LECTURE 1
Madott paid investors with income of new investors, now in jail.
Depending on how you present information people tend to prefer one of the graphs, because of
other graphs you show (too good to be true, you cannot beat the market if you show market return
next to returns).
Mandatory: Wall Street and the Housing Bubble, Cheng et al, 2014:
Analyze whether midlevel managers in securitized finance were aware of large-scale housing bubble
using personal home transaction data.
Introduction: once prices started rising, wishful thinking among agents in the financial sector may
have led to contagious over optimism, collective willful blindness, and groupthink (Benabou, 2013),
effect may be exacerbated by cognitive dissonance (Barberis, 2013). To enhance liquidity, widely
used near-riskless debt securities may have been designed to reduce any agent’s incentive to
acquire information about risk, thus disarming the financial system’s self-correction mechanisms
(Dang, Gorton, and Holmstrom, 2012).
Research question: were experts evil, was it bad luck, or were they biased/ignorant?
• Look at behavior of experts (=mid-level managers in securitization (ASM)).
Test: take a look at relevant private behavior of ASM: their own housing decisions and compare to
that of others of similar social class/wealth, with:
• Different information
• Similar information, but different incentives.
Possible outcomes/hypotheses:
• Evil: fully aware that large-scale housing crisis was likely and imminent, ASM knew it was a
bubble, but sold because they had an incentive to do it (financial reward, bonuses) à
popular view.
o Homeowners were avoiding losses in own homes: divest home before 2007-2009
crash, non-homeowners don’t enter market during bubble of 2004-2006 à attempt
to time own housing market.
• Unlucky: no one could have seen it coming, neither incentives nor biases played a role à
rational expectations view.
o ASM did no worse than others with similar access to information, but not in the
business.
• Ignorant: sellers did not see warning signs, because of biases (overconfidence) à big short
view.
o Expect timing of private housing decisions is no better than that of similar others
with less potential inside knowledge (no private information about housing and
securitization markets) à they didn’t know, and worse than that of outsiders with
similar knowledge (people in other business) à they could have known.
§ Similar others with less information: financial analysts, equity analysts (not
in housing), random sample of lawyers (not in real estate).
§ Others with similar access to information: real estate lawyers.
Conclusion: empirical evidence supports the bias/ignorance/not noticing view.
• Securitization managers didn’t do no better than financial analysts outside the business.
• Securitization managers did do worse than real estate lawyers.
Some groups of agents were particularly aggressive in increasing exposure to housing suggesting job
environments that foster groupthink, cognitive dissonance, or other sources of over optimism.
, Biases:
• Overconfidence: subjective confidence is greater than objective accuracy.
• Extrapolation/confirmation: if you have done things for years in a certain way and it was
always successful then you tend to believe that it will go on (in every area).
• Motivated blindness: tend up to look for information that confirm opinion. Incentive not to
look up all information, because information can be negative for you (without realizing).
• Salience effect/disaster myopia: people tend to estimate probability of something in line
with how easy they find it to imagine. If you see a car crash, you drive carefully the rest of
the day. Banks didn’t default à think probability is zero, don’t pay attention of signals if
probability is going up.
o Unfortunately, also present among policy makers.
Mandatory: behavioral evidence, DellaVigna, 2009: individuals deviate from the standard (standard
= rational) model in three respects:
• Non-standard preferences:
o Time preferences: self-control problems: ability to control oneself, in particular
emotions and desires, especially in difficult situations.
o Risk preferences: reference dependence.
o Social preferences: help explain giving to charities, striking workers to wage cuts,
giving to fund-raisers.
• Non-standard beliefs/judgment:
o Overconfidence: consumers overestimate performance in tasks requiring ability,
including precision of information.
o Law of small numbers: consumers expect small samples to exhibit large-sample
statistical properties.
o Projection bias: consumers project current preferences onto future periods.
• Non-standard decision-making:
o Limited attention: neglect (or overweighting) of information.
o Menu effects: suboptimal heuristics used for choices out of menu sets.
o Persuasion: excess impact of beliefs of others.
o Social pressure: explicit pressure by others.
o Emotions.
Behavioral economics: starts from observing what people do, not from assumption that they are
rational (prebehavioral economics). Looks for patterns in behavior. Finds consistent patterns that
are different from what prebehavioral economics assumes.
Rational: based on reasoning, doesn’t necessarily equal wise good.
Irrational: not based on reasoning, doesn’t necessarily equal stupid/bad.
1. Impossible to make people rational.
2. Undesirable to make people rational.
Mandatory: cognitive reasoning test (CRT), Shane Frederick
People with higher cognitive ability differ from those with lower cognitive ability in variety of
important and unimportant ways (live longer, earn more, have larger working memories, faster
reaction times). Decision researchers may neglect cognitive ability, because they are more
interested in average effect of experimental manipulation.
Two types of cognitive processes, dual process theory:
Madott paid investors with income of new investors, now in jail.
Depending on how you present information people tend to prefer one of the graphs, because of
other graphs you show (too good to be true, you cannot beat the market if you show market return
next to returns).
Mandatory: Wall Street and the Housing Bubble, Cheng et al, 2014:
Analyze whether midlevel managers in securitized finance were aware of large-scale housing bubble
using personal home transaction data.
Introduction: once prices started rising, wishful thinking among agents in the financial sector may
have led to contagious over optimism, collective willful blindness, and groupthink (Benabou, 2013),
effect may be exacerbated by cognitive dissonance (Barberis, 2013). To enhance liquidity, widely
used near-riskless debt securities may have been designed to reduce any agent’s incentive to
acquire information about risk, thus disarming the financial system’s self-correction mechanisms
(Dang, Gorton, and Holmstrom, 2012).
Research question: were experts evil, was it bad luck, or were they biased/ignorant?
• Look at behavior of experts (=mid-level managers in securitization (ASM)).
Test: take a look at relevant private behavior of ASM: their own housing decisions and compare to
that of others of similar social class/wealth, with:
• Different information
• Similar information, but different incentives.
Possible outcomes/hypotheses:
• Evil: fully aware that large-scale housing crisis was likely and imminent, ASM knew it was a
bubble, but sold because they had an incentive to do it (financial reward, bonuses) à
popular view.
o Homeowners were avoiding losses in own homes: divest home before 2007-2009
crash, non-homeowners don’t enter market during bubble of 2004-2006 à attempt
to time own housing market.
• Unlucky: no one could have seen it coming, neither incentives nor biases played a role à
rational expectations view.
o ASM did no worse than others with similar access to information, but not in the
business.
• Ignorant: sellers did not see warning signs, because of biases (overconfidence) à big short
view.
o Expect timing of private housing decisions is no better than that of similar others
with less potential inside knowledge (no private information about housing and
securitization markets) à they didn’t know, and worse than that of outsiders with
similar knowledge (people in other business) à they could have known.
§ Similar others with less information: financial analysts, equity analysts (not
in housing), random sample of lawyers (not in real estate).
§ Others with similar access to information: real estate lawyers.
Conclusion: empirical evidence supports the bias/ignorance/not noticing view.
• Securitization managers didn’t do no better than financial analysts outside the business.
• Securitization managers did do worse than real estate lawyers.
Some groups of agents were particularly aggressive in increasing exposure to housing suggesting job
environments that foster groupthink, cognitive dissonance, or other sources of over optimism.
, Biases:
• Overconfidence: subjective confidence is greater than objective accuracy.
• Extrapolation/confirmation: if you have done things for years in a certain way and it was
always successful then you tend to believe that it will go on (in every area).
• Motivated blindness: tend up to look for information that confirm opinion. Incentive not to
look up all information, because information can be negative for you (without realizing).
• Salience effect/disaster myopia: people tend to estimate probability of something in line
with how easy they find it to imagine. If you see a car crash, you drive carefully the rest of
the day. Banks didn’t default à think probability is zero, don’t pay attention of signals if
probability is going up.
o Unfortunately, also present among policy makers.
Mandatory: behavioral evidence, DellaVigna, 2009: individuals deviate from the standard (standard
= rational) model in three respects:
• Non-standard preferences:
o Time preferences: self-control problems: ability to control oneself, in particular
emotions and desires, especially in difficult situations.
o Risk preferences: reference dependence.
o Social preferences: help explain giving to charities, striking workers to wage cuts,
giving to fund-raisers.
• Non-standard beliefs/judgment:
o Overconfidence: consumers overestimate performance in tasks requiring ability,
including precision of information.
o Law of small numbers: consumers expect small samples to exhibit large-sample
statistical properties.
o Projection bias: consumers project current preferences onto future periods.
• Non-standard decision-making:
o Limited attention: neglect (or overweighting) of information.
o Menu effects: suboptimal heuristics used for choices out of menu sets.
o Persuasion: excess impact of beliefs of others.
o Social pressure: explicit pressure by others.
o Emotions.
Behavioral economics: starts from observing what people do, not from assumption that they are
rational (prebehavioral economics). Looks for patterns in behavior. Finds consistent patterns that
are different from what prebehavioral economics assumes.
Rational: based on reasoning, doesn’t necessarily equal wise good.
Irrational: not based on reasoning, doesn’t necessarily equal stupid/bad.
1. Impossible to make people rational.
2. Undesirable to make people rational.
Mandatory: cognitive reasoning test (CRT), Shane Frederick
People with higher cognitive ability differ from those with lower cognitive ability in variety of
important and unimportant ways (live longer, earn more, have larger working memories, faster
reaction times). Decision researchers may neglect cognitive ability, because they are more
interested in average effect of experimental manipulation.
Two types of cognitive processes, dual process theory:
