Basil Halperin

Essays — Basil Halperin


Behavioral biases don't affect stock prices
Aug 15, 2017

Confidence level (?): Low

Most people are probably somewhat overconfident. Most people – myself surely included – probably typically overestimate their own talents, and they (we) are overly confident in the precision of their estimates, underestimating uncertainty.

This bias has viscerally real, important consequences. Governments are overconfident that they can win wars quickly and easily; overconfident CEOs have a higher tendency to undertake mergers and issue more debt than their peers.

I claim, however, that this bias does not matter for asset pricing in particular. That is, stock prices (and other asset prices) are not affected by overconfident investors.

In fact, I claim that any kind of behavioral bias cannot in and of itself affect stock prices.

The idea that behavioral biases, on their own, can affect asset prices is one of if not the most widely held misconceptions about financial markets. Just because most people (myself included!) are blinded by cognitive biases – overconfidence, status quo bias, confirmation bias, etc. – does not mean that stock prices are at all affected or distorted.

If this seems crazy, let me try putting it another way: just because behavioral biases exist does not mean that you can get rich by playing the stock market and exploiting the existence of these biases.

The trick is that it only takes the existence of one rational unconstrained arbitrageur to keep prices from deviating away from their rational level.

To see this, consider two extremes.

All it takes is one
First, suppose everyone in the world is perfectly rational and unbiased, except for one poor fellow, Joe Smith. Joe is horribly overconfident, and thinks he's smarter than everyone else. He invests all of his money in Apple stock, insisting that everyone else is undervaluing the company, and pushing the Apple share price up.

Of course, since every other investor is perfectly rational and informed, they will notice this and immediately race to go short Apple, betting against it until the price of the Apple stock is pushed back to the rational level.

Now, consider the inverse situation. Everyone in the world is systematically biased and cognitively limited, except for one rational informed Jane Smith. Perhaps more realistically, instead of Jane Smith, the one rational agent is some secretive hedge fund.

Now, billions of irrational investors are pushing prices away from their rational value. However, as long as Rational Hedge Fund LLC has access to enough capital, this one rational agent can always buy an undervalued stock until the price gets pushed up to its rational level, or short an overvalued stock until the price gets pushed down to the rational level. Rational Hedge Fund LLC profits, and prices are kept at their rational levels.

Even more realistically, instead of a single hypervigilant rational hedge fund keeping all stocks at their respective rational levels, there could be many widely dispersed investors each with specialized knowledge in one stock or one industry, collectively working to keep prices in line.

The marginal investor
The real world, of course, is somewhere between these two extremes. Most people have a host of cognitive biases, which leads to "noise traders" randomly buying and selling stocks. However, there is also a small universe of highly active, often lightning fast rational investors who quickly arbitrage away any price distortions for profit.

It is these marginal investors who determine the price of stocks, not the biased investors. This is why I say that "cognitive biases don't matter for stock prices" – the existence of any unconstrained rational investors ensures that biases will not flow through to asset pricing.

The important caveat: the "limits to arbitrage"
There is an extremely important caveat to this story.

Note that I quietly slipped in the requirement that Rational Hedge Fund LLC must have "access to enough capital." If the rational investors cannot raise enough money to bet against the noisy irrational traders, then prices cannot be pushed to their rational equilibrium level.

(The importance of access to capital is more than just the ability to apply price pressure. It's also important for the marginal investor to be able to withstand the riskiness of arbitrage.)

This assumption of frictionless access to leverage clearly does not hold perfectly in the real world: lending markets are troubled by principal-agent problems, moral hazard, and other imperfections.

This (very important) friction is known as the "limits to arbitrage."

Summing up
It is irrationality in conjunction with limits to arbitrage which allow for market prices to diverge from their rational levels. It is important to acknowledge that cognitive biases alone are not a sufficient condition for market inefficiency. Irrationality and limits to arbitrage are both necessary.

More pithily: Peanut butter alone is not enough to make a PB&J sandwich, and behavioral biases alone are not enough to make the stock market inefficient.

⇒ Back to essays ⇐