Why prices move

Price is set by orders, not by value

A company has some true worth: the cash it will produce for its owners over its whole future, which chapter 1 talked about. Nobody knows that number exactly. Everyone has an estimate, the estimates disagree, and they change as news arrives.

The price on your screen is not anybody's estimate. It is the last trade, which happened because one person was willing to sell at that price and another was willing to buy. If a lot of people decide at once that the company is worth more, they send buy orders, those orders eat through the asks in the book, and the price climbs until enough sellers appear. Nothing else moves a price. Not news on its own, not an analyst's opinion, not a company announcement: only orders, and the quotes market makers choose to post in response.

That gives a useful way to think about every price move. Ask: who just changed their mind, what did they learn, and how badly did they want to trade?

How information gets into a price

There are three main routes, and the simulation below has all of them.

Public news. An earnings report, a court ruling, an interest rate decision. Everyone sees it at roughly the same moment, and market makers move their quotes immediately, often before anyone has traded. For big public announcements in large US stocks, most of the price adjustment happens within seconds to minutes. Professional firms read machine-readable news feeds and economic releases with software precisely so they are not the slow quote that gets traded against.

Informed trading. Some information is not announced; it is worked out. A fund notices that a retailer's car parks are emptier, an analyst rebuilds a company's numbers and decides it is undervalued, a trader sees a related stock move. They buy or sell, and their orders push the price toward what they believe. They don't move it all the way at once, because buying aggressively costs money (more on that below), so their information leaks into the price over hours, days or weeks. Trading on material information that is not public and was obtained in breach of a duty, such as a company insider trading ahead of an announcement, is insider trading and illegal. Doing your own research is not.

Learning from the flow. Market makers can't see who is informed, but they can see the orders. If buyers keep lifting the ask, the maker infers that someone probably knows something and raises their quotes, even without knowing what the news is. This is the adverse selection from chapter 4 working in reverse: the same process that costs makers money is what moves the price toward the truth.

Things to try

• With news set to only informed traders, press Good news. The value jumps, the price doesn't. Then watch it climb as informed buyers take the asks and makers raise their quotes after being hit. Now set informed traders to 0% and do it again: the price barely moves, because nobody who knows anything is trading. Set it to 40% and the gap closes in a few seconds.
• Switch news to everyone and press Bad news. Makers drop their quotes straight away and the price gaps down, with few trades needed. That is what a public announcement looks like.
• Pull the market makers, then send a big seller. Watch the spread blow out and the price fall far below value as the seller works through a thin book. Then bring the makers back.

Efficient, mostly

The idea that prices quickly reflect available information is called the efficient market hypothesis, associated with the economist Eugene Fama. In its everyday form it says something modest and well supported: public information is absorbed so quickly by so many professionals that you can't expect to profit from reading the news, and stock prices are very hard to predict.

It doesn't say prices are always right. Researchers have documented patterns that persisted for years: prices tend to keep drifting in the direction of an earnings surprise for weeks after the announcement, and stocks that have risen over the past year have tended, on average, to keep rising for a while (the momentum effect). Some of these patterns shrink once they are published and traded on; others survive because they are risky or costly to exploit. The practical version: the market is not perfect, but it is very competitive, and the obvious mistakes are found and fixed by people with more data, faster computers and lower costs than you.

Liquidity and price impact

Even an order with no information in it moves the price, simply by using up the book. Buy 100 shares and you take the best ask. Buy 5,000 and you walk up several price levels, and the market makers who sold to you raise their quotes because they now hold less inventory and can't be sure you aren't informed. This is price impact.

Studies of large institutional orders have found a consistent shape: the impact grows with the size of the order, but less than proportionally, roughly with its square root. Doubling the size of a big order raises the cost per share by about 40%, not 100%. It also grows with the stock's volatility and shrinks with how much of it normally trades in a day. That is why large funds slice big orders into many small ones over hours or days, and why a stock that trades a few thousand shares a day is a very different thing to own from one that trades fifty million.

When liquidity disappears

Most liquidity comes from people who are free to leave. Electronic market makers keep quoting because the expected spread outweighs the risk; when the risk of being picked off rises faster than the spread, they widen out or step away. Other buyers hesitate too. A book that held thousands of shares within a few cents can, within minutes, hold almost nothing.

On 6 May 2010, US markets had what became known as the Flash Crash. On a day already nervous about the European debt crisis, a large fund began selling a very big position in S&P 500 futures through an algorithm that sold according to trading volume, without regard to price or time. High-frequency traders bought some and quickly sold to each other; liquidity in futures and then in stocks drained away. Major indexes fell several percent in minutes, with the Dow Jones Industrial Average down nearly 1,000 points at the low, and mostly recovered within about half an hour. Some individual stocks traded for a penny and others at $100,000, because the only orders left in their books were placeholder quotes set absurdly far from the price. Many of those trades were later cancelled.

The reforms that followed are still in place. Placeholder "stub" quotes were banned. Individual stocks now have limit up-limit down bands: if a stock tries to trade outside a band around its recent average price (5% for large stocks priced above $3 during most of the day; wider for smaller and cheaper stocks and near the open and close), trading in it pauses briefly. And there are market-wide circuit breakers: if the S&P 500 falls 7% or 13% from the previous close, all US stock trading halts for 15 minutes (unless it happens late in the day); at 20% it stops for the rest of the day. The market-wide breakers were triggered four times in March 2020, as the pandemic hit.

Bubbles and momentum

Prices can also move far from any sensible estimate of value for a long time, and here order flow explains a lot. When a stock rises, people who own it feel richer, people who don't feel left out, stories spread, and some traders buy simply because it has been going up. Their buying pushes it up further, which attracts more. The simulation has a small group of momentum traders who do exactly this.

Selling it short (betting on a fall, chapter 7) is the natural brake, but it is risky and costly: the price can keep rising long after it looks silly. So bubbles can inflate until buyers simply run out. The Nasdaq Composite, full of internet companies, roughly quintupled in the five years to March 2000 and then lost about three quarters of its value by late 2002. Many of the companies were worth nothing. Some, like Amazon, turned out to be worth far more than even the peak price, but their stock still fell more than 90% on the way down.

Next: what happens when people trade with borrowed money, bet on falls, or simply hold something volatile.