Two volunteers of the CFA Societies Australia, Hugh Lam, Investment Analyst at Lonsec, and Nga Pham, CFA, Senior Research Fellow, Monash University sat down with Mike Aked, CFA, Senior Investment Strategist at Scientific Beta to discuss his career pathway and some key skills to succeed in quantitative investment management.

Question. What is quantitative investing, and how did you get involved?

I did my undergraduate degree in Sydney before quantitative finance courses started. I ended up doing an arts degree because that was the only way I could do both economics and mathematics. There were only 2 of us at the time. I did not understand investments and finance but hoped that if I did what I loved, then whoever wished to hire me would want me to do more of the same. Luckily the industry had better hiring practices than mine, and Sturt Piper hired me at SBC Investment Management.

Question. Could you walk us through a day in your life and what quant skills you regularly use?

Lots of coding, reading and reproducing papers, writing and presenting complex ideas with a simple narrative. I have learnt over time that simplicity is complicated to get to. You need to understand the complexity of something before you can derive a simple representation. You will therefore be willing and comfortable to get into the details, but most of the time, you will advance your ideas through simple representations.

Also, you will learn something from everyone you talk to, as we are all experts in an area. I suggest you be open to all conversations and discover everything you can when with a person. ‘Why do you think that is?’ is a great question.

Question. Who should learn these skills and why? What is the best way people can learn? Post-graduate degrees/online courses? Mike, I also note you are currently doing your Master in Mathematics and Statistics. What made you pursue further education?

I believe we should never stop learning; we do on the job daily. The two most favourite phrases in my professional vocabulary are “Why is that?” and “I don’t know”. After my undergraduate degree, I attended the first Master of Quant Finance class at the University of Chicago. About ten years ago, I got a Master’s in Statistics from the University of Virginia. The theoretical basis of statistics and quant models are invaluable, but knowledge of the markets is the other side of that coin. The CFA Program is a great way to get the market experience to combine with a more structured skill-based education.

Question. How is quantitative investing different from traditional/discretionary investing? Do you still incorporate fundamentals into your investment process?

The era of fundamental investing is not what it used to be. Working hard at knowing a company better than the marginal investor has become harder now that information has become more widely distributed and regulators have been policing this aggressively, particularly in the US or Europe. Back in the 1990s, analysts could gain information from the ‘C-suite’ themselves and be protected under the mosaic approach to information gathering. Reg-FD (Fair Disclosure) has closed this avenue down, and it is hard to see the edge that these managers now have.

The academic literature on asset pricing has shown us that there are many compensated risks that can be harvested to deliver a return over the cap-benchmark, across equities, bonds and in asset allocation. These will turn up as alpha in many judgements of investing.

We also have some reasons to determine the skill or effort in alternative investing. Private equity (not mega buyout) is about buying small family businesses and institutionalising them. They are already private (so can have a longer view than the reporting cycle). It’s about putting in experts or outsourcing to service providers, which a PE manager has a network of and levers across many target companies. This return is more about running a business than it is about investing.

Question, Where does Machine Learning (ML) / Artificial Intelligence (AI) fall within quantitative investing? Do you use ML algorithms?

I was really into neural networks and evolutionary algorithms in the early ‘90s. Unfortunately, we didn’t have the computer power to make a real dent, and they fell away. The emergence of the ML and AI approaches, and the exciting resurgence in statistics, is solely because we can now analyse real-world problems requiring substantial computer power for their optimisation.

When fast and accurate computation became accessible, as this is what computers are, it took whole industries out. The need for accountants and actuaries fell by an order of magnitude. Humans are not great at exacting calculators, and outsourcing these to computers has led to massive productivity gains.

The benefit of the human mind is that it is not-exacting. We make mistakes, and the same input can lead to different outputs when we run it through the system repeatedly. That is a benefit for learning and creativity but is not great for when we need accuracy – ask the average high school student what they think about math class. Recreating these characteristics and calling it AI is not where great productivity gains will be made. Continue finding the route processes that humans currently do and codifying them will be a great outcome. Eventually, we might be able to “train” a computer to replace nearly everything we do, but as a first step, let’s use them to do what they have a large competitive advantage on, being exacting.

Data mining is a real issue in finance. ML or AI approaches are at their core non-linear investigations. Where we have problems with data mining when we use linear models, ML/AI will make this matter worse.

Question. What do you find most rewarding about your role as a quant?

By combining my fundamental market knowledge with modern-day quantitative techniques, we can approach investing using the scientific method. By this, I mean that if we have a theory of investing, it needs to be supported by empirical data. Let’s leave the investment industry in a place where capital is allocated more efficiently. Our industry extracts a lower amount of wealth, and individuals can save when working and consume when retired. We have done our job.

Question. What advice do you have for students wanting to transition to a quant role from a more traditional finance background?

I think we should all approach investment as a science and be critical of what we do. If we can open our methods to peer or independent review, we will learn more about our industry and improve our processes. We can do this by taking traditional science-based methods and statistics and applying them to finance.