Essays on risk preferences, time preferences, and credit risk contagion

Abstract

This cumulative dissertation comprises two contributions on behavioral finance and one contribution on credit risk management. The first contribution examines the impact of investors’ probability distortion on the stock market and future economic growth. The empirical challenge is to quantify the optimality of today’s decisions in order to study its impact on future economic growth. Risk preferences can be estimated using stock prices. We use a time series of monthly aggregated stock prices from 1926 to 2015 and estimate risk preferences via an asset pricing model using cumulative prospect theory agents and compute a recently proposed probability distortion index. This index negatively fore- casts future GDP growth, both in-sample and out-of-sample, with stronger and more reliable predictability as the time increases. Our research results suggest that distorted stock prices can lead to significant welfare losses. The second contribution establishes empirical relation- ships of risk and time preferences on academic success. Subjects of our experiment are fourth-semester undergraduate economics students at Leibniz University Hannover. We measure academic success via the points achieved in a business exam in the 4th semester as well as the grade point average of the academic progress so far. Our methodology is based on Tanaka et al. (2010), who use a multiple price list to estimate time preferences and lotteries for the preference parameters of cumulative prospect theory. We find empirical evidence for quasi-hyperbolic discounting and a relationship between higher academic success and lower time discounting. No empirical evidence is observed for a link between risk preferences and academic performance. In the final contribution, we examine contagion effects in credit default risk defined as co-movement in the distances-to-default of U.S. firms, which we estimate from the model of Campbell et al. (2008). We quantify financial, inter-industry, and intra-industry contagion effects based on Fama and French’s 12 sectors and document significant co-movement across sectors during times of crises. We also find that a firm’s size and average share of total sales in each sector are significantly related to intra-industry contagion. Our results are robust to different crisis definitions and index weighting methodologies. Moreover, our results suggest that the probability of default increases in times of crisis due to contagion effects, which may lead to an underestimation of the risk measures of individual loans or portfolios and ultimately of economic capital.