Smartphone usage is likely to be tracked someday soon in the U.S. to determine whether you qualify for a loan or what interest rate lenders should charge.
Already, some lenders in a few African nations are using smartphone apps to monitor how often a user texts or charges a phone battery, then to correlate such information with the user's ability to repay a loan. The apps, in some cases built by U.S. startups, are being downloaded by users who opt-in to have their usage habits monitored.
It's only a matter of time before that practice comes to the U.S., analysts said. Some companies are already using social network behaviors to determine credit risk for thousands of U.S. consumers, although that data may not directly rely upon smartphone habits -- yet.
"Much of our lives can viewed from our smartphone usage—it's an extension of someone's persona—and can inform a lender about a person's creditworthiness," said Gartner analyst Avivah Litan. "This is especially useful for lending money to people who are either 'unbanked' and therefore have no established and publicly-available financial records, such as is often the case in developing countries."
Litan also said smartphone usage habits can and will be applied to reviewing creditworthiness of residents in the U.S. or U.K. or other mature economies. That would be especially true for students or immigrants with 'thin' credit files who have not had the time or experience to establish a credit record with banks.
"These [smartphone usage] models can be useful for assessing the creditworthiness of anyone—banked or unbanked, rich or poor, in any country," Litan added.
Litan and Patrick Moorhead, an analyst at Moor Insights & Strategy, said the practice could help more people get needed loans and improve the accuracy of lending decisions, potentially even lowering the cost of lending money.
"Risk scores based on a person's smartphone usage is sure to make lending decisions more accurate so that more good borrowers are accepted and less 'bad' borrowers are rejected," Litan added.
"Lenders and insurers have used scoring systems for decades to determine risk probability," Moorhead said. "Smartphone behavior is just another set of data points to assess the probability of something negative happening. The good thing about this is that lenders could get a much better view into what activity the applicant is really engaged in, which could be much better than looking at a [typical] credit report. This is good for lender and applicant."
How it works
Correlating a person's smartphone behavior with credit risk isn't so far-fetched; that kind of connection is already part of the accepted science of credit risk analytics -- itself a subcategory of business intelligence. Credit risk analytics has been around for decades, although using smartphone use behavior is a fairly new add-on for judging risk.
In one example, a person might be found to recharge a phone often, which could undermine creditworthiness. The logic behind that decision would be based on correlating frequent battery charging with other factors, such as how old the handset is, how often it is used, what apps are used and how much battery power those apps consume, Litan noted.
"A lot of battery charging alone isn't necessarily correlated with bad credit, but it could be a case of an old phone with a battery that had to be charged a lot and therefore a conclusion is reached that the user couldn't afford a new phone," Litan explained. "No lender would rely on just that single variable to evaluate creditworthiness."
Based on usage patterns of 150,000 smartphone customers in Kenya and Tanzania, as collected by lenders there and reported by the Wall Street Journal, good credit risk customers are correlated with those that make calls in the evening to avoid daytime prices and with those that receive more texts than send them (which would make sense if it costs more to send than receive a text, as is often the case abroad).
Bad risk customers drain their phone battery more quickly than average or send more texts than they receive, the data on users showed.
The data also correlated smartphone gamblers and frequent travelers with those likely to repay loans. Some analysts theorized that known gamblers might be familiar with the harsh penalties (such as getting knee-capped) for missing a loan payoff. Frequent travelers, in turn, might be more familiar with the concept of borrowing money, and paying it back, when making transportation and housing arrangements in advance.
Branch app for Kenya borrowers, built by Silicon Valley startup
Branch International offers a free Android for use in Kenya that allows for loans of up to 50,000 Kenyan Shillings (about $490 U.S. dollars). According to Branch's website, a user can download the app and the funds are delivered to a mobile money account in under five minutes.
Also on its site, Branch says it "collects data from your phone when log in to our app. This includes information about your device, SMS logs, call logs and contact lists. We use this information to assess your creditworthiness and provide a seamless experience when applying for a loan."
Branch also tells users it it will not sell such information to third parties or share it unless for certain business purposes like reporting to banking authorities. The app has been downloaded more than 50,000 times.
The average branch loan is reportedly just $30, charged at between 6% to 12% interest, based on a borrower's creditworthiness.
Branch is a Silicon Valley startup founded in early 2015 by Matt Flannery, who now serves as its CEO. He was co-founder of the non-profit Kiva.org, a microlending group.
Analysts said it is likely that Branch and other startups will expand the use of their apps beyond Kenya, probably to other developing nations. But it is possible that such apps could be used for lending money to needy residents in the U.S. or other developed countries. Officials at Branch and other institutions couldn't be reached to comment on their expansion plans.
For nearly a decade, residents of Kenya have been able to transfer money with cheap cellphones, but the monitoring of smartphone behaviors to determine creditworthiness is fairly new.
A recent 60 Minutes report described the widespread use of cheap cell phones for transferring funds in Kenya without the need for a bank or credit history. Safaricom, a wireless company, launched M-Pesa in 2007 in Kenya for peer-to-peer texting of money with the backing of Vodafone; the service has since expanded to Tanzania, Afghanistan, South Africa, India and parts of Eastern Europe.
Santa Monica, Calif.-based InVenture Capital Corp. has made loans in Kenya and Tanzania via smartphones. The company uses more than 10,000 data points per mobile user to determine credit risk, according to its website.
InVenture is headed by CEO Shivani Siroya, a former officer for the United Nations Population Fund. The company lists several investors, including Google Ventures.
Other companies in the space include Saida, which provides quick loans via smartphone in Kenya and relies on a Google Play app to monitor smartphone behaviors. That app was developed by Greenshoe Capital in Los Angeles, founded by Kenneth Ngetha and Kyale Mwendwa.
It's not just smartphone behaviors being monitored
While recent startup activity has focused on smartphone behavior tracking, other companies focus on user behaviors on social networks like Facebook. One example is Lenddo, which works by loading code onto a desktop or mobile device with the user's permission.
Other companies like Affirm, LendUp and ZestFinance have long used social media and other online behavior, as well as data from data brokers, to determine creditworthiness of U.S. consumers. Analysts believe these kinds of companies would be interested in adding in smartphone behaviors outside of social network activity to determine risk.
Litan noted that major U.S. wireless carriers have been monitoring smartphone behaviors for five years or longer to find the customers most likely to buy products and services. The business intelligence applied by carriers to smartphone behaviors helps determine who are the "influencers" who buy products and services and who can influence their peers to buy more, she said.
That kind of business intelligence can also be applied to creditworthiness, although it isn't clear how interested the carriers are in engaging in such analysis.
Questions of privacy
Because smartphone users are consenting to have their device behaviors monitored when they download apps like Branch, most experts don't see major concerns about a person's privacy.
In many cases, the people wanting loans via smartphones need money badly enough that they will let a lender see how often they are talking, texting or browsing, experts said.
"The obvious reason people do this is that they have no alternative," said Marc Rotenberg, president of the non-profit Electronic Privacy Information Center in Washington. "Yes, they are desperate, and the data mining firms exploit their vulnerability just like loan sharks. The obvious question is what is done with all the data obtained from the applicant."
In the case of Branch, the company says on its website that it will not sell a user's information to third parties. With other lenders, it is less clear how data is used once a loan has been given or refused. Litan said it is likely that the lenders aggregating the data are not selling it and are probably bound contractually not to do so. Because the lenders and carriers technically own the data they gather, it is also unlikely a user would have the ability to delete it, she said.
Jack Gold, an analyst at J. Gold Associates, said the practice of analyzing smartphone behaviors for judging creditworthiness is another example of how personal privacy is fading away. "In the long term, there is no privacy," he said. "In a completely connected world where every electron passes through some sort of broker, the amount of privacy you'll have is close to zero, unless governments step in to change that."
Gold was harsh on the practice of measuring smartphone use to evaluate credit risk. "It makes no sense to me," he said. "There are many people who are good credit risks that don't use their smartphones all that much, and the contrary is also true."