and Business Resources
Emotion, neuroscience and investing: Investors as dopamine addicts
By James Montier
Global Equity Strategy: January 20, 2005
What goes on inside our heads when we make decisions? Understanding how
our brains work is vital to understanding the decisions we take. Neuroeconomics
is a very new field that combines psychology, economics and neuroscience.
That may sound like the unholy trinity as far as many readers are concerned,
but the insights that this field is generating are powerful indeed.
Before I head off into the realms of neuroscience I should recap some
themes we have explored before, but that provide the backdrop for much
of the discussion that follows. One of the most exciting developments
in cognitive psychology over recent years has been the development of
dual process theories of thought. Alright, stay with me now, I know that
sounds dreadful, but it isn't. It is really a way of saying that we tend
to have two different ways of thinking embedded in our minds.
Spock or McCoy?
For the Trekkies out there, these two systems can, perhaps, be characterised
as Dr. McCoy and Mr. Spock. McCoy was irrepressibly human, forever allowing
his emotions to rule the day. In contrast, Spock (half human, half Vulcan)
was determined to suppress his emotions, letting logic drive his decisions.
McCoy's approach would seem to be founded in system X. System X is essentially
the emotional part of the brain. It is automatic and effortless in the
way that it processes information. That is to say, the X-system pre-screens
information before we are consciously aware that it even made an impact
on our minds. Hence, X-system is effectively the default option. X-system
deals with information in an associative way. Its judgements tend to be
based on similarity (of appearance) and closeness in time. Because of
the way X-system deals with information it can handle vast amounts of
data simultaneously. To computer nerds it is a rapid parallel processing
unit. In order for the X-system to believe something is valid it may simply
need to wish that it were so.
System C is the "Vulcan" part of the brain. To use it requires deliberate
effort. It is logical and deductive in the way in which it handles information.
Because it is logical, it can only follow one step at a time, and hence
in computing terms it is a slow serial processing unit. In order to convince
the C-system that something is true, logical argument and empirical evidence
will be required. The table below provides a summary of the main differences
between the two systems.
This dual system approach to the way the mind works has received support
from very recent studies by neuroscientists. They have begun to attach
certain parts of the brain to certain functions. In order to do this neuroscientists
ask experiment participants to perform tasks whilst their brains are being
monitored via elector-encephalograms (EEG), positron emission topography
(PET) or most often of late functional magnetic resonance imaging (fMRI).
The outcomes are then compared to base cases and the differences between
the scans highlights the areas of the brain that are being utilised.
The table below lays out some of the major neural correlates for the two
systems of thinking that were outlined above. There is one very important
thing to note about these groupings - the X-system components are much
older in terms of human development. They evolved a long time before the
C system correlates.
The primacy of emotion
This evolutionary age edge helps to explain why the X-system is the default
option for information processing. We needed emotions far before we needed
logic. This is perhaps best explained by an example using fear. Fear is
one of the better understood emotions1.
Fear seems to be served by two neural pathways. One fast and dirty (LeDoux's
low road), the other more reflective and logical (the high road). The
links to the two systems of thinking outlined above are hopefully obvious.
Imagine standing in front of a glass container with a snake inside. The
snake rears up, the danger is perceived, and the sensory Thalamus processes
the information. From here two signals emerge. On the low road the signal
is sent to the amygdala, part of the X-system2,
and the brain's center for fear and risk. The amygdala reacts fast, and
forces you to jump back.
However, the second signal (taking the high road) sends the information
to the sensory cortex, which in a more conscious fashion assesses the
possible threat. This is the system that points out that there is a layer
of glass between you and the snake. However, from a survival viewpoint
a false positive is a far better response than a false negative!
Emotions: body or brain?
Most people tend to think that emotions are the conscious response to
events or actions. That is, something happens and your brain works out
the emotional response - be it sadness, anger, happiness etc. Then your
brain tells your body how to react - tear up, pump blood, increase the
breathing rate etc.
William James, the grandfather of modern psychology, was amongst the first
to posit that actually true causality may well flow from the body to the
brain. In James' view of the world, the brain assesses the situation so
quickly, there simply isn't time for us to become consciously aware of
how we should feel. Instead the brain surveys the body, takes the results
(i.e. skin sweating, increased heart beat etc) then infers the emotion
that matches physical signals that the body has generated.
If you want to try this yourself, try pulling the face that matches the
emotion you wish to experience. For instance, try smiling (see we aren't
always miserable and bearish despite our reputations). If you sit with
a smile on your face, concentrating on that smile, soon enough you are
likely to start to feel the positive emotions that one associates with
An entertaining example of the body's impact upon decisions is provided
by Epley and Gilovich4
(2001). They asked people to evaluate headphones. Whilst conducting the
evaluation, participants were asked to either nod or shake their heads.
Those who were asked to nod their heads during the evaluation gave much
more favourable ratings than those asked to shake their heads.
In the words of Gilbert and Gill5,
we are momentary realists. That is to say, we have a tendency to trust
our initial emotional reaction and correct that initial view "only subsequently,
occasionally and effortfully." For instance, when we stub a toe on a rock
or bang our head on a beam (an easy thing to do in my house), we curse
the inanimate object despite the fact it could not possibly have done
anything to avoid our own mistake.
Emotion: Good, bad or both?
However, emotion may be needed in order to allow us to actually make decisions.
There are a group of people who, through tragic accidents or radical surgery,
have had the emotional areas of their minds damaged. These individuals
did not become the walking optimisers known as homo economicus. Rather,
in many cases, these individuals are now actually incapable of making
decisions. They make endless plans but never get round to implementing
any of them6.
Bechara et al7
devised an experiment to show how the lack of emotion in such individuals
can lead them to make sub-optimal decisions. They played a gambling game
with both controls (players without damage to the emotional centres of
the brain) and patients (those with damage to the emotional parts of the
brain). Each player was sat in front of four packs of cards (A, B, C and
D). Players were given a loan of $2000 and told the object of the games
was to avoid losing the loan, whilst trying to make as much extra money
as possible. They were also told that turning cards from each of the packs
would generate gains and occasional losses. The players were told the
impact of each card after each turn, but no running score was given.
Turning cards from packs A and B paid $100, whilst C and D paid only $50.
Unpredictably, the turning of some cards carried a penalty. Consistently
playing packs A and B led to an overall loss. Playing from C and D led
to an overall gain.
Performance was assessed at various stages of the game. Four different
periods were identified. The first involved no loss in either pack (pre-punishment);
the second phase was when players reported they had no idea about the
game, and no feeling about the packs. The third was found only in the
controls, they started to say they had a hunch about packs A and B being
riskier, and finally, the last phase when (conceptual) players could articulate
that A and B were riskier.
The table below shows the average number of rounds in each phase, and
the percentage of players making it through each phase of the game. The
patients were unable to form hunches, and far fewer survived the game.
Now cast your eye over the two charts below. The first shows the number
of cards drawn from packs A and B (Bad) and C and D (good) in each phase
by the controls. In the pre-hunch phase they are already favouring the
good packs marginally. In the hunch phase, controls are clearly favouring
the good packs.
Now look at the performance of the patients. In the pre-hunch phase they
kept choosing the bad packs. As noted above there was no hunch phase.
And perhaps most bizarrely of all, even when they had articulated that
packs A and B were a bad idea, they still picked more cards from those
decks than from C and D! So despite "knowing" the correct conceptual answer,
the lack of ability to feel emotion severely hampered the performance
of these individuals.
However, similar games can be used to show that emotions can also handicap
us. Bechara et al8
play an investment game. Each player was given $20. They had to make a
decision each round of the game: invest $1 or not invest. If the decision
was not to invest, the task advanced to the next round. If the decision
was to invest, players would hand over one dollar to the experimenter.
The experimenter would then toss a coin in view of the player. If the
outcome was heads, the player lost the dollar, if the coin landed tails
up then $2.50 was added to the player's account. The task would then move
to the next round. Overall 20 rounds were played.
Bechara et al played this game with three different groups: 'normals',
a group of players with damage to the neural circuitry associated with
(target patients who can no longer feel fear), and a group of players
with other lesions to the brain unassociated with the fear neural circuitry
The experimenters uncovered that the players with damage to the fear circuitry
invested in 83.7% of rounds, the 'normals' invested in 62.7% of rounds,
and the patient controls 60.7% of rounds. Was this result attributable
to the brain's handling of loss and fear? The chart below shows the results
broken down based on the result in the previous round. It shows the proportions
of groups that invested. It clearly demonstrates that 'normals' and patient
controls were more likely to shrink away from risk-taking, both when they
had lost in the previous round and when they won!
Players with damaged fear circuitry invested in 85.2% of rounds following
losses on previous rounds, whilst normal players invested in only 46.9%
of rounds following such losses.
Bechara et al also found evidence of just how difficult learning actually
is. Instead of becoming more optimal as time moves on, normal players
actually become less optimal! (See chart below) For the record, a rational
player would, of course, play in all rounds.
So emotion can both help and hinder us. Without emotion we are unable
to sense risk, with emotion we can't control the fear that risk generates!
Welcome to the human condition!
Camerer et al10
argue that the influence of emotions depends upon the intensity of the experience.
At low level of intensity, affect appears
to play a largely "advisory" role. A number of theories posit that emotions
carry information that people use as an input into the decisions they
.... At intermediate level of intensity, people begin to become conscious
of conflicts between cognitive and affective inputs. It is at such intermediate
levels of intensity that one observes ...efforts at selfcontrol...
...Finally, at even greater levels of intensity, affect can be so powerful
as to virtually preclude decision-making. No one "decides" to fall asleep
at the wheel, but many people do. Under the influence of intense affective
motivation, people often report themselves as being "out of control"...
As Rita Carter writes in Mapping the Mind "where thought conflicts with
emotion, the latter is designed by neural circuitry in our brains to win".
Camerer et al (2004)
It is also worth noting that we are very bad at projecting how we will
feel under the influence of emotion - a characteristic psychologists call
hot-cold empathy gaps. That is to say, when we are relaxed and emotion
free, we underestimate how we would act under the influence of emotion.
For instance, Loewenstein et al11
asked a group of male students to say how likely they were to act in a
sexually aggressive manner in both a hot and cold environment. The scenario
they were given concerned coming home with a girl they had picked up at
a bar, having been told by friends that she had a reputation for being
"easy". The story went on that the participants and the girl were beginning
to get into physical genital contact on the sofa. The participants were
then told they had started to try and remove the girl's clothes, and she
says she wasn't interested in having sex.
Participants were then asked to assign probabilities to whether they would
(1) coax the girl to remove her clothes (2) have sex with her even after
her protests. The chart below shows the self reported probability of sexual
aggressiveness (defined as the sum of the probabilities of 1+2). Under
the no arousal condition there was an average 56% probability of sexual
aggression. After having been shown sexually arousing photos, the average
probability of aggression rose to nearly 80%!
Self-control is like a muscle
Unfortunately a vast array of psychological research12
suggests that our ability to use self-control to force our cognitive process
to override our emotional reaction is limited. Each effort at self-control
reduces the amount available for subsequent self-control efforts.
A classic example of Baumeister's work concerns the following experiment.
Participants are asked to avoid eating food for three hours before the
experiment began (timed so they were forced to skip lunch). When they
arrived they are put into one of three groups.
The first group were taken into a room which cookies had recently been
baked, so the aroma of freshly made chocolate chip delights wafted around.
This room also contained a tray laid out with the freshly baked cookies
and other chocolate delights, and a tray full of radishes. This group
were told they should eat as many radishes as they could in the next five
minutes, but they were also told they weren't allowed to touch the cookies.
A second group was taken to a similar room with the same two trays, but
told they could eat the cookies. The third group was taken to an empty
All the food was then removed and the individuals were given problems
to solve. These problems took the form of tracing geometric shapes without
re-tracing lines or lifting the pen from the paper. The problems were,
sadly, unsolvable. However, the amount of time before participants gave
up and the number of attempts made before they gave up were both recorded.
The results were dramatic. Those who had eaten the radishes (and had therefore
expended large amounts of self control in resisting the cookies) gave
up in less than half the time that those who had eaten chocolate or eaten
nothing had done. They also had far less attempts at solving the problems
before giving up.
concludes his survey by highlighting the key findings his research has found:
- Under emotional distress, people shift toward
favoring high-risk, high payoff options, even if these are objectively
poor choices. This appears based on a failure to think things through,
caused by emotional distress.
- When self-esteem is threatened, people become
upset and lose their capacity to regulate themselves. In particular,
people who hold a high opinion of themselves often get quite upset in
response to a blow to pride, and the rush to prove something great about
themselves overrides their normal rational way of dealing with life.
- Self-regulation is required for many forms of
self-interest behavior. When self-regulation fails, people may become
self-defeating in various ways, such as taking immediate pleasures instead
of delayed rewards. Self-regulation appears to depend on limited resources
that operate like strength or energy, and so people can only regulate
themselves to a limited extent.
- Making choices and decisions depletes this same
resource. Once the resource is depleted, such as after making a series
of important decisions, the self becomes tired and depleted, and its
subsequent decisions may well be costly or foolish.
- The need to belong is a central feature of human
motivation, and when this need is thwarted such as by interpersonal
rejection, the human being somehow ceases to function properly. Irrational
and self-defeating acts become more common in the wake of rejection.
When I read this list it struck me just how many of these factors could
influence investors. Imagine a fund manager who has just had a noticeable
period of underperformance. He is likely to feel under pressure to start
to focus on high risk, high payoff options to make up the performance
deficit. He is also likely to feel his selfesteem is under threat as outlined
in 2 above. He is also likely to begin to become increasingly myopic,
focusing more and more on the short term. All of this is likely to be
particularly pronounced if the position run resulting in the underperformance
is a contrarian one. Effectively pretty much all the elements that lead
to the psychology of irrationality are likely to be present in large quantities.
Hard wired for the short term
Having explored the role of emotions and our ability to moderate their
influence, it is now time to turn to some examples of how powerful neuroscience
can be in helping us understand investor behaviour.
The first example suggests that we may be hard wired to focus on the short
term. Economists are all brought up to treasure the concept of utility14
- the mental reward or pleasure experienced. Traditionally, economists
view money as having no direct utility, rather it is held to have indirect
utility, that is, it can be used to purchase other goods and services,
which do provide direct utility.
Neuroscientists have found that money actually does have "utility", or
at least the brain anticipates receiving money in the same way that other
rewards are felt such as enjoying food or pleasure inducing drugs15.
The trouble is that the reward system for the brain has strong links to
the X-system. The anticipation of reward leads to the release of dopamine.
Dopamine makes people feel good about themselves, confident and stimulated.
Cocaine works by blocking the dopamine receptors in the brain, so the
brain can't absorb the dopamine, and hence nullify its effects. Because
the brain can't absorb the dopamine, it triggers further releases of the
drug. So when one takes coke, the dopamine release is increased, taking
the user to a high. Neuroscientists have found that the larger the anticipated
reward the more dopamine is released.
McClure et al16
have recently investigated the neural systems that underlie decisions
about delayed gratification. Much research has suggested that people tend
to behave impatiently today but plan to act patiently in the future. For
instance, when offered a choice between £10 today and £11
pounds tomorrow, many people choose the immediate option. However, if
asked today to choose between £10 in a year, and £11 in a
year and day, many people who went for the 'immediate' option in the first
case now go for the second option.
In order to see what happens in the brain when faced with such choices,
McClure et al measure the brain activity of participants as they make
a series of intertemporal choices between early and delayed monetary rewards
(like the one above). Some of the choice pairs included an immediate option,
others were choices between two delayed options. The results they uncovered
When the choice pair involved an immediate gain the ventral stratum (part
of the basal ganglia), the medial orbitofrontal cortex, and the medial
pre-frontal cortex were all disproportionately used. All these elements
are associated with the X-system. McClure et al also point out that these
areas are also riddled by the midbrain dopamine system. They note "These
structures have consistently been implicated in impulsive behaviour, and
drug addiction is commonly thought to involve disturbances of dopaminergic
neurotransmission in these systems". Since money is a reward, the offer
of money today causes a surge in dopamine that people find very hard to
When the choice involved two delayed rewards, the pre-frontal and parietal
cortex were engaged (correlates of the C-system). The more difficult the
choice, the more these areas seemed to be used. Given the analysis of
the limits to self-control that was outlined above, perhaps we shouldn't
hold out too much hope for our ability to correct the urges triggered
by the X-system. All too often, it looks as if we are likely to end up
being hard wired for the short term.
Keynes was sadly right when he wrote "Investment based on genuine long-term
expectation is so difficult to-day as to be scarcely practicable".
Hard wired to herd
In the past, we have mentioned that there is strong evidence from neuroscience
to suggest that real pain and social pain are felt in exactly the same
places in the brain. Eisenberger and Lieberman17
asked participants to play a computer game. Players think they are playing
in a three way game with two other players, throwing a ball back and forth.
In fact, the two other players are computer controlled. After a period
of three way play, the two other 'players' began to exclude the participant
by throwing the ball back and forth between themselves. This social exclusion
generates brain activity in the anterior cingulate cortex and the insula.
Both of which are also activated by real physical pain.
Contrarian strategies are the investment equivalent of seeking out social
pain. In order to implement such a strategy you will buy the things that
everyone else is selling, and sell the stocks that everyone else is buying.
This is social pain. Eisenberger and Lieberman's results suggest that
following such a strategy is really like having your arm broken on a regular
basis - not fun!
To buy when others are despondently selling
and sell when others are greedily buying requires the greatest fortitude
and pays the greatest reward
Sir John Templeton
It is the long-term investor, he who most promotes the public interest,
who will in practice come in for the most criticism... For it is in the
essence of his behaviour that he should be eccentric, unconventional and
rash in the eyes of average opinion
John Maynard Keynes
Plasticity as salvation
All of this may make for fairly depressing reading. With emotions we can't
control ourselves, and without them we can't make decisions. We appear
to be doomed to chase short-term rewards and run with the herd. When we
try to resist these temptations we suffer subsequent declines in our ability
to exercise self-control. Not a pretty picture.
However, all is not lost. For many years it was thought that the number
of brain cells was fixed and they decayed over time. The good news is
that this isn't the case; We are capable of generating new brain cells
pretty much over our lifetime.
In addition, the brain isn't fixed into a certain format. The easiest
way of thinking about this is to imagine the brain as a cobweb. Some strands
of that cobweb are thicker than others. The more the brain uses a certain
pathway, the thicker the strand becomes. The thicker the strand, the more
the brain will tend to use that path. So if we get into bad mental habits,
they can become persistent.
However, we are also capable of rearranging those pathways (neurons).
This is how the brain learns. It is properly called plasticity. We aren't
doomed, we can learn, but it isn't easy!
Largely thanks to the work of Joseph LeDoux, see his wonderful book the
Emotional Brain for details.
2 Also know as the limbic system
3 For more on this see Paul Ekman's Emotions
Revealed. It is also worth noting that some developmental psychologists
have designed programs to teach children to recognise the physical signs
of emotions (such as anger) and then use thought to control those emotions.
See Mark Greenberg's work on PATHS (www.prevention.psu.edu/projects/PATHScurriculum.htm).
Much of the work has focused on teaching children to constrain their anger
- a modern day equivalent of counting to ten.
4 Epley and Gilovich (2001) Putting adjustment
back in the anchoring and adjustment heuristic, Psychological Science
Vol 12 No. 5
5 Gilbert and Gill (2000) The momentary
realist, Psychological Science, Vol. 11, No. 5
6 For more on this see Damasio (1994)
7 Bechara, Damasio, Tranel and Damasio
(1997) Deciding advantageously before knowing the advantageous strategy,
Science Vol 275
8 Bechara, Damasio, Damasio, Loewenstein
and Shiv (2004) Investment behaviour and the dark side of emotion, unpublished
9 Technically speaking this group had
suffered lesions to the amygdala, orbitofrontal and insular/somatosensory
cortex - all parts of the X system
10 Camerer, Loewenstein and Prelec (2004)
Neuroeconomics: How neuroscience can inform economics, Journal of Economic
11 Loewenstein, Nagin and Paternoster
(1997) The effect of sexual arousal on expectations of sexual forcefulness,
Journal of Research in Crime and Delinquency, Vol. 34 No. 4
12 Muraven and Baumeister (2000) Self-regulation
and depletion of limited resources: Does self-control resemble a muscle?
Psychological Bulletin, Vol. 126 No. 2 Or Baumeister (2003) The psychology
of irrationality: Why people make foolish, self-defeating choices, in
Brocas and Carrillo (2003) The Psychology of Economic Decision Volume
I: Rationality and Well-Being
13 Op cit
14 In fact psychologists have recently
argued that there is no single utility. Instead we have experienced utility
(actual liking from an outcome), remembered utility (memory of liking),
predicted utility (expected liking for the outcome in the future) and
decision utility (the actual choice of outcome).
15 Knutson and Peterson (2004) Neurally
reconstructing expected utility, forthcoming
16 McClure, Laibson, Loewenstein and
Cohen (2004) Separate neural systems value immediate and delayed monetary
rewards, Science Vol. 306
17 Eisenberger and Lieberman (2004)
Why rejection hurts: a common neural alarm system for physical and social
pain, Trend in Cognitive Sciences, Vol 8 No. 7
© Dresdner Kleinwort Wasserstein Securities Limited 2005
Top of page.