Mathematics in Finance: How to Bridge the Gap Between Science and Business

In the dynamic world of finance, where precision and predictive analytics are paramount, the collaboration between mathematical optimization and quantitative research becomes a powerful force for innovation and business. This is why swissQuant places a strong emphasis on continuous learning and research to ensure that we stay ahead of the curve in delivering solutions that anticipate and address future challenges and client needs.
Photo of Jianzhe Zhen in the swissQuant office during the interview about the impact of research in finance and how he manages the balancing act between academia and business.

We had the privilege of speaking with Jianzhe (Trevor) Zhen, Optimization Expert and Quant Engineer at swissQuant, whose groundbreaking work at the intersection of mathematics and finance has left an indelible mark.

He was recently appointed Adjunct Associate Professor at one of China’s top universities in Beijing, the University of Chinese Academy of Sciences. Trevor has also recently published several papers in prestigious journals. Read more about them at the end of this article.

Join us for an interview with Trevor, where we ask him 5 questions about the transformative impact of quantitative research in finance and how he manages the balancing act between academia and business.

Jump directly to the 5 questions:

  1. How does research and development impact personal growth and client solutions?
  2. In what ways is specialized research being translated into unique client solutions?
  3. How is scientific research balanced with business goals?
  4. How are research uncertainties managed in business decisions?
  5. What are the future outlooks for integrating research insights into business strategy?



Research and Development at swissQuant: Impact on Personal Growth and Client Solutions

How does swissQuant’s emphasis on continuous learning and research contribute to your personal and professional growth? And how does this impact the solutions provided to clients, especially in addressing the evolving challenges of the financial landscape?

At swissQuant, the focus on continuous learning and research has significantly contributed to my personal and professional development.

It has given me the opportunity to collaborate with leading experts in my field, access to cutting-edge tools and resources, and the support to pursue innovative research projects.

Collaborating with industry experts and utilizing cutting-edge tools has enabled me to stay updated with the latest trends, fostering the development of innovative solutions for complex client problems.

This is particularly vital in the ever-evolving financial landscape, where my involvement in research on mathematical optimization and model predictive control directly translates into forward-thinking, efficient solutions for our clients, keeping us ahead of industry advancements.

Translating Specialized Research into Unique Client Solutions

Can you discuss how your specialized expertise in mathematical optimization and model predictive control translates into unique advantages for swissQuant’s clients? 

My expertise in these specialized fields has led to the development of advanced algorithms and models, such as an optimization tool for complex systems, which is now a key component of swissQuant’s offerings.

This not only positions us uniquely in the market but also ensures our solutions are tailored to meet our clients’ specific needs, providing them with a competitive edge.

Balancing Scientific Inquiry with Business Goals

How do you balance rigorous scientific research with practical business objectives? And how does interdisciplinary collaboration contribute to this?

Identifying research areas that align with organizational goals requires understanding both the industry’s challenges and advancements. Regular interactions with colleagues and stakeholders, coupled with active engagement in the academic community, are crucial.

This collaborative approach, involving simplifying complex findings and developing user-friendly tools, ensures our solutions are both scientifically sound and practically applicable. Additionally, fostering interdisciplinary collaboration among colleagues enhances our ability to tackle complex challenges innovatively.

Managing Research Uncertainties in Business Decisions

How do you manage and communicate the risks and uncertainties inherent in research findings to stakeholders to facilitate informed decision-making?

Effective management and communication of research risks involve a transparent approach, clarifying assumptions, limitations, and potential impacts.

When introducing innovative research-based approaches, I make sure to present a strong case backed by empirical evidence, including case studies, simulations, and real-world examples that demonstrate the effectiveness of the approach.

Involving stakeholders throughout the research process and addressing their feedback builds trust and ensures alignment with their needs, enabling informed decision-making even in the face of uncertainties.

Future Outlook: Integrating Research Insights into Business Strategy

Looking forward, how do you envision the integration of research-based insights shaping swissQuant’s offerings and business strategy?

The ongoing integration of research insights is crucial for the future of swissQuant. Staying adaptive and innovative is key in the dynamic financial sector.

Research-driven methodologies will lead to the development of advanced algorithms, predictive models, and solutions, potentially leading to new and enhanced services. Furthermore, these insights will inform our business strategy, helping us identify trends and opportunities, thereby driving long-term success.


Mathematical Foundations of Robust and Distributionally Robust Optimization

The paper “A Unified Theory of Robust and Distributionally Robust Optimization via the Primal-Worst-Equals-Dual-Best Principle,” authored by Prof. Daniel Kuhn of Ecole polytechnique fédérale de Lausanne, Prof. Wolfram Wiesemann of Imperial College London, and Jianzhe Zhen, represents a seminal advancement in optimization theory, with significant implications in practical domains such as finance and machine learning. Featured in Operations Research, this work introduces a groundbreaking theoretical framework in robust and distributionally robust optimization.

At the heart of this paper is the ‘primal-worst-equals-dual-best’ principle. This principle creates a strong duality between complex semi-infinite and non-convex formulations in optimization. The framework developed by the authors not only showcases an innovative theoretical approach but also paves the way for its application in simplifying the complexity of robust optimization problems.

The real-world implications of this research are considerable. While the paper primarily contributes to theoretical understanding, the principles it outlines could have broad applications in various fields. For instance, in finance, where managing risk and uncertainty is crucial, and in machine learning, where algorithms must be resilient to data variability, the insights from this research could be pivotal.

The theoretical framework lays a foundation that could significantly influence decision-making processes and the development of solutions in these and other areas where robustness against uncertainty is crucial.

This work presents a significant leap in the mathematical foundations of optimization, offering a vital tool for developing robust and reliable solutions in various complex and uncertainty-driven environments.

Advancing Robust Convex Optimization: A Unified and Extended Approach

In a notable collaboration with Prof. Dimitris Bertsimas of the Massachusetts Institute of Technology, Prof. Dick den Hertog of the Universiteit van Amsterdam, Dr. Jean Pauphilet of the London Business School, and Jianzhe Zhen, co-authored the paper “Robust Convex Optimization: A New Perspective That Unifies and Extends,” featured in Mathematical Programming, a leading journal in Mathematical Optimization.

This paper heralds a novel approach to addressing robust convex constraints, effectively bridging and expanding upon various existing methodologies in the field. Notably, our method surpasses previous solutions and tackles problem classes previously unaddressed.

At the heart of our research is a sophisticated solution utilizing an advanced form of the Reformulation-Linearization Technique. This method is particularly adept at managing general convex inequalities and diverse convex uncertainty sets.

One of the key breakthroughs of our work is the development of a series of refined approximations. These approximations are crucial for accurately estimating both the upper and lower bounds of the optimal objective value.

By applying our approach to robust control and robust geometric optimization scenarios, the paper underscores the significant practical and numerical benefits of our advanced methodology, marking a considerable leap in the realm of robust optimization.

Enhancing Data-Enabled Predictive Control Through Robust Optimization

A significant paper published in IEEE Transactions on Automatic Control titled “Robust Data-Enabled Predictive Control: Tractable Formulations and Performance Guarantees” represents a collaborative achievement in the field of model predictive control. The authors, Dr. Linbin Huang, Prof. John Lygeros, Prof. Florian Dörfler from the Automatic Control Lab of ETH Zurich, and Jianzhe Zhen, introduce a novel framework for handling robust predictive control in linear time-invariant systems, leveraging imprecise data.

This research is notable for its innovative approach to computing robust and optimal control sequences. By adopting a receding horizon methodology, it adeptly addresses the challenges associated with data uncertainties. Central to this framework is a min-max optimization problem that ensures the robustness of the control sequences against a spectrum of data uncertainties.

The team’s findings are further validated through comprehensive simulations, particularly on a power converter system, underscoring the practical applicability and significance of their work.

This collaborative effort marks an important advancement in model predictive control, addressing the critical challenge of data imprecision in real-world scenarios and demonstrating the potential of robust, data-driven control solutions.

Are you interested in more research topics? Find an overview of all Trevor’s publications on his Google Scholar Profile page.


Interested in our Company Values and our Roots in Science?

Table of Contents

Reading Progress

Master complexity, amplify impact:

Let's work together


Download our SNB Case Study

The SNB Case Study will be sent to your email. Check your spam folder too.


Download our LLB Case Study

The LLB Case Study will be sent to your email. Check your spam folder too.