The Effects of Probability Ambiguity and Trust

"Take the probability of loss times the amount of possible loss from the probability of gain times the amount of possible gain. That is what we're trying to do. It's imperfect, but that's what it's all about."

—Warren Buffett, speech given at the Berkshire Hathaway Annual Meeting in 19891

In 2000, during the run-up to the June Mexican presidential elections, an outsider was poised to overturn the PRI political legacy that had ruled Mexico for more than 80 years. Vicente Fox, candidate of the PAN party, was polling ahead of the candidate of the incumbent PRI political party, Francisco Ochoa. The PRI had retained power through a corrupt system of vote buying, ballot-box stuffing, nationalist rhetoric, and intimidation. The PRI party bosses were angered by the possibility of their candidate's defeat. Two weeks before the presidential election, Ochoa expressed his hope that there would be no "riots" or "blood on the street" if the PRI lost the election.

Ochoa's words were interpreted as a veiled threat, though he denied any ill intent. The Mexican stock market (Bolsa) and the Mexican peso dropped 20 percent over the next week. Political violence, which had seemed unlikely, suddenly was in the realm of possibility, even probability.

Two weeks after Ochoa's comment, the Sunday of the elections passed uneventfully, with Fox declared the victor by a 6 percent margin in the popular vote. Immediately after the official results were announced, Fox held a press conference. In his victory speech, he expressed faith in Mexico's market economy and vowed to continue economic and political reforms.

No violence came to pass. Within two days the Bolsa and the peso surged through their levels of two weeks prior.

Investors in Mexico were frightened by the possibility of political violence. As we'll see in the following pages, investors' risk probability assessments are distorted by their feelings—in this case, fear. Risk perceptions are often excessively low (especially after recent gains) or too high (such as following losses), creating opportunities for astute investors.

Risk perceptions are formed from three characteristics of a decision situation: (1) the perceived timeline of the outcome (prolonged outcomes appear less risky), (2) the associations the scenario induces (e.g., fears of bankruptcy), and (3) one's evolutionary preparedness for certain emotional reactions (e.g., is one constitutionally nervous or optimistic?).2 Each of these factors can lead to distortions in probability assessment.

In this section, we look at how investors ought to assess probabilities of potential outcomes and how they actually do it. In particular, we'll examine how investors misjudge expected values and overreact to ambiguity.


"Every day, investors must translate investment opportunities into probabilities—indeed, this is an essential skill. So we need to think carefully about how we come up with probabilities for various situations and where the potential pitfalls lie."

—Michael Mauboussin3

According to traditional economic theory, it makes no sense for someone to buy a lottery ticket. It's not rational. The "expected value" of a lottery ticket is about $0.40 for every dollar spent, meaning that the average lottery ticket buyer is losing $0.60 on their $1.00 investment.

Yet people do play the lottery, and they go to casinos, and they day-trade stocks and currencies—even when they consistently lose. As seen in this book, over the past 30 years behavioral economists have begun investigating why people decide to pursue negative-expected-value gambles.

Most decisions in which potential outcome sizes and probabilities are known are made according to the guidance of expected-value calculations. These calculations are functions of the brain's analytical (not the intuitive) decision-making system. Analytical decisions consider mathematical outcome characteristics like the exact sizes, probabilities, and time delays for potential gains and losses.

In an expected-value calculation, each decision option is examined separately to determine which has the highest average return. For every choice, the probability of each possible outcome is multiplied by its likely magnitude. Then the various products are summed into an overall expected return. Given that most investment decisions produce uncertain outcomes of unknown size, such formulas must be used cautiously.

Many people gamble and trade stocks because they like the markets, they enjoy discussing strategy with their colleagues, and they enjoy the emotional thrill. Liking something, regardless of mathematical outcomes, is called utility. Decisions made using "expected utility" rely on considerations of what people value qualitatively. Because the market future is fundamentally uncertain, most investors make decisions based on expected utility rather than expected value calculations.

Utility is a largely subjective phenomenon. Decision theorists say that a sense of utility comes from: (1) current feelings (moment utility), (2) the feelings one expects to have after receiving an outcome (outcome utility), (3) the process of making the decision itself (decision utility), and (4) one's recall of past similar experiences (experienced utility). The theme that runs through these various types of utility is feelings—utility is boiled down to how one feels about a decision and the expected outcomes. In the end, investors often pursue the decision that they expect to make them feel the best.

In a challenging decision environment such as the stock market, most people make systematic errors in their outcome probability and size estimations. These errors are often due to the biasing influences of feelings on the analysis of uncertain or ambiguous information. Perhaps surprisingly, even in perfect circumstances where all outcome information is known, investors still demonstrate biased decision making. This chapter will first address the biases of finite expected value calculations (such as are possible at casinos), with biases related to ambiguous outcome information following.


In casinos the game with the lowest expected value is keno. In keno, players select six numbers between 0 and 50. If any of their selected numbers is a match with six randomly generated numbers, then they win a jackpot. Casinos take approximately $0.29 for every $1.00 invested in keno, giving players the lowest expected return per dollar of any casino game. It's the keno jackpot, displayed in large numbers over the keno seats, that is so appealing to gamblers.

Brain imaging studies show a disproportionate effect of outcome size versus probability on decision making. As the size of a potential monetary gain increases, so too does reward system activation (specifically the nucleus accumbens).45 Potential reward size is more emotionally arousing than proportional changes in probability.6 7

When I worked in Professor Knutson's lab at Stanford in 2003, we designed an experiment to measure the neural effects of changes in a reward's expected value. Equivalent expected rewards caused radically different brain responses depending on how the potential reward characteristics (size and probability) were displayed.

We designed an experiment to sort out the different effects of reward size (magnitude) and probability on the brain. In our preliminary results, the reward system's response when anticipating a $5 gain was much more intense than when anticipating a $1 gain. However, when we controlled for (eliminated) the effects of reward size from the brain's activation map, the results were much different.

FIGURE 13.1 The brain's activation when playing for a $5 gain, contrasted against the anticipation of a $1 gain.
FIGURE 13.2 The brain's activation pattern when playing for an 80 percent chance of winning $5 ($4 equivalent) versus a 20 percent chance of gaining $5 ($1 equivalent). These images were recorded at the same significance threshold, and they show the brain's preferential activation to size.

If we looked at people's brains while they were anticipating rewards of different expected values, with the magnitude of reward held constant at $5 but the probability of receiving it varied, the brain did not activate much differently. For example, an 80 percent chance of winning $5 (expected value $4) was only slightly more activating than a 20 percent chance of winning $5 (expected value $1). The largest absolute reward ($5) was extraordinarily more exciting than the absolute $1 reward. That is, subjects were watching primarily the size of the potential reward, not the probability, even when they had equivalent expected values.

In Figures 13.1 and 13.2, equal statistical thresholds were used to generate the images. That is, these activations were recorded as if with a camera whose shutter speed and aperture were held constant. In the top image, one can see the profound, and widely distributed, effect of gain size ($5 versus $1) on brain activation. In the lower image, the highest probability of gain (80 percent) is compared against the lowest probability (20 percent) with the reward size of $5 held constant. The greater absolute magnitude is much more activating.

Please understand the above results as illustrative but not watertight. This study was never published due to problems with the "high" and "low" probabilities—some subjects' "low" probabilities were actually 50 percent due to their better skill than others. After a redesign that ensured constant reward probabilities regardless of skill, a similar subsequent study was published in the Journal of Neuroscience. However, Figures 13.1 and 13.2 are derived from the first study, and the probability confound may account for some (but not all) of the dramatic activation difference between the two brain images.

Does this increased activation to size lead to psychological effects? In fact, in previous studies subjects report more "happiness" (positive activation) when awaiting larger rewards.8

In some laboratory experiments, participants trade money in simulated stock markets. Many times, bubbles and crashes appear in the prices of the stocks traded during these experiments. Researchers investigated this phenomenon by testing for the presence of probability judgment errors in traders during the life cycle of an experimental market. They found that bubbles are correlated with traders' preference for high-value, low-probability gains in their investments. This finding makes intuitive sense. Bubble investors are making probability errors by seeking high rewards at low probabilities (overweighting low probabilities). They are overweighting the possibility of large gains.9

Professor Paul Slovic has found that emotion plays a leading role in distorting probability judgments. According to the "affect heuristic," when the outcome of a gamble has a strong emotional meaning, people mentally overweight the size of the potential reward (or loss) versus its actual probability in their decision making.10 Slovic's finding may explain why, in my experiment with Knutson, I saw so much reward system activation for college students (presumably poor) playing a game for a series of potential $5 payouts (outcomes that are very exciting for them).


"It is principally at games of chance that a multitude of illusions support hope and sustain it against unfavourable chances."

—Simon Laplace, 1796

Misjudgments of probabilities are systematic and universal. Recent experimental evidence shows precisely how people misestimate probabilities when facing simple gambles.

Probability Bankruptcy With Ambiguity

FIGURE 13.3 The probability weighting curves for most assessed risks. The different curves represent probability functions for gambles framed as gains, losses, and other permutations. The straight line represents the actual probability. Source: Prelec, D. 1998. "The Probability Weighting Function." Econometrica 60: 497-528.

FIGURE 13.3 The probability weighting curves for most assessed risks. The different curves represent probability functions for gambles framed as gains, losses, and other permutations. The straight line represents the actual probability. Source: Prelec, D. 1998. "The Probability Weighting Function." Econometrica 60: 497-528.

When an outcome is possible but not probable, people tend to overestimate its chance of occurring. This is called the possibility effect.

When an outcome is likely, people tend to underestimate its odds. This bias has been named the certainty effect.11

Events of probability less than 40 percent are susceptible to the possibility effect. Outcomes with greater than 40 percent probability are in the realm of the certainty effect12 (see Figure 13.3).

For example, long shots in pari-mutuel wagering (horse and dog racing) tend to be preferred bets of amateurs. Due to amateurs excessively betting on long shots, their expected payoffs are lower than their actual odds of winning suggest they should be (based on how tracks figure odds). The more people who bet on the long shots, the weaker their odds and thus the lower the payout if they win.

Because bettors mentally underweight the probability of high likelihood winners actually winning, they avoid these bets. As a result, high-likelihood winners are often a good bet, having a better chance of winning than predicted by most bettors.13

For the most part, there is a dissociation between intellectual judgments of risk and emotional feelings about risk. Emotions in uncertain or risky situations are more sensitive to the possibility rather than the probability of strong consequences, contributing to the overweighting of very small probabilities.14 In general, naive investors think that very low probability but emotionally loaded events (such as potential market crashes) are much more likely than they actually are.15 High-likelihood, emotionally weighted outcomes, such as bull markets, are assumed to be less likely than they actually are The probability weighting curve shows exaggerated misestimations of probability when the potential outcomes are emotionally weighted (see Figure 13.4).

Probability assessments are biased not only by the emotions associated with the event outcome, but also by one's internal emotional state. People who are generally happy are more optimistic, judging higher probabilities for positive events and lower probabilities for negative events.16 The inverse pattern was found for subjects in a negative emotional state.17

The misjudgment of probabilities described above occurs for choices that are not repeated over time (except in the case of pari-mutuel wagering). The tendency to overweight small probabilities is true for single choices. This pattern does not hold true, however, for choices based on experience. Investors with experience can often identify when amateurs as a group are over or under-weighting probabilities.

FIGURE 13.4 The probability weighting curve when emotion (affect) is aroused by the potential outcomes. People further overestimate low probabilities and underestimate high probabilities when an emotional outcome is involved. Source: Hsee, C. K., and Y. Rottenstreich. 2001. "Money, Kisses, and Electric Shocks: On the Affective Psychology of Risk." Psychological Science 12: 185-190.

FIGURE 13.4 The probability weighting curve when emotion (affect) is aroused by the potential outcomes. People further overestimate low probabilities and underestimate high probabilities when an emotional outcome is involved. Source: Hsee, C. K., and Y. Rottenstreich. 2001. "Money, Kisses, and Electric Shocks: On the Affective Psychology of Risk." Psychological Science 12: 185-190.




"Our probability assessments shift based on how others present information to us."

—Michael Mauboussin18

Emotionally, investors have stronger reactions if possible outcomes are more vivid or imaginable. Before the Mexican elections, anticipation of rioting in Mexico City brought to mind images of violence, chaos, destruction, and bloodshed. Likewise, a vividly imagined possibility of bankruptcy, personal poverty, job loss, or market panic will generate the desire to sell any assets that carry such risks. The possibility of wealth and material success in an impulsive and inexperienced investor will lead to a strong drive to buy securities that appear promising.

Feelings of fear or worry in the face of decisions under risk or uncertainty have an all-or-none characteristic: they are more sensitive to the possibility rather than the probability of negative consequences. For example, the thought of receiving a painful electric shock is enough to spark intense fear in experimental subjects. Changes in the precise likelihood of being shocked have little impact on the degree of fearfulness. The increasing vividness of a potential reward or catastrophe exaggerates the possibility effect.19

People who have more control over their imaginative faculties are more physiologically responsive to images. For example, researchers found that these people salivate more when thinking of their favorite food, become more sexually aroused in fantasy exercises, and have a greater ability to increase their heart rate using visual imagery. In all likelihood, the possibility and certainty effects are exaggerated in these individuals. It's tempting to speculate that imaginative people are thus more motivated by fantasies of imminent financial wealth or ruin.


"The fundamental law of investing is the uncertainty of the future."

—Peter L. Bernstein

Most people are averse to uncertainty and ambiguity. A classic experiment into ambiguity aversion studied subjects' choices between two urns of red and black balls. Participants were asked if they would rather bet on the drawing of a black ball from the urn in which the proportion of red and black balls was unspecified, or one in which the ratio was 50:50. Most subjects avoided ambiguity, preferring to bet on the urn with the known ratio of red to black balls.20

People shy away from decisions about which they have insufficient information, especially when an alternative decision with more information is available or the decision maker has less information about a choice than others. People prefer to make decisions where they know the odds and the possible outcomes, and they avoid investing where the odds are unknown. With experience, they learn to infer the approximate odds in a situation.


"The future is never clear, and you pay a very high price in the stock market for a cheery consensus. Uncertainty is the friend of the buyer of long-term values."

—Warren Buffett

Ambiguity aversion is reflected in the returns of stocks of companies with ambiguous information on their balance sheets (e.g., research and development spending) that cannot traditionally be correlated with future valuations. Stocks with poorer earnings quality (having more ambiguous information) have greater long-term returns than those with better (more transparent) accounting.21 As a result of aversion to ambiguous items on accounting statements, investors mistakenly avoid such stocks and miss out on greater long-term returns.

The effect of accounting ambiguity may be to bias how investors feel about the stock in question. Investors with limited information about a stock are more likely to rely on their feelings when judging whether to buy or sell shares. In the case of poor earnings quality, those feelings are likely to be excessively negative. However, there are some interesting exceptions.

Typically, investors prefer to avoid stocks with ambiguous accounting, but when optimism is high, investors may project an unfounded positive spin onto the uncertain data. In the late 1990s, initial public offerings (IPOs) of Internet companies with negative cash flows had relatively higher initial trading prices. According to researchers, "The role of earnings in valuation of these firms ... completely disappeared in 1999."22 Presumably, investors considered negative cash flows to be an investment in the future. The lack of concrete earnings information led to an exaggerated projection of investors' optimism. In such cases of "irrational exuberance," investors are ambiguity loving. In summary, investors are usually ambiguity averse, especially when they are pessimistic, but during periods of optimism, they become ambiguity loving.

Investors' sentiment influences both their purchasing behavior and stock price performance. To investigate how market sentiment affects the returns of stocks with ambiguous or subjective information on their balance sheets, researchers compared a composite measure of investor sentiment with their subsequent performance. Highly subjective stocks were considered to be small stocks, young stocks, high volatility stocks, unprofitable stocks, non-dividend-paying stocks, extreme growth stocks, and distressed stocks. The researchers discovered that when investors are optimistic, stocks with ambiguous, uncertain and generally subjective information underperform over the subsequent period (one year). When investors are pessimistic, the opposite occurs. A contrarian approach to sentiment improves returns in more ambiguous and uncertain stocks.23

Emotion also plays a role in how investors feel about the "idea" or story behind a stock. Researchers found that emotion and imagery, aroused by the concept of a stock, bias predictions of stock performance. According to researchers, emotion and imagery may be the only judgmental bases on which individuals are able to rely when information about a financial offering is vague.24 Stocks with exciting stories cause people to forecast higher stock returns.

Interestingly, during times of greater uncertainty in the markets, when prices are volatile and trading volume high, investors tilt their portfolios toward familiar local stocks.25 Investors prefer the familiar and the known when market price action indicates potential danger. Perhaps this preference is driven by a need for certainty and comfort, which are more likely to be found closer to home.

Additionally, investors' behavioral biases are increased by uncertainty. When stocks are more difficult to value using objective accounting data, investors exhibit a stronger disposition effect. Overconfidence is increased by market-wide uncertainty.26 Even among executives, ambiguity in corporate accounting heralds the presence of behavioral biases such as loss aversion and overconfidence among management.

An interesting reversal of ambiguity aversion occurs when people feel the need to make more money than that offered in a certain outcome. In one gambling experiment, when participants' financial need was greater than the known option's expected average payout, subjects preferred the ambiguous (high volatility) option.27 People are more willing to take a chance when other options won't give them the payoff they want. During the Internet bubble, as investors and portfolio managers scrambled to catch up with their perceptions of others' wealth, they may have felt a more acute need to take a chance for profit.

This need-based pursuit of ambiguity supports loss aversion and the disposition effect. When investors are painfully losing in an investment, they would rather take an unknown chance for a comeback in the stock than sell it for a guaranteed loss.

In other scenarios, psychologists have found that people tend to believe they can have what they want, regardless of evidence to the contrary. The intensity of one's desire for an outcome increases one's estimation of the likelihood he or she will actually get it.28 Their wanting creates over-confidence in their skills, which translates into an expectation of success.

In another study, participants told to expect an outcome that conflicted with their desires gave no weight to the evidence. They remembered information that supported their ability to get what they wanted, rather than the explicit information telling them to expect otherwise.29 If investors want to believe that a money-losing Internet stock is more valuable than one with earnings, then they will value it higher, regardless of the evidence. These studies demonstrate the "confirmation bias," in which information supporting one's desires is accepted, but evidence contradicting them is ignored.


"Markets are constantly in a state of uncertainty and flux and money is made by discounting the obvious and betting on the unexpected."

—George Soros

Caltech researchers designed a monetary experiment to identify the different brain regions directing risky decision making (where probabilities are known) and uncertain decision making (where information about probabilities and outcomes is absent). As seen above, in prior experiments on risky decision making, participants were more likely to take risk when the probabilities were known versus when the odds were ambiguous.30

The researchers, including economist Colin Camerer, found that higher levels of ambiguity led to increasing activation in the amygdala (fear and memory) and orbitofrontal cortex (emotion and reason integration) and decreasing activation in the nucleus accumbens (reward). That is, participants appeared to find ambiguity fearsome, and tried to integrate their consideration of unknown risks with their prior experiences in such situations. Ambiguity caused decreased activity in the nucleus accumbens, thus implying that ambiguity arouses disappointment. However, higher-expected-value decisions (without ambiguity), led to increasing activation in the nucleus accumbens.

When patients with lesions in the orbitofrontal cortex (the excessive risk takers famously identified by Damasio) were tested in the Caltech task, they were found to be insensitive to the level of risk and ambiguity in choices. That is, they made the same decisions regardless of whether the outcome probabilities were known or not. Thus, it appears that the orbitofrontal cortex governs the integration of emotion and reason— weighing information about actual probabilities and payoffs against one's feelings about risk. Impairment of the orbitofrontal cortex leads to difficulty integrating ambiguity aversion (fear) into decision making.31


Investors are uniquely susceptible to distortions in probability assessment, especially when events are novel or rare. The market future is uncertain, and information is too often ambiguous. Emotions play a large role in biasing low probability assessments. The following is a list of conditions in which investors are vulnerable to overweighting low probabilities and becoming biased by the possibility effect:

• Vivid or easily imagined results.

• Minimal awareness about the event's likely outcomes.

• Minimal investor conditioning or experience with such events.

• Event is represented as a relatively novel/unique phenomenon.

• Desiring, wanting, or needing the outcome to occur.

• Feeling a personal or emotional stake in the outcome.

• Feeling excited or fearful about the event happening.

• Ambiguous information about the event.

While investors can try to be aware of the above conditions during their decision making, it is extremely difficult to do so. The list is long and complex, and some of the effects are subconscious and quite subtle. Nonetheless, I hope that this list will provoke some thought about how your own odds assessments might be impacted. The next sections address the problem of trust. What role does trust play in investors' interpretations of market information or interactions with clients?


When people hear ambiguous or uncertain information and must make a rapid decision, how do they assess if it is truthful? If they trust the source, then they are likely to take the information at face value.

A Securities and Exchange Commission (SEC) filing from a respected corporation is one of the most trustworthy types of corporate information, in part because a fraudulent SEC filing leads to prison time for executives. Both a history of honesty from the source and harsh consequences for misinformation are crucial support for the veracity of provided information.

Among the least believable sources of stock information are anonymous faxes, e-mails, cold calls, and Internet message boards. A promotional fax about a small gold-mining stock in Alberta is untrustworthy for obvious reasons. There are no consequences for misinformation, and one has no history of experience with the promoter.

Investment information is always somewhat suspect. Yet most investors believe dubious information if it appears under the banner of a trusted news source. Researchers in Muenster, Germany, found that subjects were more likely to believe an ambiguous headline if it was in the banner of a trusted brand-name magazine. Their trust activated the medial prefrontal cortex (MPFC).32 The MPFC is activated by emotionally satisfying events and learning about rewards. It appears that trust feels good, and people trust a news source when it agrees with their underlying political beliefs and biases.


One of the simplest ways to elicit trusting (or distrustful) behavior in economic experiments is to use a paradigm called the ultimatum game. In each round of the ultimatum game, participants have the opportunity to send money provided by an experimenter to another participant. This money is tripled, and the receiving player can return all, some, or none of this money back to the player who sent the money to him. The "proposer" is given a sum of money, for example, $10. Then the proposer offers some percentage of that $10 to a "responder," say $5. The amount of money transferred to the responder is tripled by the experimenters, so it is now $15. The responder then has an opportunity to reciprocate the transfer by sending some of the $15 back to the proposer. Obviously, some proposers and re-sponders never send money, some transfer a little, and some transfer the majority. There are tremendous individual differences in money transfers.

Money transfers are also affected by "trust signals," namely how much money one player sends to the other over each round.

In a study using the ultimatum game, some fascinating disparities among age groups and genders were noticed. In the experiment, people tended to preferentially trust others in their own age group. Young participants were significantly more trusting than older subjects, sending 24 percent more money to their partners than did older players. There was also a very strong gender effect in this study, where women were both more trusting of others and perceived to be more trustworthy than men.33

In a variation of the trust game called the "ultimatum game," proposers offer to split a sum of money with the responders. If the responders reject the proposal, then neither party receives money. In a fMRI study, respon-ders with high activation in the insula when they contemplated unfair offers were more likely to reject them.34 By refusing the unfair offer, responders were acting based on their emotions (disgust and pain) and forgoing guaranteed profits. In the markets, such a rejection of "unfair"events can be damaging. For example, after a company's negative earnings surprise, investors should suppress the urge to punish it (by selling stock) if their research shows that it continues to be fundamentally undervalued.

British fMRI researchers asked subjects to rate pictures of faces in terms of their "trustworthiness." Activity in the bilateral amygdala and right insula increased when faces judged to be untrustworthy were presented. An area of the temporal lobe was activated when trustworthy faces were viewed. The researchers noted that the temporal cortex activation implied a cognitive intention to trust, while mistrust was an automatic negative emotional reaction.35

If untrustworthy individuals have an "honest face," then they can avoid provoking immediately negative emotional assessments. Of course, intentional facial manipulation, such as former Enron CEO Ken Lay's generous smiles, can disarm the automatic negative amygdala and insula reactions that would have promoted avoidance. Additionally, now there is some indirect evidence that it is important for financial advisers to be both physically presentable and financially generous (e.g., fee discounts) with clients in order to encourage trust and prevent emotional misunderstandings.


Paul Zak, a professor at Claremont Graduate University in southern California, has taken a keen interest in the cultural role of trust in economic activity. Zak has run several studies examining the role of a hormone called oxytocin in facilitating trust. Zak first became interested in oxytocin when he read a Science magazine article describing the role of oxytocin in voles.

Voles are a genus of rodent living in North America. Prairie voles live on the Great Plains, while the mountain vole inhabits the Rocky Mountains. These two species are identical in all respects except sexual behavior. The mountain vole is very promiscuous, while the prairie vole selects one mate for life. When researchers examined the voles' hormone levels, they found that the prairie vole has very high oxytocin levels relative to its promiscuous mountain cousin.36

Oxytocin in humans facilitates bonding between mother and baby. Breast-feeding releases oxytocin in both the mother and infant. Additionally, orgasm increases a woman's oxytocin levels, perhaps allowing her to bond with her mate. Zak wondered whether oxytocin might affect trusting behavior during a "trust game" (the ultimatum game), so he set up a study to find out.

Zak found that responders who were given higher levels of money by the proposer had increased oxytocin following the transfer.37 Additionally, responders who returned more money to the proposer were found to have higher levels of oxytocin prior to reciprocating. Zak believes it was the responders' perception of trusting behavior from the Proposer that raised their oxytocin levels. Simply receiving money without implied trust did not raise oxytocin levels.38

Zak, in collaboration with Swiss researchers, found that proposers given an oxytocin medication (via intranasal squirts) were more likely to transfer a larger amount of money to responders39 at the beginning of the experiment. In fact, approximately 45 percent of the proposers given oxytocin transferred the entire initial sum, versus 20 percent of those given placebo.

Zak may have uncovered a biological mechanism (oxytocin level) for such investor biases as the endowment effect (loving a stock), the urge to reciprocate financial adviser fee discounts, confusion between loving a company and seeing its stock as a good investment, and the home bias.


Investors should calculate the expected value of every investment option before selecting the optimal choice. However, judgments of expected value are biased by risk perception. The fear of "riots" before the 2000 Mexican presidential election raised risk perceptions and sent the markets tumbling 20 percent.

Because the future is intrinsically uncertain and market dynamics change, the past is a poor guide to the future. When assessing ambiguous or uncertain information, investors overestimate the probability of danger. They often utilize the emotional defense mechanism of projection, which drives them to overestimate risk in ambiguous data due to their own fear.

Emotions such as fear also lead to the overestimation of the likelihood of low probability events that are vivid or emotionally weighted. Ego attachment to particular outcomes increases the severity of such biases. Self-awareness of risk-perception biases is challenging due to the thought distortions that arise from unconscious emotions. One will truly think that the danger exists and will search for confirming evidence to justify their fearful risk aversion.

Trust appears to be strongly related to perceptions of risk, and investors perceive less risk in situations where they trust both the information they receive and the source delivering it. In the case of unfamiliar stocks or markets, investors will usually be less trusting and will thus have greater risk perceptions. This may explain why most developed-world investors prefer local or national, rather than international, investments, even when greater returns are available overseas.

The hormone oxytocin promotes trust. Its levels rise when people are shown generosity, and it correlates with increased attachment, investment, and reciprocal behavior. On a broad national and cultural level, trust appears to correlate with market returns. Professor Paul Zak found that OECD countries with higher trust levels had higher stock market returns during his sample period (1990-2000).

0 0

Post a comment