Diversification by banks affects the systemic risk of the sector. Importantly, Wagner (2010) shows that linear diversification increases systemic risk. We consider the case of securitization, whereby loan portfolios are sliced into tranches with different seniority levels. We show that tranching offers nonlinear diversification strategies, which can reduce the failure risk of individual institutions beyond the minimum level attainable by linear diversification without increasing systemic risk.

What market features of financial risk transfer exacerbate counterparty risk? To analyze this, we formulate a model which elucidates important differences between financial risk transfer and traditional insurance, using the example of Credit Default Swaps (CDS). We allow for (heterogeneous) insurer insolvency, which captures the possibility that relatively risky counterparties may exist in the market. Further, we find that stable insurers become less stable as the price of the contract decreases. The analysis includes insured parties that have heterogeneous motivations for purchasing CDS. For example, some may own the underlying asset and purchase CDS for risk management, while others buy these contracts purely for trading purposes. We show that traders will choose to contract with less stable insurers, resulting in higher counterparty risk in this market relative to that of traditional insurance; however, a regulatory policy that removes traders can, perversely, cause stable counterparties to become less stable. We conclude with two extensions of the model that consider a Central Counterparty (CCP) arrangement and the consequences of asymmetric information over insurer type.

Taking advantage of unique longitudinal data, we provide the first characterization of what college students believe at the time of entrance about their final major, relate these beliefs to actual major outcomes, and provide an understanding of why students hold the initial beliefs about majors that they do. The data collection and analysis are based directly on a conceptual model in which a student's final major is best viewed as the end result of a learning process. We find that students enter school quite optimistic about obtaining a science degree, but that relatively few students end up graduating with a science degree. The substantial overoptimism about completing a degree in science can be attributed largely to students beginning school with misperceptions about their ability to perform well academically in science.

CEOs with substantial general managerial ability (generalist CEOs) possess a substantial share of organization (human) capital and have different risk-taking incentives than do their counterpart specialist CEOs. Using an index increasing in CEO general managerial skills as a proxy for general managerial ability, we find that investors require higher returns from firms featuring CEOs who have profuse general managerial ability. Furthermore, expected returns are significantly increasing with CEO general managerial ability in firms with high organization capital, that belong to M&A-intensive industries and that have complex operations, high agency problems and high anti-takeover provisions. These findings are consistent with arguments that organization (human) capital has significant expected return implications and that CEOs with higher general managerial skills may lead to higher agency problems, feature different risk-taking incentives and be more costly to retain in times of need.

Consistent with the predictions of Brennan and Thakor's (1990) model of shareholder preferences, we find that, on average, institutional shareholders are net sellers during share repurchases. After controlling for liquidity provision and characteristics investing, we find that a one standard deviation increase in share repurchases during a given quarter is associated with a 0.11 standard deviation decrease in institutional investor demand. We estimate that 37% of the inverse relation is attributed to institutional investors executing liquidity provision strategies, 8% is explained by institutions reacting to the investment characteristics signaled by a repurchasing firm. We attribute the majority, 55%, to the information asymmetry between institutions and individual investors. This work is one of the first to exploit the SEC mandate requiring firms to report the actual number of shares they repurchase each quarter, beginning in 2004. Using actual number of shares repurchased, we find evidence of institutional investors increasing their selling as firms increase their repurchasing. This finding is robust to models of endogeneity and autocorrelation in share repurchases and institutional investor trading.

Though common stocks are one of the most important assets in an economy, little is known about their demand curves. I estimate demand curves for 144 NYSE stocks using a unique data set of all orders, including off-equilibrium orders, during three months in 1990 and 1991. Connecting asset pricing with industrial organization, I find that stocks of firms in less competitive industries are more elastic because they have closer substitutes than stocks in more competitive industries. Tests that exploit the 1991 Gulf War shock and S&P 500 Index additions confirm these results.

This paper conducts an investigation of volatility transmission between stock markets in Hong Kong, Europe and the United States covering the time period from 2000 up to 2011. Using intra-daily data we compute realized volatility time series for the three markets and employ a Heterogeneous Autoregressive Distributed Lag Model as our baseline econometric specification. Motivated by the presence of various crisis events contained in our sample, we detect time-variation and structural breaks in volatility spillovers. Particularly during the financial crisis of 2007, we find effects consistent with the notion of contagion, suggesting strong and sudden increases in the cross-market synchronization of chronologically succeeding volatilities. Investigating the role of mean breaks and conditional heteroskedasticity in the realized volatilities, however, we find the latter to be the main driver of breaks in volatility spillovers. Taking the volatility of realized volatilities into account, we find no evidence of contagion anymore.