How alternative data quality can differ
January 5, 2022
Alternative data is defined as information taken from non-traditional data sources, like mobile devices, satellites, social media history, credit card transactions, and so much more. Modern usage of alternative data stands in stark contrast to the reliance financial institutions once had on traditional sources of data, such as banking history and administrative records—which many unbanked, underfinanced people simply do not have.
Alternative data gained popularity because of its ability to paint a more holistic picture of a potential client/loanee. With alternative data, lenders and other financial institutions gain more insight into how likely someone is to successfully pay back their loans, allowing them to expand their services while simultaneously minimising the risks that they take/significantly improving their capacity for credit risk management.
Yet not all sources of alternative data are created equal. For example, data sourced from rental payments are both widely available for borrowers with no pre-existing credit file and can be a reliable indicator of a borrower’s capacity to pay back loans. The problem is that the rental market is fragmented and there is no uniform template for data reporting. Its indiscriminate use will pose problems in the future if used for fintech credit scoring or even vetting someone’s background.
What makes a good source of alternative data
Lenders that utilise the right alternative data sources first will gain a vital advantage in reaching and successfully serving new markets.
According to a report by management consulting firm Oliver Wyman, the best alternative data sources should have broad and consistent coverage rather than being too niche. For instance, mobile data would meet this criterion as world mobile penetration stands at nearly 80%, even in developing countries. At the same time, however, this data should be detailed enough to glean insights from—it won’t work if the data is so vague and nonspecific that it says nothing about the person’s habits and behaviour.
Good data is also up-to-date, better if it has a system of ongoing verification and management. This also helps improve the predictive power of the data, relevant to the insights a lender is trying to glean about a borrower. Additionally, it’s best if the alternative data used complements any pre-existing data already collected by the institution in question. By combining the two, the predictive accuracy of any new score is made better by improving the signal-to-noise ratio.
Finally, alternative data sources must comply with pre-existing regulations for consumer credit and data handling in their jurisdiction (Fair Credit Reporting Act, Equal Credit Opportunity Act, the General Data Protection Regulation Act, etc.). Ensuring their compliance not only saves institutions from future problems with regulations but also acts as a good way to filter out unreliable sources.
The rising use of alternative data
The global alternative data market was expected to grow from $1.70 billion in 2020 to $2.41 billion in 2021 at a compound annual growth rate (CAGR) of 41.4%. Yet the demand wasn’t entirely driven by evolving credit risk management or even fintech credit scoring: growing interest in stock market trading was a major factor, with investors looking to leverage alternative data in order to make better investment decisions.
It’s clear that the use of alternative data is here to stay. One example of a fintech company helping institutions leverage it is credolab, which provides smartphone metadata-based digital scorecards that take the strictest alternative data standards into account. Credolab, for example, makes a point of only accessing 1st party and privacy-consented smartphone metadata to ensure data privacy compliance. By using credolab’s solutions, institutions can be sure they are using properly sourced alternative data.
Alternative data is a powerful tool but not all of it is made equally. Institutions that accomplish due diligence and search for solutions (such as those provided by fintech companies) that vet data sources will clearly be at an advantage.
To learn more about how credolab’s alternative credit scoring solutions can help your company achieve quality results, speak to us today and schedule an appointment here.