This thesis aims to explore possible conflicts of interest between the activities of agents in the Brazilian financial market, especially banks, investment funds and analysts. In the first chapter, the focus is the possible lack of a Chinese Wall between the investment banking activities and the asset management unit of the same bank during the bond underwriting process. The conflict of interest analyzed is the possible use of funds under management of the same bank to allocate bond issuances that would not have sufficient demand in the market. The paper proposes an empirical study of bonds issued in Brazil between 2005 and 2015. The results show that bond 'excess return' at the time of issuance is negative when there is allocation of funds associated with the underwriter, supporting the thesis of conflict of interest. However, when the associated funds are for Institutional Investors, the 'excess return' is positive indicating that the better monitoring of the fund's activities could serve as mitigation for the conflict of interest. In the second chapter the assessed conflict of interest is the funding activities of banks and the rates obtained by investment funds from fixed income assets issued by banks. The hypothesis tested is that affiliated funds obtain a lower yield than non-affiliated funds on the same assets. To test this hypothesis, an empirical study was conducted in Brazil with the rates of Letra Financeira, from the beginning of its emergence in 2010 until the end of 2015. The results point that funds, when investing in Letras Financeiras of the controlling bank, obtain a lower yield than that obtained by funds not affiliated to the same bank. However, like the results obtained in the previous chapter, this effect is not observed in funds for Institutional Investors. Corroborating the idea that better monitoring can mitigate the conflict of interest involved. Finally, the third chapter analyzes two possible conflicts of interest derived from the relationship between the opinions of stock analysts and the investments of the funds affiliated to the same bank. The first hypothesis is about optimism of the analysts as the affiliated funds have a large volume of investment in the stock. The second hypothesis is about Front Running, where funds anticipate changes in analysts' opinions of the same bank. The database used in this study was the opinion of all analysts covering Brazilian stocks for 10 years, from 2005 to 2015. The results suggest that: (i) there is an optimism of analysts associated with Brazilian banks related to the volume invested by affiliated funds, and (ii) the evidence does not support the hypothesis of front running by the funds.