Behavioral finance FAQ / Glossary (Anomaly)
This is a separate page of the A section of the Glossary
Dates of related message(s) in the Behavioral-Finance group (*):
Year/month, d: developed / discussed, i: incidental
(market) Anomaly
01/5i - 02/7i - 04/2i,5d,6i - 05/5i,6i + see behavioral biases,
price anomaly, EMH, inefficiency, collective bias, fair value,
asymmetry.
Something wrong with the price if compared to the norm?
Or something wrong with the norm?
Definition (general): An anomaly is as a state of things that does not fit an usual norm.
Definition (finance):
What is commonly called a
financial market
anomaly (*) is
an asset price or return (**) that
diverges from some standard paradigm.
Typically, such a discrepancy contradicts the EMH / Efficient Market Hypothesis (see that phrase).
(*) Also called inefficiency (see that word) or imperfection, or
distortion
or also, as a more neutral and factual world, effect (see also that word).
That last appellation tries not to give opinion on how weird
either the phenomenon or the paradigm might be.
(**) Anomalous price rises or falls translate into anomalous (negative or positive) returns.
The main component of an asset return is its price change
(except maybe for bonds or, in some degree, for real estate).
What are the usual criteria to call "anomalous" a
financial situation?
Unrealistic ...realities?
Or abnormal ...norms?
There is a paradox in qualifying real situations as anomalous,
and even more if those purported anomalies are frequent.
That ambiguity might come from the criteria used.
What are usually considered as market anomalies are
discrepancies / divergences between current market realities and:
What is given as the "fair value" (*), or "fair return"
by theoretical economic / financial
models
designed to take into account the assets' "fundamentals".
In the economic field, the
optimal allocation of resources.
(*) At least in the economic sense of that phrase (see the "value" articles in the glossary),
as there is some semantic ambiguity: "fair value" is also used as an accounting term
based mostly on the market price.
The EMH / RWH references, the theoretician's shop window.
In financial asset markets, the usual reference is the EMH / Efficient market hypothesis-based models
(see "efficient market hypothesis").
Anomalies are then defined as:
Differences of returns - and the related mispricing (see price anomaly)
when comparing those market returns and prices
to what models based on the EMH (for example the CAPM) predict.
Another common reference (closely related to the EMH)
is the
RWH / Random walk hypothesis .
It states that things evolves according to a normal random distribution law (see distribution).
The flaw is that, in economics and finance, evolutions do not fully follow such laws.
This is shown in data clusters, fat tails, asymmetries, percolation thresholds, cumulative bifurcations
and other phenomena explained in this glossary.
Well, those various discrepancies can here also be called anomalies,
if we take that random hypothesis as fully valid.
To sum it up, are those criteria fully relevant to spot anomalies?
Can they help you
detect, on the market crime scene,
evidences of financial anomalies and find a use for them?
Anyway, in reality, anomalies, unless blatant, are hard to define, to spot and to
measure (*).
This is because the theoretical models used as references (**)
might be themselves ...anomalous, at least partially.
(*) And therefore hard to be used for one's advantage!
They do not always help to detect anomalies with certainty
so as not to invest in overpriced assets (see overpricing).
It is even less easy to correct them to find what is the "true" value
(**) Even behavioral finance refers to the EMH instead or looking for its own valuation theory
(with a few exceptions: see "behavioral pricing").
This can be considered as one of weaknesses of BF in its present state.
Are return and price anomalies common or rare? Temporary
or persistent?
Just occasional market gaffes?
Or the ordinary and repetitive market way of life?
The idea behind the word "anomaly" is that the famous "invisible hand" suffers occasional clumsiness.
But the question arises about those purported gaffes: Occasional? Or common and recurrent?
1) Occasional anomalies (short term anomalies)
In some periods, or in some types of assets, anomalies are rare and/or get automatically and quickly corrected,
as soon as some traders see an opportunity of
"arbitrage" (see that word).
Those mild and transient discrepancies do not contradict fundamentally the EMH paradigm behind the model.
They also fit what most of those models consider as simple "volatility" (see that word).
without much explanation about its origin,
2) Persistent anomalies (long term anomalies)
Some anomalies are recurrent and / or persistent,
or are corrected in ...anomalous ways.
In those cases, which seem rather frequent and widespread, some distorting phenomena
researched by behavioral finance are at play. This includes:
Some so-called persistent "market effects" (see P/E effect for example),
Many market trends
(uptrends, downtrends), which durability raises
questions about the quality of previous - or future - anticipations,
All
extreme price evolutions (bubbles, crashes), showing exaggeration
(positive feedback) instead of correction / regulation (negative feedback)
Some price stickiness
, on the other hand,
Various other phenomena due to imperfect markets, or to individual and collective /
cognitive and emotional "biases" (see below),
Behavioral finance tries to describe, and as far as possible to explain those phenomena
that diverge, in a persistent way, from random distribution laws (see distribution)
In reality, anomalies seem to be all over the place.
Sometimes we might wonder if a lack of anomalies
would not be ...anomalous.
Thus, the EMH might be a theory that only describes idealized, nearly never reached, cases.
If total efficiency / inefficiency happens only in borderline cases,
we should talk about "degrees
of efficiency".
In that case, the EMH is a tentative approximation that reflects and explains the market only in some degree.
What causes market anomalies?
Officer, what was wrong? The map? The car? The driver?
Anomalies, defined again as market deviations from standard theoretical value, seem to have
three main causes
:
![]()
1) Imperfect market structures.
Clogged pipes
In general, nearly all markets are imperfect, with minor or big built in obstacles
that prevent a "perfect competition".
In none of them, the law of supply and demand works without any glitch.
None really has the ideal structure which meets the conditions that textbooks
exact of "perfectly competitive" markets.
But some markets, get nearer that blissful state than others.
* Organized financial assets markets, the big stock exchanges for example,
are said to be closer to perfect competition than other economic markets.
* The main commodity markets are also considered as such.
But even so, all those markets have some structural imperfections.
Those markets' main structural imperfection lies in
"asymmetries", the fact that investors have
different information and different relative financial strengths.
Examples: big hands vs. little hands, institutions vs. the general public, etc.
Another imperfection can be a lack of transaction volume and liquidity.
![]()
2) Collective / contagious behavioral biases.
Blind crowds
One or a few person's behavioral biases cannot really change the market course.
The incidences on markets becomes sizeable only
when those biases become collective, and affect
a large or powerful enough mass of investors or traders.
Those behavioral biases (see that phrase), whether cognitive or emotional, individual or collective,
are the bread and butter of this glossary.
Herding is the most typical of those collective biases that distort market prices.
Well, the words anomaly and bias are sometimes interchangeable.
You are also welcome to use "flaws", "blunders" or more picturesque terms.
![]()
3) Not to be neglected, oversimplifications
in the standard market theories used as the comparison basis.
Kids in the lab
Those theories and models, whatever the Greek letters
on the blackboard, might be more or less biased:
In the best cases by poorly realistic assumptions,
In the worst cases by framing, reductionism, representativeness heuristics (see those phrases)
or whatever cognitive bias.
Yes, models themselves might be built on cognitive biases. See "model".
(*) To find those messages: reach that Behavioral-Finance group and, once you are there, 1) click "messages", 2) enter your query in "search archives".
Members of the Behavioral Finance Group, please vote on the glossary quality at Behavioral-Finance/polls
This page last update: 16/11/11
Disclaimer / Avertissement légal
![]()
A section of the Glossary
![]()
Behavioral-Finance Gallery main page