Alternative data is loosely defined as information that is collected outside of the traditional sources of companies, public agencies and associations, when it pertains to investing.
Long the realm of quantitative and hedge funds, alternative data is making inroads into other investment groups. Last January, STOXX Ltd. launched an index that runs on an artificial-intelligence algorithm to select companies with relevant patents awarded in the field of AI.
Two surveys by Greenwich Associates last year found, respectively, that 80% of investors want greater access to alternative data sources in general,1 while 90% of alternative data users said they were satisfied with the returns.2
PULSE ONLINE caught up with Dan Connell, managing director for market structure and technology at Greenwich Associates, to discuss how alternative data is being used by, and changing, the asset management industry.
Dan, is big data a key focus at the moment?
Absolutely, and more and more so as data and computing power come together and facilitate the work of each other. Right now you’ve got so many different types of data, often unstructured and in very large quantities. It may seem difficult to manage it but the technology is there to allow investors to work with that data.
And how exactly is all that data, and in particular alternative data, being used in asset management?
Data has been at the heart of investing for eons. Until recently, the focus has really been on what we would call fundamental data, regularly disseminated to the investor community: historical pricing, income statements and balance-sheet information, earnings projections, etc. That’s been around forever; but we have gotten more efficient at it. The data has gotten better and there are more and more suppliers that have combined the data and the analytics so that process has gotten smoother.
But as the search for the ability to beat the benchmarks continues, folks have gotten into what is called alternative data. The concept is not new. I was working with satellite imagery back in the late 1990s, for example. But in the last couple of years the availability of alternative data that has come to the market is extraordinary. You and I are alternative data sources at the moment. Apps on your phone can transmit your location, and the app provider can turn that information to a central source and produce terabytes of data every day to learn all kinds of things about you. They can provide information to investors about how many people frequented a given Whole Foods store before Whole Foods reports earnings.
Is this data accessible only to sophisticated investors?
The investor community is very mixed on their capabilities to handle that data. There are two ends of the spectrum and everything in between. So at one end of the spectrum you can think of big quantitatively focused hedge funds that have the technical talent. They just want the raw data and they’ll do the processing. At the other end of the spectrum, there are folks that don’t have those capabilities. They may monitor a particular segment and are happy getting a one-page analysis brief from the data source at the end of the day. These investors may find out the data signal a few hours later than everybody else using the alternative data, but still before the broader market.
In the middle, you will have folks that will ask the provider for the data with specific signals, or some analytics on given securities.
What challenges are there in using alternative data?
The big challenge with big data is proving its alpha-generating capability. The back-testing capabilities that exist in the large quant shops and in the analytics providers is key. The bigger firms will take in datasets and will work with them for weeks or maybe months to determine its usefulness. It’s a continuous process.
One of the things that folks look at is the applicability across their portfolio. There may be a particular data set that is valuable but if it is valuable for only 1% of the portfolio, well, that’s not very interesting.
Another issue is that there is so much data that it’s hard to figure out where to go, what to use, how to process it and how it applies to your portfolio. Using data requires a very significant skillset and that skillset is becoming harder to find. I would say there is a talent issue and a cost issue.
The other question is on the regulatory side, particularly around personal data. Data coming from applications that track your location, for example, is a very unregulated area.
And do you foresee alternative data growing further in coming years?
Yes. We expect it to grow tremendously. Like I said, everybody with an exhaust in their business is now an alternative data source. There are aggregators currently tracking up to 700 data sources.
Looking forward, what role will be taken by alternative data and which by traditional data?
The emphasis on data as an informational advantage is critical to investors. I think alternative data as a concept will be around for a long time. Not to the exclusion of fundamental data. That’s why the term quantamental investor has taken hold where you’re combining quantitative analytics with traditional fundamental investing processes to try to generate alpha. I don’t think the fundamental part of that will ever go away.
As the data source becomes more ubiquitous then it moves from being categorized as alternative data to just being data. If you go back to the 1960s, it used to be that the ability to get a real-time stock price was alternative data.
1 Greenwich Associates, ‘Alternative Data for Alpha,’ Q1 2017.
2 Greenwich Associates, ‘Putting Alternative Data to Use in Financial Markets,’ Q3 2017.