Reproduction permitted only if source is stated.ISBN 978-3-95729-010-6 (Printversion) Non-technical summary A risk premium is the compensation demanded by investors for holding a financial asset with risky payoffs that exceeds the risk-free rate. A recent strand of literature believes that the fear of large negative shocks is a component that drives asset prices because investors expect compensation for the risk that such a rare event occurs. This paper aims at estimating equity risk premia that are due to the compensation for rare events. Rare events, such as the collapse of Lehman in September 2008, trigger large price jumps and are seldomly observed in the data which makes it hard to estimate the distribution of such events. I use a newly developed method to extract price jumps from option data and high frequency futures price data of the S&P 500 for estimating the distribution of rare events and the equity risk premia for the US stock market. In addition, the method allows for constructing an investor fear index. I replicate the method expanding the data sample to include more recent years. Furthermore, I apply the method to German data using the DAX as the proxy for the German stock market.The compensation for rare events accounts for a considerable part of the equity risk premia in both stock markets. The results are much higher than the results of similar analyses. The investor fear index works very well, as it spikes at all significant events that moved the stock markets. But the correlation of the fear index with commonly used volatility indices such as the VIX for the US market and the VDAX for the German market is about 90%. Moreover, in the financial crisis the fear index spikes only after the Lehman default, whereas indicators based on credit spreads increase sharply much earlier. Therefore, the fear index appears to be inappropriate as an early-warning tool but describes the prevalent situation in the stock markets well. Bollerslev and Todorov (2011b) to estimate risk premia for extreme events for the US and the German stock markets. The method extracts jump tail measures from high-frequency futures price data and from options data. In a second step, jump tail distributions are approximated using the extreme value theory. Applying the method to German data yields very similar results to the ones shown for the US data. The risk premia for rare events constitute a considerable part of the total equity and variance risk premia for both markets. When using the results to build an investor fear index for the US and Germany, I find that the correlation of the fear index for the US with the VIX is 89.5% and that of the fear index for Germany with the VDAX is 90.6%.
Nicht-technische Zusammenfassung