Two volunteers of the CFA Societies Australia, Hugh Lam, CFA, Investment Analyst at Lonsec, and Nga Pham, CFA, Senior Research Fellow, Monash University sat down with Andre Roberts, a Senior Portfolio Manager at Invesco, to discuss his pathway to a career in quantitative investment management.
Nga: Hi Andre, thank you for joining us today. Let’s start with a basic question to uncover the differences between fundamental and quantitative investing.
Andre: Quantitative investing primarily involves using quantitative methods (e.g. statistical modelling, mathematics, programming) in the investment process. However, the investment thesis is still grounded in an economic/financial intuition where you’re looking for the same quality companies with growing earnings that are priced attractively. The difference comes in the how rather than the what. Quantitative (or systematic) investing is simply another approach to identifying alpha-generating securities through numerical techniques with a fundamental orientation.
Hugh: How did you begin your career?
Andre: I started in the 90s, when academic studies on momentum and quality factors joined the fray alongside valuation-type work that started a little earlier. There were other prominent academic publications released, notably by Grinold and Kahn, which helped me get a broad perspective of how the characteristics were being applied quantitatively. I started as a generalist quant in a fundamental team where I integrated data into the investment process using screens and risk analysis to model factor risk. Throughout the role, I read a lot about characteristics that could drive returns and built simple models which would eventually form the foundation of the investment process I use.
Hugh: How are these numerical methods applied?
Andre: They are applied in a number of areas.
- One area is determining whether a shock (such as the SVB bank collapse) is systemic or not. Numerical methods are used to determine the characteristics driving returns throughout the shock.
- Principal components analysis (PCA) is another method used to determine what areas of the market have been moving in a correlated way and what are the focal points that have changed in recent days.
- Regression is used to decompose returns and determine which characteristics have been driving markets, and
- Researching new ways to improve existing models.
You know most of the skills you have gained from an undergraduate stats course? These techniques become the bedrock of your research. And I know I have touched on the more recent machine-learning techniques. So whether it’s old school or new school, you would test in a historical context and simulate the strategy in an as realistic world as you could build.
Nga: Are you a believer of new school numerical techniques?
Andre: The systematic implementation of techniques, whether it is machine learning (ML) or natural learning processing (NLP), are the two that I find helpful. NLP is an exciting tool where you can identify themes or controversies in companies through news, company reports and other text sources online. It helps you consume information very efficiently and cover a wide breadth of information sources that you couldn’t do manually.
Machine learning is a new extension of more ‘old-school’ statistical techniques. For example, do two characteristics combine non-linearly? My feeling is that ML is not always used to come up with brand-new insights but often to improve the implementation of insights that have been around for a while.
Nga: Can you please share with us an example of an interesting factor you’ve identified and how you’ve implemented that in your investment process?
Andre: “12-1 month” momentum has been an effective factor, given that this factor has been one of the best ways to outperform the market systematically. One of the criticisms is that it is prone to poor performance during turning points (such as new regimes). Macro-specific factors prone to this change can be ‘stripped out’ by using regression techniques which will leave you with a more idiosyncratic style of momentum. This is a more robust measure of a standard 12-month momentum factor.
The “12-1 month” factor is implemented in the investment process by bucketing names with these factors together (forming a peer group). Return forecasts are then built around these names alongside risk, costs, and tax considerations. We can then use an optimiser to identify ideal portfolio weights.
Nga: What do you find rewarding about your career?
Andre: The skills that you learn 30 years ago are still very relevant today, and you see outcomes for investors that are very solid and reliable. Quant methods allow for better risk control, utilising breadth of information and broader diversification than the fundamental world. This holistic framework of quant investing allows you to generate great outcomes for investors, and that’s what we are here for.
Hugh: Do you have any advice for students or young professionals who want to pivot to quantitative investing?
Andre: Refresh the statistics and programming techniques you learnt in coursework. Find internships and dive into a project to see how these quantitative techniques are being applied in real life.
Please appreciate the breadth of opportunity you can find by leveraging the technical skills you have in finance, statistics, mathematics and programming. This is an excellent and relevant toolkit for your working life. Speak to people in the industry, have coffee with people you find interesting and inspire you, and start building relationships. Learn how to marry your skill sets in this industry that is so rich.
Nga: Your sharing is very inspiring. Thank you for your time!
Andre’s Bio: Based in Melbourne, André Roberts is a Senior Portfolio Manager in the Invesco Quantitative Strategies (IQS) team. He is responsible for managing IQS Australia’s equity portfolios (Core, Tax-aware, 130/30 long/short and Smaller Companies) and represents IQS global capabilities in the APAC region. André also coordinates Invesco’s ESG integration for Australia.
André has over 29 years’ experience in Australian equities markets (fundamental, active quant and index), working in quantitative portfolio management, research, investment solutions and systems development. Prior to joining Invesco in February 2018, André was with IFM Investors, BlackRock and County Investment Management.
André is a CFA® charterholder and board member of the CFA Society of Melbourne, of which he is a past president. He holds a Bachelor of Science in statistics and a post-graduate diploma in econometrics from Monash University and a CFA certificate in ESG Investing.
Date: 29 May 2023
What finance electives should I take? >