• Daniele Grassi, CEO and Co-founder at Axyon AI


• 08.03.2024 01:30 pm
#AI #datamanagement #data

Historically, asset managers struggled to find AI solutions that accommodated the detail in their work and the objectives they were trying to achieve. For all their computational power and speed, attempting to integrate older AI systems into firms sometimes created its own set of challenges.

However, the pressing need for real-time, insightful market analysis and consultation within the industry has led to the development of a new generation of bespoke AI models. For institutional investors, these represent better data management, forecasting, portfolio management, and predictive capabilities.

Data management and processing

Wherever there is alpha to be gained by drawing insights from larger bodies of data, the industry is pivoting to these new AI models which are capable of intelligently parsing through data from various sources, such as regulatory filings, market data, and news, as well as unstructured data like social media trends. AI can clean, categorise, and convert this data into actionable information, in essence enabling asset managers to base their decision-making on a more holistic view of the market and gain an informational edge that was previously not possible.

Data mining and real-time forecasting

Another challenge faced by asset managers in the past, has been AI’s heavy reliance on historical data for model derivation, which increased the risk of overfitting. To address this issue, institutional investors are now turning to predictive models that learn from current data in real-time, adapting to market changes as they happen. This is particularly vital for short-term trading.

These models need very little human supervision and can autonomously decide where to invest capital. Cutting-edge AI can serve as a manager of its own strategies, understanding market signals and deciphering what they mean for future trends.=

Predictive technology and tailored integration

Previously AI’s dependence on historical information proved to be a limitation in unchartered or rapidly changing market conditions. However, bankruptcies, geopolitical tensions, and unprecedented events like the COVID-19 pandemic are occurring more often, and while portfolio managers are interested in thinking differently about the world and finding original investment ideas, AI’s predictions around these events tended to converge as most firms referred to the same models.

Today, institutional investors require investment picks that align with their specific risk appetites, fund strategy, and even ethical values. This means firms must look beyond off-the-shelf solutions, embrace the significant leap forward AI has made, and – as few yet have their own AI departments set up – connect with teams of experts who can offer external support for integration.