Why Stock Prices Move Before Earnings Do — The Market Caches the Future

A few weeks ago a junior colleague shoved a phone screen at me over lunch. Record earnings, the headline said. The stock was down. How does that make sense, he asked.

That's the first wall anyone hits when they start looking at stocks. The year a company makes the most money, the stock drops.

It runs the other way too. A company bleeding cash for years, barely anything coming in the door, and the stock keeps climbing.

By common sense, that's backwards.

Companies that make money should go up. Companies that don't should go down.

The market doesn't work that way.

I've spent a long time in infrastructure, so when something doesn't match my expectations my reflex is to reach for a systems analogy. Model the stock market as a system that reads current data and reacts, and it keeps contradicting you. Model it as a system that caches future data ahead of time, and things start to line up.

Earnings are a log of the past

Say you've been handed a server.

Yesterday: 30% CPU, no incidents. That's useful information.

But it doesn't let you claim the box will be fine tomorrow.

If a huge event kicks off tomorrow and traffic comes in at 10x, yesterday's log is just background reading.

Nobody who actually runs systems looks only at yesterday's log. You size for the traffic you expect and scale up before it arrives.

Same with the market.

Revenue and operating profit matter. But they're results that already happened. They're the log the company left behind.

The person buying the stock is calculating what that company will earn from here.

So the price moves before the earnings do.

Record earnings, falling stock

A company earned 1 trillion won last year. This year the market expected 1.5 trillion. It came in at 1.3 trillion.

Look at the company alone and that's a great year. Profit up 30%, from 1 trillion to 1.3.

The stock falls anyway.

Because the number already sitting in the price was 1.5 trillion.

The earnings weren't bad. The expectations parked on the future were too high.

The IT version: you provisioned a pile of servers expecting a traffic spike, and the traffic didn't show up at that size.

So you recalculate down to what you actually need. In the market, that recalculation shows up as a price correction.

And why a money-losing company goes up

Now take a company that's still in the red.

Look at today's numbers and there's nothing to like.

But say the new product is growing fast, the market itself is widening, and you can sketch a picture where serious cash flow shows up in two or three years.

Investors don't stop at today's losses. They start pricing in future profit. That calculation lands in today's price.

So you get this:

Losing money, stock going up.

Nothing strange about it.

You're not buying a past report card. You're buying and selling a claim on cash flows that haven't been generated yet.

There's a catch attached.

The future hasn't happened. Forecasts are wrong.

The bigger the expectation, the bigger the possible return — and the wider the surface where it can break.

The price is an expected value, not a current value

Most people picture it like this:

Good company → good earnings → stock goes up

The real loop has an extra turn:

Forecast the future → expectations form → priced in now → actual results land → expectations revised → price re-adjusts

It's far more often the gap between earnings and expectations that moves the stock hard, not the earnings themselves.

Once you can see that gap, the news reads differently.

"Revenue up 30%, stock plunges" stops being a contradiction.

You just check one more thing. What was the market expecting?

Which means a good company and a good stock aren't the same thing

That opens up another problem.

Say there's a genuinely excellent company.

Revenue growing, strong technology, big market share. Good company by any measure.

What if everyone already knows that?

What if the expectation that it'll grow enormously over the next decade is already fully in the price?

The company is still excellent. Whether it's a good thing to buy is a separate question.

Buying a server that performs at 100 for a price of 100 and buying the same box for 1,000 are completely different transactions.

"The server is great" and "the price I paid makes sense" have to be evaluated separately.

Stocks work the same way.

Analyzing the company gets you halfway. You need the company's value and the price the market is already paying, side by side.

The market is one giant expected-value engine

I tend to look at the stock market as a big distributed system.

Every participant holds different information.

Someone's watching rates. Someone's watching earnings. Someone's watching the industry outlook, someone the technology, someone the policy environment, someone a competitor's next move.

Each of them decides and trades.

All of those judgments get compressed, continuously, into one number: the price.

That number isn't always right.

People are in the loop, so there's fear, there's greed, there's asymmetric information.

The market price is less a correct answer than a collective expected value of the future, produced by whoever happens to be participating at that moment.

The trouble starts when the cache is wrong

Anyone who's worked in IT knows both sides of caching.

When it's right, it's fast. When a stale value is sitting there, things get ugly.

Same in the market.

Say an expectation of major growth is cached into the price.

If the growth shows up, no problem. Sometimes more expectation piles on top.

Then some quarter, the growth rate comes in below forecast.

The market throws out the old expectation.

Cache invalidation begins.

The stock falls much faster than you'd think.

The company may not have broken. The number the market was using to compute the future just changed.

Which is why this keeps repeating.

The company is still good and the stock is cut in half. Or the company is still struggling and the stock has doubled.

Look only at today's numbers and there's no explanation. Look at how expectations shifted and you get at least part of one.

Investors watch the future and the price at the same time

I think that's where the difficulty in investing actually lives.

Finding a good company isn't the finish line.

How much will this company earn from here, how far has the market already priced that in, and is the current price expensive or cheap against that expectation — all three are running at once.

So stocks aren't a game of guessing the company right.

It's closer to a game of judging the distance between the company's future and the future everyone else is imagining.

I used to study only the company. Tear apart the financials, read the annual report, and I figured the answer would fall out. Looking back, I was seeing half of it. I never once counted what was already sitting in the price. My trade log from those years is still a little embarrassing to read.

Today's numbers matter. But they're a log that's already been written.

The price is out there on top of that log, continuously predicting the next state.

So the first question is probably this one:

Not "is this a good company?" but "what future is already priced in here?"

Next post goes a level deeper.

Why a good company and a good stock are different things. Why you can buy a stock with strong technology, growing revenue, and a bright future, and still lose money. The problem of price.