Behavioural finance has a semantics problem

Clients are rightly sceptical when we talk about applying behavioural finance to analysing market behaviour. Behavioural finance seems very “woolly” at the best of times and most of the well-known literature focuses on the decisions of individuals, not how this applies to the markets as a whole.

At the same time, terms like ‘irrational behaviour,’ and ‘risk appetite’ tend to be mis-used in a way that often leaves them feeling counterintuitive and flawed. This is more important than it may seem: a failure to understand what is driving market moves makes it harder to assess whether or not they represent investment opportunities.

This week, a piece written by Nassim Nicholas Taleb, as well as market commentary around some positive performance from global equity markets illustrated these points well.

Loss aversion and irrationality

Last December I wrote a piece (“Are today’s bond investors irrational?”) which highlighted the work of Nassim Nicholas Taleb and others on the idea of ‘loss aversion.’

Taleb argues that it is not necessarily irrational to be ‘loss averse’ i.e. to refuse to take a bet even though the expected return is positive. In a recent piece on the nature of risk taking Taleb makes the point clearly: when the order of outcomes matters (think Russian roulette), using the average outcome over repeated bets is not a good guide to whether a decision is rational or not.

This may seem like a direct contradiction of behavioural finance, which is often held up as identifying ‘irrational’ behaviour among investors. However, when discussing loss aversion in his book: “Thinking, Fast and Slow” Daniel Kahneman makes a similar point:

In fact another line from that book comes close to some of Taleb’s recent arguments about the importance of survival as a criterion for judging what is and is not rational:

The reality is that, as is often the case in academia, the behavioural finance literature is using a different definition of ‘rational’ to the rest of us. Kahnemann’s work simply showed that the economists’ previous model of rationality (‘expected utility theory’) did not reflect how people actually behaved in the real world.

However, behavioural finance never claimed that economists’ definition of ‘rational’ represented the way people should behave in the first place.  One last quote from ‘Thinking, Fast and Slow’ makes this point explicitly:

This may seem unimportant, but such confusion over definitions of ‘rational versus irrational’ can also crop up in descriptions of price moves at an aggregate market level. When such confusion is dominant, it is easy to distort what is really driving prices.

This is important, because Investors who want to try to ‘beat the market’ need to have a clear sense of how they are going to do it. Flawed models of why prices are moving, especially when they imply that aggregate market behaviour is somehow ‘foolish’ or ‘neurotic’, can lead to equally flawed decision making. Descriptions of this week’s price action illustrate this.

“Risk appetite” doesn’t explain aggregate market behaviour

The term ‘risk appetite’ can often be found in the media to provide a description of market behaviour. Here is a selection of online headlines as equity markets generally rose on Monday.

The term ‘risk appetite’ is associated with behavioural finance because it provides explanations of market moves which are routed in human behaviour. It also resembles the concept of ‘risk aversion’ which seeks to assess the extent to which individuals may prefer a ‘sure thing’ to an uncertain gain.

However, does it properly explain what happened on Monday? Is it the case that investors were suddenly happier to take a greater risk for the same gains than they were the day before?

This is not what behavioural finance would suggest. The evidence is that investors’ ‘appetite for risk’ (insofar as anyone thinks in those terms) is very unlikely to have changed overnight. Instead price moves will reflect a belief that either risk has fallen, or the potential gains have got bigger (or, when moves are as small as Monday’s, largely random forces which tell you very little at all). The idea of changing risk appetites, particularly if they are seen as irrational, is not a helpful explanation of what is going on in markets.

So what can we say about market moves, and how does this relate to behavioural finance?

Rational opportunities – explaining aggregate market behaviour

Like Kahneman’s observations on individual decisions, Robert Shiller’s behavioural finance work suggested that markets in aggregate were ‘irrational’ in that they did not conform to the definition of rationality put forward by believers in efficient markets (prices are more volatile than rational expectations theories say they should be).

This does not mean that markets are necessarily ‘unreasonable’ or stupid. It could be very rational (in the common sense of the word) to chase a bubble (or employ a stop loss) if you are a short term investor or a professional manager benchmarked against peers. For these shorter term investors, going against the crowd is a bigger threat to your survival (in terms of career risk) than going with it. Taleb’s observations that many of us tend to overlook non-linearity and ignore unexpected events when assessing probability may have similar origins.

As I discussed in May, the evolutionary elements of Andrew Lo’s efficient markets hypothesis provide a possible explanation for why the time horizon of the average market participant could also change over time, while an excellent piece by Morgan Housel in June clearly sets out a model of bubbles which does not need ‘irrationality’ but is simply due to ‘people with different objectives thinking they’re playing the same game.’

These are just possible reasons for why different investors can take the same information and come up with different forecasts about future prices without any of them being ‘irrational’ (this is the basis of Modecai Kurz’s theory of ‘rational beliefs’).

Importantly, though the origins of price moves may be rational, this doesn’t mean that they aren’t opportunities for other investors, or that emotion has no role to play. Our view is that the biases identified by behavioural finance and psychology (anchoring to past experience, base rate neglect, extrapolation) often have an emotional origin, and that these in turn can influence how investors come up with their forecasts. Moreover, these emotions will not always mean poorer decision making.

The role played by these various forces are necessarily difficult (if not impossible) to assess quantitatively in complex systems and so there will be little comfort here for those who view behavioural finance as lacking rigour. However, the examples above show that behavioural finance, when properly understood, still has some important insights when considering how prices are determined (particularly relative to efficient markets theories). Understanding these is essential to trying to formulate any kind of investment process.