Robust statistical modeling of portfolios

Atypical points in the data may result in meaningless e±cient frontiers. This follows since portfolios constructed using classical estimates may re°ect neither the usual nor the unusual days patterns. On the other hand, portfolios constructed using robust approaches are able to capture just the dynamics of the usual days, which constitute the majority of the business days. In this paper we propose an statistical model and a robust estimation procedure to obtain an e±cient frontier which would take into account the behavior of both the usual and most of the atypical days. We show, using real data and simulations, that portfolios constructed in this way require less frequent rebalancing, and may yield higher expected returns for any risk level.

Toward convergence? The brazilian stock market concentration and the role of private equity in its diversification

Using quantitative data obtained from public available database, this paper discusses the difference between of the Brazilian GDP and the Brazilian Stock Exchange industry breakdown. I examined if, and to what extent, the industry breakdowns are similar. First, I found out that the Stock Exchange industry breakdown is overwhelming different from the GDP, which may present a potential problem to asset allocation and portfolio diversification in Brazil. Second, I identified an important evidence of a convergence between the GDP and the Stock Exchange in the last 9 years. Third, it became clear that the Privatizations in the late 90’s and IPO market from 2004 to 2008 change the dynamics of the Brazilian Stock Exchange. And fourth, I identified that Private Equity and Venture Capital industry may play an important role on the portfolio diversification in Brazil.

Mídias digitais, viralização e reputação corporativa: a percepção de executivos de agências de mídias digitais

As mídias sociais vêm ganhando grande importância nos últimos anos, e transformando a maneira como as pessoas se articulam, se engajam ou simplesmente trocam informações a respeito de todos os assuntos. A evolução das tecnologias móveis de comunicação, cada vez mais robustas, e a disseminação de smartphones, aparatos modernos e completos para a convergência de voz e imagem, têm cumprido um papel importante no contexto de conexão permanente das pessoas, com tudo e com todos. Essa pesquisa se propõe a debater como campanhas de boca-a-boca (eWOM) no Facebook (a maior mídia social de todas) vêm impactando a gestão de reputação das corporações e de imagem de marcas, a partir de pesquisa de campo que capturou a visão de executivos de agências de mídia digital, complementada por pesquisa secundária para a análise de experiências vividas por algumas empresas de grande visibilidade. Os resultados demonstram que as mídias sociais tornaram mais complexo o processo de gestão de reputação, que está cada vez mais fora do controle absoluto das organizações e mais compartilhado com os seus públicos de interesse. Indicam, ainda, que as mídias sociais podem representar mais oportunidades para as organizações que se prepararem para elas e mais ameaças para as que forem em sentido contrário.

Convergence analysis of sampling-based decomposition methods for risk-averse multistage stochastic convex programs

We consider a class of sampling-based decomposition methods to solve risk-averse multistage stochastic convex programs. We prove a formula for the computation of the cuts necessary to build the outer linearizations of the recourse functions. This formula can be used to obtain an efficient implementation of Stochastic Dual Dynamic Programming applied to convex nonlinear problems. We prove the almost sure convergence of these decomposition methods when the relatively complete recourse assumption holds. We also prove the almost sure convergence of these algorithms when applied to risk-averse multistage stochastic linear programs that do not satisfy the relatively complete recourse assumption. The analysis is first done assuming the underlying stochastic process is interstage independent and discrete, with a finite set of possible realizations at each stage. We then indicate two ways of extending the methods and convergence analysis to the case when the process is interstage dependent.