Behavioral Finance

1
Table of content

Table of content .............................................................................................................. 2

Chapter 1 : Bounded Rationality & Heuristics ................................................................. 7
Section 1 : The Conceptual Product ................................................................................. 7
1.1. Conceptual Introduction ........................................................................................ 7
1.2. Bounded Rationality: Understanding the Mechanism ............................................ 7
1.3. Heuristics and Biases: A Logic of Simplification .................................................. 7
1.4. Application: Analysis of a Financial Bubble ......................................................... 7
Section 2 : Les biais et leurs mécanismes......................................................................... 7
2.1. Comprendre les biais comme des mécanismes, pas des erreurs.............................. 7
2.2. Overconfidence: A Progressive Construction ........................................................ 8
2.3. Illusion of Control: Giving Meaning to Randomness............................................. 8
2.4. Heuristics: Between Efficiency and Distortion4 .................................................... 9
2.4.1 Representativeness .......................................................................................... 9
2.4.2 Availability ..................................................................................................... 9
2.5. Reading a Case as a Researcher ............................................................................ 9
Level 1: Observation................................................................................................ 9
Level 2: Cognitive interpretation ............................................................................. 9
Level 3: Economic consequences ........................................................................... 10
2.6. Linking Biases Together ..................................................................................... 10
2.7. Position critique .................................................................................................. 10
• Classical approach (EMH) ................................................................................... 10
• Behavioral approach ............................................................................................ 10
Section 3: Understanding the Deeper Logic of Biases in Times of Crisis ....................... 11
Chapter 2: Prospect Theory ........................................................................................... 14
2.1. Why Prospect Theory is a Breakthrough ............................................................. 14
2.2. The Value Function: A Formalization of Psychology .......................................... 14
2.3. Understanding λ (Lambda) as a Behavioral Principle .......................................... 14
2.4. Probability Weighting: Distorting the World ....................................................... 15
2.5. Understanding a Decision as a Combination of Two Distortions ......................... 15
2.6. The Lottery Effect: A Logical Consequence, Not an Anomaly ............................ 15
2.7. What Researchers Must Truly Understand .......................................................... 16

2
 2.8. Moving to an Advanced Level ............................................................................ 16
Chapter 3: Market Anomalies as Evidence Against Pure Rationality ............................. 17
3.1. Why Anomalies Are Not Exceptions................................................................... 17
3.2. Momentum: Collective Cognitive Slowness ........................................................ 17
3.3. Bubbles: When Beliefs Replace Fundamentals.................................................... 17
3.4. Calendar Anomalies: Psychology Over Time ...................................................... 18
3.5. The Case of Entrepreneurs: A Different Rationality ............................................ 18
3.6. The 2007–2008 Crisis: A System of Biases ......................................................... 19
3.7. Confirmation Bias: An Institutional Bias............................................................. 19
3.8. Moving to Advanced Reasoning ......................................................................... 19
3.9. Critical Perspective ............................................................................................. 20
Chapter 4: Behavioral Portfolio Management as a Departure from Classical Optimization
...................................................................................................................................... 21
4.1. Why Classical Portfolio Theory Fails Empirically............................................... 21
4.2. The Portfolio as an Extension of Identity ............................................................ 21
4.3. The Disposition Effect: A Time-Biased Management ......................................... 21
4.4. Familiarity Bias: An Illusion of Control .............................................................. 22
4.5. The Real Portfolio: An Unstable Compromise .................................................... 22
4.6. The Barberis et al. Model: Linking Psychology and Prices .................................. 23
4.7. Risk: A Perception, Not a Given ......................................................................... 23
4.8. Toward Behavioral Portfolio Management .......................................................... 23
4.8.1 Corrective Approach ..................................................................................... 24
4.8.2 Behavioral Approach .................................................................................... 24
4.9. Expected Reasoning ............................................................................................ 24
Chapter 5: Individual Investors in Crisis: From Perceptions to Behavior ....................... 25
5.1. Why Crises Are Privileged Moments for Analysis .............................................. 25
5.2. A Central Question: How to Link Perception and Action .................................... 25
5.3. A Major Break: Perception ≠ Behavior ............................................................... 25
5.4. Understanding This Divergence .......................................................................... 26
5.5. The Central Role of Disagreement ...................................................................... 26
5.6. Rethinking the Link Between Information and Decision ..................................... 26
5.7. The Problem of Measuring Risk.......................................................................... 27
5.8. Learning and the Limits of Experience ................................................................ 27
Chapter 6: Social Influence, Media, and Market Dynamics: Understanding the GameStop
Episode ......................................................................................................................... 28

3
 6.1. A Fundamental Transformation: From the Market to the Community ................. 28
6.2. Attention as an Economic Resource .................................................................... 28
6.3. Coordination Without Centralization................................................................... 28
6.4. The Role of Norms and Identity .......................................................................... 29
6.5. From Discussion to Action: A Phased Dynamic .................................................. 29
6.6. Anticipation or Causality?................................................................................... 29
6.7. The Role of Market Microstructure ..................................................................... 30
6.8. Reading GameStop as a Case of Advanced Behavioral Finance .......................... 30
6.9. Moving Toward Research Reasoning .................................................................. 30
Chapter 7: Herding, FOMO, and the Social Construction of Market Behavior ............... 32
7.1. The Paradox of Herding: Individually Rational, Collectively Inefficient ............. 32
7.2. The Fragility of Cascades ................................................................................... 32
7.3. Herding or Mere Correlation? ............................................................................. 32
7.4. FOMO: An Emotional Transformation of Decision-Making ............................... 33
7.5. FOMO and Risk-Taking ..................................................................................... 33
7.6. Herding and FOMO: Two Complementary Mechanisms ..................................... 33
7.7. The Role of Platforms and Influencers ................................................................ 34
7.8. Measuring Herding and FOMO........................................................................... 34
7.9. From Narrative to Research ................................................................................ 35
Chapter 8: Emotions, Sentiment, and the Quantification of Market Behavior ................. 36
8.1. A Shift in Perspective: From Psychology to Measurement .................................. 36
8.2. “Risk as Feelings”: Another Logic of Decision-Making ...................................... 36
8.3. Fear: A Dynamic of Protection ........................................................................... 36
8.4. Greed: An Expansion of Risk .............................................................................. 37
8.5. Surprise: An Informational Shock ....................................................................... 37
8.6. Building an Index: A Theoretical as Much as an Empirical Act ........................... 37
8.7. The Fundamental Problem: Endogeneity ............................................................. 38
8.8. What Empirical Studies Show ............................................................................. 38
8.9. From Signal to Interpretation .............................................................................. 38
8.10. Moving Toward Research Reasoning ................................................................ 39
8.11. Final Synthesis: From Individual Emotions to Market Dynamics ...................... 39
Chapter 9: Fintech, Artificial Intelligence, and the Transformation of Behavioral Finance
...................................................................................................................................... 40
9.1. An Initial Illusion: Technology as a Solution to Irrationality ............................... 40
9.2. HFT: Speed Without Understanding ................................................................... 40
4
 9.3. The Robo-Advisor: Rationality or Rigidity? ........................................................ 40
9.4. AI and Sentiment: Capturing Emotion… or Noise? ............................................. 41
9.5. The Major Risk: The Algorithmic Loop .............................................................. 41
9.6. The Illusion of Algorithmic Performance ............................................................ 42
9.7. The Central Question: Who Controls the Decision? ............................................ 42
9.8. Toward an “Augmented” Behavioral Finance ..................................................... 42
9.9. Moving Toward Advanced Reasoning ................................................................ 43
9.10. Final Synthesis: From the Human to the Algorithm ........................................... 43
Chapter 10: Behavioral Regulation and Nudges: Designing Markets for Real Humans .. 44
10.1. A Break in the Philosophy of Regulation .......................................................... 44
10.2. The Nudge: A Minimal but Structuring Intervention ......................................... 44
10.3. The Paradox of Libertarian Paternalism ............................................................ 44
10.4. The Central Role of Biases in Policy Design ..................................................... 45
10.5. Automatic Enrollment: A Fundamental Example .............................................. 45
10.6. The Prospectus: Regulation Through Form ....................................................... 45
10.7. The Fundamental Problem: The Effectiveness of Nudges .................................. 46
10.8. The Limits of Behavioral Regulation ................................................................ 46
10.9. The Ethical Question: Manipulation or Support? ............................................... 46
10.10. Toward an Economy of “Assisted Decision-Making”...................................... 47
10.11. Moving Toward Research Reasoning .............................................................. 47
10.12. Final Synthesis: From the Individual to the Institution .................................... 47
Chapter 11: Recent Trends and Research Directions: From Knowledge to Inquiry......... 49
11.1. A Transformation in the Role of the researcher ................................................. 49
11.2. The “Third Generation”: Expanding the Scope of Analysis ............................... 49
11.3. Neurofinance: Understanding the Biological Mechanism of Decision-Making .. 49
11.4. Big Data and Real-Time Behavior .................................................................... 50
11.5. AI and the Transformation of Behavior ............................................................. 50
11.6. ESG: The Introduction of Values into Financial Decision-Making .................... 50
11.7. The Central Challenge: Transforming an Idea into a Research Question............ 51
11.8. Designing an Experimental Protocol ................................................................. 51
11.9. The Limits of Behavioral Research ................................................................... 51
11.10. The Role of Interdisciplinarity ........................................................................ 52
11.11. Transition to the Research Level ..................................................................... 52

5
6
Chapter 1 : Bounded Rationality & Heuristics

Section 1 : The Conceptual Product

1.1. Conceptual Introduction

Classical finance assumes that agents are perfectly rational and capable of processing all
available information. However, research in behavioral finance has shown that this
assumption is often unrealistic. Individuals make decisions in an uncertain environment,
with limited cognitive capacities, which leads to systematically biased behavior.

1.2. Bounded Rationality: Understanding the Mechanism

Bounded rationality, introduced by Herbert Simon, implies that individuals do not seek to
optimize their decisions but rather to achieve a satisfactory level of outcome. This approach
is based on two fundamental constraints: the limitation of available information and the
limitation of cognitive processing capacities. As a result, financial decisions are often made
on the basis of mental simplifications.

1.3. Heuristics and Biases: A Logic of Simplification

Heuristics are fast mental strategies that enable decision-making without a complete
analysis. They are efficient but can lead to systematic errors known as cognitive biases.
For example, an investor may overestimate the probability of a recent event (availability
heuristic) or extrapolate a past trend (representativeness).

1.4. Application: Analysis of a Financial Bubble

Speculative bubbles perfectly illustrate bounded rationality. In the case of the Tulip Mania
or the 1929 crash, investors followed collective behaviors amplified by psychological
biases. Rising prices reinforce the belief in their continued increase, creating a self-
sustaining dynamic.

Section 2 : Les biais et leurs mécanismes

2.1. Comprendre les biais comme des mécanismes, pas des erreurs
In classical approaches, investors’ errors are often viewed as isolated anomalies.
Behavioral finance adopts a different perspective: it considers that these “errors” are in fact
predictable, systematic, and structured.

Overconfidence, the illusion of control, and heuristics are not accidents.

They constitute regular cognitive mechanisms arising from the normal functioning of the
human brain.

The objective, therefore, is not merely to identify them, but to understand:

7
• The conditions under which they emerge

• How they shape decision-making

• Why they persist even among experts

2.2. Overconfidence: A Progressive Construction

Overconfidence does not arise suddenly. It is often built from an initial positive experience.

An investor who experiences a series of successes:

• Attributes these results to their own skill (rather than to chance)

• Gradually reduces their level of doubt

• Increases their exposure to risk

This process creates a cumulative dynamic: success → confidence → risk-taking →

exposure → amplified error

This mechanism explains why overconfidence is particularly dangerous:

it does not produce losses immediately—it prepares them.

Empirical evidence confirms this phenomenon: the most confident investors are often those
who trade the most and perform the worst.

2.3. Illusion of Control: Giving Meaning to Randomness

The illusion of control is based on a fundamental need: to reduce uncertainty by giving
meaning to events.

Faced with complex and uncertain markets, individuals:

• Search for patterns

• Interpret coincidences as causal relationships

• Develop “personal methods”

A trader who believes they “understand the market” after a few successes is not irrational
in the strict sense.

They are engaged in a natural cognitive process: transforming randomness into a

comprehensible system.

The problem arises when this interpretation:

• Becomes rigid

8
• Is no longer questioned

• Leads to the disregard of contradictory information

2.4. Heuristics: Between Efficiency and Distortion4

Heuristics are not errors. They are adaptive solutions to a complex environment. However,
their effectiveness relies on an implicit condition:

→ that the environment is stable and repetitive.

In financial markets, this condition is rarely met.

2.4.1 Representativeness
The investor extrapolates:

• “This company resembles a success story”

• “This trend resembles sustained growth”

They replace an objective probability with a mental resemblance.

2.4.2 Availability
The investor overweights:

• Recent events

• Salient events

• Highly publicized events

They replace an objective probability with ease of recall.

2.5. Reading a Case as a Researcher

When analyzing a case (manager, trader, or market), we should avoid a purely descriptive
reading. Instead, we should adopt a three-level analytical framework:

Level 1: Observation
What does the agent do?

• Trading frequency

• Reaction to information

• Level of risk

Level 2: Cognitive interpretation

Which bias is at play?

9
 • Overconfidence?

• Availability?

• Illusion of control?

Level 3: Economic consequences

What is the impact?

• Market inefficiency?

• Misallocation of capital?

• Excessive volatility?

It is this third stage that distinguishes a good researcher from an excellent one.

2.6. Linking Biases Together

A key point (often overlooked): biases do not operate in isolation.

A typical example:

• Overconfidence → increases risk-taking

• Illusion of control → justifies this risk-taking

• Representativeness heuristic → reinforces the conviction

Result: a decision that appears coherent… but is deeply biased.

2.7. Position critique

At this stage, we must be able to address a more challenging question:

→ Do these biases make markets persistently inefficient?

Two perspectives exist:

• Classical approach (EMH)

Biases exist but are corrected by the market

• Behavioral approach
Biases are collective → they can durably distort prices

Our task is not to choose one answer, but to argue using mechanisms and examples.

10
Section 3: Understanding the Deeper Logic of Biases in Times of Crisis

3.1. Why crises reveal biases

Cognitive biases become truly visible only under certain conditions.
Periods of crisis, such as 2008, play a particular role: they disrupt the usual points of
reference.
Under normal conditions, individual errors may remain invisible.
Under extreme conditions, they become massive, synchronized, and amplified.

This point is essential: → biases are not merely individual; they become collective.

It is this collective dimension that transforms psychological errors into market phenomena.

3.2. The disposition effect: a coherent emotional logic

At first glance, selling winners and holding losers appears irrational.
Yet from a psychological perspective, this behavior is perfectly coherent.

When an investor sells a winning asset:

• They validate a gain
• They experience immediate satisfaction

When they sell a losing asset:

• They materialize an error
• They suffer a strong psychological loss

Thus, the decision is not guided by expected profitability, but by a different logic:
→ minimizing immediate psychological pain

This mechanism explains why the disposition effect persists despite its documented
economic costs.

3.3. Loss aversion: A fundamental asymmetry

Loss aversion is not simply one bias among others. It is a fundamental structure of human
decision-making.

It introduces an asymmetry:
• Gains → moderate satisfaction
• Losses → intense pain

This asymmetry completely transforms decision-making: an investor no longer seeks to

maximize an expected gain. Instead, they seek to avoid a certain loss, even at the cost of
taking greater risk.

11
It is here that classical rationality fails:
preferences are not stable; they depend on the reference point.

3.4. Mental accounting: fragmenting in order to cope

Mental accounting allows individuals to manage complexity, but at the cost of economic
inconsistency.

An investor does not reason in terms of their total wealth. Instead, they divide it into
categories:
• Savings
• Investment
• Security
• Speculation

Each category becomes a distinct psychological space.

This mechanism serves a function:

→ making decisions emotionally bearable

But it also has a consequence:

→ it prevents the portfolio from being optimized as a whole

Thus, a loss in one “mental account” may be tolerated or rejected independently of the
overall situation.

3.5. Cognitive dissonance: protecting the coherence of the self

Cognitive dissonance arises when:
• A past decision
• Comes into contradiction with new information

In this case, two options exist:

1. Acknowledge the error

2. Modify the interpretation of the information

In most cases, individuals choose the second.

The investor does not say: “I was wrong.”

They say:
“The market is wrong”
or
“The situation will turn around.”

12
This mechanism protects personal identity and coherence, but it prevents:
• Learning
• Rational adjustment

3.6. Reading a crisis as a system of biases

A researcher should not analyze biases separately. A crisis must be read as an interaction
of mechanisms:
• Loss aversion → refusal to sell
• Disposition effect → misallocation
• Mental accounting → incoherent decisions
• Cognitive dissonance → persistence of error

Together, these biases create a dynamic:

→ inertia + poor adaptation + amplification of losses

3.7. Moving toward advanced reasoning

At this level, the question is no longer: “Which bias is present?”
but rather:
“How do these biases produce market dynamics?”

Examples of expected reasoning:

• Why does the disposition effect slow down price correction?
• How can loss aversion amplify a crisis?
• In what way does cognitive dissonance delay investor capitulation?

It is this ability to move from individual behavior → aggregated phenomenon that is

expected.

3.8. Critical perspective

We must be able to discuss a central tension:
• If all investors are biased → persistent inefficiency
• If some are rational → arbitrage is possible

So:
Are biases sufficient to explain crises?

Or must we also incorporate:

• Institutional constraints
• Information asymmetries
• Leverage effects

Our role is to construct an argued response, not merely to recite the course.
13
Chapter 2: Prospect Theory

2.1. Why Prospect Theory is a Breakthrough

Expected utility theory is based on a simple idea: individuals evaluate choices according
to their average value weighted by probabilities.

Prospect Theory introduces a fundamental shift: individuals do not evaluate final states,
but rather changes relative to a reference point.

In other words, what matters is not total wealth, but:

• Whether one gains or loses
• Relative to an initial situation

This shift in framework completely transforms the analysis.

2.2. The Value Function: A Formalization of Psychology

The value function is not merely an equation. It is a mathematical representation of

psychological experience.

Its S-shaped form captures three essential ideas:

a) Diminishing sensitivity
A gain of €100 is significant. An additional gain of €100 is less so.
Value does not increase linearly; it flattens.

b) Gain/loss asymmetry
A loss equivalent to a gain does not have the same impact. The slope is steeper for losses:
losing hurts more than gaining brings pleasure.
This is where the parameter λ takes on its full meaning.

c) Reference point
Individuals do not react to absolute levels of wealth, but to changes relative to a given state.
This explains why two objectively identical situations can be experienced very differently.

2.3. Understanding λ (Lambda) as a Behavioral Principle

The parameter λ is not merely a coefficient. It expresses a deeper reality: individuals are
structurally averse to regret and loss.

14
When λ ≈ 2, this means:
• A loss of €100 “weighs” as much as a gain of €200

Direct consequence:
An individual may refuse a fair gamble, not because they are irrational,
but because their value function distorts the evaluation.

2.4. Probability Weighting: Distorting the World

Prospect Theory does not only modify value—it also modifies the perception of
probabilities. Individuals do not treat p as an objective datum.

They transform it into a subjective probability.

Two major distortions arise:

• Small probabilities are overweighted
• Large probabilities are underestimated

This point is crucial because it explains apparently paradoxical behaviors:

• Playing the lottery
• Buying insurance
• Overreacting to rare events

2.5. Understanding a Decision as a Combination of Two Distortions

A decision under Prospect Theory is never simple.

It results from two simultaneous transformations:

1. Transformation of gains and losses (value function)

2. Transformation of probabilities (weighting)

Thus, a decision depends on:

v(x) × w(p)

This means that even a very unlikely event

can become attractive if:
• Its value is high
• Its probability is psychologically amplified

2.6. The Lottery Effect: A Logical Consequence, Not an Anomaly

15
The purchase of lottery tickets is often considered irrational. Prospect Theory offers a
different interpretation:

• The potential gain is extremely high → v(x) very large

• The probability is low → but overweighted → w(p) increases

The product of the two can become subjectively attractive.

The behavior is not absurd. It is consistent with a different psychological rationality.

2.7. What Researchers Must Truly Understand

It is essential to understand that Prospect Theory is a model that explains why individuals:
• Take excessive risk in the domain of losses
• Avoid risk in the domain of gains
• Overreact to rare events

2.8. Moving to an Advanced Level

Expected reasoning:
• Why do investors become more risk-seeking after a loss?
• How does Prospect Theory explain speculative bubbles?
• How does probability weighting amplify crises?

Model → behavior → market phenomenon

16
Chapter 3: Market Anomalies as Evidence Against Pure Rationality

3.1. Why Anomalies Are Not Exceptions

In the classical approach, anomalies are often presented as isolated irregularities. In reality,
they pose a much deeper problem.

If markets were perfectly efficient, these anomalies:

• Should not exist
• Or should disappear quickly

However, empirical literature shows the opposite: they persist, sometimes for decades.

This implies a fundamental idea:

• Anomalies are not marginal errors
• They reveal a systematic structure of investor behavior

3.2. Momentum: Collective Cognitive Slowness

Momentum is often presented as a simple statistical regularity.

But its true meaning is deeper.

It reveals that investors:

• Do not immediately incorporate information
• Gradually adjust their beliefs

In other words, the market is not instantaneously rational.

It is cognitively slow.

This phenomenon is explained by:

• Conservatism (resistance to belief updating)
• Mimetic behavior (social validation)

Thus, momentum is not just an anomaly: it is an observable trace of the psychological

functioning of the market.

3.3. Bubbles: When Beliefs Replace Fundamentals

A speculative bubble is not simply a valuation error.

It corresponds to a moment when expectations become self-fulfilling.

Prices rise because investors expect them to rise.

17
This mechanism relies on several combined biases:
• Mimetic behavior → following others
• Optimism → believing in continuous growth
• Confirmation bias → ignoring negative signals

In this context, fundamental value becomes secondary.

The market operates through a logic of coordination of beliefs.

3.4. Calendar Anomalies: Psychology Over Time

Anomalies such as the “January effect” or the “weekend effect” are particularly interesting.

Why?

Because they show that even time is not neutral.

Financial decisions are influenced by:

• Tax constraints
• Psychological cycles
• Mood variations

This challenges a key assumption of classical finance:

that preferences are stable over time.

3.5. The Case of Entrepreneurs: A Different Rationality

The behavior of entrepreneurs is particularly revealing.

From a classical perspective, their strategy appears irrational:

• Lack of diversification
• Extreme exposure to risk

But from a behavioral perspective, it becomes understandable.

Entrepreneurs:
• Overestimate their abilities
• Believe they control uncertainty
• Interpret risk as an opportunity

✔ This is not an absence of rationality

✔ It is a subjective rationality, shaped by perception

18
3.6. The 2007–2008 Crisis: A System of Biases

The housing crisis is not merely a technical failure.

It is the result of a system of interconnected biases:

• Banks → overconfidence
• Rating agencies → confirmation bias
• Investors → optimism and mimetic behavior

Each actor behaves in a locally coherent way,

but globally, the system becomes unstable.

This is where behavioral finance provides a unique contribution:

• It explains how “plausible” decisions
• Lead to collective catastrophes

3.7. Confirmation Bias: An Institutional Bias

Confirmation bias is often presented as individual.

But during crises, it becomes institutional.

Rating agencies:
• Select favorable information
• Ignore contradictory signals

This behavior is not only cognitive. It is reinforced by:

• Economic incentives
• Market pressures

✔ This shows that biases are not only psychological

✔ They are embedded in the structure of the financial system

3.8. Moving to Advanced Reasoning

At this level, the question is no longer:

→ “Which anomaly do we observe?”

But:
→ “What does this anomaly reveal about market functioning?”

Examples of expected reasoning:

• Why does momentum persist despite arbitrage?

19
• In what sense are bubbles rational from an expectation’s perspective?
• Are anomalies exploitable, or do they disappear once identified?

3.9. Critical Perspective

We must be able to discuss a central tension:

• If anomalies exist → inefficiency

• If they persist → limits of arbitrage

But then:
Why do rational investors not eliminate them?

Possible answers:
• Arbitrage costs
• Synchronization risk
• Uncertainty regarding anomaly duration

This is where behavioral finance intersects with institutional finance.

20
Chapter 4: Behavioral Portfolio Management as a Departure from Classical
Optimization

4.1. Why Classical Portfolio Theory Fails Empirically

Markowitz’s theory is based on remarkable mathematical elegance.

It assumes that the investor selects a combination of assets by rationally arbitrating between
return and risk.

However, this view relies on a strong assumption: that the investor processes information
in a coherent and stable manner.

Empirical observations, however, show otherwise:

• Portfolios are often concentrated
• Decisions change over time
• Investors do not rebalance optimally

This gap is not accidental. It reveals that the portfolio is not merely a financial object:
→ it is a psychological object

4.2. The Portfolio as an Extension of Identity

A fundamental point (often underestimated): investors do not build a portfolio solely to

optimize returns.

They construct a portfolio that reflects:

• Their beliefs
• Their experience
• Their confidence
• Their relationship to risk

This is why two investors facing the same data may make radically different decisions.

4.3. The Disposition Effect: A Time-Biased Management

The disposition effect is not merely a one-time decision bias.

It structures the dynamic management of the portfolio.

By selling winners too early:

• The investor reduces exposure to high-performing assets

21
By holding onto losers:
• They lock capital into inefficient positions

Thus, the portfolio evolves in a direction opposite to optimality.

This phenomenon is directly linked to Prospect Theory:

• Gains are secured
• Losses are avoided

The portfolio becomes a management of emotions over time.

4.4. Familiarity Bias: An Illusion of Control

Familiarity bias is based on a simple intuition: “I understand better what I know.”

However, this intuition is misleading.

Investing in familiar assets does not reduce actual risk.

It only reduces the feeling of uncertainty.

This bias leads to:

• Excessive concentration
• Under-diversification
• Exposure to idiosyncratic risk

The Enron case is emblematic: employees associated proximity with safety, even though
the risk was maximal.

4.5. The Real Portfolio: An Unstable Compromise

In practice, a portfolio is never optimized once and for all.

It is the result of a permanent compromise between:

• Financial objectives
• Emotional constraints
• Cognitive biases

This explains:
• Inconsistent adjustments
• Overreactions to losses
• Procyclical behavior

22
The portfolio becomes a dynamic system influenced by psychology.

4.6. The Barberis et al. Model: Linking Psychology and Prices

The model by Barberis, Shleifer, and Vishny represents a major contribution.

It shows that:
• Belief errors are not random
• They follow identifiable cognitive patterns

Two key mechanisms:

• Conservatism → slow adjustment
• Representativeness → excessive extrapolation

These mechanisms generate:

• Short-term momentum
• Long-term reversals

Individual behavior is transformed into market structure.

4.7. Risk: A Perception, Not a Given

In classical theory, risk is measured by variance.

However, for investors, risk is experienced differently.

It depends on:
• Emotions
• Context
• Recent experiences

The same asset may be perceived as:

• Risky after a loss
• Safe after a rise

Risk becomes a subjective construct.

4.8. Toward Behavioral Portfolio Management

The central question becomes: should biases be corrected or integrated?

Two approaches exist:

23
4.8.1 Corrective Approach

Correct biases in order to move closer to the optimum

4.8.2 Behavioral Approach

Construct portfolios that are compatible with investors’ psychology

Recent literature favors a hybrid approach:

• Maintain diversification
• Integrate psychological preferences

4.9. Expected Reasoning

A good researcher does not merely describe biases.

They are able to address questions such as:

• Why do investors remain under-diversified despite available information?
• How does the disposition effect undermine long-term performance?
• Can an “optimal” portfolio be designed if preferences are unstable?

24
Chapter 5: Individual Investors in Crisis: From Perceptions to Behavior

5.1. Why Crises Are Privileged Moments for Analysis

Under normal conditions, it is difficult to clearly observe the limits of rationality.

Individual errors exist, but they often remain dispersed. Crises completely transform this
situation.

They create an environment in which:

• Uncertainty is maximal
• Reference points disappear
• Decisions must be made rapidly

In this context, behaviors become more visible. The crisis acts as a revealer of cognitive
mechanisms.

5.2. A Central Question: How to Link Perception and Action

One of the major contributions of the analyzed study is to raise a simple yet fundamental
question:
Do investors act according to what they perceive?

In other words:
• If perceived risk increases
• If expectations decline

Should they logically:

→ reduce their exposure to risk?

The intuitive answer is yes.

The empirical answer is far more complex.

5.3. A Major Break: Perception ≠ Behavior

The central result is counterintuitive:

During the crisis:

• Investors become more pessimistic
• Their perception of risk increases

25
Yet:
• They continue to trade
• They do not significantly reduce the risk of their portfolios

There is therefore a disconnect between cognition and action.

5.4. Understanding This Divergence

This divergence should not be interpreted as simple inconsistency.

It can be explained by several deeper mechanisms.

a) Behavioral inertia
Portfolios are not continuously adjusted. Inaction is often the norm.

b) Belief in a rebound
After a decline, investors may anticipate a recovery. The downturn becomes an
opportunity.

c) Dispersion of perceptions
Not all investors think the same way.

This disagreement creates:

• Buyers
• Sellers

And therefore, trading.

5.5. The Central Role of Disagreement

One of the most important findings is the following:

✔ Trading increases not despite the crisis,

✔ But because of the divergence in perceptions

When beliefs become heterogeneous:

• Some perceive risk
• Others perceive opportunity

The market becomes an active space of exchange.

5.6. Rethinking the Link Between Information and Decision

26
Classical theory assumes:
information → updating → coherent decision

However, what this study shows is different:

information → multiple interpretations → divergent behaviors

Information does not produce a single reaction

→ it produces a dispersion of reactions.

5.7. The Problem of Measuring Risk

A fundamental point for a Master’s-level student:

measured risk is not perceived risk.

In the study:
• Risk is proxied by volatility
• Perception is measured through surveys

These two dimensions may evolve differently.

This raises a central methodological question: What is “true” risk in finance?

5.8. Learning and the Limits of Experience

A common intuition would be to say:

“investors learn during crises.”

However, reality is more nuanced.

Some investors:
• Improve their decisions

Others:
• Continue to make mistakes
• Or exit the market

The observed improvement may be due to:

• Genuine learning
• Or selection (the least successful participants disappear)

27
Chapter 6: Social Influence, Media, and Market Dynamics: Understanding the
GameStop Episode

6.1. A Fundamental Transformation: From the Market to the Community

Classical finance is based on an impersonal vision of the market.

Prices result from the aggregation of individual information.

The GameStop case profoundly challenges this representation. The market no longer
appears as a simple allocation mechanism, but as a social space structured by interactions,
narratives, and collective identities.

Investors do not respond only to financial information.

They also respond to:
• Discourses
• Group norms
• Dynamics of visibility

The market becomes a coordinated social phenomenon.

6.2. Attention as an Economic Resource

One of the major contributions of the recent literature mobilized in this session is to show
that:

✔ Attention is not neutral

✔ It structures market activity

In the case of GameStop, several elements converged:

• Intense activity on Reddit
• Media amplification
• Rapid circulation of content

This attention acts as a selection mechanism:

• Some assets become visible
• Others remain ignored

It is not information in itself that matters

→ but its collective visibility.

6.3. Coordination Without Centralization

28
A central feature of the GameStop case is the absence of a central authority.

Coordination does not occur through a formal organization, but through an emergent
dynamic:
• Sharing of strategies (“buy and hold”)
• Collective validation
• Reproduction of behaviors

Recent studies refer to an endogenous formation of consensus within online communities.

This consensus does not imply perfect unanimity, but a sufficient degree of convergence
to orient collective action.

6.4. The Role of Norms and Identity

The observed coordination does not rely solely on financial incentives.

It is also driven by:
• A group identity (retail vs. institutions)
• Norms (“hold,” “do not sell”)
• Narratives (justice, revenge, collective opportunity)

These elements transform individual decision-making:

✔ The investor no longer acts solely for themselves

✔ They act as a member of a collective

6.5. From Discussion to Action: A Phased Dynamic

Empirical analyses suggest that the GameStop episode did not unfold uniformly.
It followed a structured dynamic:

1. A discussion phase → accumulation of attention, emergence of the theme

2. An action phase → increased trading, stronger coordination
3. A visibility phase → media amplification, large-scale diffusion

This structuring is essential because it shows that:

✔ The market does not react instantaneously

✔ It evolves through progressive social dynamics

6.6. Anticipation or Causality?

29
A central methodological point for a Master’s-level student: some analyses show that
Reddit activity may anticipate trading volumes in certain phases.
However, this anticipation should not be confused with causality.

Several alternative mechanisms may exist:

• A common factor (media, external events)
• A simultaneous reaction to information
• Feedback loops

The analytical challenge is to distinguish between :

• Correlation
• Predictability
• Structural causality

6.7. The Role of Market Microstructure

The GameStop case cannot be explained solely by psychology.

It also involves elements of market structure:
• High short interest
• Brokerage constraints
• Options dynamics
• Temporary trading restrictions

Official sources emphasize this confluence of factors.

This implies that:

• Individual behaviors
• Social dynamics
• Institutional constraints

→ together produce the observed phenomenon.

6.8. Reading GameStop as a Case of Advanced Behavioral Finance

At this level, the expected analysis goes far beyond description.

A student should be able to answer questions such as:

• Is the market still efficient in a context of social coordination?
• Can attention be considered a pricing factor?
• Do social networks create new forms of inefficiency?

6.9. Moving Toward Research Reasoning

30
The right approach is not to retell the GameStop story.
It is to transform it into a research problem.

Examples:
• “To what extent does Reddit activity influence trading volumes?”
• “How can coordination within a social network be measured?”
• “Are the observed effects generalizable?”

Each question must be associated with:

• Variables
• A method
• Limitations

31
Chapter 7: Herding, FOMO, and the Social Construction of Market Behavior

7.1. The Paradox of Herding: Individually Rational, Collectively Inefficient

Herding is often perceived as irrational. Yet the foundational models show exactly the
opposite. In a sequential framework, each individual observes the decisions of others.
These decisions then become a source of information.

Thus, it may be rational for an agent to:

• Not follow their own signal
• But instead imitate the observed decisions

This process leads to a paradoxical situation:

✔ Individually rational decisions
✔ Produce a collectively inefficient dynamic

This phenomenon is known as an informational cascade.

7.2. The Fragility of Cascades

An essential characteristic of cascades is their instability. They rely on an accumulation of

indirect information, rather than on solid fundamentals.

Thus:
• New information can reverse the trend
• A loss of confidence can trigger a sudden reversal

The market then becomes extremely sensitive to weak signals.

This point is fundamental for understanding:

• Bubbles
• Crashes
• “Meme stock” phenomena

7.3. Herding or Mere Correlation?

A major analytical challenge (at the Master’s level) is to avoid a common confusion.

Observing that “everyone is buying” is not sufficient to conclude that herding is taking
place.

32
Two situations must be distinguished:
• Correlated trading: everyone reacts to the same information
• Herding: agents follow others while disregarding their own information

The entire empirical analysis consists precisely in distinguishing between these two cases.

7.4. FOMO: An Emotional Transformation of Decision-Making

FOMO (Fear of Missing Out) introduces a different dimension. It is not merely imitation,
but an emotional pressure linked to anticipated regret.

The investor no longer reasons in terms of expected return. They reason in terms of:
“what I will feel if I do not act.”

This shift transforms the decision:

• Not investing becomes a potential loss
• Investing becomes a way of avoiding regret

7.5. FOMO and Risk-Taking

Recent studies show that FOMO is associated with:

• A higher probability of investing
• Greater tolerance for risk
• Sensitivity to opportunities perceived as rare

This mechanism is particularly powerful in environments characterized by:

• High visibility (social networks)
• Highly publicized extreme gains
• Collective narratives

7.6. Herding and FOMO: Two Complementary Mechanisms

It is essential not to analyze these phenomena separately.

In actual markets:
• Herding provides the structure (imitation)
• FOMO provides the energy (emotional urgency)

Together, they produce: a dynamic of market acceleration.

33
The more prices rise:
• The more FOMO increases
• The more herding intensifies

This mechanism is self-reinforcing.

7.7. The Role of Platforms and Influencers

Contemporary markets introduce a new element: the digital infrastructure of influence.

Platforms:
• Amplify certain content
• Create visibility effects
• Facilitate imitation (copy trading)

Influencers:
• Simplify decisions
• Reduce perceived uncertainty
• Reinforce confidence (or the illusion of confidence)

However, these mechanisms also introduce risks:

• Conflicts of interest
• Manipulation
• Dissemination of biased information

Regulatory authorities explicitly warn against these distortions.

7.8. Measuring Herding and FOMO

A central challenge for an advanced student is measurement. Herding and FOMO are not
directly observable. They must be operationalized.

Examples:
• Social activity (posts, tweets)
• Attention indices (Google Trends)
• Synchronization of transactions
• Dispersion of returns

But each measure raises problems:

✔ Imperfect measurement

34
✔ Selection bias
✔ Uncertain causality

7.9. From Narrative to Research

A good researcher is not satisfied with saying: “investors follow the crowd.”

They must be able to formulate:

• A testable hypothesis
• An empirical model
• An identification strategy

Example:
“An increase in attention (Google Trends) at t-1 leads to an increase in volume at t.”

But immediately:
✔ What are the confounding factors?
✔ What is the actual causal relationship?

35
Chapter 8: Emotions, Sentiment, and the Quantification of Market Behavior

8.1. A Shift in Perspective: From Psychology to Measurement

One of the major changes introduced by behavioral finance is the following: emotions are
no longer merely internal states. They become indirectly observable variables, through:

• Attention
• Volatility
• Volumes
• Collective behavior

Thus, the question is no longer: “Are investors afraid?”

But rather:
→ “How does this fear manifest itself in the data?”

8.2. “Risk as Feelings”: Another Logic of Decision-Making

Classical theory assumes that risk is evaluated objectively.

But in practice, decisions are often guided by feelings.

Fear, greed, and surprise do not merely alter preferences.

They alter:
• What individuals pay attention to (attention)
• What they consider risky (perception)
• What they do (collective action)

The market becomes a space in which emotions structure decisions.

8.3. Fear: A Dynamic of Protection

Fear rarely translates into simple inaction. It generates specific behaviors:

• Seeking protection
• Increased volatility
• Shifts toward assets perceived as safe

Indicators such as the VIX capture this dynamic indirectly.

They do not measure the emotion itself, but rather its consequences for option prices.

36
It is therefore essential to understand: an indicator is never the emotion itself; it is a trace
of it.

8.4. Greed: An Expansion of Risk

By contrast, greed corresponds to a phase of expansion.

Investors:
• Increase their exposure
• Reduce their perception of risk
• Seek high returns

This dynamic is incorporated into composite indices such as the Fear & Greed Index, which
aggregates several market signals.

However, these indices must be interpreted with caution:

✔ They do not say “buy or sell”
✔ They describe a state of the market

8.5. Surprise: An Informational Shock

Surprise plays a particular role.

It corresponds to a gap between:

• What was anticipated

• And what actually occurs

It manifests itself through:

• Price jumps
• Volume spikes
• Changes in volatility

Unlike fear or greed, it is not observed in an average, but in discrete events.

8.6. Building an Index: A Theoretical as Much as an Empirical Act

One of the essential contributions of this session is methodological.

Building a sentiment index is not a neutral technical operation.

It is a theoretical choice.

37
Choosing:
• Keywords
• A period
• A normalization method

amounts to defining what we call “sentiment.”

✔ The index does not reveal reality

✔ It constructs a representation of the phenomenon

8.7. The Fundamental Problem: Endogeneity

A central point that students must master:

• Sentiment does not necessarily cause the market

• The market may also generate sentiment

When prices rise:

• Attention increases
• Sentiment becomes positive

There are therefore feedback loops.

This implies that any analysis must distinguish between:

• Correlation
• Predictability
• Causality

8.8. What Empirical Studies Show

Recent research shows that:

• Sentiment may be associated with volatility
• It may sometimes anticipate short-term market movements
• These effects are unstable and context-dependent

This leads to an important conclusion:

• Sentiment is informative
• But rarely sufficient

8.9. From Signal to Interpretation

38
A sentiment index should never be interpreted in isolation. It must be placed within a
framework:

• Macroeconomic context
• Market dynamics
• Investor structure

Thus, a high level of “fear” may signify:

• A real risk
• Or a contrarian opportunity

Interpretation depends on the theoretical framework.

8.10. Moving Toward Research Reasoning

At this level, a student must be able to formulate not a conclusion, but a testable question.

Example:
“Does an increase in attention measured by Google Trends precede an increase in
volatility?”

But immediately:
• Which variable?
• Over what horizon?
• Which controls?
• Which biases?

The quality of the reasoning is more important than the result.

8.11. Final Synthesis: From Individual Emotions to Market Dynamics

This session makes it possible to understand a fundamental point: markets are not guided
solely by information. They are guided by the way that information is felt.

Thus:
• Emotions influence individual decisions
• These decisions become coordinated
• And they produce market phenomena

39
Chapter 9: Fintech, Artificial Intelligence, and the Transformation of Behavioral
Finance

9.1. An Initial Illusion: Technology as a Solution to Irrationality

One of the most widespread claims in contemporary finance is the following: algorithms
make it possible to eliminate human biases.

This idea rests on an implicit assumption: that biases are solely linked to human emotions.

However, this assumption is incomplete. Biases do not disappear with technology.

They change form.
→ They become encoded, amplified, or displaced.

9.2. HFT: Speed Without Understanding

High-frequency trading does not make “intelligent” decisions in the human sense.
It executes strategies at extremely high speed.

Under normal conditions, this produces:

• Apparent liquidity
• Narrower spreads

But under stress conditions, this liquidity can disappear instantly.

Algorithms withdraw their orders simultaneously, creating what is referred to as fragile

liquidity.

This phenomenon is crucial:

• Technology does not eliminate crises
• It can accelerate their dynamics

9.3. The Robo-Advisor: Rationality or Rigidity?

Robo-advisors are often presented as tools of rationalization.

They allow for:
• Discipline
• Diversification
• A reduction in emotional decision-making

But this rationality is based on parameters that are:

• Historical

40
• Simplified
• Standardized

Thus, the problem shifts:

• It is no longer the investor who is biased
• It is the model itself

A robo-advisor may, for example:

• Underestimate extreme events
• Impose unsuitable mechanical decisions
• Ignore the diversity of actual investor profiles

9.4. AI and Sentiment: Capturing Emotion… or Noise?

AI systems now make it possible to measure sentiment on a large scale.

They analyze:
• Tweets
• Articles
• Forums

And they produce a signal.

But a fundamental question arises: what are they actually measuring?

• A real emotion?
• Media amplification?
• A sampling biases?

Social data are:

• Non-representative
• Noisy
• Manipulable

✔ AI does not capture “market reality”

✔ It captures a partial and biased representation

9.5. The Major Risk: The Algorithmic Loop

The most important point (and often the most underestimated) is the following:

✔ Algorithms are not neutral.

✔ They interact with one another.

41
For example:
• A model detects pessimism
• Triggers sales
• This movement reinforces pessimism

The result is an algorithmic feedback loop.

This mechanism can amplify:

• Volatility
• Bubbles
• Crashes

9.6. The Illusion of Algorithmic Performance

A model may perform very well… on past data. But this guarantees nothing about the
future.

The problem of overfitting is central: the model learns regularities that no longer exist.

This creates an illusion:

• High accuracy
• But low robustness

9.7. The Central Question: Who Controls the Decision?

In a system dominated by technology:

• The investor delegates
• The algorithm executes
• The platform structures the environment

Decision-making becomes distributed.

It becomes difficult to identify:

• Who is responsible
• Where rationality is located

9.8. Toward an “Augmented” Behavioral Finance

Fintech does not replace behavioral finance.

It makes it more complex.

42
It introduces:
• Algorithmic biases
• Platform effects
• Human-machine interactions

✔ Behavior is no longer solely human

✔ It becomes hybrid

9.9. Moving Toward Advanced Reasoning

At this level, students must be able to address questions such as:

• Do algorithms reduce or amplify biases?
• Can a “neutral” model be designed?
• Do markets become more efficient or more unstable?

These questions have no single answer.

They require structured argumentation.

9.10. Final Synthesis: From the Human to the Algorithm

With this session, the transformation is complete.

The course trajectory is the following:
• Individuals are biased
• Biases structure markets
• Technologies intervene
• But they do not eliminate these biases

→ They transform and redistribute them.

43
Chapter 10: Behavioral Regulation and Nudges: Designing Markets for Real Humans

10.1. A Break in the Philosophy of Regulation

Classical finance is based on an implicit idea: investors must be protected through

information.
Providing more information → better decisions.

Behavioral finance challenges this principle.

Individuals:
• Do not always read
• Do not always understand
• Interpret information in a biased way

Information is not sufficient. Regulation therefore evolves toward another logic: modifying
the way choices are presented.

10.2. The Nudge: A Minimal but Structuring Intervention

The concept of a nudge is based on a simple yet powerful idea:

✔ Decisions can be influenced

✔ Without restricting choice

This mechanism relies on choice architecture:

• Order of options
• Default choices
• Presentation of information

Within this framework, the individual remains free, but their environment is structured.

10.3. The Paradox of Libertarian Paternalism

Nudges fall within what Thaler and Sunstein call libertarian paternalism.

This concept contains a fundamental tension:

• Protecting individuals
• While respecting their freedom

But this tension does not disappear.

It is reformulated as:

44
✔ Guiding without manipulating
✔ Steering without constraining

10.4. The Central Role of Biases in Policy Design

Nudges are not arbitrary. They are explicitly based on the biases identified in previous
sessions:

• Inertia → default options

• Framing → reformulation of information
• Loss aversion → highlighting risks

Thus, behavioral regulation: uses biases to correct biases.

10.5. Automatic Enrollment: A Fundamental Example

The case of retirement systems is particularly revealing.

Moving from:
• Opt-in (voluntary enrollment)

to:
• Opt-out (automatic enrollment)

produces a massive transformation in behavior.

Why?

Because individuals:
• Avoid decision effort
• Prefer the status quo

✔ This is not a constraint

✔ It is an exploitation of inertia

10.6. The Prospectus: Regulation Through Form

Financial documents are not neutral.

The way information is presented influences:

• Understanding

45
• Risk perception
• Final decision

Regulation (e.g., MiFID II) imposes:

• Clarity
• Transparency
• Hierarchization

But this raises an essential question: can framing bias truly be neutralized?

10.7. The Fundamental Problem: The Effectiveness of Nudges

A nudge is not an idea—it is an intervention.

And every intervention must be evaluated.

This implies:
• A counterfactual (what happens without the nudge?)
• Impact measurement
• Sustainability analysis

A nudge that is effective in the short term may fail in the long term.

10.8. The Limits of Behavioral Regulation

Several limitations quickly emerge:

a) Adaptation of individuals
Individuals may circumvent or ignore nudges.

b) Heterogeneity
The same nudge does not have the same effect on everyone.

c) Saturation
Too many nudges may reduce their effectiveness.

Behavioral regulation is not a universal solution.

10.9. The Ethical Question: Manipulation or Support?

This is where the discussion becomes fundamental.

46
A nudge can be seen as:
• A decision aid
• Or a subtle form of manipulation

The difference depends on:

• Intention
• Transparency
• Benefit for the individual

Behavioral regulation is inseparable from ethical reflection.

10.10. Toward an Economy of “Assisted Decision-Making”

With this chapter, a complete transformation emerges.

The role of regulation is no longer only:

• To correct markets

But to:
accompany human decision-making

This implies:
• Integration of psychology
• Active design of choice environments
• Continuous empirical evaluation

10.11. Moving Toward Research Reasoning

At this level, students must be able to address questions such as:

• Are nudges more effective than financial education?
• Can the “well-being” generated by a nudge be measured objectively?
• Can nudges be diverted for commercial purposes?

These questions require:

• Theory
• Data
• Critical thinking

10.12. Final Synthesis: From the Individual to the Institution

This session concludes the progression:

47
• Individuals are biased
• Markets reflect these biases
• Institutions intervene

But this intervention:

✔ Does not eliminate biases

✔ It frames and redirects them

48
Chapter 11: Recent Trends and Research Directions: From Knowledge to Inquiry

11.1. A Transformation in the Role of the researcher

Until now, the course progression has led us to understand:

• Biases
• Markets
• Technologies
• Institutions

But at this stage, a new requirement emerges:

✔ We must no longer only understand

✔ We must produce new questions

Behavioral finance is not a fixed set of results. It is a constantly evolving field.

11.2. The “Third Generation”: Expanding the Scope of Analysis

Recent work, particularly that of Statman, proposes a major evolution.

Behavioral finance is no longer limited to correcting individual errors.

It now incorporates:
• Values (ESG)
• New assets (crypto)
• The diversity of investor profiles

This implies a shift in perspective:

✔ From individual rationality

✔ Toward a situated, social, and normative rationality

11.3. Neurofinance: Understanding the Biological Mechanism of Decision-Making

Neurofinance introduces a radical question: can biases be observed directly in the brain?

Research uses:
• Functional MRI (fMRI)
• EEG
• Biomarkers

49
It shows that:
• Fear activates certain regions (amygdala)
• Reward activates other circuits

However, these approaches raise important questions:

• External validity (laboratory vs. real market)
• Interpretation of signals
• The risk of overinterpretation

Neurofinance opens new perspectives, but requires strong methodological caution.

11.4. Big Data and Real-Time Behavior

A major transformation in contemporary research is access to data.

Researchers can now observe:

• High-frequency decisions
• Social interactions
• Real-time reactions

This makes it possible to move:

✔ From “static” finance

✔ To dynamic and continuous finance

However, this abundance of data also introduces challenges:

• Noise
• Causality
• Overfitting

11.5. AI and the Transformation of Behavior

Artificial intelligence does not only change tools. It changes behavior itself.

Investors:
• Rely on algorithmic recommendations
• Delegate certain decisions
• Interact with automated systems

This creates a new type of behavior: a hybrid behavior (human + machine).

11.6. ESG: The Introduction of Values into Financial Decision-Making

50
Traditional finance assumes that decisions are guided by returns.

ESG introduces another dimension: moral and social preferences.

Investors may accept:

• Lower returns
• In exchange for a positive impact

This transforms the very notion of rationality: decision-making becomes multidimensional.

11.7. The Central Challenge: Transforming an Idea into a Research Question

At this level, the main challenge is not finding a topic. It is transforming an intuition into
a testable question.

A good research question:

• Is precise
• Is measurable
• Can be tested empirically

Example:
❌ “Do social networks influence markets?”
✔ “Does an increase in Twitter activity on an asset at t−1 predict an increase in volatility
at t?”

11.8. Designing an Experimental Protocol

Robust research relies on a clear structure:

1. A hypothesis
2. A design (experiment, data, or simulation)
3. Measurable variables
4. An analytical strategy

But above all: credibility depends on the ability to isolate an effect.

11.9. The Limits of Behavioral Research

At this level, it is essential to be critical.

The main limitations are:

• Sampling bias

51
• Lack of reproducibility
• Difficulty accessing real-world data
• Confusion between correlation and causality

Methodological rigor is central.

11.10. The Role of Interdisciplinarity

Modern behavioral finance is a hybrid field.

It draws on:
• Economics
• Psychology
• Neuroscience
• Data science

The best research often lies at the intersection of these disciplines.

11.11. Transition to the Research Level

A researcher must be able to:

• Formulate a clear question
• Propose a method
• Discuss limitations
• Interpret results

But above all: justify why the question is relevant.

Behavioral finance becomes a tool for exploring the world.

52