Payday loans. Kind of. Not really. But really…Backed by Mathemagic Algorithms
On the face of it, payday loans are simple. And toxic.
Desperate borrowers, often those that live from paycheck to paycheck or those that can only go from a paycheck to a few days before the next paycheck borrow money from payday lenders for smallish-sounding convenience fees (backed by serious default penalties of course).
But on an annualized basis, these “convenience fees” are anything but:
The fees for payday loans are extremely high: up to $17.50 for every $100 borrowed , up to a maximum of $300. The interest rates for such transactions are staggering: 911% for a one-week loan; 456% for a two-week loan, 212% for a one-month loan.
The problem is compounded by the fact that the borrowers are caught up in an endless cycle of borrowing (and repaying). The payday loan industry of course believes that it is offering “choices” to otherwise choice-less consumers and fights tooth and nail to preserve these “choices“:
We believe consumers deserve choices with simple, understandable loan terms and to be treated fairly throughout the process. Whether on Capitol Hill or at the state houses, CFSA has led the way in working with legislators and regulators to pass strong consumer protection laws.
Anyway.
A new breed of lenders that target these same consumers is now rising up, backed by algorithms – those all powerful mathemagic-backed thingies that are promising to change a lot of consumer activities these days. One such lender is ZestFinance:
Using data-crunching skills polished at Google, (Douglass) Merrill says ZestFinance (Merill started it) analyzes 70,000 variables to create a finely tuned risk profile of every borrower that goes far beyond the bounds of traditional credit scoring. The more accurately a lender can assess a borrower’s risk of default, the more accurately a lender can price a loan. Just going by a person’s income minus expenses, the calculus most often used to determine credit-worthiness, is hardly enough to predict whether a person will pay back a loan, he says.
So these companies use everything from a borrower’s Facebook profile to their owning a cell phone, among other things, to assess risk and make loans.
Admittedly, while these new breed of loans are cheaper, they are still much more expensive compared to traditional securitized or non-securitized credit card debt. What they do have going for them is transparency about interest rates, a much better user experience (translation: no bullet proof glass in their loan centers), borrowers’ ability to get discounts on future loans, potentially improving their credit scores by repaying payday loans (not an option today), etc. (The article in Wired from which I excerpted above is well worth a read.)
That, is certainly a great step forward and qualifies as social-entrepreneurship in my world.
70K variables — I dont know more than 10 variables about myself
Ha ha. I know. That is a lot…I wonder how they gather that much data…perhaps the answer lies somewhere between exaggeration and extrapolation.