Part 9. The Behaviour Biases of Individuals Flashcards
2 forms of behaviour biases:
- Cognitive bias - corrected/eliminated through better information, education and advice.
- Emotional bias - harder to correct as stem from impulses and intuitions.
2 categories of cognitive errors:
- Belief perseverance bias = the tendency to cling to ones previously held beliefs by committing statistical, info-processing or memory errors.
e. g. conservatism, confirmation, representativeness, illusion of control, hindsight - Processing errors = describe how information may be processed and used illogically/irrationally in financial decision making.
e. g. anchoring and adjustment, mental accounting, framing and availability.
Cognitive dissonance
Belief perseverance biases result from mental discomfort that occurs when new information conflicts with previously held beliefs or cognitions.
Conservatism bias
A belief perseverance bias in which people maintain their prior views or forecasts by inadequately incorporating new, conflicting information.
- in Bayesian term - tend to overweight their prior probability of event and underweight new information, resulting in revised beliefs about probabilities and outcomes underreact to new information.
Consequences of conservatism bias:
- Maintain or be slow to update view or forecast, even when presented with new information.
- Maintain prior belief than deal with mental stress of updating beliefs given complex data – difficultly in processing new info.
How to overcome conservatism bias?
- Properly analysing and weighting new information.
- Be aware bias exists, especially for technical, abstract and/or statistical information, as cognitive cost involved in processing info is higher.
- FMPs should conduct careful analysis incorporating new info and respond appropriately.
- Updating prior beliefs in light of new information, in consistency with Bayes Rule.
- If difficult, FMPs should seek guidance in explaining how to interpret the information/explain its implications.
Confirmation bias
This refers to tendency to look for and notice what confirms prior beliefs and to ignore or undervalue whatever contradicts them.
This reflects a predisposition to justify ourselves and what we want to believe.
e.g. clients insisting on a particular investment in a portfolio, when adviser recommends otherwise.
Consequences of confirmation bias:
- considers only positive information about investment, ignoring any negative information.
- develop screening criteria while ignoring info either refutes the validity of criteria or supports other criteria.
- under-diversify portfolios - FMPs may be convinced of the value of a single company’s stock, building larger position than appropriate.
- Hold disproportionate amount of their investment assets in their employing company’s stock, as believe in their company, and convinced of its favourable prospects.
How to overcome confirmation bias?
- By actively seeking out information and challenging existing beliefs.
- To corroborate an investment decision, e.g. if investment selections are based on criteria confirming existing belief, its advisable to confirm decision from another perspective/source.
Representative bias
This refers to tendency to classify new information based on past experiences and classifications.
2 types:
Base rate neglect = a phenomenon’s rate of incidence in larger population, its base rate is neglected in favour of specific information.
Sample size neglect = FMPs incorrectly assume that small sample sizes are representative of populations.
Consequences of representative bias
- adopt a view/forecast based almost exclusively on individual, specific information or small sample.
- update beliefs using simple classifications than deal with mental stress of updating beliefs given high cognitive costs of complex data.
How to overcome representative bias?
- Ask the question “What is the probability that X (the investment under consideration) belongs to Group A (the group it resembles or considered representative of) vs Group B (the group it is statistically more likely to belong to)?
- Questions that help FMPs think whether they are failing to consider base-rate probabilities or neglecting law of small numbers, thus inaccurately assessing particular situation.
- More research to obtain base-rate information.
- More research whether to widen the sample size of observations.
Illusion of control bias
When people tend to believe that they can control or influence outcomes when in fact they cannot.
i.e. people perceived themselves as possessing more control than they did, inferred causal connections where none existed, displayed surprisingly great certainty in the predictions for outcomes of chance events.
Consequences of illusion of control:
FMPs may do the following:
- Inadequately diversify portfolios - some investors prefer to invest in companies they feel they have control over such as those they work for, leading them to hold concentrated portfolios.
- trade more than is prudent - portfolio turnover is negatively correlated with investment returns.
- construct financial models and forecasts that are overly detailed - belief forecast from models control uncertainty of investment outcomes and inherent risk.
How to overcome illusion of control bias?
- The need to recognise that investing is a probabilistic activity, with even the largest IM firms having little control over outcomes of investments they make.
- It is advisable to seek contrary viewpoints; what are the downside risks? What might go wrong?
Hindsight bias
This refers to believing past events as having been predictable and reasonable to expect, whereby outcomes that did occur are more readily evident than outcomes that did not.
e. g. people tend to remember their own prediction of future as more accurate than actually were as they are biased by knowledge of what actually occurred.
* poorly reasoned decisions with positive results are remembered as brilliant tactical moves.
* poor results of well-reasoned decisions may be described as avoidable mistakes.
Consequences of hindsight bias:
FMPs may do the following:
- Overcome the degree to which they correctly predicted an investment outcome, or predictability of an outcome generally, related to overconfidence bias.
- Unfairly assess money manager or security performance, where performance is compared against what has happened as opposed to expectations at time investment was made.
How to overcome hindsight bias?
- FMPs should carefully record their investment decisions, and key reasons for making those decisions in writing at or around time decision is made.
- Consulting written records than memory will often produce a far more accurate examination of past decisions.
Processing errors
This refers to information being processed and used illogically or irrationally.
These are less related to errors of memory or in assigning and updating probabilities, rather relates more closely to flaws in how information itself is processed.
Anchoring and adjustment bias
This refers to relying on an initial piece of information to make subsequent estimates, judgements and decisions.
Estimating a value with unknown magnitude, people begin with initial default number (anchor) which they adjust up or down.
Regardless of how initial anchor was chosen, people tend to adjust anchors insufficiently, producing end approximations that are consequently biased.
Consequences of anchoring and adjustment bias:
- FMPs may stick too closely to their original estimates when learning new information, that potentially has downside and upside risk.
How to overcome anchoring and adjustment bias?
- Ask questions “Am I holding onto this stock based on rational analysis, or am I trying to attain a price that I am anchored to such as purchase price or high water mark?
- Awareness that company’s revenues and earnings given period reflect conditions in that period.
- Awareness security prices reflect investors perception of the future at given point in time, a given investor cost basis, past market levels etc.
- FMP should look at basis for any investment recommendation to see whether it is anchored to previous estimates or some default number.
Mental accounting bias
This refers to mentally dividing money into “accounts” that influence decisions, even though money is fungible.
Rather that FMPs considering their entire portfolio holistically in risk-return context, investors construct portfolios layered pyramid format, with each layer addressing a specific financial goal.
Consequences of mental accounting bias:
FMPs may do the following:
- Neglect opportunities to reduce risk by combining assets with low correlations, as offsetting positions across various portfolio layers/mental accounts can lead to suboptimal aggregate performance.
- Irrationally distinguish between returns derived from income, and those from capital appreciation - focus chasing income streams unwittingly risking principal in process.
- Irrationally bifurcate wealth/portfolio into investment principal and investment returns - belief greater risk can be taken with returns than principal initially contributed.