Innovative scientific financial market research is the cornerstone of our success. Our proprietary research work is based on modern econometric approaches as well as cutting edge methods and techniques.
Our research group is headquartered in Berlin, Germany. Berlin, in particular, stands out due to its excellent scientific environment. Three universities, several Max-Planck institutes and other research facilities guarantee a pool of talented researchers. Moreover, we maintain a lively exchange with selected institutes on a regular basis.
The research of GA Asset Management has a clear focus on the development of investment strategies and can be divided into three primary fields:
The main goal of our research is to develop market models with the ability to accurately reflect various statistical properties of empirical financial data. With the help of a market model we can make statements about the probability of the direction of the drift of a market or describe the temporal development of the volatility. Depending on the model we might additionally be able to estimate the likelihood of very unlikely events, a market crash for example. An accurate market model allows the comprehensive risk assessment or the construction of a truly efficient portfolio.
- Enhancement of standard market models (HMM, GARCH)
- Detection of structural changes
- Adaptive adjustment to non-stationarities (switching models)
- Intraday models
- Modeling of jump processes
- Modeling of derivative prices
A second essential aspect of our research is the improvement of risk measures (e.g. CVaR) to obtain better estimates of severe losses. In addition to heavy-tail modeling of return distributions we focus on modeling the interdependencies of financial instruments.
- Risk models with generalized distribution assumptions (Monte-Carlo, heavy-tail distributions)
- Improved estimation of covariances (shrinkage) and generalized dependencies of risk factors (copula)
- Robust and non-linear risk factor analysis
- Stress tests and scenario modeling
- Long-memory effects
A third aspect of our research is the optimal construction of efficient portfolios. Here we focus on model risk and the optimization of alternative risk measures.
- Optimization under generalized risk measures and risk preferences (CVaR, utility functions)
- Multi-period optimization (stochastic dynamic programming)
- Optimization subject to constraints (investment constraints, SOCP)
- Valuation involving model uncertainty and prior knowledge (Bayes, Black-Litterman)