The first essay examines the impact of investor protection, market monitoring, and liquidity on the firm-level and country-level earnings management using a sample of 432 firms from 34 countries cross-listed in the U.S. The major findings are as follows: First, cross-listed firms from countries with strong legal system, strong outside investor rights, more institutional investors, and higher financial transparency are less likely to engage in earnings management. In addition, in countries with strong investor protection or market monitoring, the level of earnings management is more pronounced for illiquid firms as compared to liquid firms. Second, cross-listed firms following IFRS have lower propensity in earnings management than those following the U.S. GAAP. Third, the degree of earnings management for cross-listed firms is greater in the home country than in the U.S. market. Fourth, cross-listed firms have higher earnings management in the pre-listing period than in the post-listing period. Fifth, foreign firms listed in U.S. major markets have lower propensity to engage in earnings management than those listed in the OTC market. The findings remain robust with the inclusion of industry fixed effects and GMM estimation. All findings are largely consistent with my hypotheses that better investor protection, greater market monitoring, and higher liquidity reduce the extent of earnings management. The second essay examines the relative contribution to price discovery process of EURO/USD currency futures traded on two major exchanges: Chicago Mercantile Exchange (CME) and Intercontinental Exchange (ICE), using the intraday data in 2010. The relative contribution to price discovery is estimated using the information share approach of Hasbrouck (1995). Empirical findings indicate that CME accounts for approximately 87% of price discovery in the EURO/USD market and its contribution is substantially larger in the morning than that in the afternoon. This study also examines the effect of trading characteristics, including volume, quoted bid-ask spread, and price volatility, on information share. CME’s price discovery leadership is attributed to its high trading activity, low transaction costs, and lower volatility. The results support the liquidity hypothesis that a market with greater liquidity contributes more to price discovery.

Submitted by Yurie Yassunaga Suzuki (yurieyassunaga@gmail.com) on 2015-09-03T02:42:33Z No. of bitstreams: 1 Dissertação - YYS.pdf: 638348 bytes, checksum: 7a31f5f0e578eed37d240d302f503e27 (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2015-09-03T16:04:18Z (GMT) No. of bitstreams: 1 Dissertação - YYS.pdf: 638348 bytes, checksum: 7a31f5f0e578eed37d240d302f503e27 (MD5) / Made available in DSpace on 2015-09-03T16:44:56Z (GMT). No. of bitstreams: 1 Dissertação - YYS.pdf: 638348 bytes, checksum: 7a31f5f0e578eed37d240d302f503e27 (MD5) Previous issue date: 2015 / This work aims to study market behavior involving Petrobras’ stock and options markets applying price discovery methodology. Using high-frequency data, provided by BM&FBOVESPA, econometric models used in this methodology were estimated and measures of Information Share (IS) and Component Share (CS) were calculated. The results of the analyzes indicated dominance of the spot market in the process of price discovery, since, for this market, were observed values over 66% for IS and above 74% for CS. Graphical analysis of the impulse response function indicated that the spot market is more efficient than the option market. / Este trabalho tem como objetivo estudar o comportamento do mercado de ações e opções de Petrobrás utilizando a metodologia de price discovery (descoberta de preços). A partir de dados de alta frequência de ambos os mercados, fornecidos pela BM&FBOVESPA, os modelos econométricos utilizados nessa metodologia foram estimados e as medidas de Information Share (IS) e Component Share (CS) foram calculadas. Os resultados das análises indicaram dominância do mercado à vista no processo de descoberta de preços, dado que, para este mercado, foram observados valores acima de 66% para a medida IS e acima de 74% para a medida CS. Análises gráficas da função resposta ao impulso indicaram, também, que o mercado à vista é o mais eficiente.

<div>In the first chapter, our empirical tests use data from three sources. First, we obtain the Loughran and McDonald’s (hereafter LM wordlist) positive/negative wordlist and from the authors’ website. Second, we obtain the monthly Fama and French (1993 and 2015) factors (i.e. SMB, HML, Rm-Rf, CMA, and RMW) and momentum factor (MOM) from Kenneth French’s website for the sample period January 1994 through December 2016. Third, we obtain the monthly stock returns, monthly index returns, month end market value from the Center for Research in Security Prices (CRSP) as well as accounting information such as annual book</div><div>I the second chapter, we utilize five main datasets in this study. The first dataset is the stock market transaction level data for S&P 500 stocks, obtained from Trades and Quote (TAQ). The second dataset is the corporate bond transaction data from Trade Reporting and Compliance Engine (TRACE) through Wharton Research Data Services (WRDS) for the S&P 500 firms. The TRACE data provides over the counter (OTC) corporate bond market real-time prices.To examine the price discovery of bonds in equity prices we use a sample period of over 1,000 trading days from January 2004 through December 2008.</div><div>Our third data source is the institutional level transaction data from ANcerno, which provides transactional level trade data for corporate bonds and stocks for the first quarter of 2006 through the third quarter of 2010. Several studies have used equity transaction dataset to examine the ANcerno institutional trading behavior. See for example Puckett and Yan (2011), Bethel, Hu and Wang (2009), Chemmanur, He and Hu (2009), Goldstein, Irvine, Kandel and Wiener (2009). Additionally, Hu, Jo, Wang and Xie (2018) provide a comprehensive review of ANcerno dataset. The fourth source of data comes from Mergent Fixed Income Security Database (FISD), which provides details of bond characteristics and credit ratings from standard and poor’s (S&P) and Moody’s. Finally, we obtain the daily stock returns data from center for security prices (CRSP) database and match it with the daily bond returns to examine the lead-lag relationships.</div><div><br></div>