Behavioral finance FAQ / Glossary (Trend)

A    B    C    D    E    F    G-H    I-L    M    N-O    P-Q    R    S    T-U    V-Z

Full list

This is a separate page of the T-U section of the Glossary

 

Dates of related message(s) in the
Behavioral-Finance group (*):

Year/month, d: developed / discussed,
i: incidental

Trend (as market fashion or momentum)



00/5d,6d,8i - 01/1i,4i,12i - 03/6i,11
- 04/7i - 06/6i - 07/4i + etrend + see
cycle, imitation, fashion, momentum,
herding, percolation,
persistence,
rational expectations, cascades,
trend following, cycle, herding

A long walk up,

Then a long walk down.

With a few setbacks for sure,
but not exactly the mythical "random walk".
 
Why the market plays the alpinist and speleologist ?

Definition (market trend)

In asset markets, what is called a trend refers usually to a

A rather long period (up to
several years) of   price rises or falls.

As usually trends revert in the opposite direction at the end of their
course, markets experience an alternation of uptrends and
downtrends
, what is commonly called the market "cycle"

(see that word).

Also, we should not to forget "neutral trends" in which prices
lazilly oscillate around an horizontal line like small waves on the
seashore.

Price trends apply to asset returns also, just because asset holders'
returns include asset price variations.

Trends have bumps.

Those persistent market moves do not follow
straight and even roads.

They show usually some drunken behaviors:

Hairpin curves, bumps and potholes: sudden chaotic moves,

Also shorter sub-trends, micro-trends, small setbacks and zigzags
  
(see volatility) inside the big zigzags.

The trends / cycles phenomena is not limited to markets, it is seen also in
other dynamical systems.

The four attributes of a market trend

Trendy square.

Four elements are obviously important to characterize a trend:

1) Its direction.

It can be a bull trend (uptrend), a bear trend (downtrend) :

see the related glossary definitions for those market
"animals"
.

There are also neutral trends : here prices move
sideways,  yes, they stagnate, except for small vibrations
(sub-trends, daily volatility...).


2) The speed / strength of its rise or fall.

The combination "direction + speed" is commonly called the
(price) momentum (see that word).


3) Its duration, since the start.

The full duration is obviously known only ...after the trend ends,
although some mathematical tools try to gauge the persistence
of a trend (see below "trend persistence").

Statisticians have noted anyway that an asset or a market
that has shown an impressive rise or fall in the previous
year
has a high chance to do the same in the following year
(see momentum effect, bandwagon effect).

As it usually creates a growing discrepancy between prices and
"fundamental values", there is often a violent correction and trend
inversion at the end.

This is the glory and danger of momentum trading / trend
following
(see those phrases)


4) Its spread and linearity: the size of the range - some

call it a "canal" -between the lows and highs, due to sub-trends /
short volatility.

Some also see a "pivot" as the middle line between the extreme
prices forming the track.

Trends, EMH, RWH, volatility

If it is the "fair price", how come it moves on when
no breeze, no new information hits the market?

Are investors playing with a yoyo?

The existence of market price trends (and cycles) contradicts the
EMH - Efficient Market Hypothesis
:

The EMH states that stock prices move only by
    sudden
jumps or falls that happen randomly.

They would be due to the unpredictable occurrence of new
information.

It admits the exception of small moves which are not directly

linked to information falling suddenly from the blue.

It calls "volatility" those small vibrations, those zigs and zags
that are circling what should look like an equilibrium price.
This "pivot" would reflect the "fair value".

But this "EMH cum volatility" paradigm does not hold water
fully, as market trends show.

It admits also some clusters / streaks of price
    rises
or falls in statistical time series, when favorable or
    unfavorable events
haphazardly coincide

When they are not too repetitive, they are compatible with the
RWH / random walk hypothesis.

But the EMH has problems to explain
   persistent
trends in which randomness 

does not seem to apply.

Optical effects? Or reality of life?

Randomness disguised as trend?


Some theoreticians stay anchored on the idea of
  market prices   randomness and efficiency.

They consider market trends as just statistical "join the dots" optical
illusions
.

It is true that such an optical bias could easily happen.

Investors who play the technical analysis (see that word) game
     should take into account those limitations.

It could be dangerous to see trends in any alignment or zigzagging
curve
as it might be just randomness in disguise.

But long alternating trends are too clear to be just statistical
ghosts
.

Evidences are strong that trends, although not ever-present, are real
recurrent phenomena
that contradict in some way "rational" hypotheses
such as the EMH and RWH.

Let us forget dogmas, let us put the cursor somewhere between the
two approaches, using the yin yang
allegory (two principles at the
same time opposed but intertwined and supporting each other).

Therefore let us talk of the "randomness cum persistence paradox"
or "persistence cum randomness paradox",
whatever your
semantic preference.

How to explain trends?

Behavioral stuff? Reactions and fashions?

Trends seem to imply cognitive and emotional factors such as:

A 3 step gradual mental perception:

1) In the short term, there is an underreaction to
    events by most investors.

2) It is followed in the medium term by a mental
    
and behavioral adjustment.

In this phase, new information confirm those events, investor
become used to the new state of affairs.

Also they start to understand better their probable
consequences.

This can be considered the Bayesian phase (see that word).

3) In the long run comes overreaction, when herding
    and
emotions kick in.


Also, social fashions.

Like for any social "trendiness", there are times when it becomes
fashionable to buy or sell a given stock, or to buy or sell
stocks in general.

Those social phenomena grow and ebb, which makes them often
called cycle-trends (see cycle).

What is the persistence of a trend?

See trend following, persistence, momentum

 

Dates of related message(s) in the
Behavioral-Finance group (*):

Year/month, d: developed / discussed,
i: incidental

Trend following / persistence




01/3i,10i,12i - 02/4i,8i - 03/5i,8i +
see trend, momentum trading,
imitation, noise trading, technical
analysis,
persistence,cascade,
herding,  self-fulfilling prophecy,
rotation, Hurst coefficient

Adonde va Vicente? Adonde va la gente.

(Where is Vincent going? Where people are going).

Definition (trend following):

Trend following (or momentum trading or riding the trend)
is the practice of
investing in asset markets in the direction
of the current price trend
(buying when it rises, selling when
it falls) under the idea that
everybody will do the
same    and that trends  persist.

In an asset market (or an market asset segment: see bandwagon, rotation...),
once a price trend is detected
and shows enough momentum and
persistence  (*)
, trend followers start chasing the market bus.

(*) Speed and duration, to give simplified definitions

Those followers consider that, in the near future, the odds are that
the
trend / fashion would go on rather than stop (see rational
expectations).

"The trend is your friend" , as a
traders' saying goes.

They stick to the - half true / half false - impression that trends
always last longer than reasonable people would think.

Might be so, but it is hard to see beforehand the critical point
when reversal will strike.

Thus the trick is "how far can you go too far".


Also, not many people would feel easy with the idea that they
"chickened out" and deserted the battlefield, even in a strategic
retreat.

So they stay in ... until a violent correction happens, and
make them lose as much or even more than they gained before

See for example a description of the pros and cons of "bubble playing"
in the "rational bubble" article

Do trends really persist? And for how long?

Settling in, or changing places?

There is much guesswork in predicting how long and how
far a trend can persist before it breaks into another trend,
following a full cycle of underreaction - adjustment -
overreaction
(see overreaction, cycle...)

Its "persistence" (see that word, and also "Hurst coefficient") might be
usually the case
once the trend "settles in", let us say after
several months or a year
.

Real "bear markets" and "bull markets" (see those phrases, and
the "momentum" and "cycle" articles) last at least several months,
and more often several years
, (*) (**)

(*) at least in financial markets (in commodity markets the law of
      offer and demand works better
and price rises attract less

deman after a while and price falls more demand).

(**) "Long cycles", lasting for decades, have also been spotted,

but of course they include multiyear interim subtrends.

A reason of trend persistence is that trend-following
helps to prolong it, in a
feedback loop 
caused by player's mimicry, bringing a kind of
self-fulfilling prophecy
, as seen in the "cascade"
definition.


The risk for trend chasing
investors
is that, when the momentum stalls / reverts
and the trend
breaks (or if their trend identification was
false positive):

The change of orientation might look fuzzy on a chart, thus hard to
   detect
beforehand  or even  immediately. Also false starts are common.

It also can bring sudden and massive losses before the investor can
    react (liquidity squeeze, funnel effect...).

It could also be too late and ineffective to take or to adjust

precautionary actions (options, stop loss orders...)

(*) To find those messages: reach that BF group and, once there,
      1) click "messages", 2) enter your query in "search archives".

Members of the Behavioral Finance Group,
please vote on the glossary quality at
BF polls

separ

This page last update: 20/07/15  

    T-U section of the Glossary
Behavioral-Finance Gallery main page

Disclaimer / Avertissement légal