2. How a piece of news gets into the price

Level 8: What is priced in and what is not

If the price incorporates public information, it is worth looking at the concrete mechanism by which it does so. Because when you look closely a distinction appears that almost nobody makes and that decides where an edge can exist: finding out about a piece of news, processing it and interpreting it are three different things, and only one of the three is still available to someone who does not compete on speed.

Three things we all call information

• Raw public information. The published fact: the sales figure, the announcement, the employment number. It is identical for everyone and reaches everyone at almost the same time.
• Processed information. That fact already converted into its mechanical implication: how much earnings per share changes, which debt covenant is triggered, which companies in the sector get dragged along. It takes work, but it is work with a single answer.
• Interpretation. What the fact means for the value of the business three or five years from now. Here there is no single answer: there is judgment, and therefore legitimate disagreement among competent people.

The first layer is exhausted for the private investor within seconds. The second is exhausted within minutes or hours, because there are thousands of professionals trained for exactly that. The third can stay open for months, and it is where almost all the edge available to someone without infrastructure lives.

The chain: who acts and in what order

When a figure is published, the order of reaction is fairly stable:

First to move are the automated systems that read the announcement in structured form and execute simple rules: if the figure beats expectations, buy; if not, sell. They understand nothing about the business and do not need to.

Then the professional desks step in, comparing the figure with their own prior estimate and adjusting positions. There is judgment here, but the judgment is applied fast and to the obvious.

Then come the analysts revising their forecasts and publishing notes. This takes hours or days.

And finally the rest of the market appears, reading the headline in a general news outlet and deciding what to do. By then, the easy part of the news has been inside the price for a long time.

How long it really takes

The figures help to size up the competition. Studies of the reaction to earnings releases find that the bulk of the price adjustment happens in the first seconds and minutes, not in the hours that follow. In today's electronic markets, the time between a figure being published and the first wave of orders is measured in milliseconds, and on some routes in microseconds.

The race for those milliseconds has reached extremes that illustrate the point well: the microwave links built between the US financial centers cut the round-trip time compared with optical fiber, going from around fifteen milliseconds to under ten. Enormous sums were invested to gain a few thousandths of a second, and that investment only makes sense if those thousandths are worth money. They are worth money precisely because the first layer of information is exhausted there.

Why the easy part is already done

Out of this comes an uncomfortable but liberating rule. If your idea of an edge is reacting quickly to a piece of public news, you are competing against participants who have spent fortunes on not losing that race and who react before your screen has finished refreshing. It is not a difficult competition: it is a competition already lost before it starts.

It is also worth disabling a false intuition. The price does not react to the figure being good or bad, but to the difference between the figure and what was priced in. That is why a company can report excellent growth and fall sharply: the market expected more. Anyone trading on the headline without knowing what was expected is betting blind on the half of the equation that matters.

What is left: the slow part

And yet the adjustment is not perfect. The literature has described a persistent phenomenon for decades: after an earnings surprise, the price tends to keep moving in the same direction for weeks, in what is known as post-announcement drift. The mechanical part adjusts instantly; the part requiring expectations several quarters out to be revised adjusts slowly, because it requires different people to change their minds, and changing your mind takes time.

That is where the crack is. Not in arriving first, but in concluding something different from what the consensus concludes and being right. That does not depend on the speed of your connection: it depends on how much you understand about the business and how long you can bear being alone in your opinion.

It is worth sizing that crack properly before getting excited. Post-announcement drift is a modest effect, it has shrunk as more people have studied it and it is concentrated above all in small, thinly followed companies, which are also more expensive to trade. It is exactly the kind of margin the previous module described: it exists, it pays for the work of looking for it and not much more.

Real case: three minutes in April 2013

On April 23, 2013, the account of a large news agency was compromised and published a false message reporting explosions at the White House. Within seconds, the US market slumped: the benchmark index fell around 1% and some 136 billion dollars of value temporarily evaporated. The agency denied the message straight away and the market recovered practically everything it had lost in about three minutes.

That episode is the best available X-ray of the mechanism. The incorporation of the news was instantaneous and automatic, prior to any human verification; the correction, almost equally fast as soon as better information arrived. Nobody with a phone in their hand had time to take part in either phase. If your edge had to be somewhere, it was not in those three minutes.