Citadel and GMI’s latest data:
In 2026, every time the S&P drops, retail investors’ net buying is a normal 3.5 times.
Institutions have net sold tech stocks by -4 standard deviations.
Institutions are selling, while retail investors are taking it.
Over the past two years, retail investors have been trained into a muscle memory routine:
Crash → Trump backs down → V-shaped rebound → the fortune of bargain hunting.
The April 2025 tariff war.
The Iran-Iraq war between Iran and the U.S. in March 2026.
Two scripts, exactly the same.
The problem, though, is right here.
Every V-shaped reversal only makes the next round’s leverage worse.
Retail wins once and adds to the position; wins twice and goes all-in; wins three times and gets on Margin.
Institutions see it for what it is, and on the next big selloff day they’ll dump even more shares into your hands.
But there is no rule that stocks always rebound into a V forever. It just happened to V twice.
The truly dangerous part is the structure: retail isn’t using cash to buy the dip. It’s using leverage.
Margin accounts, margin buying, options buyer positions—every layer is debt.
If the S&P drops 5% and the margin isn’t enough, brokers will place calls.
Add margin.
Retail has no money.
Liquidation.
Forced liquidation sell orders smash into the market; the S&P drops another 3%.
Another batch of margin accounts gets punched through. Another round of forced selling.
This is what a stampede is.
Once a stampede starts, it has nothing to do with retail’s beliefs. The liquidation algorithm doesn’t care about belief. It only cares whether there’s enough margin.
The current setup: leveraged retail accounts are piled up on top, with the memory of two V-shaped reversals cushioning underneath.
Being able to catch it is called contrarian investing.
If you can’t catch it, it becomes a chain liquidation.
As long as it doesn’t fail to rebound into a V just once?
Tech stocks saw net inflows of more than 20 billion, turning into net outflows of 15 billion—this is only the beginning.
Institutions are still selling. If earnings in July’s reporting season can’t hold up the P/E, then the orders where retail adds leverage to charge in this time will all be left hanging on top.
What do you do?
First, reduce leverage. Do it now.
Don’t wait until the Margin Call arrives to passively cut positions.
Cut leveraged positions to zero.
For options buyer positions, the premium must be no more than 5% of the total position size.
This position isn’t worth taking leverage.
Only consider borrowing for assets with a P/E below 20; don’t touch anything with a P/E above 35.
Second, hold cash. Cash isn’t wasted—it’s an option.
When the market collapses, cash is the most expensive asset.
For every extra dollar of cash you have now, during a stampede you can buy three dollars’ worth of assets.
In March 2020, check what stock price levels the people with cash bought.
Third, buy volatility.
Now the VIX is being pushed too low by the past two V-shaped reversals.
The market treats "it must rebound into a V after a crash" as the pricing baseline.
VIX calls or VXX calls—spend a small amount of money to buy insurance that pays 3x to 5x when a stampede happens.
This isn’t speculation—it’s fireproofing.
Fourth, rotate out of assets.
Cut tech, cut high P/Es—cut everything whose valuation is propped up by the "AI narrative."
Add cash, add short-term U.S. Treasuries, and add a bit of gold.
Gold also gets smashed in the early stages of a liquidity crisis—but it’s the first among all the smashed assets to rebound.
You don’t necessarily need to do everything. But at least do one thing.
Because on the trading table, the most dangerous situation isn’t just losing money in a row.
It’s believing the rules belong to you after winning in a row.
#Megadrop VIX