Could what you do online have an effect on your credit score in the future? It’s a distinct possibility, according to the International Monetary Fund (IMF). Recent research from the organization claims lenders may soon start using data from browsing, search, and purchase history to make loan decisions. Read on to learn more about how this would work and the potential pros and cons.
What the Researchers Say About Using Online Data
So, how would using online data work? In theory, a lender would be able to see a credit report that uses an algorithm to show a variety of a consumer’s online shopping and browsing behavior, as well as traditional financial data like income, credit scores, and payment history. The information would likely be collected using a combination of artificial intelligence and machine learning.
The IMF isn’t the only organization that thinks lenders will use our digital footprints; a 2018 study by the Frankfurt School of Finance & Management also researched lenders looking at online data along with traditional data from credit bureaus. What they found was that even very simple information from online behaviors matched the information found in traditional credit reports and that in many cases, the information was complementary.
In addition, Frankfurt researchers found that combining credit scores along with digital footprints “further improves loan default predictions.” Their findings also showed that the use of digital footprint scoring would allow some “un-scoreable” consumers to gain access to credit, while consumers who have a low-medium credit score could either lose or gain credit access.
The Pros and Cons of Using Online Data for Lending
The idea of using online data for lending purposes is somewhat Orwellian and has some interesting pros and cons.
On the one hand, researchers from IMF suggest that using online data will help more borrowers who have been denied by traditional financial institutions, particularly those who have fallen on tough times. For example, even with mortgage rates hitting historic lows during COVID-19, many lenders became more choosey about who qualified for those lower rates.
If lenders were to take a consumer’s browsing and purchase history into account, it could give them more confidence to approve the person for a loan, even if their credit score had suffered a few dings. Ultimately, using online data could be good news for those who have had trouble getting approved in the past, up to two billion adults worldwide, according to the Frankfurt study.
Another benefit to online data usage in credit scoring, according to IMF, is that it circumvents two significant issues with “hard” credit scores. The first issue is that banks tend to reduce credit availability in times of financial downturn. Unfortunately, this is often the time when people need credit most. The second issue is that it can be difficult for companies and individuals with no credit history to establish one. According to the study, the use of online data could be supplementary in providing people with more access to credit in either of these cases.
Now, onto the cons. What about privacy or security concerns? IMF acknowledges that these are valid, but effectively an “efficiency-privacy trade-off” that would require the government to set official standards for data collection and use. For example, fair lending laws in the U.S. prohibit the use of information about gender or race in lending decisions. Would those rules still apply when it comes to using information from a digital footprint?
What about other factors that might cause discriminatory decisions, such as religious beliefs or political views? This proposed credit scoring system is very similar to the new one implemented in China, where saying the wrong thing online or visiting the wrong website can mean being denied access to loans or even some social events.
Another issue is that there’s no evidence that AI is currently capable of this task or will be in the near future. According to Microsoft AI researcher Kate Crawford, “AI is neither artificial nor intelligent. It is made from natural resources and it is people who are performing the tasks to make the systems appear autonomous.”
In addition, the fact that people are actually performing these tasks brings up the growing problem of bias and discrimination in AI. One example of this is women being offered less credit than men based on credit-worthiness algorithms. Unfortunately, it’s not just how the systems are applied, but also how people build and train AI to see the world. Gartner, a marketing research and consulting firm, predicts that 85% of AI projects through 2022 “will deliver erroneous outcomes due to bias in data, algorithms, or the teams responsible for managing them.”
And another important question: how will your information be protected from data breaches?
Traditional Financial Data Still Rules
Fortunately, the idea of using online data to determine lending decisions is still fairly speculative at this point. However, it brings up many questions about how this could affect the rental housing industry as a whole, including Fair Housing laws and tenant screening. It also brings into question the growing use of AI for tenant screening and making rental decisions.
AI is still a young technology, and as we’ve discussed, one that can be subject to bias and discrimination. In contrast, traditional screening methods remain the most objective way to evaluate a prospective tenant and avoid potential discrimination issues. Another thing to consider: would you want to reject potentially good tenants who were blocked from creditworthiness because of online data—or, conversely, accept tenants that have would have otherwise been denied?
While there’s no way to know for sure if online data will be used to determine creditworthiness in the future, it does give some food for thought.
Landlords Property Managers Contact TSCI