Conditioned Multi-Asset Models

The greatest value creation takes place at the asset allocation level.

Strategic asset allocation and its tactical management over economic cycles and changing risk regimes on the capital market determine more than 90 % of the long-term performance of an investment. Although investors are increasingly including illiquid asset classes in their portfolios, the international equity and bond markets are still the main basis for harvesting systematic risk premiums in investments. The relative performance of equity and bond markets in different regions, currencies and economic areas is determined by a variety of macroeconomic variables and fundamental valuation ratios. Investors use these as a guide when investing new funds or switching between asset classes in their portfolios. Thus, the management of a multi-asset portfolio requires a continuous evaluation of the countless factors influencing the expectations of market players, embedded in a disciplined tactical allocation process.

Our models learn the tactical allocation preferences of investors.

Our models analyze the complex, constantly changing interactions between valuation-relevant data and price developments on international stock and bond markets. The asset classes with the highest risk premiums are continuously identified as the basis for tactical portfolio decisions. Machine learning is carried out in the context of the conditioned valuation models that have been tried and tested in practice for many years, whereby the data space for conditioning risk premiums is larger by dimensions than in the context of econometric approaches. In addition, non linear relationships between influencing variables as well as possible structural breaks in cause-and-effect relationships are captured by the learning algorithms.