Here at the Faculty of Economics we're always looking at the latest cutting-edge economic research, particularly when it comes to the financial markets which may appear rational, but how about the traders in those markets - are they as Professor Bossaerts has argued, imperfectly rational?

As he glances up from his desk, it’s clear that Professor Bossaerts is immersed in just how, and why, we make complex decisions. Many of his published papers are in his office, along with evidence of his current research.

He has used brain scanners to examine what is going through a person’s mind when they are presented with complex decisions, and now he is a core member of the Faculty of Economics, he continues his pioneering work into controlled experimentation in the study of financial markets, and the use of decision and game theory in cognitive neuroscience.

To study that field, he has been looking at how the brain makes decisions, which has helped to establish the novel fields of neuroeconomics and decision neuroscience.

The Leverhulme Trust Board, established by philanthropist Lord Leverhulme in 1925, enabled his move from Australia to the University of Cambridge, where he investigates human and market behaviour in complex situations.

“I'm working on the intersection between economics, economic theory, decision theory, game theory. And on the other hand, cognitive neuroscience and decision neurobiology. And that's a mouthful,” he admits. “In effect it all boils down to combinatorial complexity.”

He says his work is very close to computer science - the theoretical computer science that was effectively started here at Cambridge by people like Alan Turing, as a relatively new part of mathematics. “What I do is study how people make complex decisions, and perhaps more interestingly, what mistakes they make. People are complex and as a result they rarely get complex decisions right first time. However, they know they're not right and they know there is uncertainty left.”

I ask if he can sum up neuroeconomics for me. Immediately it’s clear there are so many areas he could discuss, in depth. “Economists very often asked - what's the problem we're trying to resolve? And the neuroscientists themselves, just like the psychologist, in the case of complexity, they were basically thinking about the brain as a stimulus response mechanism. Just like a computer.”

One fashionable problem in behavioural economics is optimal inattention which addresses neuroeconomics uncertainty, which appears to be  a simple question – ‘What should I not pay attention to?’, in particular if a decision-maker has no control, even if the person making that choice believes otherwise. As he explains this is more common than many believe, and many people make the choice easier by deciding what not to pay attention to. “However, that is also a complex problem”, he says. “It turns out inattention is kind of a discrete thing you're doing, for example to pay attention to me.”

Instead, he gives me an example of work by David Marr at MIT. “Say you want to investigate how the visual cortex works, you first have to ask the question ‘What is vision for?’. We realised that vision is not meant to ‘see’ as such. Your eye is not interested in a photographic record of what's going around the world, but it is actually trying to do something else. It wants to survive, it wants to have rewards, and it must avoid losses. To make the task easier, anything that is irrelevant, you will not see.”

He says we often pay attention to maybe two or three things at the same time, but more than four or five is pretty much impossible. “You have to decide. Which is difficult in a modern world, where everyone wants to get attention.” Optimal inattention is what you decide not to pay attention to. “However, mathematically that is a very complicated problem, to decide what to ignore.”

He says this explains many cognitive biases understanding complexity. He extrapolates this argument to the world of complexity in economics. “It's one thing to say that all the world is complicated, but the question is how complicated is it? And can you do something about this?”

He explains that many of us can experience complexity, but just not know it, with everyday financial transaction. He cites the example of a person who goes into a supermarket to buy dinner. “It seems trivial, but it’s a complicated problem, not a split-second decision.”

A few people may have a precise meal in mind and buy whatever they want, but for many they balance cost versus reward, and have decisions swayed, for example if seafood is discounted. “You have to completely rearrange the rest of what you buy, and juggle in your head the cost of it, where the items are, and suppose there are other combined discounted too – that slews your decision even more. We may argue many people are not financially literate, but this is a complex problem people are skilled at; so complicated that most often you really won't find the optimal solution, but you will do a pretty good job at it.”

He explains he is working with Cambridge neuroscientist Professor Wolfram Schultz (Churchill College) to understand food choice. “You're indirectly driven by the energy in the food, and the nutrients you desire. Not all nutrients give you the same amount of energy, so you have to balance this issue and solve a very complex problem. Our hunch is that some of the eating problems some people suffer is because they are solving a very complex problem, and a few people get really daunted by that, can’t find a solution, and so they eat badly.”

They are planning experiments. “If you want to simplify the problem, people will never go for any foodstuff that does not provide energy, no matter how good the nutrient balance is. That is something that we observe on daily level. I think people tend to overdo the energy side, but we’ll see where the research leads.”

His research also looks at some of the fundamental models within economics, such as the theoretical model of a person’s behaviour which suggests a positive relationship between expenditure and income.

“The basic production and consumption model in economics is also a combinatorial hard problem – but you can’t just assume away the problems. You certainly can’t just assume that you can buy a half or quarter of a car, and then you go to a rationally performing market. Some models assume I need only half an apple.” He gesticulates with his hands, explaining that real world doesn’t work like that. “Once you have these this lumpiness, it becomes a very difficult a problem and you can't really approximate the solutions, so you have a continuum.”

This has a real-world impact. “Production problems are like that as well. You can use four or five materials, but not seven or eight materials to put together a battery. It’s inefficient, and a manufacturer just won’t do that.”

Even in finance investments, you have to focus on certain asset classes. Which one are you going to take? This turns out to be, as far as Peter’s concerned, the most difficult problem in finance.

He moves onto his pioneering work into controlled experimentation in the study of financial markets. For workers in London’s Square Mile, just down the train line from Cambridge, this is research that is directly relevant to them. He has caried out experiment on what he calls Perfectly Rational Markets, and come up with some theories that he calls ‘Imperfectly Rational Traders’.

“In the past, economists worked with models of finance that don’t work in hindsight. For example, claiming that people always choose optimally with continualattention. Sure, in computing, things either work or they don’t, but we’re dealing with imperfect humans. That’s why, in economics, the concept of rationality has been weakened as theory has developed.

He explains that markets seem to have a kind of intelligence. “When I'm talking about rational markets or perfect markets and imperfect people, I am talking about the fact that markets can generate prices to a particular high, that require knowledge that none of the individuals have. It is kind of comparable to what you see in biology with swarm intelligence. However, it's a lot more complicated because that requires dumb behaviour at the individual level relative to say a flock of birds. Whereas in markets you have very intelligent people that can see the entire flock, and you can see them model what's in the market, to take advantage. We can predict what's going to happen, but not perfectly.”