Mark van der Kroft: “Quant investing is a critical part of what Robeco does and stands for. Ever since our first director said that “every investment strategy should be research-driven”, we’ve put empirical research at the core of everything we do. And, ultimately, that research is what supports our quant investing strategies.”
“So, while the quant industry may be going through tough times, we are convinced that it should remain one of our key capabilities. You know, we’ve gone through many market cycles already, and we’ve experienced disappointing periods in the past. Our commitment to go through these cycles with our clients has not changed.”
Pim van Vliet: “Also, remember that our quant offering is very broad and diversified, and includes for a large part also fixed income and emerging markets strategies. Other quant asset managers are less diversified and more focused on US equity markets, which is where quant investment approaches have been suffering the most in terms of performance.”
MvdK: “The main change has to do with sustainability. Clients increasingly ask for solutions that combine a quant approach with specific sustainability targets, especially regarding climate change. And I expect this to become the norm. The future of quant will likely not only be about delivering pure alpha solutions anymore, but about providing both of factor exposures and achieving sustainability targets.
PvV: “And this is great news for us because we have been pioneers in both quant and sustainable investing. And we have developed an efficient and flexible quantitative portfolio construction platform, that also happens to be very effective at translating sustainability-related client wishes and preferences into portfolios.”
PvV: “Indeed. Imagine for instance that you want to address some specific sustainable development goal (SDG) or carbon reduction with your portfolio, but do not want to deviate too much from the market index. Our approach can help you find the stocks that will help achieve your SI targets while avoiding excessively skewed portfolios and potentially high tracking error.”
As more data becomes available, we need to think of ways to take advantage of that
MvdK: “Another important change is that quant solutions must increasingly look beyond the most conventional factors, such as those proposed by Fama and French for example, for the simple reason that more and more data is becoming available. Clearly, you can think of countless potential factors based on alternative data.”
“I mean, our factor definitions already differ greatly from standard academic definitions. For more than two decades, we have been using proprietary sophisticated approaches, that combine sets of complementary variables and avoid the typical pitfalls of simplistic factor definitions. But as more data becomes available, we need to think of ways to take advantage of that.”
“Then, of course, the challenge is to figure out what kind of data you should look at, and what you should look for, which my sometimes feel like looking for a needle in a haystack. Moreover, these non-conventional factors are likely to be a much more short-term in nature. But that’s something we really believe will end up boosting long-term performance.”
PvV: “Yes, these are good examples. If you look at conventional equity factors, it all started with Fama and French’s size and value. But Robeco was very early in promoting a new one: low volatility. And low volatility is now widely accepted among investors. Then, we were also early in adopting momentum. In fact, we have been using analysts’ revisions momentum for over a decade.”
“Over the past couple of years, a lot of innovation has been taking place regarding signals on the more short-term end of the spectrum, thanks to an increased use of alternative data sources and innovative techniques such as machine learning. And Robeco, once again, has been at the forefront, investing considerable resources to find additional return drivers to enhance our existing models”
PvV: “Another example would be our recently implemented short-term factor momentum signal. In this case, we use short-term trends in equity investment styles, not to try to chase short-term factor performance – which is notoriously difficult to achieve – but to enhance our trade timing process and therefore enhance long-term returns. In addition, we are working on applying machine learning in order to predict distress risk, which would be useful for our credit and low-risk equity strategies. In early stages, but promising in our view.
MvdK: “One thing I want to insist on is that although we are dedicating considerable resources to innovation, our basic principles – that all our decisions should be evidence based, with a clear economic rational and prudent – still apply, even when we look at thinks like machine learning or alternative data.”
We do not think that human behavior – which is at the root of most anomalies our quant strategies exploit – has changed
MvdK: “The world may look very different today, but I don’t think it has changed that much for quant investors. We do not think that human behavior – which is at the root of most anomalies our quant strategies exploit – has changed. So, while we may need to innovate, using new data sources or more powerful tools, this does not mean the basic principles supporting what we do have changed.”