What is machine learning?
Machine learning is a field within artificial intelligence in which a computer imitates the way humans learn from experience. It involves training a computer to make predictions or decisions using data, rather than explicitly programming it to follow instructions. This training is repeated to gradually improve the computer’s performance.
As individuals, we are already benefitting from machine learning on a daily basis, when listening to music with smart speakers, picking a film thanks to streaming recommendations or choosing the best route via our favourite application.
1. We believe that over short time-horizons, investor behaviour rather than company fundamentals is the main driver of company share price returns. | 2. To identify and forecast the sentiment from this behaviour, a broad range of higher frequency data is required. | 3. By enabling the combination of a wide range of features, artificial intelligence can be a powerful tool to quickly and efficiently assess investor behaviour and make more accurate forecasts. |
How do we invest?
The Quest investment team has developed a proprietary and innovative stock selection model that uses AI to forecast relative stock returns.
- This model uses hundreds of features engineered from a broad range of data such as fundamentals, analyst sentiment, prices and market activity, short interest and calendar effects.
- Forecasts are regularly updated and combined in a proprietary optimiser with risk and cost estimates that portfolio managers oversee to better reflect market conditions.
- We also apply Pictet wide ESG exclusions and a positive ESG tilt based on a proprietary ESG scoring model.
Index equity approaches | Enhanced Index | Traditional active equity portfolio | |
---|---|---|---|
Desired tracking error | Very low | Low | Medium to high |
Active return | Similar to the benchmark | Aiming at compounding outperformance | Potentially higher but less consistent returns |
Absolute risk profile | Similar ti the benchmark | Similar to the benchmark | Medium to high |
Cost | Low | Low | Medium to high |
Concentration | Low | Low | Medium to high |
Our team
Our Quest team consists of 20 professionals with an average industry experience of 15+ yearsAs of 01.04.2024 . They have a skill-set spanning model building, feature engineering, AI/ML, risk management, portfolio construction and implementation.
Their expertise is grounded in academic research, including a focus on machine learned stock selection as a ‘crystal box’ rather than a ‘black box’, which is demonstrated by the publication of several peer-reviewed research papers“Performance attribution of machine learning methods for stock returns prediction”, Stéphane Daul, Alexandra Nagy and Thibault Jaisson, published in the Journal of Finance & Data Science in April 2022, and “Deep differentiable reinforcement learning and optimal trading”, Thibault Jaisson, published in the Quantitative Finance journal in June 2022. published in well-known finance and economic journals.
The lead Portfolio Manager for Quest AI-Driven is Stéphane Daul.