WORK IN PROGRESS:
Stochasticity and Term Risk Premia
We compute durations of stock indices in the Eurozone, the UK, and Japan. This enables us to decompose expected excess returns on an index into term and stochasticity risk premia. The term risk premium is an expected excess return on a government bond which has its duration equal to the duration of the equity index and compensates for risks related to shifting a fixed amount of money in time. The stochasticity risk premium is the difference between expected return of an index and the respective term risk premium, and it compensates for randomness in the size of dividends, once the timing effect has been controlled for. The stochasticity risk premium is negative, highly volatile and has a stronger predictability by dividend-price ratio than the standard equity risk premium. We reconcile the negative sign of the stochasticity risk premium with theory by documenting that stocks provide a hedge against realized dividend growth. Our approach implies that the current practise of benchmarking equity returns to short-term yields does not provide a fair investment holding-period metric as it compares a long-term risky asset with a shot-term one. Instead, one should compare a risky long-term risky asset with a long-term bond to get a fair metric of the risk-reward, i.e., one should compare equity to a government bond of the same duration. We also analyse the implications for time and risk preferences in a Long-Run Risk model.
Foreign-dominated banking sectors, such as those prevalent in Central and Eastern Europe, are susceptible to two major sources of systemic risk: (i) linkages between local banks, and (ii) linkages between a foreign parent bank and its local subsidiary. During and after the global financial crisis, the second source of risk has been stressed by local regulators. Using a nonparametric method based on extreme value theory, we analyze interdependencies in downward risk in the banking sectors of the Czech Republic, Poland, Slovakia, and Turkey during 1994–2013. We find that the risk of contagion from a foreign parent bank to its local subsidiary is substantially smaller than the risk between two local banks.