Does Managerial Overconfidence Change with Market Conditions? Risk Management for Financial Institutions

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The paper have merit, I am suggestion some things to improve the paper. 

1. When using abbreviation for first time i.e., GFC, SOX should be define. GFC is defined for the first time in line 110, mentioned dosen times before that  and SOX in 145 line. 

2. How the sample was chosen? Which sampling method was used?  - this part is missing. More information on students is also missing, University, Faculty, Study program, Level of program (undergraduate, graduate, postgraduate...). Authors used non-probability sampling but which technique (convenience; snowball?) I guess?

3. In line 331-341 authors should describe more in detail anchoring process. What kind of anchor was used (for example Visual, textual, Verbal etc.)?

4.  The abstract should contain the main purpose of the paper, the research method used in the research and the main contributions. The abstract must contain the main purpose of the paper, the research method used in the research and the main contributions more precisely and clearly.

5. It would be very useful to add in the "Introduction" section and the purpose, objectives and hypothesis of the research. I consider that a weak point of the paper is that the authors did not show the novelty of the paper compared to other works. That is why, I consider that the introduction should specify the novelty of the paper compared to other papers published in this area.

6. In methodology is missing part of time frame of the laboratory experiment.

7. The authors need to add more discussion and policy implications to the conclusion section. 

8. The results provided by authors weren’t compared to the similar studies – this is recommendable. 

9. Authors did not include several critical regression diagnostics that are essential for assessing the adequacy of the model. These diagnostics ensure the validity and reliability of the regression analysis (test should be presented for: Homoscedasticity, Linearity, Normality, Multicollinearity etc.) 

 

Comments on the Quality of English Language

The text should be proof read. 

Author Response

Thank you very much for your comments and please find the point-to-point feedback below:

1. Abbreviations such as GFC and SOX are defined in the early part of the article, in line 31-32.
2. The details of the psychological study have been added to the Study 2 section.
3. The kind of anchor was textual and it is now specified in the lines.
4. The abstract is modified and it includes the main purpose of the paper (stated-based managerial overconfidence, a new concept to be happened in finance literature), the research methods (regression analysis and t-test for the two studies), the main contribution (stated based CEO overconfidence and its implication to the credit market under the external economic/environmental shocks, Global Financial Crisis in 2008 and Sarbanes Oxley Acts in 2002).
5. Amendments are made in the introduction section.
6. The time frame is added (Autumn 2018)
7. The discussion and policy implications have been extended.
8. The relevant tests were conducted and not shown in the main text. Breusch-Pagan test has been conducted for homoscedasticity and the p-value is very small that we can infer the error variance is non-constant. For normality, we used Shapiro-Wilk test and the results show it was drawn from a normal distribution. For Multicollinearity issues, variance inflation factor (VIF) has been used here, and there’s no evidence for multicollinearity. With respect to linearity, Ramsey RESET test has been done, and we don’t see there’s a concern for adding a polynomial term.

Reviewer 2 Report

Comments and Suggestions for Authors

The topic is interesting and the paper should provide a good contribution for the research community. 

The objectives of the reserach were well-provided. However, the term "overconfidence" should be clearly defined by disucssing past studies more extensively. The reserach methodology was acceptable, but the author should provide greater details on data selection and data screening more clearly. In addition, the selection criteria for the factors in the study should also provide additional explanation to demonstrate the suitability of the research. The data analysis was fine and the research results were well-presented. The discussion of the findings with the past research should be expanded with more details to highlight the contribution of the study. The rest of the paper was fine. 

 

Comments on the Quality of English Language

Minore revision of the paper should be done. 

Author Response

Dear Reviewer,

 

Thank you very much for your comments and please find the feedback below:

 

As there are two different studies, study 1 using real market data and study 2 using a psychological lab experiment data, the overconfidence with respect to the two studies are discussed inside the body of the two body. Following an article published in Journal of Finance by Malmendier and Tate in 2005, study 1 mainly focused on the market approach by treating the deliberate delay of top executives vested with option compensations in exercising their in-the-money option as the determination of the overconfidence level of the CEOs.

 

In Study 2, the definition of overconfidence is from the psychology perspective, it would be classified as three distinctions, namely over-estimation, over-placement and over-precision. Over-estimation emphasizes on the people’s feelings on their performance, abilities, level of control and success rate. For example, students may over-estimate the test score. Over-placement refers the judgment of your own performance when comparing to the others. Usually, the analysts being interviewed may not be the most outstanding ones, but they will show their confidence and they will regard themselves as the top tier analysts during the conversation. Over-precision reflects the confidence of people knowing the truth or reality. To sum up, it reveals that the overconfidence is a result of emotional illusion, cognitive bias and social factors.

 

The selection criteria and the findings’ part are amended and please see the updated version.

Reviewer 3 Report

Comments and Suggestions for Authors

The researchers have worked on important topic having high relevance. This paper argues that managerial overconfidence can be state-contingent, meaning that a financial crisis can affect it. This paper introduces state-based managerial overconfidence to finance literature. 

There is need to add latest papers in review. 

Why in real market data the period 1996-2011, was used. Is this the latest data available. Please justify.

Add research questions to enhance depth. 

The paper is exhaustive and the researchers reported two empirical studies. In Study 1, they analysed real market data and found macroeconomic changes affected managerial overconfidence. 

Study 2 was a lab experiment that simulated how external manipulations could affect conference level. 

Both empirical studies showed state-contingent overconfidence. This is important finding. 

 

The research methods used are okay and analysis is fine.  

 

To capture the evidence of state-contingent overconfidence while mitigating the potential endogenous effects, the researchers adopted the following model for the regression: 

Loan Size or loan Duration = f (Overconfidence level*GFC (or SOX); firm controls; CEO controls; loan characteristics; error term)

 

The study suggests that if the state-contingent overconfidence hypothesis is ignored, creditors may reduce loan amounts or durations by more than the efficient level during an economic downturn.

 

The study has some important implications, which may be revised and theoretical and practical implications may be two heads for reporting these. 

Comments on the Quality of English Language

Moderate editing needed

Author Response

Dear Reviewer,

 

Thank you very much for your comments and please find the feedback below:

 

Regarding the period of the data coverage, our study tried to include at least 3 years’ observations before and after the treatment effect (Global financial Crisis and Sarbanes-Oxley Acts). Therefore, the period 1996-2011 would be considered as the ideal range of data coverage. Research questions are addressed in the main text of Introduction part in the article. The other parts are revised accordingly.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Thank you for accepting my suggestions.