Matt's Blog

Psychology of Behavioural Finance: Past and Future

Tue Aug 1 16:51:01 BST 2006

These are some notes I took of Professor Daniel Kahneman's 2005 Nomura Centre for Quantitative Finance Lecture. Professor Kahneman won the Nobel Prize for Economics Science 2005 (prior to this lecture). He is a psychologist by training, not an economist. Lecture was on 5th May 2005.

  • Achievements of behavioural finance:
    • Search for anomalies in standard theories eg assumption that market is perfect
    • Leads to more realistic view of human agents

Developments in behavioural finance:

  • Study done 20 years ago indicated that people's predictions of future events tend to be extreme

    • Found that investment in stocks that are long term losers led to excellent returns
    • Peopls's reaction to news is imperfect
  • Explanation to "equity-premium puzzle" (stocks return about 6% more than bonds)

    • Explanation: two assumptions
      • People are loss averse, reaction to losses is much greater than reaction to gains
      • "Narrow framing" - instead of global attitude people consider stocks one at a time, which lead to systematic errors
  • Explanation of "diposition effect"

    • if people hold portfolio with winners and losers they tend to sell winners preferentially (2:1)
    • Explanation is that people selling losers are "punishing" themselves so tend not to do it. This is a bad idea for eg tax reasons
  • (Most significant) Individual investors perform badly in financial markets

    • Can see when an investor has an "idea": they sell one stock and buy another
    • Can judge how good this idea is by comparing stock values one year later
    • On average 3.5% improvement (before transaction cost!)
    • Women do better than men because they don't have as many "ideas" (ie don't trade as much)

These things aren't supposed to happen in a perfect market, which implies that we need a more complex model. These are not small effects, individuals lose money relative to market, insitutions make money. Approximately 2.2% of GDP in Taiwanese study.

Now studying the behaviour of individuals (who make mistakes) and institutions (which exploit the mistakes of individuals).

Summary: individual investors are overconfident, non-regressive, loss averse and frame problems too narrowly.

People take risks because they don't know that they are taking them, but don't take other risks because they are loss averse (the two effects don't quite cancel).

Traders for institutions take more risks in the afternoon if they have lost in the morning.

Where from here? Three topics:

  1. Attention
  2. Emotion
  3. Intuition


Explain why same stimulus doesn't invoke same response - figuring out what attracts people's attention is important.

Low probability events are hard to consider, people tend to either ignore them or exaggerate them (ie give them an increased perception of risk).

News items attract attention of individuals, which makes them tend to buy whereas institutions don't do this.


(Decision making) When studying the mistakes that people make it is clear that emotion plays a critical role, for example emotional reactions such as unease (made a note about "DeCartes Area" book, probably refers to "Descartes' Error: Emotion, Reason, and the Human Brain" by Antonio R. Damasio).

  • Neuro-economics:

    • Study of brain processes when people make decisions
    • Conflict between emotion and reason
  • Two examples:

    • Two player game called the Alligator game.
    • Player 1 is the alligator and is given 20GBP which he must divide between himself and player 2.
    • Player 2 has two choices: accept or reject the offer. If he rejects the offer then neither player receives anything.
    • The standard prediction is that player 1 offers 1GBP and player 2 accepts (ie P1 gets 19GBP, P2 gets 1GBP)
    • In practice on average player 1 offers 10GBP, and if the offer is less than 25% of the total it is usually rejected.
    • Front part of the brain: computation, back part of the brain disgust. Can predict from relative activity.
    • Discounting
    • Can choose smaller amount to be recived soon or large amount later.
    • Immediate prize excites region in the back of the brain.
  • Examples of some surprising results:

    • Litigation of surgeons. Can predice from a phone converstation which surgeons will tend to get sued (more authorative speech tend to).
    • People are more optimistic about industries that they like, which leads to different participation of individuals and institutions in some areas of the market.


Contrast two ways that ideas come to mind:

  1. 17x24. Reason leads to going through the rules of multiplication to get the answer.
  2. "Vomit". Intuition leads to an immediate emotional response (and physiological).

Intuitive thinking is fast, associative, doesn't rely on concious rules and has the same characteristics as perception.

  • However it is:
    • susceptible to illusions
    • emotional response
    • "more primitive" (backbrain response)

Highly skilled behaviours migrate from reason to intuition eg chess, driving, tennis, (decision making?).

There are flaws in intuitive thinking, eg if offered travel insurance when you are going travelling which is either 100k GBP for death by accident or 100k GBP for death by terrorism, you are likely to pay more for the latter due to a fear response.

Intuition versus statistical rules (eg marriage counselling prediction of marriage failure)

  • statistices wins in 90% of test cases. People don't assign proper weight to factors, tend to bias towards data taken personally.
  • interviewing is detrimental to getting correct result as it makes people ignore other data, implying that it is better to let the formula decide. For example professional bank officers were asked who would default on business loans given a detailed set of information. They were beaten by a single index prediction, the ratio of debts to assets.

Why do we keep faith with intuition?

  • Sometimes get things right in a big way but also can get strong intuitive feelings when we have no expertise eg political forecasts by "experts".
  • Expertise increases confidence but not necessarily accuracy, for example there is little indication of "skill" in financial markets (but traders think that there is anyway) (see "A random walk down Wall Street" by Burton G. Malkiel).

Rapid, immediate, unequivocal feedback allows skill development, for example trading. Other forms of feedback which are slower or less equivocal do not lead to skill development, eg individual investors.


There is a psychological influence of behaviour on finance, and the argument covered in this lecture should stand the test of time.

Other notes:

  • see "Geography of (Culture or of the Mind)", can't work out to which book this referes.
  • Sidenote at the end of the lecture about the boycot of Israeli Universities that was happening when he gave the lecture: He was not happy with it, used to go to Hafir University and is an Israeli citizen. A boycot of liberal academic institutions actually helps isolationist forces within the Israeli government.


This lecture ties in well with Matt Ridley's book "The Origins of Virtue", which basically says that all creatures currently living have got here through evolution, and a common evolutionary task is the iterated prisoners' dilemma, so we have evolved to (a) identify untrustworthy "defectors" and (b) be altruistic to each other by default. The second conclusion is a bit surprising, but seems reasonable when you consider that to take advantage of someone reduces the probability of that person helping you if you are in need in the future, and up until recently it was usual for people to spend their entire lives interacting with the same group of people (there is less reason for altruism if you are constantly leaving social groups for new social groups).

This gives a good explanation of the result of the aligator game discussed in the lecture (and from memory it is discussed in the book as well). If you offer less than a "fair" split of the money there is an instinctive reaction to treat you as a defector and punish you, rather than the rational decision of taking any money offered as you will be better off. This is an instinctive reaction to behave for the good of society.



code (24)

erlang (5)
ideas (19)
lisp (1)
me (11)
notes (4)
ocaml (1)
physics (45)
qo (7)
unix (6)
vim (3)