Latest newsBack to articles
Is the Future 'Syscretionary'?
I read about the recent demise of Informed Portfolio Management (IPM) with interest. IPM, a once successful systematic macro fund manager had been finding things tricky for a while and the Covid-19 crisis aggravated things further. And IPM wasn’t the only firm that failed to respond effectively to the pandemic, as quants Renaissance Technologies, Winton and Two Sigma all reported significant losses.
AI is aiding systematic trading
Having had a lot of conversations with systematic fund managers recently, I started to think about the systematic approach and how despite a few glitches, it seems a sensible approach. In today’s data driven world it makes absolute sense that systematic funds would be a popular choice for investors, who see the value in removing human instinct and biases and trade based on data, patterns and predictions. AIML (AI and machine learning) funds particularly gained popularity in the years leading up to 2019 with a quarter of all systematic funds launched that year using AIML. The problems started in 2020 with the pandemic uncertainty and on-off lockdown situation, which meant that AIML systems that analyse historical and trend data were thrown completely off balance. Many systematic funds found they were completely off the mark, cutting their exposure to stocks as markets were soaring.
Discretionary traders' agility was a blessing during Covid-19
In contrast, the FT reported that discretionary funds and human traders were able to react quickly to the sell-off in risky assets in February and March 2020 and the subsequent huge rebound following central bank and government stimulus. But it doesn’t always follow that humans perform well in times of great volatility, as the 2008 crisis demonstrated, when many funds underperformed against the market.
So which approach is right?
The accuracy and consistency of some types of systematic trading, namely quantitative and HFT, means that when you hit on a successful formula, the results can be repeated. The ability to back-test is also usually an extra level of assurance that the formulas work. When automated, trades are executed at the best possible prices, at the right time, instantly and accurately. This can be a huge plus.
Can investors achieve excess returns?
Discretionary trading however, relies on the knowledge and expertise of the portfolio manager, which can be both a blessing and a curse. There is a perception that big returns can be made by a skilled portfolio manager who uses stories and ‘magic’ to make decisions, but in fact, much of the research shows that there are low correlations between discretionary trading and excess returns. The last decade or so has been a challenge for active discretionary managers with the rise of passive investing but there is still a space in allocators portfolios for non-correlated active managers who can show real skill in their chosen sector.
Other research shows that neither systematic or discretionary managers are inherently superior. Each style can deliver good investment outcomes under the right circumstances.
Could ‘syscretionary’ be the way forward?
Certainly for investors, it can be beneficial to incorporate both types into their allocations, but many managers themselves utilise models that are guided by data but augmented by humans.
The two approaches are not as polarised as they might seem at first glance. Far from relying solely on machine data, systematic managers use strategies that employ human judgement in their design and implementation. Many use fundamental inputs, traditionally associated with discretionary managers.
When looking at active equities for example, the approach usually deployed by discretionary managers is to understand a company’s fundamentals by looking at various factors such as the financial data, value for money, safety in businesses that demonstrate good resilience, strong management teams or those that have momentum or a catalyst that could change their value in the future. Systematic managers often use the same metrics, employing data analytics and intelligence tools to sift through large quantities of data to find the right targets.As the approaches seem to be converging, it might be wise for investors to consider less whether the manager employs a discretionary or systematic approach and more whether the process is repeatable, sustainable and has appropriate independent risk management and oversight.
Operationally we have seen some frustrations from systematic managers in implementing cost-effective operating models with the ability to trade directly via APIs etc. It is an area we are continuing to come across options and are happy to share our knowledge if you are interested to discuss. Systematic Strategies, rather than HFT, should be reasonably straight forward operationally if set-up optimally in the initial stage. We support managers across the spectrum and from my perspective, I’m agnostic whether they utilise a systematic or discretionary approach so long as they have a robust operating model, watertight compliance and strong service providers.
Get in touch if you want to discuss this in more detail.Back to articles