Submitted by Cristiane Shirayama (cristiane.shirayama@fgv.br) on 2011-06-02T19:26:02Z No. of bitstreams: 1 66080100240.pdf: 531933 bytes, checksum: 1949b5f6328018b30890063d38b4c352 (MD5) / Approved for entry into archive by Vera Lúcia Mourão(vera.mourao@fgv.br) on 2011-06-02T20:13:10Z (GMT) No. of bitstreams: 1 66080100240.pdf: 531933 bytes, checksum: 1949b5f6328018b30890063d38b4c352 (MD5) / Approved for entry into archive by Vera Lúcia Mourão(vera.mourao@fgv.br) on 2011-06-02T20:37:11Z (GMT) No. of bitstreams: 1 66080100240.pdf: 531933 bytes, checksum: 1949b5f6328018b30890063d38b4c352 (MD5) / Made available in DSpace on 2011-06-03T17:31:45Z (GMT). No. of bitstreams: 1 66080100240.pdf: 531933 bytes, checksum: 1949b5f6328018b30890063d38b4c352 (MD5) Previous issue date: 2010-12-13 / The main purpose of this study is to identify the determinants of a Receivables Fund (FIDCs) spread. As the previously published papers were about debentures, this one is pioneer in terms of FIDCs. Indeed, understanding the determinants of a FIDCs’ issuance is mandatory for companies which access capital markets to obtain funds. Between 2002 and 2009, the study analyzed 113 public issuances indexed to CDI that had a pre-defined maturity and at least two investors. Those issuances were subject to four Ordinary Least Squares (OLS) regressions. The first regression focused on the variables that affected the rating. The second one excluded all rating related variables and had spread as the dependent variable, while all the others were independent variables. The third one tested a potential non-linearity relation between spread and rating, while the fourth regression tested the spread against all other variables, including those ones related to the rating. The results pointed out that rating is indeed a good determinant for the spread, as well as the issuance volume, the financial institution responsible for the FIDCs’ custody and the macroeconomic environment. These three variables also contributes to determine the rating, along with the number of investors and issuers and if the underlying assets were eventually delivered. The study also concluded that there is no evidence of linearity between spread and rating in FIDCs. The major contribution of this study is to present on first hand a thorough analysis on variables that affect a new type of fixed income investment, the FIDCs, such as the asset nature, credit risk diversification levels and quantity of issuers. The main purpose of this study is to identify the determinants of a Receivables Fund (FIDCs) spread. As the previously published papers were about debentures, this one is pioneer in terms of FIDCs. Indeed, understanding the determinants of a FIDCs’ issuance is mandatory for companies which access capital markets to obtain funds. Between 2002 and 2009, the study analyzed 113 public issuances indexed to CDI that had a pre-defined maturity and at least two investors. Those issuances were subject to four Ordinary Least Squares (OLS) regressions. The first regression focused on the variables that affected the rating. The second one excluded all rating related variables and had spread as the dependent variable, while all the others were independent variables. The third one tested a potential non-linearity relation between spread and rating, while the fourth regression tested the spread against all other variables, including those ones related to the rating. The results pointed out that rating is indeed a good determinant for the spread, as well as the issuance volume, the financial institution responsible for the FIDCs’ custody and the macroeconomic environment. These three variables also contributes to determine the rating, along with the number of investors and issuers and if the underlying assets were eventually delivered. The study also concluded that there is no evidence of linearity between spread and rating in FIDCs. The major contribution of this study is to present on first hand a thorough analysis on variables that affect a new type of fixed income investment, the FIDCs, such as the asset nature, credit risk diversification levels and quantity of issuers. / Este trabalho tem por objetivo identificar fatores que influenciam o spread de cotas seniores de Fundos de Investimento em Direitos Creditórios (FIDC). Trata-se de um estudo pioneiro no segmento de renda fixa, uma vez que os anteriores focaram o spread para debêntures. Entender componentes do spread em FIDCs é muito importante para as empresas que captam recursos no mercado de capitais por meio deste novo instrumento. A análise contemplou 113 ofertas públicas indexadas ao CDI entre 2002 e 2009 que tinham prazo determinado e pelo menos 02 investidores adquirentes das cotas. Foram elaboradas quatro regressões múltiplas pelo método dos Mínimos Quadrados Ordinários (MQO). A primeira visava identificar quais variáveis afetavam o rating. A segunda tinha como variável dependente o spread e as independentes eram todas, excluindo as que afetavam o rating. A terceira equação testou uma possível relação de não-linearidade entre spread e rating, enquanto a quarta testou o spread contra todas as demais variáveis, incluindo as que afetavam o rating. Os resultados apontaram que o rating é sim um bom determinante para o spread, assim como o volume da emissão, ambiente econômico e instituição financeira que faz a custódia do fundo. Da mesma forma, são fatores que explicam o rating: ambiente econômico, quem são o administrador e o custodiante do fundo, o número de investidores, volume, se o fundo é multi ou mono-cedente e se os ativos são performados ou não. Não há linearidade entre spread e rating em FIDCs. A maior contribuição deste estudo foi apresentar pela primeira vez variáveis que afetam um novo tipo de investimento em renda fixa, os FIDCs, tais como tipo de ativo, níveis de concentração e fundos mono e multi-cedentes. Tudo isto por meio de um cuidadoso tratamento da base de dados.

Submitted by Cristiane Shirayama (cristiane.shirayama@fgv.br) on 2011-06-03T16:54:39Z No. of bitstreams: 1 66080100253.pdf: 329789 bytes, checksum: c3439e3a0ed91b11125a608ae7104c30 (MD5) / Approved for entry into archive by Vera Lúcia Mourão(vera.mourao@fgv.br) on 2011-06-03T16:58:22Z (GMT) No. of bitstreams: 1 66080100253.pdf: 329789 bytes, checksum: c3439e3a0ed91b11125a608ae7104c30 (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia(suzinei.garcia@fgv.br) on 2011-06-03T17:02:51Z (GMT) No. of bitstreams: 1 66080100253.pdf: 329789 bytes, checksum: c3439e3a0ed91b11125a608ae7104c30 (MD5) / Made available in DSpace on 2011-06-03T18:11:30Z (GMT). No. of bitstreams: 1 66080100253.pdf: 329789 bytes, checksum: c3439e3a0ed91b11125a608ae7104c30 (MD5) Previous issue date: 2011-08-08 / This study aims to examine empirically some of the main characteristics of securitization structures in the Brazilian market. Based on the Gorton and Souleles (2005) article and analyzing characteristics of companies that use securitization in Brazil, we test whether firms with higher credit risk tend to securitize more and, assessing the characteristics that make these structures feasible, if there is evidence of an implicit contract between the sponsor companies and investors. The database used in the empirical tests was built using 59 FIDCs emissions recorded from January 2005 through July 2010 and a base with the accounting data from 86 banks in 8 semesters. The methods used in the empirical tests were panel data regressions including estimators for fixed and random effects. In the empirical tests we have found evidence that the rating of the sponsor companies has influence in the spread charged in FIDCs, which can be understood as a evidence of the existence of an implied contract. We also find evidence, albeit less convincingly, that companies with worse ratings securitize more. These results are aligned with those found in the paper by Gorton and Souleles (2005). / Este trabalho tem por objetivo analisar algumas das características principais das estruturas de securitização no mercado brasileiro. Utilizando como principal referência o artigo de Gorton e Souleles (2005) e, através da análise das características das empresas que utilizam securitização no Brasil, testamos se empresas com maior risco de crédito tendem a securitizar mais. Avaliando as características que fazem com que tais estruturas sejam viáveis, testamos também se há evidências de um contrato implícito entre as empresas cedentes e os investidores. Na base de dados utilizada para os testes foram consideradas 59 emissões de FIDCs registradas de janeiro de 2005 a julho de 2010 e uma base com dados contábeis de 86 bancos em 8 semestres. Devido à estrutura dos dados, foram utilizados métodos para regressão com dados em painel utilizando estimadores para efeitos fixos e aleatórios. Nos testes empíricos foram encontradas evidências de que o rating das empresas cedentes influencia o spread cobrado nos FIDCs, o que pode ser entendido como uma forma de evidenciação da existência de um contrato implícito. Também evidenciamos, embora de forma menos contundente, que empresas com piores ratings securitizam mais.Tais resultados mostram-se alinhados com os encontrados no artigo de Gorton e Souleles (2005) .