Seleção ótima de portfólio em renda fixa: combinações de estruturas a termo da taxa de juros e fatores macroeconômicos

• Ano de defesa: 2019
• Autor(a) principal: Gimenes, Lucas Dreves
• Orientador(a): Eid Júnior, William
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Renda fixa; Portfólios
• Palavras-chave em Inglês: Fixed income; Portfolios
• Link de acesso: https://hdl.handle.net/10438/28149

Markowitz's (1952) study of optimal mean variance portfolios are widely used for portfolio formation in equities. However, they are little employed for fixed income. Korn and Koziol (2006) present that one of the reasons for this fact is the changing characteristic of titles over time. They suggest that, to mitigate this issue, term interest rate structure models should be used. For Christensen, Diebold and Rudebusch (2011), asset pricing, portfolio allocation and risk management are key tasks in the financial markets. For fixed income securities, the more efficient term structure modeling of interest rates (ETTJ) tends to yield better pricing, higher portfolio returns and satisfactory risk management. Similar design can also be found in Bolder (2015). Also, according to Christensen, Diebold and Rudebusch (2011), the term structure models of interest rates based on Nelson and Siegel (1987) are remarkably successful in adjusting and forecasting the yield curves. For the Brazilian case, Almeida (2009) presents a superior methodology based on Nelson-Siegel-Svensson models, demonstrating that one more factor for curvature generates better fit and prediction. Macroeconomic variables as potentialization factors for better predictions gained strength after the work of Ang and Piazzesi (2003) and Diebold, Piazzesi and Rudebusch (2005). Rudebusch and Wu (2008) suggest as an improvement, the importance of analyzing fiscal variables as a way of better adjusting or forecasted rates. For the Brazilian case, Almeida and Faria (2014) demonstrate the importance of including macroeconomic factors to forecast the Brazilian yield curve. Vieira et al. (2017) explore the gains by including variables that try to capture expectations. Thus, the inclusion of macroeconomic factors in the ETTJ estimation has become important for both monetary policy makers and debt holders as it assists in decision making with a view to maximizing profits. Thus, the present paper joins the above points and seeks to study the importance of macroeconomic expectation variables, focusing on fiscal expectation variables as possible instruments to improve the adjustment and forecasting of estimated yield curves via Nelson Siegel Svensson and subsequent use in great fixed income portfolios. The data used are DI futures with maturities of 1, 2, 3, 4, 5 and 6 months combined with those of 1, 1.5, 2, 2.5, 3, 4, 5, 7 and 10 years. The main results suggest that, for the Brazilian case, there is relevance in future expectation fiscal variables together with product and inflation expectations for a better forecast of the yield curves. And that optimal portfolio strategies (mean variance) in fixed income, using forecasted data for future interest rate behavior, can lead to consistent, even better absolute returns on the Sharpe Index, and these results hold up compared to Brazilian Fixed Income fund industry.