Thesis (Ph.D.)--Hong Kong University of Science and Technology, 2007. / Includes bibliographical references (leaves 130-134). Also available in electronic version.

This thesis presents three empirical analyses on the systematic risk exposure that global and domestic asset holders face. Each paper investigates a distinct source of macroeconomic risk, but they all stem from the premise that holding human capital and financial assets is risky. This ultimately affects agents' optimal consumption choices. The first paper, The International CAPM Redux, proposes a novel empirical model to price international assets. Building on recent advances in asset pricing research on currency markets, it documents that investors are compensated for bearing exposure to currency risk when investing in foreign equity, either directly or via delegated portfolios of international assets. The second paper, One Central Bank To Rule Them All, shows that while global investors demand a premium to bear risks associated with Federal Reserve decisions, there is no comparable result for other major central banks. This puzzling finding points to the uniqueness of the Federal Reserve for global investors, that does not simply stem from the size and importance of the U.S. economy. The third paper, Human Capital, Unemployment Risk and Asset Prices, relates the riskiness of human capital to uncertainty in the labour market and documents a role for unemployment as a determinant of human wealth. Sorting U.S. industry-level portfolios by differential exposure to unemployment risk yields a novel cross-section of average excess returns. High risk-premia are consistent with less income-constrained highly educated workers willing to take financial risk beyond hedging their labour income risk. Taken together, these three studies contribute to the empirical asset pricing literature and open the door to further theoretical and empirical research in the field.

Banking has a unique role in the well-being of an economy. This role makes banks one of the most heavily regulated and supervised industries. In order to strengthen the soundness and stability of banking systems, regulators require banks to hold adequate capital. While credit risk was the only risk that was covered by the original Basle Accord, with the 1996 amendment, banks have also been required to assign capital for their market risk starting from 1998. In this research, the impact of the market risk capital regulations on bank capital levels and derivative activities is investigated. In addition, this study also evaluates the impact of using different approaches that are allowed to be used while calculating the required market risk capital, as well as the accuracy of VaR models. The implementation of the market risk capital regulations can influence banks either by increasing their capital or by decreasing their trading activities and in particular trading derivative activities. The literature review concerning capital regulations illustrates that in particular the impact of these regulations on bank capital levels and derivative activities is an issue that has not yet been explored. In order to fill this gap, the changes in capital and derivatives usage ratios are modelled by using a partial adjustment framework. The main results of this analysis suggest that the implementation of the market risk capital regulations has a significant and positive impact on the risk-based capital ratios of BHCs. However, the results do not indicate any impact of these regulations on derivative activities. The empirical findings also demonstrate that there is no significant relationship between capital and derivatives. The market risk capital regulations allow the use of either a standardised approach or the VaR methodologies to determine the required capital amounts to cover market risk. In order to evaluate these approaches, firstly differences on bank VaR practices are investigated by employing a documentary analysis. The documentary analysis is conducted to demonstrate the differences in bank VaR practices by comparing the VaR models of 25 international banks. The survey results demonstrate that there, is no industry consensus on the methodology for calculating VaR. This analysis also indicates that the assumptions in estimating VaR models vary considerably among financial institutions. Therefore, it is very difficult for financial market participants to make comparisons across institutions by considering single VaR values. Secondly, the required capital amounts are calculated for two hypothetical foreign exchange portfolios by using both the standardised and three different VaR methodologies, and then these capital amounts are compared. These simulations are conducted to understand to what extent the market risk capital regulations approaches produce different outcomes on the capital levels. The results indicate that the VaR estimates are dependent upon the VaR methodology. Thirdly, three backtesting methodologies are applied to the VaR models. The results indicate that a VaR model that provides accurate estimates for a specific portfolio could fail when the portfolio composition changes. The results of the simulations indicate that the market risk capital regulations do not provide a `level playing field' for banks that are subject to these regulations. In addition, giving an option to banks to determine the VaR methodology could create a moral hazard problem as banks may choose an inaccurate model that provides less required capital amounts.

The main focus of this thesis is the analysis of channels used by Czech banks to increase their capital ratios. We identify the increase in capital as the main channel used. Further, within these channels we find retained earnings to be the main tool used. In addition, growth in the loans volume was the dominant tool within the channel of asset volume. Furthermore, we observe an increase in the use of more advanced capital requirements quantification methods, especially in larger banks. We also identify several factors, other than capital regulation, that might have contributed to the observed developments.

Submitted by Noeme Timbo (noeme.timbo@metodista.br) on 2018-02-23T19:50:11Z No. of bitstreams: 1 Giovana Garcia Firmino.pdf: 1487977 bytes, checksum: 1b975dfb33dc625a7208e9514026843c (MD5) / Made available in DSpace on 2018-02-23T19:50:11Z (GMT). No. of bitstreams: 1 Giovana Garcia Firmino.pdf: 1487977 bytes, checksum: 1b975dfb33dc625a7208e9514026843c (MD5) Previous issue date: 2017-08-23 / This study investigates the relationship between cost of capital and results management, a measure related to earnings quality in Brazilian publicly traded companies listed in B3 - Brasil Bolsa Balcão in the period from 2009 to 2016. The general objective of the present study is to analyze if there is a negative association between the management of results and the cost of capital and third parties. To measure the cost of equity, beta-β is used as a measure of the systematic risk, and as a measure of the cost of third-party capital, the variable Ki obtained from the ratio between financial expenses and the average of the onerous liabilities. As a proxy for results management, we use the Earning Quality variable derived from the measurement of results management by discretionary accruals obtained from three models for greater robustness of results: the Modified Jones model (1995), the model Kang and Sivaramakrishnan (1995) ) and the model McNilchols (2002). By means of a fixed effects panel and application of MQO models treated by means of robust standard errors in order to eliminate possible biases of heteroscedasticity and multicollinearity, EQ measurements were obtained for each of the proposed models and then they were tested in relation with the cost of capital and third parties by means of econometric models composed of control variables for the year 2016. The obtained results do not confirm the hypotheses proposed that there is a negative relation between EQ and cost of capital, contrary to previous works as that of Persakis and Iatridis (2015), opening the way for future research on the influences of the economic context on this relation. / Este estudo investiga a relação entre o custo de capital e o gerenciamento de resultados, medida relacionada com a qualidade do resultado (earnings quality) nas empresas brasileiras de capital aberto listadas na B3 – Brasil Bolsa Balcão no período de 2009 à 2016. O objetivo geral do presente trabalho é analisar se existe uma associação negativa entre o gerenciamento de resultados e o custo de capital próprio e de terceiros. Para mensuração do custo de capital próprio utiliza-se o beta – β, medida do risco sistemático, e como medida do custo de capital de terceiros a variável Ki obtida da razão entre as despesas financeiras e a média do passivo oneroso. Como proxy de gerenciamento de resultados utiliza-se a variável Earning Quality derivada da mensuração do gerenciamento de resultados pelos accruals discricionários obtidos a partir de três modelos para maior robustez dos resultados: o modelo Jones Modificado (1995), o modelo Kang e Sivaramakrishnan (1995) e o modelo McNilchols (2002). Por meio de painel de efeitos fixo e aplicação de modelos MQO, tratados por meio de erros padrão robustos a fim de eliminar possíveis vieses de heterocedasticidade e multicolinearidade, foram obtidas as medidas EQ para cada um dos modelos propostos e em seguida as mesmas foram testadas na relação com o custo de capital próprio e de terceiros por meio de modelos econométricos compostos por variáveis de controle para o ano de 2016. Os resultados obtidos não confirmam as hipóteses propostas de que existe uma relação negativa entre EQ e custo de capital, contrariando trabalhos anteriores como o de Persakis e Iatridis (2015) abrindo campo para futuras pesquisas sobre as influências do contexto econômico sobre esta relação.