How AI, machine learning and human investment expertise identify potential alpha sources
Systematic investment strategies use AI and machine learning to pursue alpha in a targeted way, even in volatile markets. The crucial factor is the combination of data-driven models with human investment expertise, explains professor of practice Gianluca De Nard in this interview.
Interview with: Gianluca De Nard
Three key takeaways on systematic investment strategies:
- Systematic, rules-based strategies show their strengths particularly in volatile market phases because they take advantage of opportunities and manage risks in a disciplined way.
- The achieved outperformance is based on the deliberate combination of proven multi-factor signals (value, quality, momentum) with dynamic models supported by artificial intelligence (AI) and consistent risk optimisation.
- AI expands analytical and discovery capabilities but remains clearly embedded in an investment process overseen by human expertise.
Systematic strategies have performed well since the beginning of the year, despite turbulence in geopolitics and markets. What explains this?
Systematic strategies typically show their strengths most clearly when conditions are unsettled, as this is when many mistakes can occur but also when significant opportunities arise. Our strategies have repeatedly demonstrated that you need to "sail" through such phases with calm and clarity. A disciplined, rules-based approach is a real advantage here. Our approach is based on proven alpha sources that are well documented in the academic literature, which we enhance with proprietary signals, many years of experience and adaptive machine-learning models. It is crucial that all of this is anchored in a tried-and-tested investment process and consistent risk optimisation.
What have been the main performance drivers of the strategies this year?
The momentum factor in particular has been very strong this year, but value has also made a noticeable contribution. Our dynamic, AI-based signals have generated clear additional value. However, the key to success does not lie in a single factor or model, but in the deliberate combination of these different signals. This diversification at signal level increases the robustness of our strategies across different market regimes and, especially in this turbulent environment, was an important driver of their strong performance.
What are the main advantages of systematic investment strategies compared with traditional, fundamentally driven stock selection?
The main advantage lies in the breadth and scalability of the systematic approach. Our models can continuously analyse large volumes of structured and alternative data across virtually the entire investment universe. This makes it possible to identify patterns that are difficult or impossible to detect using conventional analytical methods – and which often prove complementary or even uncorrelated with analysts’ recommendations.
Systematic investing is therefore also a numbers game: instead of relying on a few large single bets, we diversify widely across many independent investment opportunities. This means we take frequent, disciplined decisions – always based on the same clear, rules-based framework. Systematic strategies can therefore provide a strong complement to both passive exposures and fundamentally active portfolios – and an additional, differentiated source of alpha.
How is the investment process structured, and how are the multi-factor signals defined and implemented?
Our investment process combines two complementary sources of information. The first is a robust, proprietary multi-factor approach based on valuation, quality and momentum. These factors in turn capture economically intuitive and empirically well documented characteristics related to long-term differences in expected returns.
The crucial aspect, however, is how these signals are defined, combined and implemented. We do not simply replicate generic factors. Instead, we have developed our own proprietary signals, which we systematically combine across a broad universe of companies.
And the second component?
The second component is our machine-learning model. It can identify complex, non-linear relationships and interactions across a large number of potential drivers. This allows us to uncover investment opportunities that are difficult to capture with traditional linear models or fundamental analysis.
The alpha signal resulting from these two sources of information is then translated into concrete positions via portfolio optimisation. This is essential: having a good return forecast is only one part of the investment process. Expected returns must also be considered in relation to risk, diversification, transaction costs and portfolio constraints. Ultimately, the objective must be to manage a portfolio that is efficient and resilient in terms of return and risk – rather than a collection of shares with attractive individual signals.
What role does human expertise play in this process?
Traditional quantitative models remain extremely valuable because they are robust, transparent and economically interpretable. Accordingly, our proprietary value, quality and momentum factors form an important foundation of the investment process. Machine learning complements these models by allowing us to analyse a much larger feature space and identify complex, non-linear patterns and interaction effects. It is equally important to stress that AI does not construct our portfolios autonomously. Responsibility continues to lie with our experts in research and our portfolio managers. People still determine the choice of data, their economic interpretation, the model architecture and the risk constraints. They also critically assess whether identified patterns are robust and economically meaningful.
In short, for us the interaction between human expertise and machines is and remains central. AI broadens the range of what we can analyse and potentially discover, while human expertise ensures that these insights are robustly translated into investable portfolios.
Where do you see the opportunities and where the risks for the systematic approach between now and the end of the year? With the interest-rate cycle and the US midterm elections, further turbulence is looming.
We are looking at an environment characterised by persistent uncertainty and periods of heightened volatility. For us, this means a field of tension comprising additional risks and attractive opportunities for systematic strategies. The risks mainly lie in the possibility that market regimes can change abruptly and repeatedly, and that short-term "noise" increases. The opportunities arise from the fact that such phases often lead to a stronger dispersion of returns between individual stocks. Our experience shows that a disciplined, rules-based process combined with a deliberate mix of robust static and adaptive dynamic signals is particularly well suited to targeting a balanced risk/return profile through such turbulence.
Our portfolios are deliberately constructed to be sector and country neutral. We do not aim to forecast macroeconomic or political developments; instead, we focus on stock selection. This reduces the risk of being caught out by sector-specific or regional shocks and enables us to concentrate on identifying relative winners and losers within individual segments.
Legal notice Switzerland
Legal notice Switzerland
This document only serves advertising and information purposes, is for distribution in Switzerland only and is not directed at persons in whose nationality or place of residence prohibit access to such information under applicable law. Where not indicated otherwise, the information concerns the collective investment schemes under the law of Luxembourg managed by Swisscanto Asset Management International S.A. (hereinafter "Swisscanto Funds"). The products described are undertakings for collective investment in transferable securities (UCITS) within the meaning of EU Directive 2009/65/EC, which is governed by Luxembourg law and subject to the supervision of the Luxembourg supervisory authority (CSSF). This document does not constitute a solicitation or invitation to subscribe or make an offer to purchase any securities, nor does it form the basis of any contract or obligation of any kind. The sole binding basis for the acquisition of Swisscanto Funds are the respective legal documents (management regulations, sales prospectuses and key information documents (PRIIP KID), as well as financial reports), which can be obtained free of charge at https://products.swisscanto.com as well as at Swisscanto Fondsleitung AG, Bahnhofstrasse 9, CH-8001 Zurich (also acting as representative of the Luxembourg Swisscanto funds in Switzerland) or in all offices of Zürcher Kantonalbank. Paying Agent for the Luxembourg Swisscanto funds in Switzerland is Zürcher Kantonalbank, Bahnhofstrasse 9, CH-8001 Zurich. Information about the sustainability-relevant aspects in accordance with the Regulation (EU) 2019/2088 as well as Swisscanto's strategy for the promotion of sustainability and the pursuit of sustainability goals in the fund investment process are available on the same website. The sub-fund referred to in the document is subject to Article 9 of Regulation (EU) 2019/2088. The distribution of the fund may be suspended at any time. Investors will be informed about the deregistration in due time. The investment involves risks, in particular those of fluctuations in value and earnings. Investments in foreign currencies are subject to exchange rate fluctuations. Past performance is neither an indicator nor a guarantee of future success. The risks are described in the sales prospectus and in the PRIIP KID. The information contained in this document has been compiled with the greatest care. Despite professional procedures, the correctness, completeness and topicality of the information cannot be guaranteed. Any liability for investments based on this document will be rejected. The document does not release the recipient from his or her own judgment. In particular, the recipient is recommended to check the information for compatibility with his or her personal circumstances as well as for legal, tax and other consequences, if necessary, with the help of an advisor. The prospectus and PRIIP KID should be read before making any final investment decision. The products and services described in this document are not available to U.S. persons under the relevant regulations (in particular Regulation S under the U.S. Securities Act of 1933).
© Zürcher Kantonalbank 2026. All rights reserved.
Legal notice international
Legal notice international
This document only serves advertising and information purposes and is not directed at persons in whose nationality or place of residence prohibit access to such information under applicable law. Where not indicated otherwise, the information concerns the collective investment schemes under the law of Luxembourg managed by Swisscanto Asset Management International S.A. (hereinafter "Swisscanto Funds"). The products described are undertakings for collective investment in transferable securities (UCITS) within the meaning of EU Directive 2009/65/EC, which are governed by Luxembourg law and subject to the supervision of the Luxembourg supervisory authority (CSSF).
This document does not constitute a solicitation or invitation to subscribe or make an offer to purchase any securities, nor does it form the basis of any contract or obligation of any kind. The sole binding basis for the acquisition of Swisscanto Funds are the respective published legal documents (management regulations, sales prospectuses and key information documents (PRIIP KID), as well as financial reports), which can be obtained free of charge at products.swisscanto.com/. Information about the sustainability-relevant aspects in accordance with the Regulation (EU) 2019/2088 as well as Swisscanto's strategy for the promotion of sustainability and the pursuit of sustainability goals in the fund investment process are available on the same website.
The distribution of the fund may be suspended at any time. Investors will be informed about the deregistration in due time.
The investment involves risks, in particular those of fluctuations in value and earnings. Investments in foreign currencies are subject to exchange rate fluctuations. Past performance is neither an indicator nor a guarantee of future success. The risks are described in the sales prospectus and in the PRIIP KID.
The information contained in this document has been compiled with the greatest care. Despite professional procedures, the correctness, completeness and topicality of the information cannot be guaranteed. Any liability for investments based on this document will be rejected.
The opinions and assessments contained in this document regarding securities and/or issuers have not been prepared in accordance with the regulations governing the independence of financial analysts and
therefore constitute marketing communications (and not independent financial analyses). In particular, the employees responsible for such opinions and assessments are not necessarily subject to restrictions on trading the relevant securities and may, in principle, conduct their own transactions in these securities.
The document does not release the recipient from his or her own judgment. In particular, the recipient is recommended to check the information for compatibility with his or her personal circumstances as well as for legal, tax and other consequences, if necessary, with the help of an advisor. The prospectus and PRIIP KID should be read before making any final investment decision.
An overview of investors' rights is available at swisscanto.com/int/en/legal/summary-of-investor-rights.html.
The products described in this document are not available to U.S. persons under the relevant regulations (in particular Regulation S under the U.S. Securities Act of 1933). Data as at (where not stated otherwise): 03.2025
© 2026 Swisscanto Asset Management International S.A. All rights reserved
