“Overconfidence bias among retail investors: A systematic review and future research directions”

This paper comprehensively evaluates the literature on retail investor overconfidence using a framework-based systematic approach to understand the various dimensions of overconfidence bias, its effect on investing choices, and market dynamics. A systematic review of 137 publications from the Scopus database have been done to detect the research trend concerning investor overconfidence bias from its inception. An integrated ADO-TCM framework has been employed to present a systematic analysis of the theory, context, and methodologies (TCM) employed in the reviewed studies. The ADO (Antecedents, Decisions, and Outcomes) framework thoroughly examines the antecedents, decisions, and results of investor overconfidence. The study identified four broad sets of factors contributing to investor overconfidence, as found in the existing literature. These factors include demographic characteristics, personality traits of investors, their knowledge and experience, and the features of investments and investor types. The Prospect theory is the most popular theory in the literature, with much research using secondary data and experiment-based analysis. The prospective study directions, based on the gaps in the existing literature, are as follows: further investigation into the decision-making processes of overconfident re-tail and professional investors is a worthwhile subject. Future research may shift their focus from financial outcome variables to non-financial outcome variables such as the impact of investor overconfidence on individuals’ stress levels, subjective financial well-being, and overall life happiness.


INTRODUCTION
Overconfidence in the financial arena remains an essential field of study in behavioral finance.The available literature has demonstrated the popularity and significance of overconfidence bias in investment and behavior (Benos, 1998;Kahneman & Riepe, 1998;Merkle, 2017;Nosić & Weber, 2010).Overconfidence is supreme faith in one's abilities (Odean, 1998a).Overconfidence is a form of cognitive bias in which an investor places greater emphasis on his or her knowledge, intuition, or strategy than is warranted by the available data or past results.Investor overconfidence can lead to underreaction and overreaction in stock prices (Daniel et al., 1998).Overconfident investors trade more frequently, and their trading expenses lower their returns (Barber & Odean, 2000;Odean, 1998b).In the financial context, overconfidence bias is critical in shaping investment decisions, risk assessment, and market dynamics.Overconfidence can play a negative role by underestimating risks (Odean, 1998a), inadequate portfolio diversification (Pak & Chatterjee, 2016), and ignoring professional advice (Hsu, 2022).Therefore, understanding and deconstructing the impact of overconfidence is critical because it may lead to poor investment decisions, market bubbles, and, eventually, financial disasters (Glaser & Weber, 2010).When examining the influence of overconfidence on the trading behavior of investors, it becomes evident that it leads to diminished returns and affects the dynamics of the market.It is imperative to ascertain the underlying factors contributing to the manifestation of overconfidence bias.

LITERATURE REVIEW
By performing a co-citation analysis, the knowledge clusters pertaining to the literature on investor overconfidence were identified and literature has been divided into three clusters.This analysis operates under the assumption that the co-cited papers share a common theme (Tabash et al., 2023).This first cluster contributes to understanding the complex interplay between human psychology and financial decision-making.Articles in the first cluster examine the influence of behavioral biases in investor decision-making under uncertainty.Tversky and Kahneman (1974) laid the foundation for the study of decision-making in uncertain situations, where they demonstrated the three heuristics for gauging probability and generating predictions.Kahneman and Tversky (1979) presented a critical analysis of expected utility theory as a framework for risky decision-making and proposed an alternative model, "prospect theory," that considers the psychological aspects of decision-making and provides a more precise description of how humans evaluate risky alternatives.Other papers in the cluster show how the disposition effect influences investment behavior (Odean, 1998a), and Baker and Nofsinger (2002) investigated common investment errors caused by an investor's cognitive and emotional weaknesses and social factors influencing financial decisions.In subsequent research, researchers have identified three distinct overconfidence indicators: overestimation, over-placement, and over-precision.Over-precision has more durability than the other two forms of overconfidence, and its existence diminishes the extent of overestimation and overplacement (Moore & Healy, 2008).These papers contribute to understanding the complex interplay between human psychology and financial decision-making, providing valuable insights into the biases and behavioral factors that shape investment choices and market outcomes.Cluster 2 mainly discusses the impact of overconfidence on people's trading behavior (Glaser & Weber, 2007;Grinblatt & Keloharju, 2009).Overconfidence can manifest as excessive trading volumes and poor return on investment for retail investors (Barber & Odean, 2000).Investors who are overconfident in their abilities or who have a previous record of outperformance are more likely to make frequent trades (Glaser & Weber, 2007;Grinblatt & Keloharju, 2009) with increased volatility (Gervais & Odean, 2001).A couple of publications contained in this group also address the significance of financial literacy (Van Rooij et al., 2011) and self-awareness (Dorn & Huberman, 2005) for investors, as these attributes could help prevent overconfidence bias and risky behavior (Nosić & Weber, 2010).The third cluster focuses on overconfident traders' impact on the financial market.Overconfidence resulting from excessive trading has been repeatedly linked to gender (women tend to be less overconfident), income, and formal education (Barber & Odean, 2001;Bhandari & Deaves, 2006;Odean, 1999).Overconfidence affects not only investment behavior but also the stock market.When traders are overconfident, they may trade more, resulting in increased market depth (liquidity) and more informative price signals (Benos, 1998;Odean, 1998b).Overconfidence may also increase market volatility and hinder market adjustment, and knowledgeable, overconfident traders may aggressively bring stock prices closer to actual values.(Odean, 1998b) Alternatively, noisy traders may push stock prices away from their actual values (Daniel et al., 1998).This compilation of scholarly articles provides valuable insights into the persistent impact of overconfidence on financial markets and decision-making when considered collectively.In summary, the process of investment decisions is influenced by psychological elements such as overconfidence, which in turn affects the trading behavior of investors and the overall financial market.
The purpose is to review and consolidate existing research on investor overconfidence, aiming to identify gaps, inconsistencies, and areas that require further investigation comprehensively and methodically.

METHOD
Systematic literature reviews are employed as a means of consolidating the information present in existing literature and identifying potential areas for further research.To explore the field, a bibliometric study is conducted which includes performance analysis, conceptual structure, and intellectual structure (Syed et al., 2023) This integrated framework is used in the analysis to identify gaps in the provided subject.Scopus was used to acquire the information since it has many double-blind peer-reviewed publications published in high-impact factor journals.The final figure of 137 articles has been determined using a systematic approach, as illustrated in Figure 1.Post-screening, the data set was triangulated for data verification.Two researchers and a group of experts opined on the suitability of data.The present study uses Biblioshiny R software and VOSviewer software for bibliometric analysis.AND TITLE-ABS-KEY ("finance*" OR "credit" OR "debt" OR "stress" OR "invest*" OR "risk" OR "literacy" OR "advice" OR "behavior*" OR "financial knowledge" OR "wellbeing") AND NOT TITLE-ABS-KEY ("CEO" OR "manager*" OR "CSR" OR "corporate social responsibility" OR "firm" OR "clinic*" OR "medical"))

Most influential authors
The identification of the most prominent authors on the subject has been facilitated through the utilization of four metrics: total citations, h-index, m-  Note: TC: Total Citations, NP: Number of Publications.

Keyword analysis
Two methods have been used to carry out keyword analysis.Table 4 has been produced to focus on the most frequent keywords in the previous ten years, and a keyword map (Figure 3) depicting the top 50 keywords with at least two co-occurrences has been shown to illustrate the overall conceptual structure and to show the latest developments/keywords in the field.The map is valuable for recognizing major themes and emerging trends and comprehending the research landscape (Singh & Malik, 2022; Syed et al., 2023).In Figure 3 (Vos viewer output), the proximity of two terms represents the closeness of the keywords in the literature, and the size of the nodes represents their frequency.The strongly connected keywords with overconfidence bias are financial literacy, gender, disposition effect, investment, risk-taking, and trading activity.
Table 3 was prepared to highlight the latest developments in the field through author keywords used in the last ten years.The table exhibits all the author keywords duplicated within the sample papers included in the review.

Interrelation of overconfidence with other cognitive biases
Investment behavior is influenced by cognitive biases, leading investors to act contrary to financial and economic theories.These biases can skew judgment, alter risk perception, and lead to ac-    2007), 43% of investors show multiple biases; nevertheless, the study also shows that, while some biases have negative correlations with other biases, it is not always the case for one bias to influence another.However, the literature has sufficient evidence on the interrelationship between these biases, like the disposition effect can be exacerbated by overconfidence (Chu et al., 2012).

Theories-context-methods (TCM) framework
This study used the TCM (Paul & Rosado-Serrano, 2019) framework to understand the theoretical underpinnings and the broader context of investor overconfidence while shedding light on its distinctive qualities.It begins by examining the theoretical landscape of overconfidence research, focusing on this discipline's commonly used theoretical foundations.Next, the geographical scope of overconfidence research is investigated, focusing on the nations where such studies have been conducted.The TCM framework has been presented in the three following paragraphs each representing theories used, context of the publications and methods used by the authors.

Antecedents in ADO framework
This section represents antecedents of overconfidence in ADO framework.Table 5 displays the percentage distribution of the antecedents of investor overconfidence in the current review.The investigation identifies 44 distinct antecedents of investor overconfidence bias from the literature.The antecedents used in the studies under consideration are divided into three groups: antecedents related to the demographic characteristics of the investors, antecedents related to the investor's knowledge and experience, and antecedents related to investment features and investor type.
The investors' demographic characteristics are covered under this category.When it comes to their investment habits, male and female investors differ from one another.The difference in trading volume between male and female investors is caused by the tendency for male investors to regard themselves as more overconfident than their female counterparts (  (Gervais & Odean, 2001).The impact of past perceived portfolio gains on investors' overconfidence can be elucidated by their inclination to concentrate on their past portfolio returns (Merkle, 2017).However, they credit these favorable results to their personal investment expertise, experience, performance, and the quality of information available (Khan et al., 2017).
The changes and growth of the overconfidence bias can also be attributed to the profession of the investor and the investor.It is observed that intraday traders are more vulnerable to overconfidence (Prosad et al., 2015) as trade frequency has a positive impact on overconfidence.

Decision variables in the ADO framework
This section represents decision variables in overconfidence literature and is one of the constituents of the ADO framework.The decisions that investors make show how they act when they are affected by overconfidence bias.

Outcome variables in the ADO framework
This section represents outcome variables in overconfidence literature and is one of the constituents of the ADO framework.Table 7 shows the percentage distribution of the outcome variables used in the studies under review.In the seminal work of Tversky Kahneman (1974), it was found that en-hanced overconfidence among individual investors had a detrimental impact on stock markets, resulting in inadequate and unfavorable outcomes.The phenomenon of overconfidence has been observed to exert various impacts within the domain of trade.There are two primary observations about the impact of overconfidence on trading behavior.Firstly, it has been shown that overconfidence tends to increase the expected trading volume.Secondly, traders who exhibit overconfidence experience a decrease in their predicted utility.This finding has been supported by Wilaiporn et al. (2021).Investors who are overly confident in their ability to time the market or choose winning stocks are prone to excessive purchasing and selling (Khan et al., 2016).Poor market timing can increase transaction costs and lower returns (Statman et al., 2006).The markets with overconfident investors have witnessed price bubbles and intense trading volumes (Michailova & Schmidt, 2016).

CONCLUSION
This hybrid systematic review uses bibliometric methods and content analysis to synthesize the vast financial overconfidence literature using the ADO-TCM framework.This review combines bibliometric and content analysis to better comprehend financial overconfidence research's important contributors, such as journals, authors, and publications.These analytical methods help one grasp overconfidence's many facets and financial decision-making implications.Furthermore, the review delves into the content analysis using the ADO-TCM framework, allowing for a systematic exploration of the prominent theories, geographical contexts, and methodological approaches employed in overconfidence research.By examining the theoretical underpinnings, geographic variations, and methodological innovations in this field, this study aims to provide a comprehensive overview that informs future research directions and practical applications.
Future research on overconfidence in financial decision-making should focus on further exploring and integrating various psychological and financial theories.Researchers should continue synthesizing and integrating psychological theories such as prospect theory, behavioral finance theory, heuristics theory, cognitive theory, and attribution bias theory.This interdisciplinary approach can provide a more holistic understanding of the factors contributing to investor overconfidence.In terms of study context (Table 4), this analysis found that most research comes from advanced nations and is based on a single nation.Future studies could concentrate on comparing overconfidence bias across countries.This will aid in investigating the impact of religious beliefs and cultural differences on overconfidence bias across countries.One can better grasp the global impact of overconfidence in financial decision-making if we conduct studies to identify the cultural factors contributing to this phenomenon.However, a promising opportunity also exists in exploring this phenomenon within the contexts of cryptocurrency, commodities, and currency markets.The review study reveals that the articles examined predominantly utilize classical statistical approaches, such as regression, ANOVA, and factor analysis, for data analysis.Furthermore, the literature is primarily dominated by experimental research.In the future, researchers can undertake big-data analytics using advanced techniques such as supervised and unsupervised machine learning and artificial neural network approaches to understand overconfidence bias better.More qualitative work (e.g., in-depth interviews, observations, discourse analysis, laddering techniques, and qualitative comparative analysis).To further clarify the causal relationship between ADO, this review calls for more longitudinal research along the life path of individuals.
Further investigation may be conducted to explore the impact of antecedents that have been relatively understudied in the existing body of literature.For example, an investigation may be undertaken to comprehend the changes in overconfidence amidst the economic downturn and geopolitical events like the Russia-Ukraine war.There is a need for more research to evaluate the impact of technologies, such as simulations and robo-advisors, on exacerbating or mitigating overconfident behavior.The data presented in Table 6 indicate that a significant proportion of the literature on overconfidence has employed risky asset allocation and excessive trading as decision variables.The distinction between the decisionmaking processes of overconfident retail and professional investors is a worthy topic for further study.Another potential domain of inquiry pertains to examining decision variables influencing investors in the cryptocurrency and commodities market.According to Table 7, most of the reviewed articles mainly assessed the impact of overconfidence on investment returns and market volatility.The focus of future studies may move from financial outcome variables to non-financial outcome variables.This covers the consequences of investor overconfidence on their stress, subjective financial well-being, and life satisfaction.
Despite following the necessary protocol and conducting an exhaustive literature review, this study has certain limitations.First, only the SCOPUS database is used for bibliometric analysis research.In subsequent studies, investigators might look at articles sourced from single or numerous databases, such as Web of Science or EBSCO.Second, there is a possibility that particular research has been omitted due to the use of filters and keywords.

Figure 2
Figure 2 represents the publication trends in research on investor overconfidence.Although 2023 was the most productive year, the field's research journey started in 2001.Less attention was paid to the overconfidence bias in the early research trends.However, overconfidence research in finance began to gain momentum in 2013, and since 2019, the number of papers published year has grown dramatically.

Figure 2 Investment
Figure 2. Publication trends

Figure 4 .
Figure 4. Percentage distribution of interrelated cognitive biases

Table 2 .
Leading authors in overconfidence research

Table 4 .
Percentage distribution of overconfidence studies in the top 10 countries

Table 5 .
Percentage distribution of the antecedents

Table 6 .
Percentage distribution of the decision variables

Table 7 .
Percentage distribution of the outcome variables