by John Coates
Were they lucky, or skilled? The question has more than an academic interest. Banks and hedge funds have to decide how to allocate capital, risk limits and bonuses among their traders, so it is crucial that they be able to distinguish luck from skill. During the credit crisis of 2007–09 bank managers found out, to their and everyone else’s dismay, that most of their star traders turned out to be like the crazy cab driver, and they lost more money in these two years than they had made in the previous five. Could the banks not have distinguished luck from skill?
Our data indicated that they could. We found that the experienced traders who consistently made money, even through the credit crisis, were ones whose Sharpe Ratios had risen over their careers. When we plotted their ratios against the number of years they had been in the business we found a nice upward slope, indicating that they had been learning to make more money with less risk. Shiller was right: training and effort do pay off in the markets. This finding led us to suggest that banks and hedge funds could determine which traders had developed a skill worth paying for merely by looking at the slope of their Sharpe Ratios plotted against the number of years they had traded. If the curve was upward-sloping, chances are this trader had developed an expertise in pattern recognition that was indeed worth paying for.
A couple of points follow from this discussion of intuition and hunches and our data on the Sharpe Ratio. First, there does appear to be such a thing as trading skill. The financial markets seem to meet Kahneman and Klein’s criteria for an environment in which intuitions can be trusted. It is not surprising therefore that discussions between traders frequently compare the current market to something they have traded through before – ‘This credit crisis feels just like the Russian default in ’98. I bet the yen rallies.’ Most of a trader’s dialogue, however, remains internal and pre-conscious, a good trader listening attentively to whispers from the past.
Second, the question that often crops up in discussions of intuition – which is more reliable, intuition or conscious rationality? – is something of a red herring. We inevitably use both. If we tried to use nothing but conscious rationality, tried to emulate Spock, we would find ourselves uncontrollably thrown back for most of our decisions on fast, online thinking. The question of whether the resulting decisions are trustworthy is not one that can be answered once and for all – ‘Yes, always trust your hunches,’ or ‘No, work out a decision tree.’ Its answer will depend on your training. We should not be asking if we should trust our intuitions; we should be asking how we can train ourselves to possess a skill that can be relied on.
HUNCHES AND GUT FEELINGS
But what, you may ask, do intuitions and pre-conscious processing have to do with gut feelings? Just because a mental process takes place behind conscious awareness does not mean it is drawing on signals from the body. Indeed, pattern recognition, even though silent, probably draws on higher brain regions, parts of the neo-cortex and the hippocampus, a brain region acting as the filing system for memories. What is the connection between these higher brain regions and the body? There is in fact a connection between pre-conscious decisions and the body, because it is gut feelings that allow us to rapidly assess whether a pattern and a considered choice will most likely lead to a pleasant or a nasty outcome, whether we like or dislike, welcome or fear it. Without such visceral colouring we would be lost in a sea of possibilities, unable to choose – a situation the cognitive psychologist Dylan Evans has called ‘ the Hamlet Problem’. We may be gifted with considerable rational powers, but to solve a problem with them we must first be able to narrow down the potentially limitless amount of information, options and consequences. We face a tricky problem of limiting our search, and to solve it we rely on emotions and gut feelings.
Such is the conclusion drawn by two neuroscientists, Antonio Damasio and Antoine Bechara. Working with patients who had damage to one particular part of the brain which integrates signals from the body, Damasio and Bechara found that these patients could have perfectly normal, even exceptional, cognitive abilities, yet make terrible decisions in their lives. Perhaps, Damasio and Bechara surmised, the patients’ IQs counted for little in making good decisions because they were deprived of help from their bodies, from homeostatic and emotional feedback. They concluded that ‘ feeling was an integral component of the machinery of reason’.
To account for the decision-making impairments of these and similar patients, Damasio and Bechara developed their ‘ Somatic Marker Hypothesis’. According to this hypothesis, each event we store in memory comes bookmarked with the bodily sensations – Damasio and Bechara call these ‘somatic markers’ – we felt at the time of living through it for the first time; and these help us decide what to do when we find ourselves in a similar situation. When we scroll through the options open to us, each may present itself with a subtle tensing of muscles, a quickening of breath, a slight shiver of dread, a brief moment of calm, a frisson of excitement – until one just feels right. These bookmarks are especially memorable, and most urgently needed, when we take a risk, for risk can hurt us, physically and financially. So it is little wonder that tales of gut feelings are so frequent and legendary in the financial markets.
Scientific research into gut feelings represents a new perspective on the brain and body. Scientists such as Damasio and Bechara argue that rationality by itself gains no purchase on the world, it merely spins its wheels, without the grit of somatic markers. They draw our attention to the physical aspect of thought, and raise the possibility that good judgement may require the ability to listen carefully to feedback from the body. Some people may be better at this than others, may have more efficient connective circuitry between body and brain, just as some people can run faster than others. On any Wall Street trading floor you will find high-IQ, Ivy League-educated stars who cannot make any money at all, for all their convincing analyses; while across the aisle sits a trader with an undistinguished degree from an unknown university, who cannot keep up with the latest analytics, but who consistently prints money, to the bafflement and irritation of his seemingly more gifted colleagues. It is possible, though odd to contemplate, that the better judgement of the money-making trader may owe something to his or her ability to produce bodily signals, and equally to listen to them. We tend to think – we want to think – that decisions are a matter of cognition, of mind alone, pure reason, a view Damasio calls ‘Descartes’ Error’. But good judgement may be a trait as physical as kicking a football.
An interesting possibility arises: could we tell whether one person has better gut feelings than another? Could we monitor feedback from their bodies? Gut feelings, like the oracle at Delphi, provide valuable insights, but are frustratingly hard to access and notoriously hard to interpret. This inaccessibility is due in part to their being processed by regions of the brain that are not fully open to conscious inspection. Could we access these signals in some way other than introspection? Could we some day hack into these communication lines between body and brain, and then use this information as a trading signal?
FEEDBACK
We all recognise that our thoughts affect our bodies. To take the most trivial example, it is your brain that tells your hand to reach for a glass of water sitting on the kitchen table. But so too does your body affect your thoughts, and here again everyday examples are easy to find. When you are hungry or thirsty, for instance, your thoughts change and you develop what is called a ‘selective attention’ to signs of food and water, and you stop paying attention to anything else, such as the book you are reading or the beauty of a sunset. Other examples of how the body affects the brain are less familiar, yet, if you stop to consider them, should be equally obvious. Like the fact that your brain, like a muscle, requires blood, glucose and oxygen to operate. In fact your brain, which constitutes only 2 per cent of your body mass, consumes some 20 per cent of your daily energy. You can verify the sobering fact that thinking is a physical process by monitoring the pulse of your carotid artery, which supplies blood t
o the brain. Engage in a taxing mental task, such as mental arithmetic, and as you do so press two fingers gently into your neck just below the corner of your jaw: you can actually feel your pulse speed up as the machine in your head draws more fuel.
In a more formal experiment, a group of radiologists in Miami measured glucose use by the brain during a verbal fluency task, which required participants to list as many words as possible beginning with a given letter in a short space of time. They found that people performing this simple task drew23 per cent more glucose into their brains than they did when at rest.
Disturbingly, a group of psychologists at Florida State, also looking at glucose levels in the brain, found that during taxing mental (as well as physical) activities our glucose reserves become depleted, and this reduces our capacity for self-control. They concluded that allocation of energetic resources during emergencies follows a ‘last in, first out’ rule, according to which mental abilities that developed last in our evolutionary history, like self-control, are the first to be rationed when fuel is low. Muscles, which draw a small amount of glucose when at rest, come to monopolise available resources during physical activity. The preferential treatment of muscles during a fight, or when playing a sport, for that matter, and the rationing of glucose to the brain regions responsible for self-control, might explain why fights so easily run out of control (ice hockey seems especially prone). Perhaps the same could be said of our self-control when we are working long hours at the office – we tend to snap more easily – or trying to stick to a diet, since the draining of glucose also drains us of resolve.
As I say, these simple facts about how the brain affects the body and the body the brain are ones we all recognise. But things get a bit stranger when we consider one of the most poorly understood and under-researched phenomena in all the neurosciences – feedback between body and brain. In the examples we have just looked at, the brain affects the body, or the other way round, but the flow of causation is one-way: brain affects body, for example, and that is the end of the story. But the situation is very different in cases of body–brain feedback. In feedback a thought affects the body, and the changes taking place in the body then feed back on the brain, changing the way it thinks.
Take a simple example of the process: when you feel depressed you may decide, in a moment of self-assertion, to pick yourself up and battle on, forcing a smile, straightening your posture, walking more briskly; and in time these changes may actually work, you may end up feeling happier. Here changes in your body – its posture, facial expression, etc. – have fed back on your brain and changed the thoughts you think. Body–brain feedback of this sort can even, under some circumstances, turn into runaway reactions which end in extreme behaviour. When you are scared, for example, your heart beats faster, you sweat, hyperventilate, and run away from a supposed threat. As you become aware of these symptoms you can start to worry even more, and the stage is set for a fully-fledged panic attack. Or an argument, to take another example, may escalate into pushing, and in the extreme to an exchange of blows. At each stage blood pressure increases, breathing accelerates and, importantly, brains lose their cool. Even a gentle person, as the physical exchange progresses, can begin to think violent thoughts. This spiralling into out-of-control thoughts and behaviour can also work in a more welcome direction, as it does for example during sex. In all these examples of feedback, the brain does not look on as a disinterested observer of a body in turmoil, it is intimately caught up in the process. It is participant, not spectator.
Nowhere has this feedback been more accurately and insightfully described than in the writings of William James, the great nineteenth-century philosopher and psychologist (and brother of the novelist Henry James). ‘Everyone knows how panic is increased by flight,’ he wrote, ‘and how the giving way to the symptoms of grief or anger increases those passions themselves. Each fit of sobbing makes the sorrow more acute, and calls forth another fit stronger still, until at last repose only ensues with lassitude and with the apparent exhaustion of the machinery. In rage, it is notorious how we “work ourselves up” to a climax by repeated outbreaks of expression.’
Words we all recognise as true; yet the familiarity of the experience James described hides a mystery: why are we built this way? If your brain wants to cheer itself up, or worry itself into a panic, why bother sending its message through the body? Why not send a signal directly from one brain region to another? These questions, in my view, take us right to the heart of the mind–body problem. For the body does indeed influence the brain, transforms its very thoughts and feelings. But again, why?
If the brain kept its signalling within the brain, surely it would save a great deal of time. Maybe. But probably not. Not if the point of the thought is to produce movement, for then the extra processing time in the brain might actually slow down our eventual action.
This was largely the conclusion drawn by James. He came to it when puzzling over a very special and powerful form of gut feeling – our emotions. He had begun to suspect that we commonly misunderstand the nature of our emotions: we tend to think that our emotional feelings come first, and then cause our emotional behaviour. But according to James, the feeling of an emotion is in some ways the least important part of the experience. In this our commonsense understanding of emotion has it all wrong. We tend to think that we cry because we are sad, run from a bear because we are scared, but James argued that the true course of events runs the other way round. We are sad because we are crying, scared because we are running away. More precisely, the course of events unfolds as follows: we perceive a bear; our brain triggers automatic escape behaviour, such as running; and this physiological change is then reported back to our brain and shows up in our consciousness as a feeling of fear. Some scientists have even argued that the feeling itself plays a very small role in an emotional event, tagging along as mere observer of action already taken. It is, as the neuroscientist Joe LeDoux suggests, no more than ‘ icing on the cake’.
James was trying to correct the belief prevailing at the time that an emotion is largely a mental event, like a thought, only with strong feelings attached. Arguing against this view, he pointed out that it missed what was most important about emotions: that they are first and foremost reflexes designed to help us behave and move, sometimes quickly, at crucial moments of our lives. As Charles Sherrington, the Nobel laureate in physiology for 1932, was later to say, ‘emotions move us’, and he meant literally. If the functional role of emotion is to promote fast behavioural reactions, then what are the emotional feelings? James thought that they emerge when we perceive the changes our body goes through during an emotional episode, changes such as a tense stomach, sweating, a pounding heart, increased blood pressure, raised body temperature and so on. In the absence of these physical sensations, emotions would be drained of feeling. ‘Without the bodily states following on the perception,’ James wrote, ‘the latter would be purely cognitive in form, pale, colourless, destitute of emotional warmth. We might then see the bear, and judge it best to run, receive the insult and deem it right to strike, but we could not actually feel afraid or angry.’
Casual support for James’s view can be found in our everyday use of emotional language, for we draw heavily on bodily metaphors: we speak, for example, of receiving bad news with a sinking stomach, a broken heart, or blood-boiling anger; we tell of a chilling moment, a tense encounter, a heart-pounding experience, of being hot under the collar or flushed with excitement.
James’s account of the stages of emotional feedback – physical reaction first, conscious feeling second – may seem counter-intuitive. But it makes perfect sense if what we need in an emotional crisis are fast reactions: we act first, feel later. Still, James received a lot of criticism for his theory, most notably during the 1920s from the great Harvard physiologist Walter Cannon. James versus Cannon was a clash of titans. Cannon argued that feedback from the body travelled far too slowly to keep up with the rapidly changing feelings you can have
in the course of an emotional encounter, one that may see you pass from anger to fear to relief to happiness, all within the space of a few seconds. For the suite of bodily changes, including fluctuations in breathing, body temperature and adrenalin levels, to keep pace with each fleeting nuance of feelings, they would have to increase and decrease with split-second timing. But they do not. Some of these physiological changes, such as adrenalin release, can take one or two seconds before they are felt, so our visceral organs would be left behind by high-speed emotional events.
There was another problem with James’s theory, or so Cannon argued. Cannon thought that feedback from the body was not unique enough to provide an individual physiological signature for each emotion. Your heart pounds and your breathing accelerates when you are scared, or angry, or joyous, or when you fall in love. In fact, whenever you experience a strong emotion, argued Cannon, you produce much the same suite of physical reactions. He reported, for instance, ‘the case of a young man who on hearing that a fortune had just been left him, became pale, then exhilarated, and after various expressions of joyous feeling vomited the half-digested contents of his stomach’. He tells of similar symptoms of nervous agitation displayed by people suffering deep sorrow or great disgust. Bodily feedback may increase your arousal during an emotional encounter, but it cannot tell you which emotion you are feeling. Physical arousal, Cannon concluded, is too ham-fisted to paint the often gentle hues colouring our emotional life.