NOBUAKI KATO

COULD YOU TELL US ABOUT YOUR BACKGROUND AND WHAT YOU ARE DOING TODAY?

I work as a portfolio manager at Amundi Asset Management, and I manage a large number of portfolios for institutional and retail investors. The company has about 1.5 trillion Euros of AUM, which ranks the top 10 globally (as of March 2020). I belong to a unique team in this organization. In this team, we offer multiple investment strategies, such as passive investment, quantitative investment, and smart beta investment. We also offer tailor-made solutions to institutional clients, such as incorporating ESG criteria, carbon emission reduction, and some other requests to the investment strategies mentioned above. I originally joined the company in Tokyo with the same role and later moved to Paris. Before joining Amundi as a portfolio manager, I worked as an IT developer at Barclays. I programmed the equity trading system, such as an order management system, a trading interface, and exchange connectivity. That was my first job. So, yes, my academic background is in engineering. I have a bachelor’s degree in Engineering at Osaka University, School of Engineering Science, and a master’s degree in Information Physics and Computing at the University of Tokyo. Since I had not studied finance when I was a student, I thought I lacked financial knowledge when I started my current job. Therefore, I worked on CFA exams because it covers broad knowledge which is essential in the industry. After earning the CFA charter, I still felt I did not have a good understanding of the economic and financial theory, so I pursued a Master of Business Administration program at the University of Chicago, Booth School of Business.

HOW MUCH DO YOU FEEL A RESEARCH DEGREE IS RELEVANT TO A PORTFOLIO MANAGER LIKE YOU TODAY?

I feel that a research degree is truly relevant. Firstly, if a portfolio manager looks after portfolios based on statistical models, it is beneficial to have an excellent statistical insight to improve the model and to implement it to the real portfolio. For example, the portfolio manager must consider which variable to include, how to check which parameter set is essential, what is a constraint, how to test, and so on. Through academic research, one can acquire such knowledge and sharpen the skills. Secondly, portfolio managers must explain the investment strategies and portfolio performance to their clients periodically. The academic insight is a great advantage to convey the idea and the rationale to the audience.

HOW DID YOU CHOOSE THIS PARTICULAR PROGRAM?

When I enrolled in the MBA program, I was fascinated by the economic and financial theories. I wanted to dig deeper into those theories, so I decided to embark on a PhD program. However, because I have a family to support, it was not possible to quit work and commit to a full-time PhD program. I searched on the internet to see if there were any part-time PhD programs in the world, and I found the EDHEC PhD in Finance program. Coincidentally, the CIO of Amundi Japan was an alumnus of the PhD program, so I contacted him to learn more about it. The CIO also referred me to one of his friends who had enrolled in the program, so I asked him about it as well. During our conversations, I found the program extremely appealing. The format (core courses and elective courses + research) is well structured. Core courses serve as necessary foundations for the research, and elective courses offer the right choice of high-quality lectures. I liked the high technicalities of the lectures and the excellent faculty and I chose the program based on this information.

THE PHD PROGRAMME IS CHALLENGING BY ITS NATURE AND FORMAT, HOW DO YOU MANAGE THE REQUIRED TIME COMMITMENT?

It is hard to say. I split most of my spare time into family and research. Whenever I have my own time, other than spending it with my son, I devote myself to research related activities, such as reading books, papers, coding, simulating, analyzing, thinking, doing assignments. I do not go out often anymore. I do not watch movies anymore. I do not read novels anymore. Also, I sleep slightly less than before. However, the most significant contribution is made by my supportive wife. She is helpful in looking after our son from time to time so that I can concentrate on the research. Thanks to her, I can manage my time more freely.

DO YOU SEE ANOTHER MAJOR CHALLENGE?

I do not see any other major challenges other than time management. Time management is the biggest challenge.

TO DATE, HAS THE PROGRAMME IMPACTED YOUR DAY-TO-DAY WORK?

One significant impact is that I am more cautious about the robustness of the statistical analysis. For example, slightly changing a parameter leading to a very different result is not robust. This fragility often happens when one deals with time series or tries to optimize a quadratic function with constraints. In practice, we need to cope with some instability due to a limited dataset, but we can mitigate it a bit if we are a bit more prudent.

HAVE YOU FOUND BENEFITS IN INTERACTING WITH OTHER PROGRAM PARTICIPANTS AND OF HAVING A MAJORITY OF PROFESSIONALS IN THE CLASS?

Yes, interacting with other program participants is extremely beneficial. Everyone has good knowledge and experience in each field. I always learn something from the cohorts. For example, I found elective courses particularly useful for interaction with others. In elective courses, the participants gather not only from the same academic year batch but also from other batches. The participants who attend the elective courses share some common interests or carry relevant knowledge from their professional backgrounds in general. There, I can find inspiration on how I should proceed with my research.

RECENTLY, YOU HAVE PRESENTED AN INITIAL DRAFT OF YOUR FIRST PAPER TO THE PHD PROGRAM CORE FACULTY AND TO PHD CANDIDATES; COULD YOU PLEASE INTRODUCE THE TOPIC AND EXPLAIN WHAT YOUR RESEARCH AIMS TO ACHIEVE?

My research topic is to specify time-varying transition probabilities from one state to another, for a Markov switching model. In general, economic phenomena we observe every day would arise from certain conditions of the underlying economy. For example, if we are in a good economic state, we may observe GDP growth, while if we are in a bad state, we may observe the opposite. So, if we know which economy we are in now, we may be able to forecast an impending economic phenomenon. However, the underlying economic conditions are not observable directly, and we need to rely on some observed values to assess them. The Markov switching model is a widely used econometric model that can filter the hidden economic states from observed time series. In the conventional model, the transition probabilities are often assumed to be constant. However, assuming the constant probabilities could be too restrictive to reflect the real economic transition. I have previous experience using the Markov switching model, but the result was not sufficiently satisfactory. I am hoping that if we allow time-varying characteristics in the transition probabilities, the model can detect the unobserved economic state more accurately than its conventional model. If that is the case, the model could be a useful tool to forecast an economic phenomenon more accurately (e.g., volatility).