Essays on systemic risk and financial market volatility

Sammanfattning: This doctoral thesis consists of four independent research papers. All papers are empirical and cover the area of financial market risk, with a particular focus on systemic risk and volatility in financial markets.The first paper analyzes the joint effect of centrality and other characteristics that are essential in determining banks’ systemic importance. Specifically, we treat centrality as a moderator variable and analyze whether characteristics such as size and Value-at-Risk become more, or less, important in determining a bank’s contribution to systemic risk. Our main finding is that a bank’s contribution to systemic risk, measured by ΔCoVaR, given a certain level of VaR, is about four times higher for a bank with two standard deviations above average estimated centrality, compared to a bank with average centrality. Neglecting this indirect moderation effect severely underestimates the importance of centrality for “risky” banks and overestimates it for “safer” banks.The second paper considers the relationship between the concept of implicit government guarantees and bank equity returns. In alignment with the risk-return trade off, riskier firms should earn higher expected returns. However, risky financial institutions also pose a threat to financial stability and can be considered ‘too big to fail’. From this perspective it can be argued that the risk-adjusted expected return should be lower for highly systemic financial institutions than for less systemic institutions. The paper examines this conjecture from an asset pricing perspective by sorting bank stocks according to the systemic risk measures ΔCoVaR and MES, and compares their risk-adjusted returns. No clear evidence is found that points towards the perception that implicit government guarantees incurred lower risk-adjusted returns during the period 1987-2013 for highly systemic bank holding companies.The third paper investigates the relationship between equity volatility and financial leverage on the firm level. We use a comprehensive dataset of large syndicated loans with a total loan amount in excess of USD12tn. This allows us to identify precisely when a company experiences a large change in leverage. In contrast to several previous studies that have relied only on accounting data, we find very clear results that increased financial leverage increases equity volatility. Our findings are robust to controlling for time trends in variance as well as the type and purpose of the loan.The fourth paper considers volatility dynamics in the Bitcoin market. Bitcoin is the world’s largest cryptocurrency by market capitalization. Bitcoin is also considered extremely volatile and predicting the volatility of any currency or asset is one of the most fundamental tasks for anyone dealing with investment decisions and risk. We study Bitcoin volatility by looking at the link between the volatility in the Bitcoin market and the volatility in other related traditional markets, as well as the general risk level in the financial system. We also consider retail investor driven search volumes on Google, as a possible proxy for investor sentiment. Our main finding is that there is a relatively strong positive link between Bitcoin volatility and search pressures on Bitcoin-related words on Google, particularly for the search word “bitcoin”. Overall, our results point at retail investors, rather than large institutional investors, being major drivers of Bitcoin volatility dynamics.