Capital Rotated Before the Selloff | Nvidia Worst Day Since April | Geneva Pushes Risk to June | February Closes the Books

Nvidia fell 5%+ on a clean beat. Asia's best February since 1998. Iran's June window jumped 8 points. February closes today.

The month had a verdict before the session opened.

Nvidia fell more than 5% on Thursday, one of the strongest quarters in its history producing its worst single session since April, and the divergence told you exactly where the pressure lives.

The Nasdaq dropped over 1%, the S&P slipped about half a percent, but the Dow barely moved, the hit concentrated in growth names rather than broad risk.

The 10-year yield broke below 4%, bonds rallying while stocks sold off, and the dollar softened alongside as the VIX ticked higher without spiking.

Gold stayed elevated and oil held near $65 as Geneva concluded without a deal, a surface that looked calm enough until you noticed what sat underneath it.

PPI drops this morning. Core PCE already sits at 3.0%, which means the first fresh inflation signal of March arrives on the last day of February.

The beat was priced. The rotation beneath it wasn't.

This is where prediction markets offer a lens traditional indicators don't.

I had to share this today.

A strange new “wonder material” just shattered two world records, and the company behind it is suddenly partnering with some of the biggest names in tech.

We’re talking Samsung, LG, Lenovo, Dell, Xiaomi… and Nvidia.

Nvidia is already racing to deploy this technology inside its new AI super-factories.

Why the urgency?

Because this breakthrough could become critical to the next phase of AI. And if any tiny stock has the potential to repeat Nvidia’s 35,600% climb, this might be it.

See the full story here.

The MSCI Asia Pacific Index gained around 7% in February, its best performance since the index launched in 1998, and Europe's benchmark advanced nearly 4%, heading for its eighth straight monthly gain while US markets finished February lower.

Around $52 billion exited US equity products year to date, the largest early-year outflow in more than 16 years, and the AI scare trade didn't just reprice individual stocks. It routed capital across continents.

South Korea's benchmark gained around 20% in February alone, copper advanced 6% on the month, and miners climbed more than 10%, all of it building weeks before Thursday's session confirmed what the flows already signaled.

The Nvidia selloff wasn't the catalyst. It was the confirmation.

Polymarket's Largest Company contract tells the rest of the story: Nvidia at over 99% on $15 million in volume to remain the top position by Friday, but Apple sitting at over 99% to hold second place on $1.7 million, a split that prices two things at once.

The infrastructure thesis held. The session-level trade didn't. The crowd reconciled both across February, pricing a structural moat that survived the quarter and a near-term multiple that no longer supports pre-earnings levels.

The Regime Read

The rotation wasn't a reaction to Thursday's session. Thursday's session was a reaction to the rotation. Capital repositioned across February into markets less exposed to disruption risk, and the limiting variable isn't whether the infrastructure thesis survives. It's whether a new catalyst is needed to re-rate US tech from here.

Nvidia fell more than 5% on a clean beat, revenue landing around $68 billion with Q1 guidance near $78 billion, a print strong enough to confirm the thesis but carrying implications that weren't sitting inside the 92% beat probability.

China revenue at zero, supply constraints through Q1, Salesforce guidance coming in soft the same night. None of it modeled. All of it repriced.

The Polymarket $190 contract confirmed the split: 85% before earnings, 20% after Thursday's close, while the $180 level held at 95%, the crowd abandoning the session-level trade without touching the structural thesis.

Chip stocks broadly followed lower, Broadcom and Lam Research dropping alongside in the Nasdaq's worst session of the week, all of it compounding on a beat that delivered exactly what consensus expected.

Nvidia still sits at over 99% to remain the largest company by Friday. The structural moat held. The crowd closed the position, not the thesis.

The Settled Inference

A clean satisfies the model. It doesn't satisfy the multiple. When 92% consensus meets 92% reality, price reprices on everything the consensus excluded, and that repricing revealed the gap between a thesis worth holding and a level worth defending. The constraint shifted from conviction to catalyst.

You may own it. Millions of Americans do. 

But while you weren't watching, Wall Street insiders started dumping this household-name tech stock months ago. It's already down 53%, and the people who move markets aren't done selling. 

40-year Wall Street Legend Marc Chaikin knows which stock it is, why it's happening, and why it may only be the beginning of a much bigger shift. 

Get the ticker here.

Talks wrapped with "significant progress" and both sides agreed to meet in Vienna next week, no deal and no breakdown, and the Polymarket Iran curve shows exactly how the crowd read that outcome.

The February 28 window dropped to 8%, down over 40% overnight, removing the immediate strike, but June 30 jumped to nearly 70%, up 8 points, which means the crowd didn't reduce escalation risk after Geneva. It pushed that risk into Q2, pricing time bought rather than tension resolved.

The shutdown is accumulating quietly alongside it, Kalshi's forecast now sitting near 50 days with at least 50 days at nearly half and at least 60 days climbing to around 30%, duration adding structural weight rather than reversing.

The Citrini scenario compressed from 12% to 10% on Kalshi on $88K in volume, the tail shrinking but holding, needing a positive productivity signal to compress further.

All three hinge on the same question: Vienna produces a framework or another deferral, each additional week of shutdown adds fiscal drag, and the Citrini tail decays only if the data cooperates.

The Deferral Premium

Geneva deferred, the shutdown extended, and the Citrini tail shrank but held, each signal resolving into a later window rather than a final answer. The crowd priced the deferral, not the resolution. When three risks push into the same future quarter rather than resolving in the current one, the premium doesn't shrink. It migrates. Vienna is where the next calculation begins.

Amateur question: the Iran contract barely moved after Geneva. Does that mean the crowd didn't react?

Professional question: which windows moved, and what does the redistribution reveal about sequencing?

The February 28 window dropped over 40% overnight, the session effectively closed, but June 30 jumped 8 points to nearly 70% while December held at over 70%.

The crowd didn't reduce escalation risk after Geneva. It pushed that risk into a later window, and that's not relief. That's sequencing.

The same logic applies to the Nvidia curve. The $190 contract collapsed from 85% to 20% while the $180 contract held at 95%, the crowd abandoning the session-level trade without touching the structural thesis.

Two different windows on the same asset. Two completely different signals. Same crowd, two different calculations, and the window is the variable that separates them.

Reading one contract gives you a number. Reading the distribution across windows gives you the crowd's sequencing, where it expects resolution and on what timeline.

The Distribution Signal

A single window tells you probability. The shift across multiple windows tells you sequencing. The odds figure isn't the signal. The direction of the volume redistribution is. That's where the edge lives.

Most traders analyze charts… patterns… indicators.

But they miss the one force that moves everything: money flow.

Institutional investors drive 80–90% of major stock price moves, and nearly half of their orders happen in secret Dark Pools.

That means the real action often happens before the chart ever changes.

TradeAlgo’s Dark Pool A.I. tracks this activity in real time and sends SMS alerts with the ticker when institutions start positioning.

If you want to trade with the current instead of against it…

Get FREE Dark Pool Alerts via SMS here.

February closes today. The scoreboard is live.

Asia gained around 7% this month, Europe gained nearly 4%, and US markets finished lower while around $52 billion exited US equity products year to date.

The capital didn't evaporate. It repositioned.

Nvidia posted one of its best quarters ever and fell more than 5% in a session. Geneva produced significant progress and the June escalation window jumped 8 points overnight.

The second inference ran at every level: session price, monthly capital flows, geopolitical risk curves.

PPI this morning is the first signal of March, and whatever it shows sets the rate path conversation for the next three weeks.

Warsh at over 90% on $739 million in combined Polymarket and Kalshi volume. Fed hold at over 95% on both platforms.

Shutdown forecast near 50 days and climbing. Vienna opens next week.

The month settled. The next window opened.

Capital moves early. Coverage catches up. The gap between the two is worth watching.