Humans are ‘most important part’ says founder of AI-run asset manager
Back in the late 90s, while the internet was gaining popularity, Miro Mitev was delving deep into the world of AI, a field that wouldn’t become mainstream for decades.
As an asset manager now, Mitev was an early adopter of AI in finance, having discovered the potential of neural networks in 1997 during his studies at the Vienna University of Economics and Business.
Fascinated by the possibilities of neural networks for financial forecasting, Mitev spent his career working for banks and tech firms like Siemens. He went on to establish SmartWealth Asset Management, where AI systems make all the investment decisions. Their latest fund, IVAC, aims to reach $2 billion in assets under management with an annualized returns target of 14-15%.
While there is no human intervention in the AI’s decision-making process, Mitev emphasizes the importance of human involvement in selecting training data, inputting variables, setting parameters, and constantly refining the model.
Mitev believes in trusting the model once it is created and warns against unnecessary interventions. He stresses the significance of ensuring data accuracy, introducing new data regularly, and avoiding the temptation to overrule AI-generated results.
Mitev highlights the importance of removing emotions from investment decisions for better results. SmartWealth Asset Management has outperformed industry benchmarks, showcasing the effectiveness of AI-driven strategies in generating gains.
While AI systems can “hallucinate” and generate false information due to overfitting or data issues, Mitev emphasizes the need for rigorous design, validation, and testing to counteract these issues. He underscores the ongoing evolution of the AI process and the crucial role of human oversight in ensuring its efficacy.
For Mitev, the key to successful AI implementation lies in continuous monitoring, data updates, and in-house development of cutting-edge technologies to stay ahead in the AI landscape.
