In April, we brought you the "FAKE-O [4]" score or, as we delicately described it, "banks' brilliant plan to lend to deadbeats."
The problem with FICO scores - or any other measure of creditworthiness where the model inputs are linked to payment history - is that they don’t end up being very high for people who aren’t good about making payments on time and that’s bad news when you’re trying to build a successful “originate to sell” model.
If you’re a subprime auto lender for instance, and you want to make as many loans as you can so that you can sell them to Wall Street where they will be put through the securitization doomsday machine, you don’t really care about the creditworthiness of the borrower, but you have to base your decisions on something because if any annoying regulators ever come sniffing around your books trying to figure out why the entire market just disintegrated into a smoldering pile of subprime ashes you can point to a number and say "see, by this metric things should have been ok."
So what do you do? Well, one thing you can do is come up with an alternative way to measure creditworthiness and preferably one which is virtually guaranteed to come back with a happy result, which is exactly what Fair Isaac decided to do. Here’s what WSJ said [5] about the new "unnamed" score at the time:
The new score, which isn’t yet named, will be calculated based on consumers’ payment history with their cable, cellphone, electric and gas bills, as well as how often they change addresses and other factors, according to Fair Isaac, also known as FICO. Traditional FICO scores that lenders use in the approval or rejection process are calculated based on the information in the credit reports from the three major credit-reporting firms, Equifax Inc., Experian PLC and TransUnion.
The new score will instead pull data from a separate database of telecommunications and utilities providers maintained by Equifax.
Got that? The "new score" will determine how creditworthy someone is by asking the utility company if the lights are still on and asking the phone company if the phone is still connected.
Now obviously, some of the last things people will stop paying as they descend into abject poverty are the electric bill and the phone bill, so if you wanted to engineer “evidence” of creditworthiness, this is a good place to start as anyone who managed to keep electricity flowing to their home last month suddenly deserves a loan for a car.
But because that wasn’t (nearly) absurd enough, Facebook has a better idea. Mark Zuckerberg will determine your creditworthiness based on your friend network.
No, really. Here’s The Consumerist [6]:
Earlier this year Facebook announced it would dip its toes into the pool of mobile payments by launching a system that allowed users to send money to friends via the Messenger app. Now it appears the company may take things a bit farther after receiving approval for a patent this week that would allow creditors to determine whether or not someone is worthy of a loan based on their circle of friends on the social networking site.
The patent — which was actually applied for by Facebook back in August 2012 — is for a system of authorization and authentication based on an individual’s social network. It could have several uses, including filtering out spam email and offensive content, and improving searches on the site.
However, it’s the use related to approving or denying a loan request based on the friends you keep that is a bit worrisome.
“When an individual applies for a loan, the lender examines the credit ratings of members of the individual’s social network who are connected to the individual through authorized nodes (connections),” the patent states. “If the average credit rating of these members is at least a minimum credit score, the lender continues to process the loan application. Otherwise, the loan application is rejected.”
According to the patent, the lender would be able to access a potential borrower’s social circle by submitting a request for information from Facebook’s databases. They would then receive a series of lists – grey, black and white – that would be used to determine the average credit score for the would-be borrower’s friends.
[7]
And while Bankrate.com's Greg McBride assures us [8] that this is "nothing to lose sleep over for people with decent credit history," it's not the borrowers we're concerned about. The question is this: what happens when billions in loans that were made based on Facebook's proprietary friend check end up packaged and sold to investors? Or maybe more to the point: how utterly insane will this one day look when everyone suddenly realizes that a wave of defaults all have one thing in common... that the underlying loans were made based on borrowers' social networks?
