Behavioral finance definitions

Plan of the whole chapter		See also
Introduction	Main concepts: BF vs. EMH [see also abstract (slides)]	Behavioral-Finance Gallery 500+ keywords BF glossary and 1800+ members informative BF forum
Part A	Individual and social behavioral biases
Part B (below)	Economic and financial incidences

Part B : Economic and financial incidences

The introduction page of this section warned about all sorts of permanent and temporary holes in the fabric of the EMH

(efficient market hypothesis) applied to capital markets:

Here is a list of the major market anomalies:

Overtrading (bringing excessive volume and volatility) are inefficiency signals.

If market where efficient, prices would adjust instantly, leaving no arbitrage opportunity.

Price trends, momentums, entailing bifurcations / alternations / cycles (although these are partly physical reactions,

their main cause is changes in beliefs),
With sometimes critical thresholds where the price behavior can change completely, a phenomenon called percolation.

Possible trend disruptions (acceleration, reversal),
    or even a sudden illiquidity (lack of buyers facing a seller rush, or the other way round

 Price reactions to information are crucial for market behavior.

The influence of past / recent memory - neglected by the EMH - and of the other cognitive and emotional biases
shown above, affects those reactions.

Thus, we get misreactions (over- / under-reactions), which are among the causes of:

Trends / momentums / fads,

Mispricing (over-pricing, under-pricing),

 

The statistical distributions of returns, prices, volatility, can differ from the standard Gauss-Laplace's "bell curve"

(narrow dispersion of values, symmetrically close to an average) by showing sometimes deviations from that "random law":

Asymmetries (skew): faster or longer or larger rises than drops, or drops than rises.   

Fat tails: prices or price variations spreading farther from the mean than is considered proper and well-behaved

   in the good old standard bell curve.

Pareto's "power law": concentration on extreme values, not on a central one, an alternation of periods

   of very high / very low prices, returns or volatility.

Congestions / clusters,

Heteroskedasticity (changes of volatility),

  Feedback loops (positive feedback, reflexivity)

A positive feedback (vicious circle) happens when the results of a previous action, accelerate the drift (addition)

instead of leading to a counter-action that stabilizes the system (negative feedback, subtraction) and makes prices

oscillate softly.

In extreme positive feedback cases, when the change in belief is long overdue, it results in bubbles / crashes.

Sometimes also the capital market evolutions can influence the economic fundamentals (reflexivity)

Even professional analysts get more optimistic / pessimistic after prices go up / down. 

They adapt their earnings consensus accordingly.

That revised consensus makes prices go up / down again.

Something that make them raise / lower again their earnings expectations, and so on.

In some cases, it could be that private information (or superior research ?) feeds the momentum.

But often the real reason seems that an analyst alone don't find it wise to issue a buy or sell
recommendation that goes against the current trend, and thus to take the risk that its results will
differ (in the bad direction) from those of the "crowd" of all other analysts.

Investors profiling. Investors (as well as each stock and other asset), can be classified in types / categories.

We can also say in clans, if we take into account the peer influence.

Each type has its own type of behavior or investment / trading style.

Stocks profiling, starts also by identifying types / categories of stocks behavior, and to see for each stock
     how closely it may fit one (or several) of them.

Stock profiling brings the notion of stock image coefficient, my operational concept described      in this site. That coefficient is obtained by dividing the stock price by its "EEV - Estimated Economic Value". (note that EEVs are a bit delicate to calculate objectively, as based on future earnings' projections). This site shows how to estimate stock price potentials by multiplying the EEV by the potential images.

Technical analysis (TA) supposes that markets have memory.

If so, past prices, or the current price momentum, can give an idea of the future price evolution.

TA is a tool to detect if a trend (and thus the investor's behavior) will persist or break.

TA gives some results but can be deceptive as it relies mostly on graphic signals that are often intertwined,
unclear or belated.

It might become a source of representativeness heuristic (spotting patterns where there are none).

Quantitative analysis (QA) is based on statistical analysis via probabilistic models.

It works in ordinary circumstances but can completely fail when it has to deal with uncharted territories

(rare events, totally new situations).

Nonlinear pricing / decision models, a field with more and more active research.

It take into accounts that sudden price changes can break trends.

The trick is to simulate the competing behavior or various category of investors (agent-based models),
or the whole market behavior.

It uses such tools as dynamical systems models (such systems behave in a more complex, and at the same time
more determinist way, than the random walk), and soft computing based on chaos theory, fractals, fuzzy logic,
neural nets, genetic algorithms... (see reader's contributions on those topics)

Value at risk, extreme risks calculations.

The real risk is not the standard price or return deviation used by the EMH, but the extreme risk that can cause

total ruin, or that is deemed not acceptable.

The above concepts of fat tails and heteroskedasticity, leading to the so called ARCH and GARCH models,

are useful to measure it.

Types of probabilities: statistical, subjective, Bayesian, imprecise (fuzzy logic), etc.

The (expected) utility concept (leading to the risk premium)

The game theory (i.e. the prisoner dilemma) supposes rational players behaviors.

Some might find similarities between the stock market and a game with money at stake, either a positive game

(more gains than losses), or a negative one (more losses).

But this is not completely relevant as the game theory considers usually the actions of only a few players.

Anyway, there are two main categories of game:

Zero-sum games, which are most of the time competitive (grab the biggest share of a limited resource),

Positive sum games, which are sometimes cooperative (make a bigger cake by cooperating).

 Experimental economy / finance, you know, the rats in the labyrinth thing in vitro or in vivo observation

of players' decisions / actions, whether alone or within groups.

This goes farther than the game theory as it shows the real behavior and its monetary consequences.

Behavioral finance draws largely from those experiments (together with other techniques, such as the analysis

of market statistics). Some examples are:

The "paradoxes" seen in the probabilities, utility and decisions page,

The prospect theory (see above): people think it worse to have lost a few money than not to have gained a bigger amount of money.

Experiments in mimicry, in collective utility and so on...

Neuroeconomics, a division of neurosciences, adds a new dimension to those experiments.

By scanning the brain it detects which of its areas correspond to specific emotions (pleasure, suffering) or on the contrary

to cognitive functions (knowledge, logic...) and are at play in money-related decision making.

Marketing techniques applied to the creation and distribution of financial (= assets / liabilities) products:

  customers segmentation, types, motivations, profiles (see above).

BF/BE are still not considered fully scientific and practical fields. They have some weaknesses, such as their
overemphasis on:

* Anomalies compared to standard finance.

BF/BE concepts refer quasi-exclusively to mental deviations. Therapeutic purpose or witch hunting?

Those references limit the main criteria of "normality" in finance (and economics) to expected monetary
returns and risks.

This contributes to consider as a "market anomaly" any divergence from those two criteria.

This seems too obvious to fit realities fully.

- What about uncertainty as a broader concept than statistical risk?

- More important, what about non monetary returns, that might also have their degree of legitimacy?

Such "soft" returns are empathy, fun, power, human challenge, search for experience and knowledge
and myriads of other goals.

* Reactions to events / information (underreaction, overreaction...).

What about the observation of the players' behavior when there is a lack of new relevant events,
just "noise" ?

There are still many things to observe and study in those fields, and more generally in what social sciences see as the
reasons, processes and effects of what is called "decision making".

We can have doubts that this would ever be a fully predictive science or technique (uncertainties will not
disappear), other than helping to find the range of possible scenarios between which deciders might choose.

Maybe we will never end to learn more and more about the human being and about human societies,

a quest that started thousands of years ago!

Other sections of the chapter		See also
Introduction	BF vs. EMH	Behavioral-Finance Gallery (*) 500+ keywords BF glossary and 1800+ members informative BF forum
Part A	Individual and social behavioral biases

(*) The place where you find the full definitions of the various BF phenomena

click scam