The 3:50 PM Market Algos Explained

If you watch the markets every day, you've probably noticed that weird, chaotic burst of activity around 3:50 p.m. Eastern. Stocks start whipping around, spreads widen, and it feels like the algos have completely taken over.

They have. But not for the reason most people think.

Let's break down what's actually happening in those final ten minutes of the trading day.

It's Due to MOC Orders

At 3:50 p.m. Eastern, the New York Stock Exchange publishes the MOC/LOC order-imbalance data.

• MOC stands for Market-on-Close.
• LOC stands for Limit-on-Close.

These are orders that execute at exactly 4:00 p.m. at the closing-auction price, regardless of where the stock is trading in the seconds before the bell.

Why does this matter? Because at 3:50 p.m., the exchange shows everyone its cards. It reveals whether there's a massive imbalance of buy orders versus sell orders waiting to execute at the close.

Imagine you suddenly learned there's $500 million more in buy orders than sell orders for Apple stock, all scheduled to hit at 4:00 p.m. What would you do? You'd probably buy Apple right now, anticipating that the closing auction will push the price higher.

Now imagine that every algorithmic trading system on Wall Street just received that same information at the exact same microsecond.

Front-running the imbalance. If the data shows a big buy imbalance, algos immediately start buying, expecting the closing auction to push prices up. If there's a sell imbalance, they sell, anticipating downward pressure.

What It Looks Like on Your Screen

• Sudden volume spikes. Stocks that were trading 100,000 shares every five minutes suddenly see millions of shares trade in seconds.
• Reversals that make no sense. A stock selling off all day might suddenly rip higher in the last ten minutes, not because anything changed fundamentally, but because the MOC data revealed an unexpected imbalance.

It's Not Manipulation

Many traders believe 3:50 p.m. is when manipulation kicks in, but it isn't. The exchanges publish order-imbalance data to promote transparency. The problem is that when you give sophisticated algorithms transparency into future order flow, they will exploit it.