In data-hungry capital markets, the rise of big data quickly yielded a bumper crop of alternative data providers, ready to search vast unstructured data stores for elusive alpha. The popularity of alternative data has been further fueled by headline-grabbing success stories, including predictions of Apple results based on social media sentiment, and accurate forecasts of Chipotles quarterly results based on Foursquare check-ins.
Fans of alternative data, see it as the secret sauce that can give investors a richer, more nuanced view of markets than standard market data quotes and benchmarks. But no matter how many successes alternative vendors log, their path to survival is fraught with pitfalls.
A Greenwich Associates study conducted late last year of 69 US and European chief investment officers, portfolio managers and traders at investment management firms, found that 80 percent of investors want greater access to alternative data sources. But those same investors say that alternative data sources are incomplete, lack quality controls, and can be hard to integrate into their workflows and internal investment analysis processes. Therein lies the rub.
A Rocky Road
The biggest investment management firms, have teams and technology to hunt for and evaluate new data sets. The problem is, only a handful of the most sophisticated firms have that capability. For many firms, its an enormous burden to even study and analyze whether new data sets add value. To begin to assess new data sources, firms have to acquire a fairly large (and often costly) data sample. Firms then must devote software engineers toward running regression testing, normalizing the data, and tailoring it to the firms specific needs. Not all new data sets firms investigate will end up adding value, so investors may also have to plan for the sunk cost of evaluating some data sets that wont work out.
Meanwhile, once successful alternative data sets are identified, they have to be stored, maintained, and fully integrated into a firms workflow. For many firms, with disparate aging infrastructures, that data integration can be costly and complex, and can greatly delay a firms time to implementation.
A Path Through the Cloud
Alternative data vendors best hope for making their data accessible and affordable to a broad cross-section of the investment management industry is through providing both evaluation and distribution data sets through a pay-per-use model. For some data firms, this is a potential strategy for greater success, for others it may be an essential path to survival.
What is attractive about alternative data, is the unique level of detail and insight it can provide into a market, whether it be energy data, geospatial data, retail transaction data, or social media sentiment. But secret sauce never stands alone. Alternative data is an enhancer that should be used in conjunction with fundamental market data to shed new light on the big market picture.
As a component of a cloud-based platform, alternative data sets would become easier for firms to access and evaluate. Investors would be able to benchmark alternative data against more traditional data sets in the cloud to assess whether it is predictive. A cloud-based platform would be able to provide tools to normalize data and maintain quality controls. Cloud-based storage would greatly cut firms data storage and maintenance costs. Meanwhile, as part of modern, cloud-based infrastructure, alternative data could be delivered via flexible technology and standardized protocols that can be seamlessly and efficiently integrated into a firms workflow.
In a world with 500 million tweets per day, more than 33 billion U.S. credit card payments per year and rapid advancements in analytics and artificial intelligence technologies, the amount of alternative data offerings is expected to continue to grow exponentially. In order to avoid being buried by the competition, alternative data vendors must make their data sets easy for investors to discover, test and deploy.
