Datamine or perish.............

Data-driven finance

Go figure

A new class of internet start-ups is trying to turn data into money

Mar 17th 2011 | from the print edition of the ECONOMIST

ASKING a friend for a loan is a quick way to tide you over till the end of the month. But Wonga, a British internet firm, is almost as fast. Those in need of cash give the service some basic information about themselves. If Wonga considers them trustworthy, the money—as much as £400 ($641) for up to 30 days for first-time users—arrives in their accounts within 15 minutes.

Consumer advocates are highly critical of the service: its annual interest rate currently exceeds 4,000%. But Wonga, which uses funding from its venture-capital backers to make loans, is attracting attention for other reasons. The firm is one of a growing number of internet start-ups that mine unconventional forms of data to offer financial services.

Klarna, a Swedish company that counts Sequoia Capital, a noted venture-capital firm, among its financial backers, allows consumers to shop online by simply typing in their date of birth, name and address. Klarna fronts up the payments and charges retailers a fee for making online purchases easier. Shoppers pay only once they have received the goods. ReadyForZero, a start-up in Silicon Valley, lets people see all their credit-card debts in one place and puts together a plan to pay them off.

Second, decisions are made very quickly. Klarna and Wonga feed all the data through elaborate algorithms which determine, almost in real time, how likely it is that a user with a certain data profile will default. Consumers who shop online at 3am may find themselves among the 20% of buyers who get rejected by Klarna. Having a mobile phone with a contract helps to get money from Wonga (which says "no" to 70% of applications). But no single factor is decisive, says Mr Damelin. "It's about how the data connect to each other." Klarna's algorithms are regularly updated to reflect new types of behaviour.Of course, conventional lenders sift through information about borrowers to decide whether someone is a good risk. But the start-ups are different. First, they combine data from many sources. Klarna started by looking at conventional credit scores, but it says that the actual behaviour of shoppers has much more "predictive power", in the words of Sebastian Siemiatkowski, Klarna's chief executive. The company receives a lot of data from online stores, including things like the time of purchase and whether the consumer's name and address were typed or copied in (the latter is more likely to signal fraud). Wonga draws on "all publicly available data", in the words of Errol Damelin, Wonga's boss, who does not want to be more specific for security reasons. ReadyForZero accesses data on users' credit-card transactions.

Third, whereas other lenders may play up their branch networks or customer service, data are the start-ups' big source of advantage. ReadyForZero's service is free: users essentially pay with their data. It is still trying to figure out how it can use this information to make money, says Rod Ebrahimi, one of the firm's founders. One idea is offering credit scores that are more fine-grained than those currently provided.

All three start-ups are already pretty popular. Wonga claims to make more than 100,000 loans a month. About 5m consumers in six European countries have used Klarna to pay for their online shopping. Users of ReadyForZero have registered nearly $30m in credit-card debt only six weeks after the service went live.

Yet the firms' success is hardly guaranteed. One risk is financial regulation. Wonga just had to increase the interest rate it shows on its site by 1,500 percentage points because an EU rule required it to calculate interest differently. Online privacy is another issue: ReadyForZero's users may balk if they find that their data, albeit anonymised, will be used elsewhere.

Even if these start-ups fail, the trend they represent is unstoppable, says Mr Damelin. As such offerings grow, "it will become less and less obvious why you would need a retail bank." That is wishful thinking. But for financial firms that do not use data smartly, the number's up...............
.