A datapoint that has been doing the rounds recently is that “$1.3bn was spent by investment management firms on alt data in 2020 and the market is expected to grow by 30% annually for next 5 years.”
However, like any headline statistic – what it reveals is interesting, but what it conceals is vital. And in this particular case, some of the probing questions behind the headline are – who is buying this data(is the market concentrated to a handful of large quantitative hedge funds)?. What types of data are being purchased?, and what are the use cases and applications?
In this article, we focus on the “last mile” question to describe how alt data is being used by investment firms – as addressing this question is the Rosetta Stone that unlocks all other answers.
Use Case 1: Quantitative Strategies
The most obvious and earliest use case for alt data has been within quantitative or black box investment strategies that identify trade ideas and size positions through algorithmic analysis. Here, the benefit of using alt data has been the availability of differentiated and faster data that can predict market fundamentals or price movements more accurately.
However, key to this use case has been finding datasets that are uncorrelated with existing inputs (e.g. have unique insights), and have limited adoption, such that the alpha potential has not been eroded. As the alt data market has developed, this has become an arms race as early adopters go “truffle hunting”.
Typically, quant funds demand direct access to the hose from which they consume data in its rawest form and develop investment signals, but still expect some basic quality (and legal) checks and back-test verification.
The challenge for providers has been to overcome the moral hazard and identify ways to demonstrate value without the risk of giving away “informational good” IP.
Use Case 2: Investment Research Support
As we move away from quantitative strategies into fundamental investing, we encounter the first use case for alt data with broader adoption – investment research. This use case divides into two distinct paths.
Firstly, research firms are using non-traditional datasets, such as credit card transactions, footfall and web-data within charts to identify KPI correlations to support and investment theses or narrative arguments.
Secondly, research firms are combining alt data with traditional datasets to develop proprietary indicators that are either sold separately as data feeds or incorporated into research reports/chart packs.
Within an investment research context, alt data has the explanatory power to help understand key price drivers and thematic trends.
To do this effectively, research firms must be able to prepare the data analysis (clean-up, tickerise etc.). Additionally, apply statistical modelling techniques and technical analysis to identify insights. Furthermore, they need the sales and operational capabilities to commercialise data products.