Hello! On behalf of the Federal Reserve Bank of Philadelphia and our partner organizations, I’d like to add my welcome to our Fifth Annual Fintech Conference. It’s great to be here today among such a great group of speakers and panelists as we discuss the latest developments in financial technologies, their impacts, and the appropriate policy response. I’m particularly delighted to delve into how fintech can promote financial inclusion. I know these conversations are of much interest to you whether you happen to be an academic, industry practitioner, regulator, economist — or, well, a Federal Reserve Bank president.
With that in mind, there is a bit of Federal Reserve policy I need to dispense with before I can continue. That is the standard Fed disclaimer: The views I express today are my own and do not necessarily reflect those of anyone else on the Federal Open Market Committee or in the Federal Reserve System.
While we are discussing cutting-edge technologies this week, in a way, many new fintech tools are in essence digitized versions of practices that existed in the past. Before the advent of the modern credit score, for instance, lenders took the kind of expansive look at borrowers that open banking today makes significantly more technologically efficient.
And versions of “buy now pay later” have been in widespread use across Latin America since long before we all became tethered to our cell phones. Or, thought about another way, buy now pay later is actually an updated version of Kmart’s old layaway system — with the main difference being you get the product at the beginning of the process rather than at the end.
And yes, I realize I’m dating myself by referring to Kmart.
But the point is that one of the reasons to be bullish about the widespread adoption of these technologies is that they have been used successfully in other, more analog, forms in the past. Fintech makes them more efficient and easier to disseminate.
Challenges and Opportunities
There are great opportunities here. New methods of assessing credit through open banking offer the possibility of promoting financial inclusion by, for example, expanding the scope of which data lenders consider and how they are used in predicting risk. Why shouldn’t a history of on-time rent or utility payments serve to make someone creditworthy? While there are a few exceptions, these are not elements included in most credit scores used by most lenders today. Many consumers trying to get their first loan find themselves in the paradoxical situation of needing to have established credit already in order to obtain new credit.
But, of course, the level of access that open banking provides also comes with significant risks. Would a lender penalize would-be borrowers when he sees they spend money at a dollar store rather than at a fancy department store? (Or perhaps vice versa.) All sorts of implicitly biased cultural judgments could come into play here, which would have the perverse effect of perpetuating, rather than mitigating, long-standing patterns of unequal treatment.
There is significant historical precedence for this. In the pre-credit score era I alluded to earlier, lenders had a much wider latitude when considering a person’s creditworthiness. But that meant all sorts of personal biases could — and often did — play a huge role in allocating credit, with particularly unfair outcomes for minority communities.
Fintech in and of itself is value neutral, in other words. It is a tool. And it is a tool that can be used to promote financial inclusion, or a tool that can be used to perpetuate long-standing biases. And that is because, fundamentally, humans are responsible for the way financial technologies operate.
Consider the question of fintech’s role in allocating credit.
Fintech and Loans
Fintech platforms are increasingly playing a role in issuing loans — both new stand-alone platforms and established lenders are using artificial intelligence to make these decisions. But rather than remove human judgment from the process, these technologies have instead shifted where in the process human judgment comes in. It used to be at the end: A lender would collect the information she deemed necessary and then render a decision. Today, it comes at the beginning of the process when algorithms are being designed and trained.
Think of all the human judgments that need to be made when training an algorithm.
First, people must be aware of how the data being used to train the algorithm were collected and cleaned. And what were their original purposes, given that most of these data were collected for purposes other than how they are being used in this particular algorithm? That answer, after all, has implications for the data themselves — especially given inherent selection biases of commission or omission.
Then there is the question of which determinations will be made by the algorithm. Is it offering a simple Yes or No on issuing a loan? Or is it also deciding on the size of the credit line, including for an account that already exists? Or is it determining the rate offered? Will it decide who to market loans to? Or what messages to stress when doing that marketing? Does it decide the strategy for managing an existing account that exhibits increasing risk or a departure from the consumer’s normal pattern of using the account? Does it determine who is eligible for a work-out program and on what terms? Is it used to determine collection strategies for accounts in default? What information is communicated to the consumer about the whys of a decision made by models that many of us view as a black box?
The humans in attendance today, in other words, can rest easy: We are not obsolete yet.
Will fintech platforms promote financial inclusion? They certainly can — and, in my view, absolutely should. But whether they do this is up to all of us — industry, academia, our elected leaders, and regulators.
That’s why this week’s series of conversations is so crucial. So thank you again for joining us.
And now, it is my distinct honor to welcome our next guests, Anthony Eisen and Larry Summers.
Anthony Eisen is cofounder and CEO of Afterpay. He has more than 25 years of experience in investing, public company directorships, and providing corporate advice across a variety of sectors. Prior to cofounding Afterpay, Anthony was the chief investment officer at the Guinness Peat Group, where he was actively involved in a number of financial services, software, and technology companies.
We’re also thrilled to have Larry Summers with us today. Larry is the Charles W. Eliot University Professor and President Emeritus at Harvard University. In addition to serving as 71st Secretary of the Treasury in the Clinton administration, he was director of the White House National Economic Council in the Obama administration, president of Harvard University, and the chief economist of the World Bank.
So, thank you again. And now let’s continue with the conference.
- The views expressed here are the speaker’s own and do not necessarily reflect those of anyone else in the Federal Reserve System.