This dissertation contains three essays on the non-pecuniary preferences pertaining to financial asset characteristics and their implications for asset pricing. The first essay considers the pricing implications of screens adopted by socially responsible investors. A model including such investors reconciles the empirically observed risk-adjusted sin-stock abnormal return with a systematic “boycott risk premium” which has a substantial financial impact that is, however, not limited to the targeted firms. The boycott effect cannot be displaced by litigation risk, a neglect effect, and liquidity considerations, or by industry momentum and concentration. The boycott risk factor is valuable in explaining cross-sectional differences in mean returns across industries and its premium varies directly with the relative wealth of socially responsible investors and with the business cycle. The second essay generalizes Fama (1996)’s concept of Multi-Factor Efficiency without being limited by additional random state variables that must affect future investment opportunities. Incorporating non-pecuniary preferences into a representative investor’s utility function generates multi-factor pricing implications. A representative investor chooses among expected returns, variances, and levels of characteristics according to their taste, which gives rise to an N-fund separation theorem with static characteristics. If a portfolio is built to maximize the exposure to the asset characteristics, the covariance between asset returns and this portfolio returns will be identical to the underlying characteristics. Such identity makes obsolete any attempts to distinguish between characteristics and risk exposures as the driving forces behind the cross-sectional variation in stock returns. The third essay develops a procedure for deriving systematic factors from characteristics, based on maximizing each factor’s exposure to a characteristic subject to a given level of factor variance. The resulting characteristic-mimicking portfolios (CMP) price mean asset returns identically as the original characteristics, irrespective of the underlying model. Accordingly, differences in the performance of mimicking factors and characteristics in explaining mean returns should be interpreted as an artifact of arbitrary procedural choices for generating mimicking factors. Factors and characteristics may be distinguished usefully only by determining if CMPs have significant explanatory power for the time series of returns. / Thesis / Doctor of Philosophy (PhD)

Submitted by Laércio Fernandes Pereira Neto (laercio_fene@hotmail.com) on 2016-02-10T17:04:16Z No. of bitstreams: 1 MPFE_Dissertacao_Laercio Neto.pdf: 920392 bytes, checksum: 0cdf978b48230f22a7a4597db4311e21 (MD5) / Rejected by Renata de Souza Nascimento (renata.souza@fgv.br), reason: Laércio, boa tarde Referente ao título de sua dissertação, houve alguma alteração? Conforme o protocolo e ata entregue na secretaria, o título é: ANÁLISE DOS RISCOS SISTEMÁTICOS E IDIOSSINCRÁTICOS, REPRESENTADOS PELO CAPM, PARA PORTFOLIOS DE DIFERENTES SETORES EM CONDIÇÕES ECONÔMICAS DISTINTAS A alteração do mesmo, só deve ser realizada mediante solicitação do orientador e não possuímos esta solicitação em Ata. Por gentileza, verificar. Att on 2016-02-11T17:52:21Z (GMT) / Submitted by Laércio Fernandes Pereira Neto (laercio_fene@hotmail.com) on 2016-02-12T19:38:57Z No. of bitstreams: 1 MPFE_Dissertacao_Laercio Neto.pdf: 920804 bytes, checksum: 4bb1317851e085d3c4542b6e2caec6eb (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2016-02-12T19:40:58Z (GMT) No. of bitstreams: 1 MPFE_Dissertacao_Laercio Neto.pdf: 920804 bytes, checksum: 4bb1317851e085d3c4542b6e2caec6eb (MD5) / Made available in DSpace on 2016-02-15T11:45:46Z (GMT). No. of bitstreams: 1 MPFE_Dissertacao_Laercio Neto.pdf: 920804 bytes, checksum: 4bb1317851e085d3c4542b6e2caec6eb (MD5) Previous issue date: 2016-02-04 / Esse estudo busca analisar os impactos causados pelo Ciclo Monetário, o Ciclo Econômico, o Nível da Indústria e a Condição do Mercado de Ações nas variáveis do CAPM para portfólios de ações de diferentes setores da indústria no Brasil. O banco de dados utilizado compreende séries temporais mensais, do retorno de ações de 17 setores da economia, no período de Janeiro de 2008 à Dezembro de 2014. Foi observado que existem relações estatisticamente relevantes entre variáveis macroeconômicas e o excesso de retorno dos portfólios de ações analisados. Além disso, foi possível notar que essas relações tem efeitos distintos sobre os riscos sistemáticos e idiossincráticos dos portfólios. A maior parte dos resultados obtidos não se mostraram estatisticamente relevantes, o que sugere que existem outras variáveis explicativas que se relacionam com as variáveis dependentes de forma mais robusta, assim como já apontado pela literatura existente, no caso do Brasil. No entanto, foi possível observar que há efeitos indiretos das variáveis macroeconômicas sobre o retorno dos ativos através do canal de retorno do mercado. / This study investigates the impacts caused by the Monetary Cycle, the Business Cycle, the Industry Level and the Condition of the Stock Market on CAPM variables for stock portfolios of different industry sectors in Brazil. The database used comprises monthly time series of stock returns of 17 sectors of the economy in the period from January 2008 to December 2014. It was observed that there are statistically significant relationships between macroeconomic variables and the excess return of the stock portfolios analyzed. Furthermore, it was noticeable that these relationships have different effects on systematic and idiosyncratic risk of portfolios. Most of the results were not statistically relevant, which suggests that there are other explanatory variables that relate to the dependent variables in a more robust manner, as already pointed out by the literature in the case of Brazil. However, it was observed that there are indirect effects of macroeconomic variables on the return of assets through the market return channel.

本研究以中國大陸上市公司為研究對象，探討股票報酬與公司治理間之關係。此外，為了檢視中國大陸公司治理是否為股票市場上之系統風險，本研究採用學術上廣為使用之四因子模式（Carhart 1997）作為基本迴歸模型，藉由控制住影響大部分股票變異之四因子，純粹探討公司治理指標對於股票報酬之解釋能力。最後，本研究進一步比較各公司治理指標對於股票報酬變異之解釋力相對強弱為何。 實證結果發現： （1）國家股股東持股比率相對較高、董事會規模相對較大、董事長兼任總經 理之公司具有較高股票報酬；而機構法人持股比率相對較高、公眾股股東持股比率相對較高、獨立董事占董事會比率相對較高之公司則具有較低股票報酬。 （2）絕大部分公司治理指標對於超額股票報酬具有顯著影響力，本研究進而推論中國大陸公司治理為股票市場上之系統風險之一。然而，公司治理指標對於提升四因子模式解釋力之程度相當有限。 （3）在原本四因子模式中加入「國家股股東持股比率溢酬因子」、「獨立董 事占總董事人數溢酬因子」二公司治理指標，會比加入「公眾股股東持股比率溢酬因子」、「董事會規模溢酬因子」指標具有較佳之模式解釋力。而在模式中加入「公眾股股東持股比率溢酬因子」、「董事會規模溢酬因子」二公司治理指標又比「機構法人持股比率溢酬因子」指標具有較佳之模式解釋力。 / This study investigates the relation between corporate governance and stock returns in China’s listed companies. Additionally, I apply four-factor model (Carhart 1997) to examine whether China’s corporate governance mechanisms are systematic risks in stock market. At last, I compare the explanation power of excess returns among all corporate governance indexes. I find that: (1) Firms, with higher level of nation ownership, larger board size and dual roles of chairman and managing director, have higher returns; firms ,with higher level of legal person ownership, of public ownership, of independent directors’ ratio, have lower returns. (2) Most Corporate governance indexes have significant impacts on excess stock returns, so we infer that corporate governance in China is one of systematic risks in stock market. However, I also find that corporate governance indexes add few margin contributions to four-factor model. (3) Governance indexes of nation ownership and of independent directors’ ratio have more impact on stock returns than the index of public ownership and of board size. Meanwhile, index of public ownership and of board size have more impact on stock returns than the index of legal person ownership.