Paulo César Chagas

Behavioral Finance: A Study of Affect Heuristic and Anchoring in Decision Making of Individual Investors

The objective of this article is to investigate the presence of affect and anchoring biases in the financial decision making of individual investors. Another parallel objective is to verify whether the gender factor (male and female) or financial knowledge interfere with the presence of these biases. Considering the results of an innovative experiment, we found support to the existence of such biases. It was verified that there was no significant difference regarding men and women with respect to being attached, anchored or attached and anchored to their investments. The women in this sample were slightly more anchored. Finally, it was observed that the knowledge factor interfered marginally in affect, in addition to partially attenuating the anchoring bias.

Behavioral Finance: The Aversion to Uncertainty Bias in Individual Financial Decisions

This research verifies the existence of the aversion to uncertainty bias in individual financial decisions. It also evaluates the effects of gender and knowledge in this bias. We considered a sample of 80 undergraduate management students from Universidade Catolica de Brasilia. The results supported the existence of the uncertainty bias amongst individuals facing financial decisions. Contrasting previous studies, men presented greater uncertainty bias than women. The students with higher financial knowledge presented lesser uncertainty bias when compared to the ones with reduced knowledge. This last result supports the classification of the aversion to uncertainty as cognitive and not emotional bias.

Information or Noise, A Study on the Accuracy of Analysts’ Forecasts in Brazil

This paper evaluates the accuracy of financial analysts’ forecasts. These predictions are very important since they are used by investors and administrators during the decision making process. We address the quality of the forecasts by comparing them to the actual numbers, using the root mean squared error (RMSE) as the accuracy measure. The data used in this research is provided by the Central Bank of Brazil in the weekly survey with the market analysts: Focus Report. Our results support the hypothesis that analysts’ forecasts are less accurate the further they are away from the date of release of the actual numbers and there is some level of information in some classes of economic indices forecasts which could be used by investors.

Central Banks Exposure to Market Risk and the 2008 Crisis

Central banks, given their relevance in the international financial market, should be aware of their risk exposure. This study analyzes the evolution of market risk exposure of the central banks of developed and emerging countries in the 2003-2011 period, which includes the 2008 crisis. The methodology adopted was the estimation and analysis of the market risk (Value-at-Risk) associated to the balance sheets exposure of central banks of 14 countries, including emerging and developed markets, highlighting the period of the 2008 crisis and its impacts. The data considered was: total assets, foreign assets and foreign liabilities. Based on the results we found that central banks of emerging countries, when compared to the ones of developed countries, presented a higher exposure to market risk during the period from 2003 to 2011, notably at the time of the 2008 crisis. In the pre-crisis period, the market risk was primarily due to increased exposure to foreign currency; and in times of crisis to greater volatility in the currencies of emerging countries.

Overconfidence in Individual and Group Investment Decisions

This study analyzes group influence on investment decisions in an attempt to determine the correlation between two financial behavior tendencies: overconfidence and herding behavior. An experiment was carried out with 92 students from the Universidade Catolica de Brasilia. The results indicated that acting in a group tended to diminish overconfidence and that women presented more overconfidence than men, even though this effect was not highly significant. It was also verified that the performance of the two genders acting together was more rational than when they acted separately.

A Behavioral Study on the Perception of Market and Credit Risk by the Investor

The objective of this study is to analyze the perception of the individual regarding market risk and credit risk. The methodology used to obtain the data was an experiment involving a questionnaire carried out with 93 business administration from the Universidade Catolica de Brasilia. The results showed that the men’s perception of risk was different from that of the women and that financial institutions should therefore offer different products to men and women.

Rating Agencies Add Value in Credit Risk Assessment

Credit rating is an index for classifying credit risk that attributes scores based on investor trust and confidence in the company issuing bonds in the financial market. This article studies rating as signs of default (insolvency) in companies. Fitch Ratings was used due to the transparency of its evaluations in its annual publications on default. The sample included a total of 268 companies that were listed as going into default in the period from 1990 to 2009. This study tested whether the rating has the power to predict bankruptcy. The results confirmed the predictive power of the rating regarding company defaults. We found a significant relation between rating and default probability: the worse the credit quality of a company according to the classification agency’s analysis, the higher its probability of default was.

The Determinants of Overconfidence: An Experimental Approach

We employ a lab experiment to test whether subjects have overconfidence. We find evidence suggesting that: in general men have a larger degree of overconfidence, and that both age and experience have some influence in the degree of overconfidence.

Illusion of Control: Does Gender Matter?

This paper tests the illusion of control in a lab experiment. We test the phenomena with undergraduate students and find little evidence of systematic differences due to gender. These results suggest that cultural differences may play a role and that generalization using lab experiments should be taken with caution.