JPMorgan beats on earnings and revenue, the stock drops 4.2%. One of these things is about the future. The other is about a president who suddenly decided credit cards are the enemy.

Deep Dive #1: TSMC Just Proved AI Demand Is Real (And Nobody Cared Yesterday)

Taiwan Semiconductor dropped earnings Thursday morning that should've been front-page material: Q4 profit up 35% year-over-year, revenue climbing to $33.7 billion, and gross margins hitting 62.3% territory most chipmakers can't even dream about. More importantly, they're not just coasting on last year's AI boom. TSMC announced plans to boost 2026 capex to $52-56 billion, up from $40 billion in 2025. That's a 37% increase in capital spending. You don't drop that kind of money unless your customers are placing multi-year orders with confidence.

The market reaction? TSM jumped 5% in pre-market, lifting Nvidia (+2%), Micron (+2%), and the entire SMH semiconductor ETF by 3%. But here's what's interesting: on Wednesday, the day before TSMC reported, chip stocks got hammered. Nvidia fell 2.6%, Broadcom dropped 4%, and the Nasdaq shed 1%. Chinese customs had reportedly blocked Nvidia's H200 chips from entering the country, and suddenly everyone remembered chips have geopolitical risk.

What TSMC's numbers actually tell you is that AI infrastructure spending isn't slowing down it's accelerating. High-performance computing (which includes AI) made up 55% of TSMC's Q4 revenue, with forward guidance projecting Q1 2026 revenue of $34.6-35.8 billion, implying 38% year-over-year growth. CEO C.C. Wei acknowledged he's "very nervous" about AI bubble concerns, but then proceeded to detail how demand is coming from consumer, enterprise, and sovereign segments. The guy who makes the chips for Nvidia's chips is nervous about sustainability yet still spending $56 billion next year.

The question nobody's asking: If TSMC is this confident about multi-year AI demand, why did the market need a full earnings report to believe it? Nvidia's been telegraphing this for months, but until TSMC showed up with receipts actual CapEx commitments the market was treating chip stocks like a trade, not a structural shift.

Deep Dive #2: Trump Discovers Credit Cards Exist, Financial Stocks Discover Pain

On Friday evening, President Trump posted on Truth Social that he was calling for a one-year 10% cap on credit card interest rates, effective January 20. No legislation, no executive order drafted, no mechanism outlined just a post. By Monday morning, Capital One had dropped 6%, Synchrony Financial fell 8%, and even diversified banks like JPMorgan (-2%) and Citi (-3.5%) were bleeding.

The average credit card APR is 22.3%, up from 16.3% in 2020. A 10% cap would cost banks an estimated $100 billion in annual revenue, according to Vanderbilt researchers. For Capital One and Synchrony banks that are credit card businesses analysts said it would "wipe out earnings." Wells Fargo estimated large bank EBT would take a 5-18% hit. Visa and Mastercard, which don't actually issue cards but process them, still fell 2% because fewer profitable cards means less transaction volume.

Here's what's odd: the banking industry lobby is one of the most powerful in Washington, and this would require Congressional action to implement. Every analyst noted the proposal's legislative path is basically nonexistent. Yet JPMorgan reported strong Q4 earnings on Tuesday revenue up, net interest income solid and the stock still dropped 4.2%. Bank of America, Wells Fargo, and Citi all reported Wednesday with similar results: fine numbers, sell-the-news reactions.

The market isn't pricing in a 10% rate cap happening. It's pricing in uncertainty about what else Trump might target. Defense contractors can't buy back stock or pay dividends. Homebuilders are under scrutiny for stock buybacks. Private equity can't buy single-family homes. And now credit cards. The thread connecting these is that they're all politically visible industries that touch consumer pain points. The market's learning that "pro-business" doesn't mean "predictably business-friendly."

Chart of the Week: Small Caps Finally Break Through

The Russell 2000 hit an all-time high Wednesday, outperforming the S&P 500 for nine straight sessions—matching the longest streak since 1990.

While megacap tech took back-to-back losses for the first time in 2026, small caps kept climbing. Wednesday's session saw 51.8% of U.S. stocks advancing despite the major indices being red—a clear sign of broadening participation. Biotech names like Viking Therapeutics (+14.8%) and energy plays rode the momentum, while the Mag 7 all finished negative.

This isn't just a "risk-on" rotation. It's the market saying: if policy uncertainty is going to whipsaw large-cap tech and financials, maybe there's alpha in companies too small to be presidential Twitter fodder. The Russell's nine-day streak matters because it signals a structural shift in where investors think 2026's returns will come from. When the most important companies can't move their own stocks with strong earnings (looking at you, JPMorgan and TSMC), maybe the answer is to own 2,000 companies instead of seven.

The Hive Mind

FinTwit went "extremely bullish" on TSMC after earnings, with Stocktwits sentiment flipping from "bullish" to "extremely bullish" and message volume hitting "extremely high." Users were predicting $350 near-term targets and calling Thursday a lift for the entire tech sector which it was. Meanwhile, the credit card drama had retail mostly confused: banks are bad for charging high rates, but also Trump can't actually do this, right? The interesting tension: retail is loud and bullish on chips; institutional positioning on banks is defensive. One of these groups is about to be right.

The Thread

Two earnings beats this week told opposite stories. TSMC proved AI demand is accelerating with $56 billion in capex to show for it. JPMorgan proved banking is profitable with $90 billion in Q4 revenue. One stock rallied 5%, the other fell 4%. The difference? One faces policy uncertainty, the other doesn't—yet. The question is: when the most predictable businesses become unpredictable overnight via Truth Social, where do you actually put capital in 2026?