This thesis analyzes the differences between venture capital (VC) funds managed by banks and funds managed by independent institutions. Banks as VC fund managers (or bank affiliates funds) contact companies that commonly require banking services such as loans, underwriting and M & A advisory. Fang et al. (2013) and Hellman et al. (2008) explore the possibility that bank affiliates seek to invest in companies that may in the future be clients of the bank to which they are associated. In this case, banks sponsor VC funds to strengthen their commercial area. In addition, bank affiliates have access to the same flow of opportunities as independent funds. This thesis explores another possibility: independent funds seek co-investment with bank affiliates to facilitate the access of their investees to advantages such as greater capital volume and obtaining credit. In this case, the flow of opportunities of bank affiliates is differentiated because they are easier to participate in co-investments. Thus, the first objective of this thesis is to seek evidence that bank affiliates have a different flow of opportunities and that they are easier to co-invest. The second objective is to investigate whether the reinterpretation of Section 20 of the Glass-Steagall Act of 1933 (or GSA) that occurred in 1989 affected the structuring of the investments of affiliated funds of banks. The GSA has in many ways limited the performance of commercial banks. Section 20 prohibited commercial banks and their subsidiaries from being the underwriters in corporate bond issues. Over the years, there have been some attempts by Congress to soften or remove the GSA, but with little success. Regulators and banks were able to effectively soften the GSA through successive reinterpretations of their content. In 1989, the Federal Reserve (Fed) allowed some financial institutions to underwrite corporate assets (including IPOs). This permission created, exogenously, two groups of financial institutions: those that could and could not do underwriting (we call the commercial banks that obtained this authorization as subsidiaries Section 20 or SS20). This constituted an exogenous shock that affected only a portion of commercial banks and thus allows us to identify whether underwriting activity affects the behavior of banks as VC fund managers. A change in the investment style after 1989 that occurs only for the SS20 would be indicative that bank affiliates adjust their investment style in function of the activities and interests of the holding company. This also suggests the conflict of interest between these two activities. Regarding the first objective, we find that the model of VC investment of affiliates of commercial banks is dependent on the round of entry into the company. When they invest in the first round, commercial bank affiliates enter rounds up to four times larger in investments with fewer co-investors and investment rounds than independent funds. Investments of bank affiliates in companies in the first round of financing are made in sectors other than the investments of independent funds and 35% of the companies invested by banks acquire a bank loan. Co-investment in companies that have passed the first round of investment represents most of investments for VC funds. Bank affiliates make 63% of the investments in this modality (55% for independents) and invest companies 400 miles away. Bank affiliates coincide with a larger number of funds, even with lower ratio numbers. Despite some differences, the sectors invested between affiliated funds and banks are similar. Investment in VC increases the likelihood of selling other banking products such as loans, underwriting and M & A advisory. Finally, companies invested in the first round by bank affiliates seeking loans from the VC bank have a spread of 90 basis points higher than companies seeking loans from other banks. In relation to our second objective, we identified a change in the investment model of funds affiliated with banks that have SS20 as compared to the VC funds of banks without such subsidiaries. There was a drop in the percentage of investments made in the first round of the company's VC by approximately 15%, and the main reason for this decline is the 19% decrease in first-round investments made alone. Early-stage investments fell by approximately 10%, while the co-investment percentage increased by 19%. There was a decrease, both in the percentage of rounds that the bank invested and, in the total, invested in the company by 20%. The distance between bank and company has decreased by 400 miles. These changes in the investment style impacted the percentage of companies that went through IPO by more than 10%. The results suggest a change in the investment model of commercial banks SS20 after 1989. The subsidiaries invested in more mature and closer companies, co-investing in more investments. One possible explanation is the search for more mature companies that can issue debentures, shares or use the new services that SS20 have been authorized.