The rise of alternative data sets is likely to create a wider dispersion of investors’ returns as the wide range of available data will lead investors to take more varying approaches than they have done so in the past, according to APG Asset Management.
Until recently, most financial strategies have relied on a handful of available data sets from a limited group of data vendors. But today, there are more than 1000 different data sets available that have relevance to investing, Gerben de Zwart, Managing Director of Quantitative Equities at APG, said at the [i3] Equities & Equity Alternatives Forum in Melbourne last week.
“Everybody has access to these data sets, but there are more than 1000 data sets, so I think that the industry will become very interesting because performance could rely or depend on the alternative data sets that are employed by the manager and it could lead to more diversity in the excess returns,” De Zwart said.
“So yes, everybody can use them, but what is the likelihood that everybody will use the same data sets when you have 1000 data sets to pick from? Investing is getting more interesting.”
What is the likelihood that everybody will use the same data sets when you have 1000 data sets to pick from?
APG, the administrator and asset management arm of Dutch pension funds ABP, bpf BOUW and other pension fund clients, manages about half of its EUR 165 billion (AUD 263 billion) equity investments through quantitative strategies, while their allocations to liquid commodities are largely quantitative.
De Zwart said his team screens a 100 data sets every year , while about four data sets are researched and the insights gained from the more promising ones are tested in paper portfolios.
“The biggest challenge is the enormous amounts of alternative data sets. In the old days, when there was a new data vendor, the whole industry was jumping on it and researching it,” he said.
“I think right now with the overwhelming amount [of data], the question is very different: ‘What information do you need to enhance your investment strategy?’ It is not led by the data vendors, but you drive your own data needs and then select the data sets. This is a paradigm shift in the industry and academia.
“Our way of thinking is that we are long-term, responsible investors, so that means that we try to avoid data sets with short-term, high-frequency information. We target to work with data sets that are low turnover and have a pretty long history,” he said.
These data sets don’t always contain new investment signals or factors, but sometimes simply form a quicker way of obtaining financial information, he said.
“Alternative data sets don’t always bring new information, but they can help to speed up factors and speed up the pace of transmission,” he said.
“If you look at earnings call transcripts and the information that you can extract from these, I think in the good old days analysts were listening in to these earnings calls, after some processing time they would adjust their forecast for the company, but if you are able to transcribe and analyse the sentiment of the earnings calls yourself, right away, you take out the time to market from the analyst.”
He said he believes corporate bonds will be the next sector to benefit from quantitative approaches and APG is about to provide two managers with seed money for such an approach.
“I see signs that the corporate bond market is getting more and more a structure like the equity markets. So you see passive indices, the first signs of smart beta and more high-conviction managers and that is exactly the structure that the equity market has today,” he said.
“The difference between the equity market and the corporate bond market is that there is no central exchange in the bond market that is clearing all the transactions. But because of regulations that have been introduced a couple of years ago, there are now good data sets becoming available on the transaction prices of corporate bonds.
Our hypothesis is that the corporate bond markets will start to behave like the equity markets and we are about to seed two external managers on multi-factor corporate bonds. That is the next phase
“Our hypothesis is that the corporate bond markets will start to behave like the equity markets and we are about to fund two external managers on multi-factor corporate bonds. That is the next phase.”
APG also has one external manager it believes has skill in timing factors, but De Zwart said it is unlikely such an approach will become a dominant strategy in the portfolio.
“Factor timing adds turnover, so apart from the skills you need to get a decent return from it because you have to make up the trading costs,” he said.
“In our case, size is an issue. Internally we run in excess of EUR 50 billion (AUD 80 billion) and it is very challenging to have additional turnover just to chase factor timing strategies.”
Gerben de Zwart, Managing Director of Quantitative Equities at APG, spoke at the [i3] Equities & Equity Alternatives Forum, which was held on 26 and 27 February 2019 at the Pullman Melbourne on the Park.