AMD Loses 17% Despite Beating Earnings, Chips Crater, Jobs Come in at Half Expectations, Eli Lilly Crushes It, and Retail Buys the PayPal Dip – Market Breakdown #171
S&P down 0.51%. Nasdaq down 1.51%. AMD worst day in 7 years. ADP jobs at 22K vs 45K expected. Eli Lilly up 8%. Alphabet beat after hours. Bitcoin back below $73K. The chip reckoning has arrived.
📊 THE MARKET BREAKDOWN
Weekly market intelligence for traders who think in systems, not headlines.
Issue #171 | February 4, 2026
🔥 Headlines & Hysteria (powered by Forked Feed)
Forked Feed says: AMD reported $1.53 EPS versus $1.32 expected. Beat by 16%. Data center revenue up 39% to $5.4 billion. Beat. Total revenue $10.27 billion. Beat. Q1 guidance of $9.8 billion, above the $9.39 billion consensus. Beat. The stock dropped 17%. Seventeen percent. Worst single-day loss in nearly nine years. Why? Because some analysts had whisper numbers above $10 billion and AMD dared to guide slightly below that. Because the bar for AI companies is no longer “beat expectations.” It’s “beat expectations so thoroughly that every possible whisper number from every analyst who didn’t publish their real estimate also gets obliterated.” Lisa Su went on CNBC Wednesday morning to explain that AI demand is “accelerating at a pace I would not have imagined” and that orders have “exploded” over the past few months. Didn’t matter. The stock closed at $200.19. Down $41. In one session. This is a company that beat on every metric, guided above consensus, and still got annihilated because the AI hype required perfection and it delivered merely excellent. Meanwhile, Jensen Huang from Nvidia appeared at a Cisco conference to say that AI won’t make software obsolete and people should calm down. He’s defending software now. The CEO of the company eating everyone’s lunch is out here playing defense for the victims. We’ve reached the part of the cycle where the apex predator has to tell the market to stop being so bloodthirsty.
The Chip Wreck Spreads From AMD to Everything That Touches Silicon
Forked Feed says: AMD wasn’t alone. Broadcom dropped 7%. Micron cratered 11%. Lam Research fell 10%. Applied Materials down 9%. SanDisk lost 16%. The Philadelphia Semiconductor Index had its worst two-day stretch since October. The Nasdaq 100 breached its 100-day moving average, which technical analysts treat like a vampire encountering sunlight. That level “is seen by some as a harbinger for more losses.” Great word choice, Bloomberg. Harbinger. Like a medieval plague doctor showing up at your village gates. Bank of America put out a note calling the selloff “internally inconsistent” and “logically impossible.” Their argument? You can’t simultaneously believe that AI capex is deteriorating to the point of weak ROI while also believing AI adoption will be so pervasive that it kills all software business models. Both things can’t be true at once. The market heard this and shrugged. Logic hasn’t governed this tape for weeks. We’re in vibes territory now. The vibes say AI is going to destroy everything and also that the companies building AI aren’t growing fast enough. Both at the same time. BofA says leading chip names are now trading below 1x projected earnings growth while the S&P trades at 1.5x to 2x. Doesn’t matter. The selling isn’t about math anymore. It’s about fear. And fear doesn’t check the PEG ratio before hitting the sell button.
Forked Feed says: The economy added 22,000 private sector jobs in January. The expectation was 45,000. That’s not just a miss. That’s arriving at a dinner party with half a bottle of wine and pretending you brought the whole thing. Nela Richardson, the chief economist at ADP, described it as a “lackluster month for hiring.” Lackluster. Like you’d describe a movie that was fine but forgettable. The number would have been negative if not for 74,000 hires in education and health services. That one sector did all the work. Professional and business services lost 57,000 jobs. Manufacturing lost 8,000 jobs, extending a streak that goes back to March 2024. Manufacturing has not added jobs in 23 months. Almost two years. Richardson also noted that 2025 saw only 398,000 total jobs added, down from 771,000 in 2024. “A continuous and dramatic slowdown in job creation for the past three years.” That’s an actual quote from the chief economist of the largest payroll processor in America. Continuous. Dramatic. Slowdown. Three words that should probably concern someone. The government’s official January jobs report, which was supposed to come out Friday, is delayed until February 11 because of the shutdown. So we get ADP and we wait. Two weeks to find out if the labor market that already looks soft is actually collapsing or just resting. Place your bets.
Forked Feed says: Eli Lilly reported $7.54 EPS versus $6.96 expected. Revenue of $19.29 billion versus $17.96 billion expected. Beat on top. Beat on bottom. Then guided 2026 revenue at $80 to $83 billion against a Street consensus of $77.64 billion. Beat on guidance. The stock jumped 8%. Mounjaro revenue up 110% year over year. Zepbound revenue up 122% year over year. The obesity drug business is so good that Eli Lilly is now expected to add $80 billion in revenue this year. That’s a number usually reserved for countries. Full-year 2025 revenue was $65.2 billion, up 45% from 2024. EPS grew 86%. The company said it reached “more than 70 million people around the world” last year with its drugs. Seventy million. And there are still hundreds of millions left to sell to. Novo Nordisk, by contrast, reported guidance on Tuesday that called for sales and profit to decline 5% to 13% this year. Same industry. Same drugs. One company is dominating so completely that the other is actively shrinking. This is what a winner-take-all market looks like. And the winner is the company that figured out how to charge $1,000 a month for what is essentially liquid discipline.
Forked Feed says: PayPal dropped 20% on Tuesday after firing its CEO, missing estimates, and guiding flat to down for 2026. On Wednesday, it dropped another 2%. And retail investors bought $79.4 million of it on the dip. That’s the largest single-day net inflow into PayPal since early 2022, according to VandaTrack. A six-fold increase from the prior session. Let me paint the picture. A company announces it fired the CEO after two years because his turnaround plan produced 1% branded checkout growth. The company guides for declining earnings. The company withdraws its 2027 targets entirely. The stock craters to $41. And retail says “you know what, this seems like a buying opportunity.” This is either extremely smart contrarian positioning or the financial equivalent of catching a falling knife while blindfolded. PayPal is now trading at less than 7x expected free cash flow. If you believe the business isn’t dying, that’s objectively cheap. If you believe digital payments is a commodity business getting crushed by Apple Pay and Google Pay and a dozen BNPL alternatives, you just bought a value trap on margin. The new CEO is Enrique Lores, formerly of HP. The printer company. His first day is March 1. His first job is explaining to Wall Street why the guy who ran a hardware company that hasn’t innovated since the LaserJet is the right choice to revitalize a payments platform competing against the most valuable company on Earth.
Forked Feed says: Alphabet reported after the close. $2.82 EPS versus $2.62 expected. Revenue of $113.8 billion versus $111.4 billion expected. Google Cloud revenue spiked 48% to $17.7 billion versus the $16.2 billion analysts wanted. Everything beat. Sundar Pichai called it “a tremendous quarter” and noted that annual revenue exceeded $400 billion for the first time ever. The Gemini app now has 750 million monthly active users. The Gemini AI model processes 10 billion tokens per minute. And then came the capex guidance. Alphabet said it expects to spend $175 billion to $185 billion in 2026. That’s roughly double the $91 billion it spent in 2025. Double. The entire 2025 capex was already considered aggressive, and they’re going to double it. CFO Anat Ashkenazi said the investments will go toward AI computing infrastructure to develop frontier models and meet demand. She also said they’ll make the investments “in a way that maintains a very healthy financial position for the organization.” That’s reassuring language that means nothing. You’re spending $180 billion. There’s no version of that sentence where the word “conservative” applies. RBC said the Cloud surge and Gemini momentum are “plenty good as proof points which warrant the higher spend.” That’s analyst-speak for “we don’t know if this is crazy or genius but the numbers look too good to argue with right now.”
Bitcoin Falls Below $73,000 Again Because Apparently Last Night’s Floor Was Just a Suggestion
Forked Feed says: Bitcoin traded at $72,176 Wednesday night. That’s below Tuesday’s intraday low of $72,884. The floor didn’t hold. The bounce didn’t bounce. The support turned out to be a trapdoor. This is now a 43% drawdown from the October peak of $126,000. The “Trump rally” that was supposed to change everything has officially given back more than everything. We’re trading below election night levels from November 2024. Analysts at Claude and ChatGPT, which is apparently how we do analysis now, both project a range between $70,000 and $85,000 for February with downside risk to $65,000 or lower. Veteran trader Peter Brandt noted that Bitcoin has “breached an important long-term support level on the weekly chart.” That’s technical for “the next floor is a lot lower than this one.” Mike Novogratz of Galaxy Digital says most of the leverage has been flushed and we’re near equilibrium. Others say we’re 20% to 30% from the final bottom. Nobody knows. CoinGlass shows $6 billion in liquidations over the past several days. That’s a lot of forced selling. The ETF inflows that were supposed to put a floor under the market have reversed into outflows. January saw $1.61 billion leave spot Bitcoin ETFs. The bid isn’t there. And until it comes back, the price action looks like a slow bleed punctuated by violent drops.
🔎 Today’s Focus — “The Chip Check
The SaaSpocalypse was yesterday’s panic. Today’s panic is the realization that if AI is good enough to kill software, maybe it’s good enough to disappoint the companies building it.
AMD’s collapse wasn’t about bad results. The results were good. Q4 data center revenue grew 39%. Total revenue beat by 6%. EPS beat by 16%. Lisa Su said demand is accelerating faster than she expected. None of that mattered.
What mattered was that the whisper numbers had gotten ahead of the guidance. What mattered was that Q1 revenue of $9.8 billion is below Q4 revenue of $10.27 billion. Never mind that Q1 always runs below Q4 in cyclical semis. Never mind that China export restrictions are taking $390 million out of the comparable period. The market saw a sequential decline and decided that was enough to sell 17% of a $394 billion company in one session.
The contagion was immediate. Broadcom, Micron, Lam, Applied, they all got dragged down. The Philadelphia Semiconductor Index had its worst two-day stretch since October. This is the index that led the entire AI trade. The one that was supposed to be immune because chips are the picks and shovels. Turns out even picks and shovels can get sold when the miners start questioning whether there’s actually gold in them hills.
Bank of America made the case that the selloff is “internally inconsistent.” Their logic is sound. You can’t believe AI spending is collapsing and also believe AI is so powerful it will destroy all software. Those are mutually exclusive outcomes. The market is pricing both.
But BofA is assuming the market cares about internal consistency. It doesn’t. The market cares about the last headline and the direction of the last tick. AMD guided below whisper numbers. Tick down. Software is getting disrupted by AI tools. Tick down. The two stories feed each other even when they contradict each other. That’s how sell panics work.
The Alphabet earnings after the close offered some relief. Cloud up 48%. Search up 17%. EPS beat. Revenue beat. Capex guidance of $175 to $185 billion for 2026 is insane by any historical standard, but it’s the kind of insane that says “we believe in this enough to double down.” That’s different from AMD’s guidance, which was good but not insane enough.
The market is now sorting the AI trade into winners and losers. Alphabet looks like a winner. Google Cloud is actually winning business. Gemini has 750 million users. The Search franchise is still growing despite everyone predicting its death. AMD looks like it might be a loser or at least not a winner big enough to justify the multiple. Data center is growing 39% but Nvidia is still taking share. The MI series chips are good but not good enough to dethrone the H100 empire.
That’s the new test. It’s not enough to participate in AI. You have to be winning at AI. And winning means exceeding not just the published estimates but the unpublished hopes. Anything less gets sold.
⚡ The Setup
SPY ~ 686.19 | BTC ~ 72,176.49 | US10Y ~ 4.272% | DXY ~ 97.719
The S&P 500 fell 0.51% to 6,882.72 as the semiconductor selloff dragged the broader market lower. The Dow gained 260 points to 49,501, diverging sharply as investors rotated into defensives and healthcare. Nasdaq dropped 1.51% to 22,904 for its worst two-day stretch since October. The 100-day moving average broke on the Nasdaq 100, a technical level that tends to invite more selling.
AMD was the day’s anchor. Down 17% despite beating on every metric because Q1 guidance came in below whisper numbers. The chip wreck spread to Broadcom (-7%), Micron (-11%), Lam Research (-10%), and Applied Materials (-9%). The Philadelphia Semiconductor Index had its worst two-day performance since October.
Healthcare provided the counterweight. Eli Lilly surged 8% after crushing estimates and guiding revenue $3-5 billion above consensus. Amgen rallied 8% on earnings. The GLP-1 drug trade remains the most reliable bid in the market.
ADP reported private payrolls of just 22,000 versus 45,000 expected. Professional services lost 57,000 jobs. Manufacturing extended its losing streak to 23 months. The official January jobs report is delayed until February 11 due to the shutdown.
Bitcoin broke below $73,000 again in evening trading after failing to hold Tuesday’s low. The entire post-Trump rally has been erased. $6 billion in liquidations over the past several days per CoinGlass.
After hours: Alphabet beat on everything. Revenue $113.8B versus $111.4B expected. EPS $2.82 versus $2.62 expected. Google Cloud up 48%. Capex guidance of $175-185 billion for 2026 signals massive AI infrastructure commitment.
🧩 Market Archetype — “The Earnings Gauntlet
This week is a stress test for every narrative that’s been driving the market.
AI spending is real, but is it real enough? AMD said yes. The market said prove it. Alphabet said yes with an exclamation point. The market might believe that one. Amazon reports Thursday. Microsoft already disappointed. The sample size is growing and the verdicts are mixed.
The labor market is cooling, but is it breaking? ADP says 22,000 jobs. That’s a whisker above zero. Education and health did all the lifting. Manufacturing is in a 23-month drought. The official numbers are delayed until February 11. We’re flying blind on the most important macro data in the economy.
Software is being disrupted, but does it matter for the disruptors? Palantir surged. Alphabet beat. The AI companies themselves are doing fine. It’s their customers, the ones buying all that software, who are rallying because their costs are about to go down. The rotation is tech to defensives, but within tech it’s “old tech” to “AI winners.” The sorting continues.
The flow story is violent rotation, not broad liquidation. The Dow was up 0.53% today while the Nasdaq fell 1.51%. That’s a 200 basis point spread. Amgen rallied 8% on earnings. Eli Lilly jumped 8%. The market isn’t risk-off. It’s tech-off. Specifically old-tech-off. Specifically software-and-now-maybe-semis-off. The money isn’t leaving. It’s just moving.
🧭 Flow Pulse
Equity flows continued the violent rotation that started Tuesday. Money left tech at the fastest pace since October, with the Nasdaq 100 down 3.2% over two days while the Dow gained ground. The 200 basis point spread between the Dow (+0.53%) and Nasdaq (-1.51%) on Wednesday tells the story. Healthcare and defensives absorbed the outflows. Amgen rallied 8% on earnings. Eli Lilly surged 8% on a blowout quarter. Merck gained 3%. The bid is real, just not for anything with exposure to AI disruption.
Semiconductor flows saw capitulation after AMD’s 17% collapse. Broadcom fell 7%. Micron cratered 11%. Lam Research dropped 10%. Applied Materials lost 9%. The Philadelphia Semiconductor Index breached its 100-day moving average for the first time since the October selloff. This was the sector that was supposed to be immune because chips are the picks and shovels. Now the picks and shovels are getting sold too.
Retail flows showed contrarian behavior. VandaTrack reported $79.4 million of net retail buying in PayPal on Tuesday, the largest single-day inflow since early 2022 and a six-fold increase from the prior session. Mom and pop bought a stock that had just lost 20% and fired its CEO. Either extremely smart or extremely painful in six months.
Crypto flows remained deeply negative. Bitcoin briefly broke below $73,000 again on Wednesday night, failing to hold Tuesday’s intraday low. CoinGlass shows $6 billion in liquidations over the past several days. No signs of institutional accumulation. ETF outflows continue. The bounce attempts keep failing because there’s no bid underneath.
Forked Feed says: The flow picture is sorting itself into two buckets. Things AI can replace are getting sold. Things it can’t are getting bought. Chips just got added to the first bucket because beating expectations by 16% wasn’t enough. Healthcare and staples are in the second bucket because you can’t automate a pill. Retail is out here catching knives. And crypto is bleeding out slowly because the Trump trade died and nothing has replaced it.
🔮 Forked Forecast
Base Case (50%): Alphabet Stabilizes the Tape
Alphabet’s after-hours beat provides the narrative reset the market needs. Cloud growing 48% reminds everyone that AI spending is working for the companies doing it right. AMD’s washout creates a contrarian opportunity that gets bid Wednesday morning. Software stabilizes near Tuesday’s lows as oversold conditions attract buyers. Eli Lilly’s blowout guides capital back to quality healthcare. Bitcoin finds a floor in the low $70Ks as the worst of the leverage flush passes. Amazon Thursday becomes the next focus. S&P holds 6,850-6,900. Nasdaq stabilizes. Dow continues to outperform as rotation continues without accelerating.
Bull Case (20%): The Buyers Return
Alphabet’s capex commitment is treated as a vote of confidence in AI infrastructure. AMD’s selloff gets called overdone and the stock bounces 5% as short-covering kicks in. Software names get a relief rally as Bank of America’s “internally inconsistent” note spreads. Super Micro’s 10% rally extends into AI infrastructure peers. Amazon beats Thursday and restores faith in cloud spending. Bitcoin reclaims $75K and the seller exhaustion narrative takes hold. S&P pushes back toward 6,950. Nasdaq snaps its losing streak. The SaaSpocalypse becomes a buying opportunity.
Bear Case (30%): The Contagion Deepens
Alphabet’s $180B capex guide gets treated as a warning sign that AI spending is unsustainable. AMD’s breakdown triggers copycat selling in Nvidia and Broadcom. The Nasdaq 100’s break below the 100-day average leads to program selling and CTA liquidation. Bitcoin breaks $70K and tests $66K as the bear cycle accelerates. ADP’s weak jobs number gets extrapolated into recession fears. Private credit stress from the software selloff spreads to other leveraged lending books. Amazon disappoints Thursday. S&P breaks 6,850 and the rotation becomes a broader selloff. VIX pushes above 20.
Triggers to Watch:
Alphabet reaction in Asia and at tomorrow’s open. The beat was clear. The question is whether the tape believes it.
Amazon earnings Thursday after close
AMD for any reversal of the 17% decline
Nasdaq 100 at 100-day MA. Below that opens technical selling
Bitcoin at $70K psychological support
Software ETF (IGV) for stabilization or further breakdown
Super Micro momentum. Up 10% today on strong guide
February 11 jobs report. The one that actually matters
February 13 DHS funding deadline
Private credit names (Blue Owl, Ares, TPG) for further stress
Dollar index. Below 97 helps. Above 98 hurts
📖 Available Now!
Before You Blow Up is a psychological reset for traders who already know the mechanics, but feel decision quality slipping when markets get loud.
This isn’t about new strategies, indicators, or setups. It’s about recognizing the moment risk starts lying to you, conviction turns artificial, and small mistakes begin stacking into real damage. Most traders don’t fail all at once. They drift, tilt, overtrade, and slowly bleed confidence away. This book exists to interrupt that process early.
Inside, you’ll learn how to spot psychological failure before it shows up in your PnL, reset your risk framework when noise overwhelms signal, and protect focus during drawdowns instead of compounding them. The goal is simple: trade less, think clearer, and stay solvent long enough for your edge to matter.
This plan also includes access to a private space tied directly to the book. I’ll occasionally add updates, clarifications, or extensions when market conditions materially change or when something needs to be said. No schedule. No noise. Only signal.
If you’ve ever felt one bad stretch turning into something bigger, this was written for you.
💬 Final Thought
AMD beat earnings by 16% and lost 17%. That’s the market we’re in now.
Not a market that rewards performance. A market that punishes insufficient excellence. A market where good isn’t good enough because great was already priced in and the price was wrong.
Lisa Su ran a company that grew data center revenue 39% and beat on every line item and guided above consensus, and she lost $41 per share in value because someone at a hedge fund thought the whisper was $10 billion and she guided $9.8 billion. A $200 million delta on a $394 billion company cost shareholders 17%.
This is what happens when expectations outrun reality. When the AI narrative becomes so baked into prices that the companies doing the actual work can’t meet the bar. When the story gets ahead of the numbers and the numbers have to catch up or get punished.
But here’s the thing. Alphabet beat after hours. Cloud up 48%. Revenue up 18%. Capex guidance that essentially says “we’re going to spend $180 billion building AI infrastructure next year and we don’t care what you think about it.” That’s not a company hedging. That’s a company swinging for the fences.
So the market has a choice. It can treat AMD’s selloff as a sign that the AI trade is broken. Or it can treat it as a company-specific overshoot in expectations that got corrected while the broader theme remains intact.
The ADP jobs number didn’t help. 22,000 jobs when you expected 45,000 looks like an economy that’s decelerating faster than the soft landing required. Manufacturing has lost jobs for 23 straight months. Professional and business services shed 57,000 positions in a single month. That’s not noise. That’s signal.
But Eli Lilly added $1.3 billion in revenue versus expectations in a single quarter. Alphabet beat by $2.4 billion. The companies that are winning are winning big. The dispersion is extreme. The market isn’t going down in a straight line. It’s sorting. Winners from losers. Disruptors from disrupted. AI beneficiaries from AI victims.
Bitcoin is back below $73,000. The Trump rally is gone. The crypto winter might not be over. But Alphabet is spending $180 billion on AI compute. Eli Lilly is growing earnings 86% selling weight loss drugs. The economy is still functioning, even if it’s functioning differently than it used to.
Retail bought $79 million of PayPal on a 20% down day. That’s either brave or stupid. It might take months to find out which.
That’s all for issue #171. AMD learned that good isn’t good enough. Chips followed semis off a cliff. The economy added barely any jobs. Eli Lilly printed money selling liquid willpower. Alphabet is going to spend more on AI infrastructure than most countries spend on everything. Bitcoin can’t find a floor. And retail traders are out here catching falling knives with both hands.
The gauntlet isn’t over. Amazon still has to report. The jobs number is still delayed. DHS funding still expires in nine days. But at least Alphabet gave the market something to hold onto tonight.
We’ll see if it’s enough.
— Forked Feed
🔗 Stay Connected
Twitter: @txwestcapital
Twitter: @theforkedfeed
YouTube: TexasWestCapital
Website: TheForkedFeed.com and ForkedFeed.ai (coming soon)