PsyFi Search

Wednesday 26 May 2010

Risk, Stone Age Economics and the Affect Heuristic

Real Risk Management

Meet Grunt, a Stone Age economist. Grunt spends his time assessing risks. Usually these involve delicate decisions that may end up having quite important consequences – like whether he’s alive at the end of the day or not. Whether it’s attempting to impress a potential mate or ingratiating himself with the alpha males or simply choosing not to eat that odd smelling fruit he’s constantly having to weight risks and probabilities.

Grunt, however, isn’t blessed with modern economic theory and probability analysis is unknown to him. Although his grasp of number theory that doesn’t run much beyond one, two, err … a whole load he has a mechanism that’s been honed over hundreds of generations. Rather than rationality Grunt relies on the Affect Heuristic.

Risk Is Not Equal to Reward

It’s perfectly obvious to the average strategically shaved simian that humans manage individual risk pretty well on the whole. Despite the aberrations in the gene pool that regularly show up in the Darwin Awards we mostly avoid killing ourselves in pointless ways and are remarkably adept at navigating our way through the incredible complexity of modern life without getting run over along the way. No one, outside the confines of academic fantasyland, really believes we do this though rational analysis of risks.

This idea of “risk” is the bugbear at the root of most modern confusion in economics and investing. Risk to the majority of us is something of which we have an instinctive sense, an idea which is infused with an emotional charge that guides us through our lives, mostly with reasonable success. When market commentators gaily refer to “risk and reward” and tell us that we can only achieve abnormal returns by taking on excess risk we don’t think of rational calculations, but of danger of some kind. Usually we’re not quite certain about what sort of danger this is. Economists are certain, but are wrong.

Rationalism – the idea that people are effectively emotionless calculators of risk – has a very long history, back at least to Plato and the ancient Greek stoics and can be traced forward through the likes of Descartes and Kant right into twentieth century psychology and economics. Oddly, it seems that the proponents of the rationalist viewpoint were themselves driven by emotion – distaste at the idea that emotions could rule human behaviour. The whole rationalist agenda can be read as a long argument about what people should be, rather than what we are.

A Whisper of Emotion

People, meanwhile, have carried on not making careful and rational judgement of risk but using “a faint whisper of emotion” to guide them through their days. This feeling is known as affect and is the quality of “goodness” or “badness” that we attach to any situation we find ourselves in. An affective response is fast, based not on any analysis of the situation but triggered by making automatic judgements based on emotional biases. We search our memories for stimuli and experiences similar to our current situation and make a snap judgement based on the pleasant or unpleasant feelings that get activated.

This idea of risk as emotion rather than risk as analysis is the dominant mode of human risk analysis, colouring our every thought and impacting our every decision. In fact, as we’ve seen before, you can’t remove emotion from our decision making processes: emotion actually guides us in making our decisions, remove it and we’re largely left rudderless, unable to judge the danger or otherwise of any situation we find ourselves in.

Risky, Not Rewarding

However, as so often, our instinctive mechanisms often guide us poorly in the modern world when we have to judge risks outside of our own little bubble of existence. Affect works rather well in our individual, direct experiences of the world but when we cast our net wider some peculiar side-effects begin to emerge. When we’re faced with a novel situation our first reaction is based on our feelings, not a rational analysis. What this implies – and what the research has shown – is that we can change perceptions of risk or benefit by presenting things in a different order.

If we like something then we judge it as less risky, and if we don’t like it we judge it as more risky. And the like or dislike we feel is an emotional reaction, which may be only faintly related to the situation we find ourselves in. Take nuclear power, for instance. Those people who regard it as a threat to their existence will generally regard it as highly risky and therefore of low benefit. Those who see it as offering the benefits of cheap and plentiful power to humanity will generally feel that it’s a good thing and therefore judge that the risk is low. In neither case is there any great attempt to analyse the actual risks involved.

The Affect Heuristic

This inverse relationship between risk and benefit is the Affect Heuristic. It’s now been shown many times most interestingly, for us, in the world of finance.

A study by Ganzach (2001), Judging risk and return of financial assets, looked at how securities analysts judged the risk-return opportunity of specific stocks. Where the stocks were familiar to the analysts then risk and reward were correlated – presumably because their affective reaction to a known stock was positive. Yet when asked to perform the same judgement for unfamiliar stocks their judgement appeared to be based on global attitudes rather than their own. Stocks with a positive aura were judged to offer high return for low risk while those with a negative one were predicted to provide low return for high risk

This implies that sectors with a buzz about them get rated higher than those with bad vibes. Value investors take note: done properly excess returns can be achieved without real extra risk:
"All these parameters are correlates of low preference, which, according to our model, leads to higher perceived risk and lower perceived expected return. These perceptions result in an unwarranted depression in price. When this unwarranted depression disappears, excessive returns are realized."
It’s emotion, not rationality we need to bet against.

Wrong Globally, Right Locally

In fact the affective system of risk assessment is pretty much guaranteed to lead us wrong outside our local experiences. As Slovic et al put it:
“…the affective system seems designed to sensitize us to small changes in our environment (e.g., the difference between 0 and 1 deaths) at the cost of making us less able to appreciate and respond appropriately to larger changes further away from zero”
This affect, known as psychophysical numbing, essentially means we can’t appreciate the kinds of vast numbers that our modern world throws at us. We respond to anecdotes not data because we can judge the personal risks and empathise with the participants in the former. The trouble is that when the Affect Heuristic plays us wrong we end up taking risks that we don’t appreciate.

This isn’t simply a set of lessons for the individual investor. The way that humans judge risk and benefits has a wide sweep, because if economists have models that work on the assumption of individual rationality while individuals are working on the basis of affective responses then the mismatch will lead to significant errors. Rationality is not bounded, it’s simply not relevant.

Ask A Politician

Ironically the people who understand this best are politicians who know full well that their voters aren’t at all rational. We’ve seen public health scares in which billions are spent to save a handful of lives when the same money could have saved thousands of patients with much more mundane illnesses. One is salient and scary, the other hidden and normal and you can’t fight the former with statistics.

There’s irony for you. Our politicians are the real Grunts, today’s Stone Age economists, because when it comes to understanding real risks politicians are far better judges than economists. Central Bank independence anybody?

Related Articles: O Investor Why Art Thou Rational?, Stocks Aren't Snakes, Get an Emotional Margin of Safety


  1. Oddly, it seems that the proponents of the rationalist viewpoint were themselves driven by emotion – distaste at the idea that emotions could rule human behaviour.

    Man, what a great sentence!

    My take is that people need to spend some time thinking through the implications of that sentence, which in my experience lead in lots of fruitful directions.


  2. On the subject of growth stocks, I read some research a while back showing that investors did indeed correctly judged the companies that were going to go up more than the market (the 'buzz' in your post) but that the trouble was they then overpaid for them.

    In other words, even when we correctly judge the risks - the saber-toothed tiger has gone for the day - we may misjudge the rewards - "What, you call *that* an antelope? It's more like a rabbit". ;)