Roundtable 'Market Cap Weighted & Factor Investing': Alternatives and New Opportunities

September 14, 2023

Are there opportunities for factor investing within fixed income and commodities? And what opportunities do alternative data and artificial intelligence offer for factor investing?

These and other questions were discussed in the Round Table 'Market Cap Weighted & Factor Investing', in which six experts, chaired by Erik Hulsegge, exchanged insights.

This is part 3 of the report.

By Hans Amesz

Are there opportunities for factor investing within asset classes other than equities, for example within fixed income and commodities?

Lesné: 'There are now a few funds that invest more systematically in fixed-income securities. We have been able to show that it may be beneficial to consider certain types of companies based on their price-to-earnings ratio, their price to book or their creditworthiness.”

Rozemuller: 'We and other providers have a so-called fallen angels strategy. Can you speak of a factor in this context?'

Lesné: 'It is a form of a factor. In the world of fallen angels you have forced sellers such as pension funds, insurance companies and asset managers who have an investment grade index. So there is a direct price impact on the bond in question when it is downgraded. Another interesting element is formed by bonds with a remaining term of less than one year. When rebalancing, we sell those bonds. If there is less liquidity – that does not apply to treasuries – there is essentially some form of premium that you can reap.'

Voermans: 'In addition to factor investing in equities, we also have a rule-based product in commodities. We have identified three key factors within commodities that we systematically address: commodity selection, curve positioning and roll timing. This strategy works very well.'

Huij: 'I think there is also some evidence that factors can be identified for treasuries, high yield and real investment trust. I recently saw a study that showed that factors are even relevant to private equity deals. So factors are indeed observed in a whole range of asset classes.'

Borovkova: 'And are they the same factors or different, new factors?'

Huij: 'The same factors, so value, momentum, small-cap effect and so on. We have been implementing it in credits for some time now, but also in emerging markets.'

More and more alternative data are available. Has this already found its way into factor portfolios?

Borovkova: 'We are very active in this area. I have been working with big data providers since 2009 in the field of 'sentiment', i.e. the news and social media sentiment. I think sentiments in particular, fast-moving data from freely accessible sources around the world, could be excellent signals regarding factors or risk, for example. In terms of ESG, there is now a movement not to rely on the big, slow ESG data providers, but on sustainability, climate-related information that is obtained in the same way as the sentiment data, so from open sources, using natural language processing , machine learning, and artificial intelligence. These types of information sources are much more difficult to manipulate than the information provided by companies themselves to the major data providers. If there is something about an oil spill on Twitter, for example, it will not immediately be visible in the communication of the companies involved. They will try to smooth things over. But Twitter can explode from such information, so to speak. I strongly believe in this, but understand why asset managers generally don't use this kind of information. It is expensive, requires a lot of computing power and requires a lot of expertise to use it properly. But I think it is an excellent addition to multi-factor strategies.'

Is this factor investing in a different way or is it something completely different?

Borovkova: 'If you look purely at correlations, it is very low correlated with other factors. For example, we thought that a sentiment-based factor, of social media use, would correlate with momentum, but that's not the case. It's about different information and I don't know how to untangle it. The trick is to convert this kind of alternative data into usable signals. We have helped several asset managers – not hedge funds – implement these types of signals in addition to their active strategies. Traditionally, hedge funds were the first to embrace it, but others have since successfully implemented these types of signals. It is clear that most large financial companies are testing artificial intelligence to see how it can be taken advantage of when it comes to creating strategies.'

I think sentiments in particular, fast-moving data from freely accessible sources around the world, could be excellent signals regarding factors or risk, for example.

Voermans: 'We do not use machine learning to compile our portfolios, but we do use it a lot to explain the positions in those portfolios. With this we explain from which factor or which restriction in your optimization a certain position has arisen. And also afterwards, when attributing performance. This way we can attribute the performance to decisions in the process. Not only on factors, but also on the exclusion list and other ESG integration. It helps us and our customers to understand where the performance comes from.'

Huij: 'We apply big data techniques to scanning voluminous documents such as reports and company descriptions in order to assess how companies are aligned with the SDGs. With machine learning you can significantly expand the universe of companies to be assessed by analysts, whose scope is obviously limited. More recently, we have also developed a metric that measures exposure to climate transition risk by scanning news articles. We try to determine whether there is a climate shock in the market and then we try to see which companies react positively and which negatively. Based on statistical analysis, we try to determine whether that response is systematic, i.e. which companies respond systematically positively or negatively to a climate shock, in order to ultimately identify companies that are benefiting or disadvantageous from the climate transition.'

Is it to make the portfolio more robust, to deliver alpha, or something else?

Huij: 'That depends. There are customers who have alpha targets and therefore a factor strategy. There are also customers who have a passive strategy, but who then want to build in sustainability. For example, these clients want to exclude certain companies from their investment universe. That basically means that we have to come up with an estimate, such as how a company is aligned when it comes to the climate SDGs. If a company is not followed by a fundamental analyst, we tried to apply an algorithm that in turn uses company descriptions and company reports.'

Borovkova: 'But how do you do that? Do you work with external suppliers or is everything built in-house?'

Huij: 'We have a large department that deals with sustainability research and software development, so it is more efficient to do it internally. An SDG classification has been developed that we believe is better than what is available on the market.'

Do you use ChatGPT?

Rozemuller: 'We are not actively using it now, but we have asked various departments to think about a good way to use ChatGPT. I think it's still too much in its infancy. I should have more proof that nothing can go wrong or misinformation can't be passed on or whatever. Social sentiment has previously been discussed as a factor or driver for investment decisions. It might be a way of using the wisdom of a crowd, or looking into people's minds, and seeing where they're going almost before they act. That can have predictive value. The big challenge is to convince investors that it can be useful to look into it. Maybe we can sell it to institutional investors at some point.'

Finally, is there anything that still needs to be addressed?

Borovka: 'I have a question. How do the various large institutional investors, for example sovereign wealth funds and pension funds, assess what we have all discussed?'

Vijver: 'It depends on what you want to achieve as an investor. The highest return, a better world or a combination of both? Pension funds may very well strive for the highest return for their participants, but for sovereign wealth funds that may be a better environment or a sustainable country. In the latter case, the social part takes precedence over the return part. By taking these fundamentals into account when compiling a portfolio, different portfolios are therefore created.'

Voermans: 'There will be a new pension system in the Netherlands based on contributions and individual accrual of pension assets. From what we can see at this point, there will be a greater need for more stable products, lower volatility and fewer drawdowns. This potentially creates room for the quality factor and low volatility, of course in combination with ESG.'

Lesné: 'I think regulation will be the driving force. In France, for example, most public pension funds have almost completely excluded fossil fuels. Institutional investors have different beliefs, behaviors and roles. There are multiple ways to discuss, explain, sell the use of a specific factor. You don't explain the same to a state fund in Norway as to a fund in the Middle East or Asia, for example.'

Read part 1 of the Round Table 'Market Cap Weighted & Factor Investing' here and part 2 here.