Standardize Consumer Financial Data

This week CFPB Director Richard Cordray stated that he was “gravely concerned” by banks' attempts to prevent screen scraping of consumer data by fintechs authorized by consumers collect their financial data.  Cordray said, "We believe consumers should be able to access this information and give their permission for third-party companies to access this information as well."  The control over consumer financial data is hugely important in terms of optimizing competition within consumer financial services, as well as being able to monetize on such data through things like cross-selling, and predictive default analytics.  

It's tempting to frame the issue in terms of who "owns" the data—is it the consumer's, to do with what s/he likes, or the bank's?  I don't think that sort of binary property ownership rubric is very helpful when thinking about jointly produced data--the consumer's inputs, but the bank's formatting and maintenance.  Instead, I think we should look at consumer financial data from another perspective—how can we maximize its value when viewed from a systemic perspective?  This means value maximization from the aggregate perspective of depositories, fintechs, consumers, and regulators.  It does not get into the distributional question about who gets what slice of that pie; I’ll leave that for Kaldor and Hicks (or at least another blog post that may or may not happen). 

There's a very simple move that could do a lot to maximize the value of consumer financial data from a systemic perspective:  standardize it.  

 

For different product categories, there should be universally defined data fields that would be the same for all institutions.  Data should also be readily exportable--that is it should be available to consumers and their authorized agents in a readily transportable format, such that if consumer X’s account moves from institution A to institution B there is no problem for institution B reading the account.  This matters for basically everyone involved:

• In voluntary account transfers, standardized, transportable data will preserve account and transaction history, which is valuable to the consumer, but also transferee institution. Behind the Rawlsian veil, financial institutions cannot tell if they will be the transferee or transferor institution, so they shouldn't worry too much about this hurting them; if they offer good products and services, they'll succeed as they should. 

• Standardized, transportable data facilitates consumers’ use of the services of fintechs like Mint that analyze and organize the data to allow the consumer to optimize financial behavior.  

• Standardized, transportable data facilitates  involuntary transfers (from the consumer’s perspective), such as transfers in mortgage servicing or among debt collectors.  Data and IT system compatibility presents more of a problem than it ought to in mortgage servicing.  If you don't believe me, one of the country's largest financial institutions argued in a recent class action suit that it couldn't produce records on a firm-wide basis from before 2012 because of non-integration of the various legacy systems of the firms that it had acquired. (Conveniently, this also frustrated damages calculations.)  

• Standardized, transportable data facilitates regulation.  Standardized data fields and formats would allow for more automation of supervision and would facilitate regulators’ ability to get aggregate information.  

• Standardized data formats would facilitate the creation of third-party add-ons (again fintech) that could benefit the institution holding the account relationship and the customer. 

So here’s what I’m thinking.  Have for different categories of product data fields A through Z pre-defined (perhaps by regulation, perhaps by industry agreement), and then allow institutions to add on their own customized fields AA through ZZ, etc.  It seems like a situation in which everyone wins.  Curious to hear thoughts.