One of these markets is that of asset management and more specifically the underlying quantitative portfolio management that drives risk-return optimization and financial risk management. The goal of portfolio management is to establish an asset portfolio that can maximize a risk-return trade-off given the allowed investment world and relevant constraints.
One of classical portfolio management’s hallmarks is that of the static mean-variance analysis based on Modern Portfolio Theory (Markowitz, 1952). In this framework, a portfolio is constructed such that the expected return is maximized given a targeted volatility. While the classic method is easy to implement, it also has many limitations such as the stability of the output, ignoring the skewness of return distributions, not discriminating between upside and downside risk, bias towards assets with high Sharpe ratio and many more. More advanced frameworks have been developed but these tend to be highly complex, non-robust to changing environments, or use-case specific. In recent years, techniques involving AI have rapidly been introduced for portfolio management and are gaining more ground.
Continue reading this article under Download.
Download
- Artificial Intelligence in Portfolio Management .pdf • 0,38 MB