L. Peterson M.D.
Managing Partner, Market Psychology Consulting
"Greed, Fear and the
It is a great pleasure to be at one of these conferences.
It has been a dream of mine.
I have been working for the last decade to figure out why emotions
cause so many problems for traders and investors. I will talk today
about how people's emotions affect decision making from within the
brain and how emotions lead to investment decision-making mistakes.
Then we will talk about how people's disposition to different emotions
(their personality) appears to affect their investment decision
making, in particular their biases. Finally, I will discuss some
software that we have developed that analyzes emotions in the media-blogs,
SEC filings, and news reports-and ranks stocks according to the
risk perceptions of the general public. This approach does appear
to generate very positive returns.
Warren Buffett has said many times, and he said it very well: Greed
and fear have been speculated to be in the markets for hundreds
of years. Everyone knows that psychology tends to drive markets,
especially in periods of uncertainty or extreme stress. Buffett
has pointed out that it is not necessarily intelligence that will
make you money. The example of Long-Term Capital Management is very
interesting-the company was run by two Nobel Prize winners, yet
the company created a disaster that could have brought down the
global financial system, so to speak. Sir Isaac Newton lost a fortune
in the South Sea Bubble. Mark Twain was a terrible investor, but
a great humorist. Over the years, it has become apparent that it
is not intelligence that will help you succeed in investing, at
least not classic intelligence. It may be more about emotional intelligence-learning
to balance your emotions with your rational inclinations. Understanding
what other people's emotions are may be key to driving a successful
In investment risk taking, there are three stages. The first stage
is excitement. We do not always identify this feeling as excitement.
Sometimes we just think of it as a "positive feeling"
or a feeling of potential success. In the brain, excitement, positive
feelings, etc., all act through the same neural pathway called the
reward system. On the other hand, when our favorite stock pick or
largest holding starts to decline, we might start to feel some fear
or doubt. If you are confident in your valuations, stock price movement
should not affect you. But stock price movement does affect most
people's emotional systems, in particular the loss avoidance system.
Achieving balance is our goal. You have emotions. You cannot avoid
them or control them. There is a book called The Brain-Damaged Investor
in which the author speculated that if you were a psychopath, perhaps
you would do better on Wall Street. The assumption was that if you
do not feel emotions, you may not be afraid of risk. Being unafraid
of risk is half of the battle, but you must also be able to identify
opportunities. In order to identify those opportunities, you must
be able to have a positive gut feeling about where they are, but
you do not want to have too much of a positive feeling. That is
where the balance comes in.
What is interesting about the stock market bubble of the late '90s
was that "irrational exuberance" was identified in 1996.
Knowing that emotions affect the market will not help you with your
timing. Intelligence will not tell you how to time turning points
in markets, especially those driven by emotion or by excessive or
inadequate risk taking.
I will talk to you about three brain regions today. I will keep
this broad and simplified. The first region is the frontal cortex,
which essentially functions as your executive decision-making center.
This area operates planning, delayed gratification and impulse control.
It serves as the steering wheel and the brakes on your two emotional
systems. The emotional systems lie in the limbic structure of the
brain. These are deeper structures in the brain, and evolutionarily
they are more primitive than the cortex. When your cortex goes offline
due to stress or uncertainty, your limbic system will be dominating
your decision making. That is why we have to work on your limbic
system to optimize its performance.
Unfortunately, because it is underneath your cortex, the limbic
system is often beneath your awareness. This is the key to understanding
why emotions mess us up. It would be easy if we could all identify
when we are feeling a little too positive or a little too negative.
In reality, we do not know how we feel most of the time. And when
we do know how we feel, we often seek confirming evidence to support
how we feel, or we rationalize to avoid any kind of introspection
that would inform us that our feelings might be wrong. Our feeling
centers are so fundamental and so basic that they will bias you
in ways that are very difficult to combat, especially intellectually.
I will talk about emotions today, but it is not a simple task to
manage your emotions intellectually. Managing your emotions requires
behavioral modifications and a number of other processes.
Within the limbic system, there are two primary motivational systems.
The reward system tells us to go forward and get things. The primary
neurochemical is dopamine. This system was originally associated
with pleasure. In 1954, professors Olds and Milner published a study
in which they had put electrodes into the brains of rats. The electrodes
were in the nucleus accumbens, and they gave the rats a lever that
would allow the rats to stimulate themselves in that area of the
brain. What they found shocked them. The rats would push that lever
repeatedly all day, all night. They wouldn't eat, drink or sleep.
They put female rats in the cage, and the male rats didn't care.
They just kept pressing that lever. More than sex, sleep, food or
water, these rats just wanted to stimulate the nucleus accumbens.
Many of the rats died and the rest collapsed from exhaustion. The
scientists concluded that the nucleus accumbens was the pleasure
area of the brain. Our understanding is now much more nuanced. We
now believe that it has much more to do with salience-significance
to your goals.
In neurosurgery in the 1960s, it was fairly common to ask brain
surgery patients (for epilepsy, for example) for permission to stimulate
various parts of their brains with an electrode to see what would
happen. Some subjects agreed. There are no sensory nerves in the
brain, so it is possible to insert filaments without causing any
pain. If they were to stimulate the motor cortex, the patient might
move a hand or a finger. When they would stimulate the nucleus accumbens,
the subjects would say how great it felt. Some people actually had
orgasms. They concluded that the nucleus accumbens was the pleasure
In fact, money activates the nucleus accumbens. Beautiful faces
do as well. So do many other things that are found rewarding. It
is not as simple as a "pleasure center," but it is associated
with your expectation of getting what you want. Your expectation
of making a financial gain is related to your feeling of positive
well-being and even the feelings that lead to orgasm and compulsive
lever-pressing in rats. Additionally, dopamine is the origin of
the word "dope" to describe drugs. Every drug of abuse
acts through the nucleus accumbens in the brain. Cocaine and methamphetamines
are both dopamine reuptake inhibitors that appear to have large
effects on the nucleus accumbens on neuroimaging scans.
When I talk about the nucleus accumbens in relation to money, remember
that this is a very important mechanism for behavior of all types.
It leads to excited motivation, confidence and perhaps greed. Dopamine
is the primary neurochemical involved in all of these feelings.
The second system is the loss avoidance system. This system is
more diffuse in your brain. This system is composed of the amygdala,
the hypothalamus (which has to do with the release of stress hormones),
the locus coeruleus (which releases norepinephrine, a derivative
of adrenaline, to promote vigilance but also leads to the panic
reaction that shuts down many other brain systems, including the
frontal cortex), and the hippocampus (which is related to emotional
memory). Some investors that I have worked with have had traumatic
losses in the markets. They have burned traumatic memories into
their hippocampus, and every time they get back into the markets,
their memories of trauma prevent them from being able to logically
and rationally play the game. Fear is the primary emotion that we
will talk about in relation to the loss avoidance system.
Brian Knutsen and Camelia Kunan at Stanford University conducted
a study last year that looked at how to predict people's risk taking
in an investment experiment. They found that they could not only
predict when people would or would not take risks, they could also
predict when people would take too much risk, irrationally buying
a stock when they should prefer a bond, for example. In this experiment,
there was a series of 10 rounds of investing. In each round, you
saw the outcome from the prior round, so you learned more about
what was happening. In each round, investors could choose stock
A, stock B or a bond. The investors did not know whether either
of the stocks was good, and one of them offered higher returns than
the other. The good stock yielded an expected value of $2.50, while
the bad stock yielded -$2.50, and the bond always yielded $1 per
trial. How many of you, in this experiment, would start out choosing
a stock? Because the values of the stocks are unknown, you will
achieve the best expected value through the first four rounds of
the experiment by choosing a bond. You will not have enough information
about the stocks to know which one is the highest yielding. When
I heard about this experiment, I thought I would choose a stock
as well. I am a risk taker, I guess. I thought I should share that
humiliation with you.
They found that excessive risk-seeking seemed to come from excessive
activation in the nucleus accumbens-the center of well-being and
positive activation. What is also interesting is that the nucleus
accumbens activation seemed to be correlated with recent gains.
If they had been in a stock that was making money, they would stick
with that stock even when that was the wrong stock. If they were
in the bond, and they saw that a stock yielded $10 while the other
stock yielded $0, they would switch to the winning stock if their
nucleus accumbens was excessively activated. This is similar to
chasing performance in the marketplace. Why is it that 90% of all
retail money in mutual funds goes into the top 10% of prior performers?
It is because people are chasing performance. People see good performance,
and they extrapolate that it will continue. Maybe it has to do with
our desire for control or our desire for predictability, but the
loss avoidance system in the brain does seem to be driving this
behavior, and it is tied to emotion and excitement.
The greed emanating from your nucleus accumbens is below your awareness.
You cannot easily control it. Experience does seem to help modify
its effects, perhaps by strengthening bonds between the frontal
cortex and the limbic system, or perhaps because of learning about
how different emotional impulses affect us. For the most part, when
we get excited about something, it biases our decision-making process
to make us want to do that. We want to do it, we think we need to
do it, and nothing will stop us. It is like the dieter who sees
a dessert and rationalizes why they should have it: "I'll start
an all-cheesecake diet!" Why is it so hard to stay on a diet?
It is difficult because your thinking center actually has very little
control over what you are doing. When you see something you want
or an opportunity you really want to pursue, your limbic system
In the case of fear, or "excessive risk aversion," the
insula is the pain system in the brain. The Functional Magnetic
Resonance Imaging (fMRI) images that you see of the brain actually
measure blood flows in the brain. When people take too little risk
(taking the bond when they should be taking the stock), they are
probably anticipating the pain of a future loss. They dread the
pain that the stock might bring them, so they hold the bond.
This probably reminds you of bear markets and clients who do not
want to invest. Clients will call you many times during the day
and tell you that they want to get out of their stocks immediately.
They think that the market is looking "shaky," but that
is also an opportunity. How do you reframe this situation for your
clients? Remember that you are not talking to them intellectually-you
are talking to their fear. You must work with their emotions if
you are going to successfully reframe the situation for them.
The anterior insula is associated with disgust, fear, pain and
many other things. Specifically, it is associated with anticipation.
In the reward system, the anticipation of making money is exciting.
It is not about the outcome (did you just make money?), except that
if you have just made money, you will become even more excited about
the next time that you will make money. Anticipation or expectation
is what drives our behavior. It is not about what just happened,
other than how what just happened introduces positive feedback and
amplifies your expectations of what will happen next. It is very
important to dissociate your future expectations from past performance.
That is one of the keys to my talk today.
Anxiety is an anticipatory emotion. If you are afraid of something
happening and it happens, you can launch immediately into action.
Fear is anticipation that something bad is going to happen. It changes
how we see threats and how we respond to our environment. Panic
is the conclusion of high fear. At some point, thresholds of fear
and stress will lead to a panic reaction. If you read Jim Cramer,
he describes his own panic in the markets very well. He talks about
times when the markets reached a certain level, and he decided to
throw the maiden in the volcano and get out of there. He recognized
his own panic, but he couldn't do anything about it.
Personality refers to your disposition to certain emotions. The
reason I bring up personality is because there are limits to the
fMRI tools. The fMRI is an excellent tool to help us understand
what is going on in people's brains when they are experiencing gains
and losses, but it is very expensive, it requires you to be in a
scanner, and it requires several hours of your time. Personality
tests are less reliable than an fMRI because you can lie, but they
allow us to measure certain propensities despite their susceptibility
to the self-report bias.
So far I have talked about several things. In your brain, there
is a reward system and a loss avoidance system that seem to drive
your risk-taking behavior. When they are excessively activated,
they lead to investment mistakes. Your personality refers to your
predisposition to the activation of one or the other system.
Personality is assessed using simple paper and pencil tests. I
actually have one of these tests online for my research (www.marketpsych.com).
I prefer if you would answer honestly, because that helps my research.
The paper and pencil tests involve between 60 and 300 questions,
and you answer each question on a scale of one to five.
There is an interesting story about how personality tests were
developed. Lewis Goldberg was a professor at the University of Oregon.
He noticed that for the 100 years leading up to 1960, there had
been a lot of research done on people's different personality styles
or temperaments. He wondered whether these differences in personality
could be measured. The easiest way to measure these differences
seemed to be to look at people's language. When they talked about
themselves, what adjectives did they use? What do they identify
with? What novels do they read? What activities do they enjoy? Goldberg
came up with a list of about 2,000 adjectives and phrases that could
be used to describe people's different behavior, things like "I
get stressed out easily," "I like to try new foods,"
or "I like to follow rules." People would respond to each
of these statements on a scale of one to five (strongly disagree
to strongly agree).
Goldberg and his research assistants knocked on the doors of over
300 houses in Eugene, Oregon, to test his questionnaire. Once they
analyzed the data, they discovered that most of the questions didn't
matter. Most of them were not correlated with anything in particular.
About 300 of the 2,000 statements, however, appeared to cluster
into about five different categories. If you rated that you get
stressed out easily and that change makes you nervous, then you
were likely to score highly on a trait that they called "neuroticism,"
referring to stress propensity. If you said that you generally have
a sunny disposition and that you liked to go to parties, then they
called you "extroverted." Each of the five categories
was independent from the others. Three of the five categories are
relevant to what we are talking about today.
Neuroticism is perhaps the most relevant. This has to do with the
activation of your loss avoidance system. As Dr. Berns found in
his study of dread, some people tend to be much greater "dreaders"
than others-some people get more stressed out by thinking about
threats. During bear markets, you see these people on CNBC. There
are people out there who are chronically focused on risk, who are
chronically pessimistic, and they have increased activation of their
loss avoidance system.
I want to emphasize that this is okay. It is okay for different
people to have different personality traits. Evolutionarily, different
personality traits are adaptive to different circumstances. The
diversity of personality traits is very important to have. On your
investment committees, you need to have someone who knows about
all of the risks. They need to know what will happen to prices in
the next oil shock. They need to know about Chinese commodity levels.
They need to know about other risks to development. It is important
for some people to have these traits. There should not be any judgment
about these traits, and it is unfortunate that Goldberg used the
word neuroticism (since Goldberg, the word has acquired a somewhat
negative connotation). There are many highly creative people who
are highly neurotic.
Extroversion refers to intrinsic levels of optimism and sociability.
fMRI studies have shown that the reward system is more active in
people who are extroverted when they anticipate making gains. People
who are neurotic have more active loss avoidance systems when they
anticipate making losses. During our studies at Stanford, we discovered
that people who are neurotic tend not to have as much reward system
stimulation when they experience monetary gains. In fact, when neurotics
have the opportunity to make money, there is no activity in their
nucleus accumbens beyond their baseline level-they don't really
care about gains. When they have a chance to lose, however, they
get very excited. That is when their nucleus accumbens gets activated.
It is somewhat paradoxical, but it makes sense if you think about
it. If you are always primed to look for threats, you are going
to get excited when you perceive a threat, because this might be
just what you have been looking for. Now is the time for action.
When you are high on the neurotic scale, it is more exciting for
you to see threats than opportunities.
Conscientiousness refers to the strength of your prefrontal cortex.
College students are notoriously low on conscientiousness-they are
slow, they are late to class, and their rooms are messy. People
who are conscientious tend to be punctual and organized, and they
follow the rules. Discipline and self-control (impulse control)
are the hallmarks of conscientiousness. Your ability to control
your emotional impulses appears to be related to conscientiousness.
Conscientiousness is critical for success. Studies have shown that
the greater your level of self-control, the greater your lifetime
level of wealth. This is related to delayed gratification: Can you
save money for retirement, even when no one is forcing you to? Can
you avoid purchasing large items? Can you be disciplined about your
spending? If you are conscientious, you are more likely to be able
to do all of these things.
Neuroticism, extroversion and conscientiousness are three independent
traits. We can also talk about combinations of traits. If you rate
highly in conscientiousness and neuroticism, then you are a perfectionist.
You follow the rules and you are disciplined, but you also get worried
about following the rules and being disciplined enough. If you do
things perfectly, that keeps you happy. In the case of extroversion
and neuroticism, we see pop icons like Woody Allen.
In our online personality test, we have 60 questions related to
personality and a few research questions related to how people rated
themselves in terms of certain investment biases. There is an inherent
flaw to this type of research, and I could not figure out how to
get around it: If you are rating yourself in such a way that you
will look good, you will continue to rate yourself to look good
when you start rating your investment biases. When asked to rate
themselves on the statement, "I dream about my investment positions
at night and it impairs my sleep," neurotic people will agree,
conscientious people will disagree (probably because they have plans
in place), and extroverts showed no correlation to this question.
If you are extroverted, you will claim to have exit plans in place
before you start investing, meaning that you've thought ahead. If
you are optimistic and you want to avoid any pain, then you will
make the plans necessary to protect yourself. This might make sense-if
you want to keep yourself in a positive state to avoid pain, then
you will be sure to protect yourself. Some research has shown that
people who are optimistic try to stay positive. Usually they take
confirming evidence to support their views, but it would make sense
for them to protect themselves so that they can continue to feel
good about themselves. Neurotic people are pessimistic, and they
tend not to put exit plans in place. Conscientious people, of course,
do put exit plans in place. They follow the rules.
Do they hold their losing positions too long? Extroverts claim
that they don't hold them too long. Neurotic people say that they
do. Conscientious people said that they don't.
So far, most of the negative attributes are falling on the neurotics.
Now, this could potentially be because they tend to rate themselves
negatively anyway. There is some literature, however, that supports
the assertion that neurotic people do tend to have these negative
Do they cut their winners short? Extroverts say that they do. This
may be because they are optimistic, and they think that their winners
could have done even better. The others were uncorrelated.
Do you double down when you're losing? This is a classic theme
from prospect theory: Do you take more risk when you are losing?
Neurotic people do double down. Conscientious people say no-they
know what they are doing so why would they mess with their dollar-cost
average when they are losing like that? The neurotics are afraid
of risk. As Michael and I have discussed, people's loss aversion
curve does tend to shift according to their intrinsic disposition
Do you tend to leave your winning positions open too long and then
lose money? This was a trick question. We wanted to see if anyone
would say that they both cut their winners short and left them open
too long. I thought extroverts might fall into this trap and show
us how they make mistakes in their self-evaluations. In fact, extroverts
say that they cut their winners short and do not leave their winners
open too long. They are consistent in their answers. I need to come
up with some more questions to try to trap people who are either
being too positive or too negative with their self-evaluations.
I didn't get them this time, but I have more questions coming. People
with neuroticism say that they leave their winners open too long.
Conscientious people say that they don't.
Are your total returns high? Only extroverts claim to have high
total returns. It makes sense that optimists would claim to do well,
but there was actually a study done in 2005 in Australia that showed
that extroverts did tend to get higher returns than others. It was
one of the few correlations that they did find because they had
a small sample (20-25 people). Also, Australia was in a bull market.
If you are optimistic in a bull market, you are likely to buy stocks
and do well with those stocks. If you were optimistic in 1996 and
1997, you would have done very well in the stock market. When it
came to 2001 and you wanted to stay the course and put more money
into the market, then you ran into trouble.
The key is to understand your personal disposition. Understand
when you are likely to go too far. This way you will be able to
put some behavioral plans in place that will be triggered when the
environment changes and you start getting into trouble. If you are
on a diet and have to walk home past the Cheesecake Factory every
day after work, a behavioral plan would be to walk on the opposite
side of the street. In this way, you have put your temptation at
a distance, but you still know it is there.
This kind of temptation goes through the reward system of your
brain. Why is addiction such a problem in this country? It is related
to the nucleus accumbens, to emotions, to stress, to loss avoidance.
Addiction is a problem because it is a complex behavior. It is driven
by emotions that people are aware of, but they cannot seem to stop
them. How do we get ourselves balanced? This is a real public health
issue, and it is an issue for investors as well.
Question: If I were to take a personality test like this,
I would be all over the place. How does your research account for
The problem with these tests is that they ask you to rate yourself
on individual items. You might circle the middle option on some
questions-you neither agree nor disagree. Answers like these do
not help me much with my data, but it also might represent your
personality. It may be that you have an objectivity that is helpful
for your performance. You may be able to go either way, depending
on the situation. This might be reflected in the performance data.
I should look into this further. How do people perform who sometimes
agree and sometimes do not?
Question: Does education have anything to do with the test?
Would someone trained in psychology answer these questions differently
because they know what the expected answers would be?
The key with this test is that lying only hurts the test taker.
The test is completely confidential, and I don't know anything about
the people taking the test. Because of this confidentiality, we
encourage people to be honest. We ask some personal questions like
investors' returns and their wealth level. Some people are willing
to answer these questions and some people refuse to answer even
the self-assessment questions. They do not seem to understand the
purpose of the test and the strict confidentiality with which we
treat this data. I think people naturally are hesitant to reveal
secrets about themselves online.
The problems that you are bringing up are the reason why fMRI will
probably provide the best personality assessment in the future.
There are correlations between fMRI scans and these results, but
I think that we will be able to detect instances where people are
fabricating a self-image that is not really there by seeing what
really happens in their brains in risky situations. There are a
lot of ethical issues associated with "mind reading,"
but it does represent a possibility for the future.
So far we have talked about emotions in your brain, how they activate
you to take too much or too little risk, and we have talked about
your emotional disposition. Do you tend to get too negative? Too
positive? And do you have the self-control to manage those impulses?
In my journey as a trader and psychologist, I discovered that there
is a lot of useful information in online discussions. During medical
school, a group of us played poker once a week. In the fall of 1998,
one of my colleagues told us that he was making thousands of dollars
trading penny stocks. While I thought it sounded risky, he told
us that it was very easy. A month later he told us that he had made
$100,000 in the last month. That caught my attention. Maybe he actually
had discovered something. I asked him to tell me about his approach.
He was going to stock message boards-pump and dump sites-where
a group would wait until someone had identified the next big stock,
and then everyone would buy it. He would buy it with everyone else,
but he would sell it the next day. Everyone else would hold onto
it for too long. He made a lot of money this way.
I thought that the language that the pump and dump artists were
using was very interesting. They used all caps. They used bold print.
They would call a stock "discounted." They would talk
about how a stock was valued at $40 but was only trading at $2.
Because I am an analytical guy, I tried to figure out what elements
of the language were driving these investors. Using this technique,
I developed some schemes of my own and even made some pretty good
money. Once my residency started, however, I didn't have enough
time to do it.
I have now automated this approach. My software scans news media,
message boards, and SEC filings for certain types of language. We
deconstruct the language to find what elements are exciting people
and what elements are conveying threats or risks to people. When
people get too excited or too risk averse, we know that their brain
makes errors. My question was whether people were making systematic
errors based on the language that is used in financial reporting.
Is there a diversity breakdown? We all have the same biological
systems that govern our gain pursuit and our loss avoidance. Do
we all collectively get activated in one way or another? We all
have different personality traits, and some are adaptive during
different time periods. In the short term, however, are we all collectively
so blinded that noise traders in the finance literature drive prices
sufficiently out of whack that you can predict the future based
on price movement and language in the media? Is this based in a
About two years ago I started running search engines on nightly
TV news transcripts. The articles have headlines like "Oracle
Posts Higher Earnings" or "The Federal Reserve Open Market
Committee Meets." These articles include a lot of business
language, but what of this language is really useful? What is useful
is language about risk. A couple of academic papers have now been
published on this topic. Paul Tetlock at the University of Texas
found that by counting the positive and negative words in the "Heard
on the Street" column in the Wall Street Journal, you could
make about a 7% annual return trading just on the sentiment in the
column. Feng [?] at the University of Michigan studies SEC filings
for the word "risk." The more that the word "risk"
appears in SEC filings, the more likely the stock is to underperform
in the coming quarter.
What I did with the nightly business report, I also did with CNN's
MoneyLine. MoneyLine was probably not a good choice initially. Their
headlines included titles like "A Large US Operation in Iraq,"
and "A New Threat to Space Shuttles Found." We see the
word "threat," but it has nothing to do with the markets.
"Search Warrants in Lacy Peterson Case." You would expect
better from MoneyLine! When you do this kind of work, it is very
messy and it is very difficult to find any meaning. It is difficult
to find out what is important. What is triggering the subconscious
systems in investors to push us beyond our rational ability to put
the brakes on? What causes people to go in excess in one direction
or the other?
With the nightly business reports and MoneyLine, there is a very
strong correlation (P value = 0.001) that negativity in terms of
both sentiment and risk predicted that the market was likely to
go up over the next four days. It was not tradable-these are rough
sources. It is also contrarian. When these sources tell you to be
scared, it is time to be buying. I have since done the same with
the Wall Street Journal. The Journal is also contrarian. What I
have found is that section C is actually tradable-you can make excess
returns simply by understanding the language that other people are
looking at. My hypothesis is that the language is working through
these brain systems to break down diversity in the markets, often
I have created some rough models in Excel to show you. On the chart,
you will see several large Internet stocks-Apple, Microsoft, eBay,
Google, Yahoo!, Intel, Oracle, etc. I scan about 60 stocks three
times per day. Every eight hours the search engines go out to search
for new news reports on these stocks. You get a smooth sense of
sentiment or risk perception for these stocks over time. The lowest
risk perception on August 4th, 7th, and 8th, was in Yahoo!. I then
smoothed the data to get a general sense of the risk perceived in
each stock. Then I look at the price trajectory for the stocks ranked
the highest or lowest. In general, there is a significant, tradable
reversal. I have had a real-time trading system running since July,
and it makes about 0.25% per trade on average. It's not huge, but
it adds up. The average trade is about 2.5 days. You can make money
trading in this way. It is a novel way to extract risk premium from
On August 8, Cisco reported their earnings after the close. At
the open the next day, the shares jumped 15%. Sure they beat expectations,
but the point is that the expectations were low. I am able to measure
the market's expectations. People were very negative on Cisco; they
were too negative. The stocks in the middle of my risk perception
chart generally have expectations well-correlated with reality.
You cannot trade those. When you get out to the extremes, however,
you can trade them. It is a labor-intensive long-short strategy,
but it does work.
Let's look at another time period in July. The expectations for
Cisco were again very negative, but through the month of July, Cisco
did not decline as much as other tech stocks did. The most positive
expectations were for Yahoo!. I found the following statement in
the business news about Yahoo!: "Analysts betting Yahoo! will
match or exceed Wall Street's rosy expectations. Investors are optimistic
about Yahoo!'s ability to grow." These statements include lots
of words that imply that there is no risk in Yahoo!. Sure enough,
our crawlers picked up on that. Yahoo! was perceived as being low
risk, which means that Yahoo! is actually very high risk. The key
is contrarianism-perceptions of low risk mean high risk, and vice
I talked to Dr. Berns during the last break, and I was struck by
how many analysts are positive on Google. Because of my data, that
makes me scared. To be fair, the literature shows that analysts
tend to be correct, though not in a tradable fashion.
What is interesting is that when Yahoo! announced that they had
fallen short of earnings expectations, they lost 20% of their value.
That is incredible based on one earnings report! I won't say that
I told you so, but the search engines showed that this loss was
all about the flawed risk perceptions in the market.
Question: You are talking now about the level of risk perception.
Have you looked at the changes in risk perception?
I have. That is also correlated, though not as strongly. There
are a lot of details that I can get into
which I will because
apparently I have time. What is interesting about the change in
the risk is that it is more significant when there are large changes.
I have not analyzed it beyond showing that it does very well with
large changes. Most of the time, investors are correct and the market
is correct. Sometimes, however, the market gets a little too excited.
In this case, it got too excited about Yahoo!. And sometimes, the
risk perception moves very quickly from too excited to too negative.
If I were to show you this chart going forward, the Yahoo! line
jumps from the least risk perception to the middle. Over the next
few weeks, it worked its way back down to extremely low risk perception.
People became positive again. There are many ways to look at this
data. I don't mind showing you this data because you really need
to get into the mathematics and statistics to see what is or is
not going to work. The lesson for you guys is that high risk is
probably actually low risk, especially at the extremes.
What are some applications? Clearly you need to be careful with
your own investment choices. I was talking with Gardiner Morris,
who is a reporter from the Harvard Business Review, about the fact
that emotions are necessary to decision making, though too much
emotion is a problem. Jack Welch's book is called Straight from
the Gut. He uses his intuition and emotion to make decisions. Why
is Jack better at doing that than so many other people? Does he
have an extremely big prefrontal cortex? Does he have more nuanced
reward and loss avoidance systems? That is really the challenge,
and that is the direction that research is now taking. It seems
to have a lot to do with experience. As people age, they become
better at calibrating their emotional impulses. Impulses that initially
indicate an opportunity are tempered by past experiences. With a
strong prefrontal cortex, you can self-reflect and self-process
Any time that you have a job that provides you with feedback over
defined time periods, you can learn. There does seem to be a learning
process-we are human, but we are all adaptable. We can use our feelings
as a guide. If you are reading a highly optimistic article about
Yahoo!'s earnings in which people are betting on their optimism
for Yahoo! to beat forecasts, you should recognize that something
is wrong. This is why Yahoo!'s price dropped so significantly. Their
price would have dropped even if they had met their official expectations
because the market had expected them to beat those expectations.
As soon as expectations get in the "whisper number" range,
you are talking about psychology and excitement.
This didn't even work during the Internet bubble. In a study from
1998 to 2001, Brett Truman at UC Berkeley studied price movements
around earnings reports. He found that on average, Internet stocks
went up about 4.8% in the five days before an earnings report, and
they went down 6.5% in the five days after an earnings report regardless
of what the report said. People got excited before the report, and
then they were disappointed afterwards. It is a lot about the dopamine
slide that Greg Berns showed us earlier. When the reward doesn't
happen, people's nucleus accumbens slows down and the dopamine slows
down, creating a disappointment signal. When you are disappointed,
you reevaluate your earlier decisions and you often sell your position.
Because there was so much liquidity in the Internet bubble, the
day traders were able to fuel this crazy price pattern for years
in the Internet stocks.
How can we use this information to evaluate portfolio managers?
Clearly, conscientiousness is key. Find people who are punctual
to meetings and who are responsible for their investment returns
(who don't give you excuses). You want someone who will tell you
that they made a bad decision and why it happened. You also want
someone who does not take credit for their performance. You want
them to tell you why a certain decision worked out well. If you
start taking credit for things, the market is going to take it back.
Find people who are objective. They will say that yes, the price
went up and now the business is valued in this way relative to these
important factors. Listen for very objective language and objective
benchmarks in their valuations of companies and components of their
You can use this information to develop a market strategy. I've
shown you how this can be used for trading systems, but how can
you apply these ideas to develop longer-term strategies? Because
I am focused on launching more trading products, I have been focused
on very short time scales for my analysis-days and weeks. Feasibly,
these same tools can be used over the long term, but it is hard
to get data old enough to track these trends. I have the Wall Street
Journal going back 20 years, the Financial Times about eight years,
and the New York Times about eight years. To develop an annual strategy,
that will not give you any statistical significance.
How can we use this for sales and marketing? A lot of you who deal
with clients deal with people who are very anxious during market
downturns. If you were to plot market downturns and client phone
calls, you probably get an inverse curve, right? People call you
when their wealth is declining. They do not call you when things
are going well, but they are bragging to their country club buddies
about it. Unfortunately, their emotions get tied up in these things.
When you are dealing with your clients, it is very important that
you address their emotions. Most planners will show their clients
all of the relevant statistics: the Monte Carlo analysis, the historical
returns, etc. It is good to do this to satisfy the client's intellect.
But you also have to address their emotions. Statistical analysis
does not deal with their impulse to panic and sell. What addresses
their emotions is an open conversation in which you tell them that
you know that they are concerned, and you talk with them about their
concerns. I won't tell you how to do counseling, but basically you
have to address their emotional concerns, and then you have to reframe
the situation as a long-term positive. "When you're afraid
is the best time to buy. Now is a great opportunity!" You need
to be prudent about this. I heard planners telling their clients
to double down in late 2000. You can tell by their level of fear
whether or not your clients are thinking long term. Stress and fear
pulls us into the short term.
What financial products have come out of this research? There is
the smart plan for retirement. There is automatic opt-in for 401(k)s.
These take advantage of people's inertia and their discomfort in
dealing with risk and volatility. Lifecycle funds account for people's
changing risk-aversion as they age. Brian Knutson has shown that
people become more risk averse as they age-they tend to irrationally
select bonds more often. As people age, the shift to bonds is probably
more appropriate for their risk tolerance in general.
I have spent this time talking to the intellectual centers in your
brains, but I hope that inside of you there is a little emotional
war going on. There are a lot of emotions involved in investing,
and the financial industry has tried to keep emotions out for a
long time. It has tried to regulate emotions out of the industry
when possible. We do need some emotion when valuating companies-we
will have a good feeling or a bad feeling. Those feelings are changing
your cognitive assessments. They are changing your outlook for the
future. And they are changing what you do to the point where there
are patterns in the markets that can be exploited. Be very careful
about your emotions.
My website is www.marketpsych.com.
You can find the personality test there. You can take a free version
of the test and see your results. I hope that the title of my book
will be Inside the Investor's Brain, although right now it is being
called Inside the Trader's Brain. We are still in negotiations.
Jason Zweig from Money magazine is releasing a book next April called
Your Money and Your Brain.
Question: What were the other two personality characteristics
that were not relevant to investing?
Answer: The other two characteristics are relevant to investing,
but I cut them out because I didn't want my talk to go on too long.
The first is agreeableness or self-interest. How much do you cooperate
and affiliate with people versus how paranoid or stand-offish are
you regarding interactions with others? How much do you trust versus
how much are you suspicious? Andrew Lo studied about 80 day traders
and discovered that most of them had very low agreeableness. They
had very high self-interest.
The second trait is called openness-openness to new experiences.
That trait is actually correlated with much of the success you see
with extroversion and conscientiousness. Openness to new experiences
refers to intellectual openness and appreciation of abstract thought
versus concreteness. On the opposite end of the scale is "traditionalism,"
which refers to people being more conventional and preferring the
tried and true. People who are open would rate themselves highly
on statements like "I like to try new and exotic foods,"
or "I like to think about complex intellectual problems."
Openness does correlate quite highly to performance.
Question: Are there any personalities that are correlated
with investment success?
Answer: The best data is on traders because there is great
short-term feedback. For long-term traders, it is difficult to get
the data. I would need to get 20 Bill Millers in the room and give
you all a personality test. That would be the way to get investor
data, and since Bill is one of a kind, it is difficult to get that
kind of data. We do have that data with traders, but the problem
with traders is that many of the ones who do well for a while tend
to blow out-look at Amaranth. They had an incentive structure called
an implied put option. Traders might make $100M one year and nothing
the next, so this incentive structure encouraged him to take greater
and greater risks to make more and more money. With traders, what
is important is confidence, conscientiousness (or self-discipline),
and openness. Neuroticism is terrible-you will be out of a job within
weeks. I suspect that neuroticism would also be negatively correlated
with success for investors as well, but I am not sure about that.
Optimism is also important for investors. Optimism makes you more
willing to look for opportunities, and makes you more interested
in what you find.
Question: The target date maturity funds have split into
two camps. In one camp, equities are at about 50% for investors
who reach their 60s. There is also a conservative camp that argues
that equities should not be above 30% because these people will
run and hide when the market declines. What thoughts do you have,
or what studies have been done?
Answer: The question is how to balance equity ratios for
target date maturity funds. How do you set them up so that the fewest
people jump out of the funds? There is no specific data on this.
My suspicion is that some basic risk-reward questionnaires will
tell you pretty well how people's assets should be allocated. People
who score highly on neuroticism should not be in a fund with a high
concentration of equities.
Question: I have a question on diversity breakdowns in the
markets. How do you go about identifying them, quantifying them
and breaking them down in the stock market?
Answer: Using the language, I look at the common usage of
certain terms and phrases. This is like a voting mechanism: 2% percent
of all phrases about Cisco were negative, while 1% of all phrases
about Yahoo! were negative. In general, most companies have about
twice as many negative comments as positive comments about them.
People tend to talk in negative terms about twice as often as they
do in positive terms about companies. Language assessment is the
tool that I use.
It is hard for us all to understand what everyone else is thinking.
This is especially hard when you're in the midst of it-when you're
in a corporate culture related to investments. All of the new opportunities
that people talk about come from within the "club." So
how do you really break out into new territory? This is very challenging.
It is almost like you need free agents out there looking for new
Michael Mauboussin: I would like to add something here.
Your work on language is fascinating. I would look for price actions.
If something is up a lot or down a lot, that should be a precursor
to some kind of diversity breakdown. Look at the popular press:
If someone is on the cover of a magazine saying that something is
great or terrible, that usually signifies a turning point of some
kind. You need to understand expectations. A big price move up or
down will revise expectations for future performance of that company
or sector. Combining these three ideas seems to be a good indicator
of whether an idea is investable.
Answer: In catastrophe theory, as volatility declines, the
risk of a large move (a fat tail event) happening increases. In
individual securities, if there is little volatility during a large
price move, that should be a warning sign. This is my intuitive
sense. I have measured it in the past, but it was not tradable.
I have a background as a statistician. I did neural network analysis
to create mathematical models to predict the S&P 500. Those
models worked for about a year, and then they started to decline
as more people started using similar tools and finding the same
patterns and arbitraging them away. I worry about sharing the language
data because I fear it might be arbitraged away, but I think that
it is so pervasive and biologically based that the arbitrage risk
is less. It is almost endemic to this industry.
Question: Is there any research into whether these personality
traits change throughout a person's life?
Answer: They do appear to change over a lifetime. People
become less neurotic as they age, and they become more conscientious.
I have had about 1000 people take the test so far, but I had to
cut out about 300 jokesters. I had to cut out everyone under 25
because their answers were not useful. Between 25 and 80 years of
age, neuroticism declines and conscientiousness rises. These results
are true in the personality literature as well. As neuroticism declines,
however, risk-aversion increases. Why is this? Perhaps my sample
reflected people who are investors, so perhaps they do not represent
the entire population very well. They may be more comfortable with
risk than the average older person. This also may have something
to do with prefrontal cortex function, which does decline somewhat
comments, opinions and any forward predictions presented about any
particular security, the economy and "the market" are
based on the analysis of the speaker. These are not necessarily
the opinion of, and should not be construed as a recommendation
on the part of Legg Mason Capital Management or any of its affiliates.