Topic: Contributed on Econometric challenges in risk management Title: Time-dependent Black-Litterman Authors:  Martin van der Schans - Ortec Finance (Netherlands) [presenting]
Abstract: Many (institutional) investors support their investment decisions with models that forecast future risk and return. In practice, it is often necessary to combine these models with views, e.g. for stress-testing purposes or because the views contain additional forward-looking information. With the Black-Litterman method, views can be combined with models in one-period setups: a density forecast for returns over the investor's investment horizon is combined with the investor's views expressed as expected returns over the investor's investment horizon. In practice, however, this setup is often too limited: the investor's views are formulated for only the near future while the investor's investment horizon is much longer. Using a time-series modelling approach, we extend the original Black-Litterman method to a time-dependent Black-Litterman method which is suitable for time-dependent setups: time-dependent views can be combined with time-dependent density forecasts varying over the investor's investment horizon. Through a stylized example, we show how the method can be used by institutional and other long-term investors in their risk management process.