What financial institutions should be careful with when it comes to alternative data

September 24, 2021

Alternative data is powerful in the quickly evolving world of financial technology. Alternative data, for instance, leverages underutilized data points such as social media use, bills payment history, and even GPS tracking to verify the identity of a potential loanee. Easier verification paves the way for alternative credit scoring, granting easier access to credit to anyone who might need a loan. Used this way, alternative data can also allow traditional lending institutions to reach out to the billions of people around the world without bank accounts, paving the way for greater financial inclusion. 


Other benefits of using alternative data include improving prediction rates, helping maximize returns, and boosting credit risk management. It’s no surprise then that a survey from the Bank of America has found that roughly 50% of all investment firms now use alternative data.


The use of alternative data is a relatively new phenomenon (59 percent of investors who said they use alternative data had been doing so for less than two years).



So with alternative data becoming an undeniably crucial asset, what should financial institutions know about using it?



Not all alternative data is created equal



Research from management consulting firm Oliver Wyman points out that there are certain characteristics that a good source of alternative data should have:



Data sources should be up-to-date and have broad and consistent coverage. Looking at data from mobile phones, for example, would have a higher coverage data considering that mobile phones have become an ubiquitous tool in developing markets such as the Philippines (mobile internet user penetration in the country is about 72.1 percent). In contrast, only a small population of Filipinos may possess traditional data sources such as a banking history. 


Data should also give specific info about a person, such as whether they belong to a specific income class or whether they’ve missed payments in the past. It also goes without saying that data should be updated and data sources should have a system for ongoing data verification and management.


Companies shouldn’t forget that the alternative data they use should be relevant to what they’re trying to predict. Doing so makes it easier to use the alternative data as an additive to traditional bureau data, and will improve the predictive accuracy of any new digital credit score.


Finally, ensure that any data used is compliant with regulations (countries all over the world have different acts regarding credit data, from the Fair Credit Reporting Act in the US to Credit Information Companies Regulations Act in India).