Multi Asset Artificial Intelligence Fund is a dynamic allocation strategy based on “Artificial Intelligence” trend and mean-reversion identification across asset classes
Objective:
The THEAM Quant – Multi Asset Artificial Intelligence Fund seeks to increase the value of its assets over the medium term, by being exposed to a diversified long/short portfolio across different asset classes (equities, fixed income and commodities), allocation of which is determined using a systematic method based on artificial intelligence. The exposure to the portfolio is adapted in order to keep the Sub-fund annual volatility at a target level of 8%.
Key points:
- Firstly, use of an Artificial Intelligence algorithm (neural networks) for an improved allocation of existing robust Quantitative Investment Strategies developed by BNP Paribas: trend following and mean reversion across asset classes.
- This allows to include indicators that go beyond the standard trend ones and thus improve the risk/return profile by aiming to capture reversal as well.
- The algorithm considers many parameters such as observation periods for trends (1-month to 1-year) and mean reversion (1-week to 1-month) – which allows to better fit optimal frequency for each component while avoiding overfitting issues.
- Machine learning techniques allow to solve non-linear optimisation across this enriched set of 378 indicators and hence to go one step further in the enhancement of the allocation process.
- A dynamic daily allocation, parameters of which are entirely reviewed on a yearly basis, to ensure reactivity to changing market conditions. It results from the collaboration between BNP Paribas Asset Management MAQS department and BNP Paribas Quantitative Investment Strategies.