Your email address, your device type and the time of day you surf the web may seem like innocuous digital details, but if a new study rings true, they could be used in determining your mortgage options—or other credit opportunities—down the road.
A recent study from the Frankfurt School of Finance & Management has found that “digital footprints”—a person’s online habits, profiles and activities—can be a good measurement of creditworthiness.
And when combined with existing FICO data, the study’s authors say it could improve lenders’ ability to predict loan default rates significantly.
The Digital Footprint
According to the study, a digital footprint is the trace information that web users leave behind when going about their day-to-day online tasks.
“Even without writing text about oneself, uploading financial information or providing friendship or social network data, the simple act of accessing or registering on a webpage leaves valuable information,” the study reported.
The study analyzed a number of digital footprint variables across e-commerce transactions, including the person’s device type, operating system, the time of day they made their purchase, their email address and how they came to the website. The analysts then compared those variables to default rates on the purchases.
Though not all variables had a significant link to creditworthiness, some—like device type and email address, for example—proved to be strong indicators of a person’s likelihood to repay and, in the future, could be used to enhance lenders’ abilities to evaluate potential borrowers.
“Our results suggest that even the simple, easily accessible variables from the digital footprint proxy for income, character and reputation and are highly valuable for default prediction,” the study reported.
Device Type and OS
One of the strongest predictors of a person’s likelihood to default was device type. According to the study, Android users are significantly less likely to repay than those on Apple devices, with Android users defaulting almost two times more than iOS ones.
Apparently, iOS device ownership is “one of the best predictors for being in the top quartile of income distribution,” according to the study.
Still, if those are iOS devices are mobile ones—not desktops or tablets—the chance of default is higher nonetheless. The study found that defaults on mobile phone orders were three times higher than on desktop orders and 2.5 times higher than tablet ones.
Email Addresses
A person’s email address—including the characters in the address as well as the account’s host platform—can also provide valuable information about a person’s creditworthiness, according to the study.
Customers who have their name in their email address? They're 30% less likely to default than other users. Those who have numbers or misspellings in their email? Their rate of default is significantly higher.
There’s also a stark difference in repayment rates between those with addresses on outdated and free platforms—like Yahoo and Hotmail—versus accounts that come at a cost, like T-Online (the German equivalent of Comcast.) In the study, those using Hotmail were more than twice as likely to default than T-online users, while those on Yahoo were nearly four times as likely.
Interestingly enough, how a user types in that email address, as well as their name and other order details, can also help predict default rates.
“Customers who use only lower case when typing their name and shipping address are more than twice as likely to default as those writing names and addresses with first capital letters,” the study reported. “We find that eponymous customers who use their first and/or last name in their email address are less likely to default.”
Access Point and Time of Day
There are dozens of ways a user can come to a website, and according to the study, each one can be used as a measure of creditworthiness. Users coming to a site from a search engine ad, for example, are twice as likely to default as those coming from a price comparison website.
In fact, those who arrive through a paid ad have the highest default rate among all users. The study found that these users are more likely to be impulse shoppers who might buy beyond their means.
“One possible interpretation is that ads, in particular, ads that are shown multiple times on various websites to a customer, seduce customers to buy products they potentially cannot afford,” the study reported.
The study also found that the time of an online transaction could indicate a user's self-control and likelihood to impulse buy. Customers buying between noon and 6 p.m. were half as likely to default on their purchases than those who bought between midnight and 6 a.m.
Digital Footprint + FICO
Though the study found that digital footprints are just as effective—if not more—than traditional FICO methods, its authors don’t advocate using it as a standalone measurement tool.
Digital footprints, instead, should be used to improve predictions from traditional credit scoring models.
“A lender that uses information from both sources can make superior lending decisions,” the study reported. “Even simple, easily accessible variables from the digital footprint provide valuable information for default prediction that helps to significantly improve traditional credit scores.”
Using digital footprints to complement FICO scores can also help lenders cut costs and speed up their processes.
“The digital footprint can also be used to process applications faster than traditional lenders,” the study reported. “Evidence suggests that a few variables from the digital footprint can (partially) substitute for variables that are otherwise more expensive to collect, otherwise take significantly more effort to provide and process, or might only be available to a few lenders with specific access to particular types of information. Almost every firm operating in the digital sphere can effortlessly track the digital footprint we use.”
China has already moved to adopt a loosely similar credit model—dubbed “social credit.” But according to Jason van den Brand, CEO of online mortgage lender Lenda, Americans shouldn’t expect to see digital footprints used on U.S. mortgages anytime soon, mostly because of how popular 30-year loans are.
"It won't be adopted by conforming mortgage for quite some time,” van den Brand said. “Alternative credit profiles, or digital footprints, are more valid for personal loans as the timeframe of repayment is short enough to see if this new credit-scoring model actually performs or beats the existing. Given that 80% of our customers select 30-year fixed mortgages, we cannot effectively test a new scoring model today as we won't learn if it actually works for over a decade. That's the challenge.”
Scoring the Unscorable
Whether it’s tomorrow or years down the line, one thing is certain: digital footprints have the power to expand credit access for billions of consumers.
“Digital footprints can facilitate access to credit when credit bureau scores do not exist, thereby fostering financial inclusion and lowering inequality,” the study reported.
Currently, about two billion adults are unbanked, unscored or otherwise lack access to financial and credit services. For those with online access, the digital footprint model could help open the door to improved financial wealth.
According to Ace Watanasuparp, vice president of residential lending at Citizens Bank, reaching these potential borrowers is where the true power of digital footprints lies.
“If in the future, the use of digital footprints to assess credit scores becomes more widely used, I believe it will be a more useful model for consumers looking to obtain a home loan that don’t have credit or have very little credit,” he said.
Note: This post has been updated to clarify the difference in default rates between Yahoo, Hotmail and T-online email users.
