In 2014, the University of Dublin stumbled upon something very interesting. A study conducted by Ciaran McAteer, submitted as a dissertation, found a strong correlation between users’ sentiments on Twitter and the exchange rate of bitcoin. McAteer concluded that “Twitter has become a valuable source of user sentiment” and could be used “as a metric for predicting financial markets.”
Two years later, in 2016, location data platform Foursquare looked at data reflecting the foot traffic in 1,900 Chipotle stores and predicted a 30% drop in Chipotle’s sales in the first quarter of 2019. The prediction, indeed, was confirmed when Chipotle reported a 29.7% Q1 drop in sales.
Then, there is Pier 1 Imports, which used satellite imagery to monitor traffic in the vicinity of malls and retailers. Pier 1 used the data to warn investors about guidance cuts.
The above are only three of the myriad ways modern market and investment research firms use non-traditional data to make investment decisions.
Although, in 2021, the moniker doesn’t seem right: non-traditional or ‘alternative’ data is not non-traditional anymore. For many investment research firms, alternative data is held in equal regard with traditional data. It’s primary and very well driving key investment decisions.
What is alternative data?
Alternative data is data that informs an investment decision but is sourced from non-traditional or alternative sources.
Traditionally, investment decisions are based on data extracted from financial statements, SEC filings, trading reports, press releases, and the ilk. The data has always been deemed trustworthy and reliable by analysts for assessing a company’s prospects.
However, markets have grown extremely complex in the last few decades, and their movements have become harder and harder to predict. As a result, investors widen their nets and look for more differentiated data to back up their claims.
Modern investors seek alternative data for a more nuanced perspective and, therefore, more accurate insights. Essentially, the extra information provides investors with an extra edge over the competition.
How is alternative data generated?
Before we look at the different types of alternative data, let’s first look at who generates it. Another way to put it is to ask, “how is alternative data generated?”
Alternative data is generated in three ways.
1. Individuals
McKinsey estimates that every day, the world generates over 2.5 quintillion bytes of data. A lot of that data is generated by individuals searching the web, posting updates on social media, consuming content, interacting with apps and websites, and leaving reviews. The data leaves trails, however complex, that illuminate their sentiments and spending patterns.
2. Corporates
Besides quarterly reports and trading records, corporates are also a source of data regarding transactions, purchases, loans, taxes, and so forth. Let’s not forget that most corporations today have an online presence. Traffic, whether on their website or application, is also a crucial source of alternative data.
Read more: 71% BFSI Firms Use Big Data Analytics to Gain Competitive Advantage – It’s Uses in Equity Research
One vital difference between the data generated by individuals and the data generated by corporates is that the latter tends to be structured and quantitative, whereas the former is unstructured and qualitative. Sentiments are difficult to reduce to numbers.
3. Data generated by ‘smart’ devices
Today, there are more than 10 billion devices connected to the internet. We call them Internet-of-Things (IoT) devices. Back when the internet was a privilege, even in the West, only a few devices could be connected to it. But as processors shrunk and the internet’s scale expanded, IoT devices exploded.
Today, most devices are smart or intelligent, from watches and TVs to thermostats and parking sensors. Given our constant interaction with them, the data they generate is also a critical source of alternative insights.