Systematic investing has recently seen increased interest as many traditional investment styles have been challenged by adverse market conditions. We sat down with Dr Silvia Stanescu and Dr Chris Longworth, Investment Directors at GAM Systematic, to discuss some of the most asked questions about how systematic investing works, some common misconceptions, as well as the role a systematic investment style can play as a constituent of a well-diversified portfolio.
For those who are less familiar with the style, what is systematic investing?
Chris: Systematic investing very simply is a rules-based approach to investment. We take data from a large variety of sources, across all the markets that we trade, and we combine it to build models that can help explain some of the interactions between this data and market moves. Ultimately, our models look for patterns that repeat over time, which have the potential to yield a return.
Silvia: As an example, let's take price momentum. Trends in prices form for various reasons that have been widely documented. These include, for example, information asymmetry and behavioural biases. Since no market participant has access to perfect information, trends can form as new information is gradually incorporated into the market price. In addition, behavioural biases such as overreaction to other market participants’ actions may strengthen and extend trends once they have formed.
Furthermore, price trends can be reinforced by activity in related, but different markets, such as futures and options markets. It could be that hedging activity in the options markets may create price momentum in the underlying futures market that other players can ultimately trade on.
Why might an investor choose a systematic investment style?
Silvia: Computers and machines are scientifically proven to be better than humans at parallelising and multi-tasking. One such example in financial markets is trading many markets simultaneously. Why is that helpful? Well, this allows greater diversification. Investing systematically enables one to invest across a broad number of markets and to trade those markets more efficiently.
Chris: Also, systematic trading has the ability to react gradually and incrementally to new information as it comes in. For example, if a price momentum model is expecting the market to go up, but instead there is a sudden market fall, the models are able to reassess their levels of confidence in their initial assumptions and adjust their market positioning appropriately.
What kind of markets do you invest in?
Silvia: Systematic investment is applicable to any market. However, the markets we tend to focus on are the most liquid macro assets across fixed income, equities, FX and commodities. This allows us to build models that can be highly scalable and that are able to adjust positions dynamically and execute these trades efficiently and at low cost.
What kind of data do you work with?
Chris: We work with a variety of data, including real time data from exchanges all around the world. Much of the data that we work with is noisy and can contain errors so a lot of skill is required to process this data efficiently to form a useful and informative input into our models.
How relevant is machine learning within a systematic portfolio?
Chris: Machine learning is just one of the tools that we have available for building robust systematic models. My doctoral research at the Machine Intelligence Laboratory in Cambridge was focused on machine learning in the context of computer speech recognition, but finance presents quite a unique problem.
One of the challenges when applying machine learning to financial applications is that for many of the problems we are particularly interested in, we have very limited amounts of data available to train models. This differs from the application of speech recognition, where it's very easy to collect large amounts of data and consequently to train and build very complex models.
For us, the challenge is not just about building very complex models, but also how to deal with situations where only limited amounts of data are available. Which techniques are most appropriate for these sparse data problems? In many cases we find that some of the most effective approaches are probabilistic in nature. These are good at dealing with sparse data but are also very effective at providing a measure of confidence in their own forecasts.
What is required to trade a systematic portfolio?
Chris: This won't come as a surprise, but systematic investing is hard and it's heavily dependent on having the infrastructure available to build and run models. We deal with a very large amount of data, which we have to process in real time and feed through our models. This requires a huge amount of code, which we've gradually written over a decade.
However, running the models and risk management processes is only a small part of our overall infrastructure. We also need to be able to track our positions and mark-to-market our performance in real time, as well as having many checks and processes in place to confirm that our systems are behaving as expected.
If the system is taking care of the trading, what are you doing?
Silvia: In short, we do research. We research new styles of investment, and we also must adapt our investment strategies as markets evolve. As new markets or new market players, and as a result, new market dynamics emerge, one has new opportunities to identify patterns and to build new investment models or adjust existing models in line with new dynamics.
Chris: Our investment team comprises people from a range of backgrounds, but all have a shared enthusiasm for research and applying the scientific method to the challenge of investment. Sometimes portfolio improvements will come from the development of entirely new models, but improvements to existing models can be just as important, for example to capture the underlying investment hypothesis more effectively, or to execute trades more efficiently and with lower costs.
How would you respond to criticisms that systematic investing is a black box?
Chris: That's a question we are often asked. People often have a perception that systematic trading is mysterious and opaque, and it can perhaps appear that way from the outside. But for us, everything that we do is rules based. We can tie every decision, every position and trade that the systems make back to the models that we built and back to the data that's fed into them.
If you were to compare this to a discretionary trader, the decisions ultimately are the result of a set of very complex interactions inside the trader's brain. In comparison, systematic approaches are much clearer about what drives each individual trading decision.
Systematic investing typically relies on historical simulation. What if the future is different to the past?
Silvia: First, while analysing and inferring patterns from past data is a big part of what we do, our trading decisions are not solely based on this. One builds a systematic portfolio by investing across a range of different investment styles or strategies, with each of these styles expected to perform well under different market conditions. The more diversified we are across these scenarios, the more prepared we are for what the future might bring.
Chris: What if this time things are different? This is a question that we've been asked for many years now. It’s a fair question because it turns out the future is always different. There's always something new happening and we purposely build models that can adapt and react to new situations as they happen. We also have to remember that many of the patterns we trade are driven by long term structural market dynamics and by human behavioural biases. These typically persist over long periods of time because ultimately human nature doesn't change.
Tell us about your experience and offering.
Chris: We've been investing systematically for well over a decade. Over that time, we've made significant investments in the infrastructure and technology that we use. We also have a long track record of research and improvement to the models that we use to trade the market. This isn't something that you can start with from day one.
We've also had an extensive investment in automation and systematisation, ensuring that many of the common tasks that have to be done are fully automated and can operate without human involvement. This ultimately frees up our time and that of the team to do what we do best, which is research and to continue to improve our strategy and trading models.
Silvia: Picking up on that point, research is part of the fabric of GAM Systematic. It allows us to keep abreast of industry developments, as well as staying in touch with academia. Our own roots are in academia, with several of the team having attained PhDs. Earlier on in my career, I was a university lecturer and I keep this close to my heart - and close to my brain! Based in Cambridge, we have longstanding ties with the University of Cambridge, particularly with the Faculty of Mathematics, with which we often collaborate on research projects.
What is your research focused on at the moment?
Chris: We are both passionate about the incorporation of sustainability considerations into systematic macro investment. This turns out to be a very interesting challenge as traditionally most of the focus of sustainability research has focused on companies. Sustainability research in macro investing is often more difficult. For example, if you are comparing Facebook against Tesla, these could both be tech companies or in different sectors but inherently they all have similar features that you can measure and compare.
If you contrast this with a macro universe where we might be comparing an investment in crude oil to one in Polish equities, the differences are much greater. So how do we think about these on comparable scales and how can we adjust our allocations in order to build a more sustainable portfolio? These are problems that don't really have an industry standard answer. So a particular area of focus for us is to build systematic approaches, which aim both to satisfy the investor’s original objectives, and to do so in a more sustainable way.
Silvia and Chris, thank you for your time.
The information in this document is given for information purposes only and does not qualify as investment advice. Opinions and assessments contained in this document may change and reflect the point of view of GAM in the current economic environment. No liability shall be accepted for the accuracy and completeness of the information. Past performance is not a reliable indicator of future results or current or future trends. The mentioned financial instruments are provided for illustrative purposes only and shall not be considered as a direct offering, investment recommendation or investment advice. The securities listed were selected from the universe of securities covered by the portfolio managers to assist the reader in better understanding the themes presented and are not necessarily held by any portfolio or represent any recommendations by the portfolio managers. There is no guarantee that forecasts and objectives will be realised.