Attention and what gets bought

A chapter about a constraint rather than a bias. Attention is finite, and the way it is finite has a specific structural consequence that most people never notice.

The asymmetry

To buy, you must first find. There are thousands of listed companies. You cannot evaluate them all, so your purchase comes from the small subset that reached your notice — through news, a price move, a conversation, a notification.

To sell, you need not find anything. You sell from what you already own, typically a handful of positions, all known to you. There is no search problem.

So attention acts as a filter on the buy side and barely at all on the sell side. Barber and Odean put it precisely: selling "poses less of a search problem and is less sensitive to attention effects", which leads individual investors to buy attention-grabbing stocks heavily. Using several measures of attention, they find individual investors execute proportionately more buy orders for more attention-grabbing stocks.

This is a structural claim, not a claim about weakness. A perfectly disciplined investor with limited time faces the same filter.

What grabs attention

The measures used in this literature are the obvious ones, and that is the point — they are observable in advance.

• Being in the news, whatever the content
• A large price move, in either direction
• Unusually high trading volume
• Hitting a price limit or another salient threshold

Note what is missing: anything about valuation, business quality, or whether the company suits your purposes. The filter that selects your candidates is uncorrelated with the things that make an investment good.

Why that is worse than it first sounds

The usual worry about following the news is that you will buy bad companies. The sharper worry is different.

Your buy list is assembled by other people's interest. Not by your analysis — by whatever the media and the market found worth remarking on that week. You then apply your judgement, perhaps well, to a set you did not choose.

The set is systematically the wrong one. Attention-grabbing stocks have just moved, just been discussed, just had volume. That is close to a definition of a stock whose news is already in the price.

Both directions of news attract buying. The survey notes individual investors in one dataset tended to be net buyers subsequent to both positive and negative extreme earnings surprises. If bad news and good news both bring retail buying, the content is doing less work than the salience.

Attention also causes neglect. Research finds the market reacts less to earnings announced on crowded days, and to Friday announcements, with more of the reaction arriving later — attention being diverted rather than absent. Information does not fail to exist because nobody looked at it; it just gets priced late.

Working the problem

Why buying is more attention-sensitive. Count the alternatives. On the buy side, thousands of candidates and no possibility of considering them all, so something must reduce the set, and attention does it. On the sell side, your own holdings — a list short enough to review in full. A constraint that bites on one side and not the other will produce asymmetric behaviour even in someone with no bias at all.

The prediction for a heavy-news day: retail buying in that stock should rise disproportionately, more than retail selling does, and roughly irrespective of whether the news was good or bad. That is what the research finds, and it is why "I only buy companies I've been reading about" is a description of a filter, not of a process.

One honest complication. The same body of work finds that stocks individuals buy tend to underperform the ones they sell over subsequent months, which makes attention-driven buying look costly. But it also finds, from a different dataset, that investors with concentrated portfolios of one or two stocks outperformed those holding three or more by about 16 basis points per month, with the effect stronger in local and non-index stocks — read by its authors as informed investors concentrating where they have an advantage. So "paying close attention to a few things" and "buying whatever is in the news" are not the same behaviour, and the evidence treats them differently. The distinction worth keeping is whether your attention was directed by you or delivered to you.

What to do about it

Choose your filter deliberately. A screen you wrote, a watchlist you built when calm, a universe defined by rules — any of these replaces the news cycle as the thing that selects candidates.

Separate finding from deciding. It is fine to notice a company because of a headline. It is not fine for the headline to be part of the case.

Distrust urgency. Attention is perishable, which is why selling pressure is applied with deadlines. A decision that cannot wait a week is usually one being made for you.

The point

Buying requires searching thousands of candidates while selling requires only reviewing what you own, so attention filters purchases far more than sales — and the things that grab attention, such as news coverage, large price moves and unusual volume, have no relationship to whether an investment is sound. Individual investors accordingly place proportionately more buy orders in attention-grabbing stocks, and in one dataset bought after both good and bad earnings surprises, which suggests salience rather than content is driving it. The remedy is to choose the filter that assembles your candidate list instead of letting the news cycle choose it.