In most emerging markets, mobile and Internet penetration has surpassed financial services or banking penetration, which creates a significant opportunity for financial technology (fintech) companies.
About one in five consumers do not have a credit history. Though, even if people do not have banking history, they do have an online footprint, a form of alternative data.
This means that not only can fintechs reach this unbanked population, but they can also analyze alternative data to serve borrowers. These borrowers might have been rejected by traditional banks and credit unions, namely due to lack of financial history, but fintechs are ready to serve this market. The world of financial inclusion is now opening up thanks to the use of alternative data.
What is alternative data
One traditional use of alternative data was by investors evaluating a company. Instead of using financial statements, SEC filings, etc., investors could look at other data such as transactional data from credit or debit cards or email receipts. This alternative data gives investors granular insight into a company’s performance, which in turn informs investment decisions.
On an individual level, alternative data can bank the unbanked population worldwide (and is already doing so in many countries). Alternative data includes bill payments such as utilities, rent, cable, and cell phones, as well as electronic payments such as transfers and remittances. It also includes public records, loan data, and more. This is all an example of structured data. There are also examples of unstructured data, such as social media use, text, email, images, and more that can be used and evaluated as well.
Why is alternative data growing in importance
After the 2008 financial crisis, many credit scores took a significant hit. A lower credit score means traditional financial institutions are less likely to work with these individuals in any capacity - credit, loans, etc. The crisis also made lenders more conservative and risk-averse. Additionally, many people have no credit at all.
In the US alone, an estimated 35-54 million people are currently outside the “credit mainstream” due to having a thin or non-existing credit file. These people still want to buy a house or start a company, for example, which means access to loans is critical. Alternative data allows financial institutions to score these individuals, giving them access to credit. The risk profile pulled from alternative data functions similarly to traditional risk profiles, giving non-traditional financial institutions a way to offer credit with minimal risk.
Alternative data for small business lending
The unbanked population gaining access to credit is a substantial reason for fintechs to use alternative data, and another reason is for small businesses to access lending. Small-to-medium sized businesses (SMEs) face a finance gap of $2.9 trillion annually. Many traditional lenders are not willing to take on the risk of financing a small business.
Alternative lenders who partner with fintech solutions are filling this gap in the lending market. Though 2016 saw a dip in small business lending, the market bounced back thanks partially to these non-traditional financial institutions, many of which rely on alternative data. This trend has become so prominent that some predict that alternative lenders will soon be the primary providers of credit to SMEs. Furthermore, traditional banks are now looking to fintech solutions to compete with this new way of providing credit.
The challenges of alternative data
Creating an alternative credit score is not exactly easy. Traditionally, a financial institution analyzes historical data such as repayment data, credit history, etc. to design and train a model. This credit score model can accurately predict future actions, such as loan repayment, based on this past data — all of this is structured data. When trying to do something similar with unstructured data, it gets much more complicated.
Companies must create new data sets in order to use alternative data, and they need a lot of data to make them accurate. Credit scoring relies on artificial intelligence, and AI relies on having a lot of structured data. The more structured data the company has, the more accurate the model will be. This is the challenge faced by many financial institutions to create models with unstructured data (or a combination of both structured and unstructured data) to accurately predict future financial behavior. However, many fintechs have met the challenge or are working hard to do so.
The numbers in small business alternative data can also prove challenging. Bank statements can show money coming in, but it might not yet be reconciled with sales and other financial spending, which means people can fool the data. While ERP data may be another option, this can be biased by accounting policies and criteria.
Alternative data becoming commonplace
The use of alternative data is becoming so prevalent that the World Bank issued a global credit information sharing standard with a section on alternative data. The World Bank noted alternative data’s increased importance and value to increase financial inclusion.
Giving credit access to those who normally would go without is important, as is bolstering SMEs globally. Alternative data can do both thanks to alternative lenders willing to take more risk and the fintech companies that make assessment possible. This trend is growing and will likely push more traditional financial institutions to become more inclusive or to partner with fintechs to offer alternative lending options.
Tagged under Financial Research, Mortgage Credit, The Economy, Feature, Financial Trends, Technology, Risk Adjusted, People, Customers, Tech Management, Online, Fair Lending, Credit Risk, Consumer Compliance, Big Data, Feature3, Mortgage/CRE, Fintech,