Washington, DC – CBAC Meeting on 04/25/2017

Welcome to the Consumer Financial Protection
Bureau’s meeting of its Community Bank Advisory Council, or the CBAC. The Consumer Financial
Protection Bureau is an independent federal agency whose mission is to help consumer finance
markets work by making rules more effective and by consistently and fairly enforcing those
rules, as well as empowering consumers to take more control over their economic lives.
My name is Zixta Martinez. I serve as the associate director for the External Affairs
Division at the Consumer Bureau. Today’s meeting is being held at the Consumer Bureau’s headquarters
in Washington, D.C. This is the Advisory Council’s first meeting of the year, and, as always,
we have a packed schedule. Today’s meeting is being livestreamed at ConsumerFinance dot
gov, and a recording will be made available on the same website. You can also follow CFPB
on Facebook and Twitter. Let me spend just a few minutes telling you
about what you can expect at today’s meeting. First, I’ll introduce the Bureau’s CBAC members,
then the Consumer Bureau’s acting deputy director, David Silberman, will provide opening remarks.
Following the acting deputy director’s remarks, CBAC Chair David Reiling will conduct the
meeting. Chair Reiling will introduce Shiri Wolf, Counsel in the Office of Regulations,
and Brian Kreiswirth, Deputy Fair Lending Director in the Office of Fair Lending and
Equal Opportunity. The two will lead a discussion about the Bureau’s recent request for information
about alternative data. Following that discussion, the CBAC will hear
from Will Wade-Gery, the assistant director for the Card Payment and Deposits Markets
Team, and Stephen Shin, Managing Counsel for the Bureau’s Office of Regulations. The two
will lead a discussion about the Bureau’s request for information about consumer access
to financial records. The meeting will then adjourn at approximately 5:00 p.m.
The Bureau established a Community Bank Advisory Council to include representatives of community
banks from across the U.S. The CBAC is charged with providing substantive information, analysis,
operational expertise, knowledge of their communities, and feedback to inform the Bureau’s
work. Today’s public meeting and discussion is in support of this important responsibility.
As a reminder, the views of the CBAC members are their views. They’re welcome and greatly
appreciated, yet they do not represent the views of the CFPB. So, let’s get started with
an introduction of our CBAC members. The chair is David Reiling. David is CEO of
Sunrise Banks in Minneapolis, Minnesota. The vice chair is Angela Beilke. She is Vice President,
Mortgage Department Manager, of American Bank and Trust in Huron, South Dakota.
Jonathan Allen is Chief Compliance Officer of Bank of American Fork in American Fork,
Utah. Melissa Ballard is Vice President and Director of First Iowa State Bank in Albia,
Iowa. Menzo Case is President and CEO of Generations Bank in Seneca Falls, New York. Kathleen Cook
is President and CEO of The Village Bank in St. Libory, Illinois.
Linda Feighery is Vice President of CRA and Fair Lending Officer at Citywide Banks in
Denver, Colorado. Jack Hopkins is President and CEO of CorTrust Bank in Sioux Falls, South
Dakota. Brenda Hughes is Senior Vice President, Director of Mortgage and Retail Lending, at
First Federal Savings Bank of Twin Falls in Twin Falls, Idaho. Dion Kidd Johnson is President,
COO, and CRO at Western Bank in Alamogordo, New Mexico.
Ricardo Leal is Community Lender and Senior Vice President of First Community Bank in
Harlingen, Texas. Cara Quick is Vice President for Compliance at First Hope Bank in Hope,
New Jersey. Cal Ratcliff is Senior Vice President and Chief Compliance Officer at Bank of North
Carolina in High Point, North Carolina. Trent Sorbe is President of Central Payments Division
at Central Bank of Kansas City in Kansas City, Missouri.
Thomas Spitz is the Chief Executive Officer of Settlers Bank in Windsor, Wisconsin. And
Samuel Vallandingham is President and CEO of the First State Bank in Barboursville,
West Virginia. We also have with us Delicia Hand, Staff Director for the Bureau’s Office
of Advisory Board and Councils. I’m now pleased to introduce David Silberman.
Prior to his current role serving as the CFPB’s Acting Deputy Director and Associate Director
for Research Markets and Regulations Division, David Silberman worked for 12 years as General
Counsel and Executive Vice President of Kessler Financial Services. David began his career
as a law clerk to U.S. Supreme Court Associate Justice Thurgood Marshall and then as a member
of the law firm Bredhoff and Kaiser. David, you have the floor.
Thank you, Zixta, and on behalf of Director Cordray, who’s, unfortunately, traveling today,
I want to welcome you to this meeting of the Community Bank Advisory Council.
Uh, we created this council almost 5 years ago because we wanted to ensure that we have
a consistent way to hear directly from community banks about what they are seeing and hearing
in their communities and about how our work is affecting their businesses. We do not supervise
banks with $10 billion or less in assets, which means that we do not conduct examinations
on any community banks. Yet, we know that the regulations we write directly impact the
institutions, as they are the first to let us know.
The Advisory Council helps — fills the gap in our day-to-day contact and helps to ensure
that the lines of communications remain open at all times. We value the dialogue we have
with our CBAC members, whose perspectives and the information they share influence our
thinking. We talk to them about our policymaking agenda, and they consistently provide essential
feedback that helps us understand the experience of their customers and how their institutions
operate. All that improves our work in many — in many ways and, I hope, demonstrates
that here at the Consumer Bureau, we’re eager to listen and deepen our understanding.
Today, I want to highlight some of the work the Bureau is doing with respect to innovation
in the consumer financial marketplace and, specifically, our wok involving the use of
data in that marketplace. We will then engage the Council in a discussion of these issues.
From early on, the Bureau has made it a priority to focus on the topic of financial innovation,
which, in today’s world, typically means technological innovation. I know and appreciate that such
innovation can both be a challenge and an opportunity for our community banks, and so
I’m particularly looking forward to today’s discussion.
One area of interest to us at the Consumer Bureau is the way in which data about consumers’
financial transactions is being accessed and utilized in the financial marketplace. We
see dramatic changes, which hold the potential to improve the way consumers manage money
and direct their financial affairs. Many of these infa — innovations rely on
access to consumer infor — to — to current information drawn from the assets, balances,
and transactions in consumers’ financial accounts. These include savings and checking accounts
and, for those who have them, prepaid card accounts, credit card accounts, and investment
accounts. In each case, the information recorded about
the consumer can be a valuable asset in helping consumers budget or obtain credit. Indeed,
it may ma — it may matter as much as the dollars they actually have in their accounts
at any given time. That is what appears to be driving efforts to access the in — this
information on consumers’ behalf. These changes in the way access — information
is accessed have not been without risk to consumers or to their financial institutions.
So, we want to understand how consumers and third parties are accessing using such data
and how it fuels new innovations. We also are deeply interested in how consumers are
exercising control over their personal financial data, including the data that is maintained
by their financial institutions. In November, we issued a request for information
to inquire about the challenges consumers face in accessing, using, and securely sharing
their financial records and about the impacts on financial institutions when consumers access
this data through third parties. We are working to assess where the barriers exist between
consumers and the personal data that their financial providers maintain about them, and
we want to hear solutions from stakeholders that can help address the risks and technological
challenges posed when consumers seek ready access to this data and seek to share it electronically.
We are keenly aware of the serious issues around privacy and security for consumers
and financial institutions alike. One pressing issue is how to satisfy the demands of consumers
without expand — exposing the institutions that maintain this data to undue costs or
risks. Another pressing issue is how to prevent consumers from subjecting themselves to undue
risks, including the possibility that that data could be misused.
Over the past few months, we’ve received about 70 comments from stakeholders across the spectrum.
We are sifting through these comments, which are extensive and thoughtful. They present
a wide range of ideas about how best to achieve the broad goals we have in mind.
Certain perspectives presented in the comments are not surprising. Banks and other financial
companies raised concerns about consumer data security and offered solutions that may address
those concerns. Aggregators and users of the data, in contrast, are recommending less — less
fettered access and greater freedom to store and use the data that consumers permit them
to access. This would give them more flexibility, they say, to enhance their services and their
business models. Almost everyone is offering arguments that
their approach will better protect the interests of consumers. At stake for us is how consumers
can control, in a safe and secure manner, what data is shared and how the data is used.
So, there’s much to digest, and we see the market moving quickly, with high stakes for
all involved. Even as we speak, vig — vigorous and spirited
negotiations are underway throughout the industry that could shape the future of information
access. We expect the interests of consumers to be at the forefront of these discussions.
We are, therefore, concerned about reports we have heard that some institutions may be
seeking to limit or restrict access unduly. In today’s meeting, we are particularly interested
in how community banks help their customers access — access their data and how they are
impacted by data aggregators accessing data on behalf of their customers. This will be
the subject of our second session this afternoon. I also want to provide an update on our latest
actions to explore the use of new types of data in accessing the creditworthiness of
consumers who are seeking credit. As you no doubt doe — no doubt know, over a period
of decades, large portions of the financial services industry have become dependent on
automated underwriting systems to make their credit decisions. These systems, in turn,
are dependent upon the data that is maintained by the three national credit reporting agencies
and that fuels the credit scoring models with which we are all so familiar.
This creates something of a catch-22. Consumers who lack a credit report and lack a credit
score are largely shut out of the mainstream credit system and, as a result, are deprived
of the opportunity to build their credit history. Some have suggested that the thoughtful and
responsible use of alternative data — that is, data that is not part of the traditional
credit reporting system — could expand the credit available to underserved consumers.
If it is possible to expand opportunity in this manner, it would benefit not only these
consumers, per — but — but perhaps would — would buoy the economy in ways that benefit
all of us. So, in February, we launched an initiative
to learn more about issues raised by alternative data, as well as by new methods of analyzing
and using the data. In particular, we issued a request for information to obtain feedback
from stakeholders about the potential benefits and risks of using unconventional sources
of information and new modes of analysis to assess people’s creditworthiness. We want
to know whether various types of such data can help more consumers build their credit
histories and gain more access to credit. Just how many consumers are we talking about
here? Well, as a self-directed, data-driven agency — as a self-described data-driven
agency, excuse me, naturally, the Consumer Bureau has dug into the data to gain a deeper
understanding. After crunching the numbers, we estimate that 26 million Americans are
credit invisible, meaning they have no credit history at all. Another 19 million people
have credit histories that, under most models, are too limited or have been in — inactive
for too long to generate a reliable credit score.
So, that means that 45 million Americans — roughly 1 in 5 — fall into one of these categories.
For every one of them, managing the ways and means of their lives usually costs more, risks
more, takes longer, and does less to build their financial future than is true for most
consumers. That is simply a tragedy in a modern economy and a modern financial system like
ours, and we all need to think harder about what we can do to address it.
Certain longstanding products, such as secured credit cards, can provide part of the answer,
and we hope and encourage those to be actively offered to these consumers. But we also want
to explore ways to enable consumers who cannot afford to put down a security deposit to obtain
responsible and affordable credit when they need it, and that is where alternative data
holds promise. For example, alternative data may draw from
sources — sources such as rent or unit — or utility or telecommunications payments, which,
in general, have not traditionally been part of the credit reporting system. It may draw
from electronic or other records of transactions, such as deposits, withdrawals, or account
transfers, and it may include other personal information, such as data generated from the
use of mobile phones or internet services. The idea is that by filling out more de — filling
in more details of a consumer’s financial life, this information may paint a fuller
and more accurate picture of their creditworthiness. So, adding alternative data into the mix may
make it possible to open up more affordable credit for millions of additional consumers.
Through the request for information we issued last month, we are looking at the pros and
cons of using the types of alternative data available today and what the future may hold
as technologies continue to evolve. We are looking at how this information is gathered
and analyzed, so that we can better understand how all this is beginning to unfold.
Now, for many of our CBAC members, all this may seem quite odd. After all, community banks
have long prided themselves — and justifiably so — for their willingness and ability to
go beyond the numbers, to understand their consumer — their customer’s life circumstance,
and to make lending decisions based on the totality of the information available to them.
In many ways, what we are explor — exploring is whether there are opportunities to combine
the rigor and objectivity of automated underwriting with community banks’ understanding that good
underwriting requires more than simply looking at a three-digit credit score.
Some of the main inquiries we posed in our RFI are these: First, can the use of alternative
data to create or augment individual credit scores actually increase access to credit
for consumers by helping lenders better assess their creditworthiness? Second, will this
lead to more complex lending decisions for both industry and consumers, and what risks
would that pose? Third, how might the use of alternative data, new modes of analysis,
and new technologies affect costs in making credit decisions? Finally, and quite significantly,
how may the use of alternative data affect certain groups in ways that might run afoul
of the Fair Lending laws or create other risks for vulnerable consumers?
We are hearing from innovators who want to expand access to credit or offer credit at
lower interest rates to borrowers whose credit scores may understate their ability and willingness
to repay, and we see promise in some consumer-friendly innovations that bring new products to the
unbanked and underbanked. These approaches also pose risks, and we want to know more
about these risks and how they can be mitigated or minimized.
On the whole, we are encouraged to benefit — we are encouraged by the potential for
using alternative data in underwriting to benefit the very consumers that the Fair Lending
laws are designed to protect. So, as we think hard about these issues, we welcome our — you,
our members of our Advisory Council, to also provide feedback and be a part of the frank
and wide-ranging discussion we have begun on this sub — subject, so that we can learn
from you and you can further inform our approach. We’re eager to hear your experiences and perspectives
today. Thank you. Thank you, David, and thank you, Zixta. Uh,
welcome to the second meeting of 2017 of the CFPB’s Community Bank Advisory Council. My
name is David Reiling. I serve as chair to the CFPB’s Community Bank Advisory Council.
I’d like to welcome the members of the public, as well as the members of our Community Bank
Advisory Council. This council has served as a resource for both the Bureau and its
participating institutions to engage in dialogue, so that the Bureau can learn more about — about
how community banks operate compared to their other peer financial institutions.
Uh, during this afternoon’s meeting, we will hear from Council members on two issues: consumers’
rights to access financial accounts, as well as account-related data, and the use of alternative
data models for credit decisions. To get our session started, we will focus on the Bureau’s
recent request for information about the potential use of alternative data and modeling techniques
in the credit process. Alternative data and modeling techniques are
changing the way some institutions do business. For some, like community banks, it is how
we have always done business. We meet, uh, our members where they are, and, sometimes,
we need to take into consideration additional information.
I’d like to now invite Albert Chang, Counsel, uh, Office of Fair Lending, and Shiri Wolf,
Counsel in the Bureau’s Office of Regulations, to provide an overview of the recent request
and engage us in confirmation. Albert and Shiri?
Thank you, uh, Chair Reiling, for the opportunity to speak before the Advisory Council. Uh,
uh, Shiri and I are pleased to discuss the Bureau’s work on alternative data, including
the recently published, uh, request for information on the topic.
Uh, I’ll plan to kick it off with, uh, a summary of the motivations for the Bureau’s work in
this area; then, I’ll turn it over to Shiri to then discuss the work of the Bureau’s Alternative
Data Working Group, of which we’re both a part, uh, as well as how the RFI itself is
organized. We’ll then both describe some of the perceived, uh, pros and cons, as described
in the RFI, associated with alternative data use, uh, at least as described in the RFI,
and then we’ll open it up for discussion by the Advisory Council on this, uh, topic.
So, going on to the next slide, in terms of — on motivations — Uh, a — as many of you
probably already know, part of the Bureau’s mission involves ensuring that consumers have
access to financial products and services in a market that’s fair, transparent, and
competitive. That being said, uh, access to traditional forms of credit, for some time
now, has depended, in part, on whether or not someone has a credit score or some form
of traditional credit history. And so, as the deputy director mentioned,
uh, one of the questions that the Bureau had, uh, a couple of years ago, was: Just how many
people potentially lack a credit score or a traditional credit history? And by the same
token, just how many people might benefit from the consideration of alternative data
or alternative information that might help demonstrate their creditworthiness?
The CFPB’s Office of Research, uh, uh, performed an analysis, and what they found in a data
point on credit invisibles that was published in early 2015 was that 26 million Americans
have no traditional credit history on file with the major credit bureaus and therefore
lack a traditional credit score. What they also found was that an additional 19 million
Americans have some form of traditional credit history but that that history is either too
thin or too old to generate a traditional credit score. And so, in totality, 45 million
Americans lack a traditional credit score and, as such, face additional hurdles in accessing,
uh, traditional forms of credit. Moving on to the next slide — Uh, the Office
of Research also looked into the composition of those folks who lack a credit score, and
what they found was that folks without a credit score tend to be disproportionately young
adults, they tend to be disproportionately those from lower-, uh, income areas, and they
also tend to be disproportionately African American or Latino.
Looking here at slide 4, which breaks things down by race and ethnicity, on the left panel,
you can see that, in terms of absolute numbers, uh, the majority of folks without a credit
score are, in fact, white. On the right panel, however, which looks at things in terms of
the fraction of each group without a credit score, you can see that the fraction is, in
fact, higher for African Americans and Latinos than it is for other groups.
These charts also break things down by, uh, exactly why these folks lack a credit score,
and so if you see, uh, in terms of the medium green shading that’s at the bottom of each
bar, which represents no credit history on file, you can see that, regardless of group,
most folks lack a credit score because they have no credit history on file, and that,
of course, reflects the overall trend that we just mentioned, with over half of the 45
million lacking any credit history whatsoever. Now, you can imagine that other groups, um,
might also be disproportionately affected by this issue — so, for example, we mentioned
young adults who are just getting started. Uh, you can also imagine that recent immigrants,
who might have a credit history, uh, from the countries from which they arrived but
not in the U.S., might be disproportionately affected. And you can also imagine that those
who are recently widowed or divorced might also be disproportionately affected by this
issue. Of course, regardless of who you are, not
having a credit score or a traditional credit history means just another obstacle in accessing
traditional, lower-cost forms of credit. And so, to the extent that alternative data at
least holds the promise of expanding access to responsible credit to at least some of
these 45 million by allowing some of these folks to demonstrate their creditworthiness
through a different means, uh — That has been one of the motivations for the Bureau’s,
uh, ongoing work with respect to alternative data.
And to speak more about some of that, uh, work, I’ll hand it over to Shiri Wolf.
So, the Bureau has, of course — Oh, excuse me. The Bureau has, of course, been interested
in the use of various types of data and modeling techniques for quite some time. It’s relevant
to our work on multiple fronts. As you all know, the Bureau has authority
over a number of U.S. laws relevant to the use of consumer data in credit underwriting,
including the Equal Credit Opportunity Act and the Fair Credit Reporting Act. Alternative
data, therefore, comes up regularly in our work in those contexts, among others.
The Bureau recently formalized its interest, in early 2016, by establishing an Alternative
Data Working Group as Albert mentioned. The Group consists of stakeholders from across
the Bureau who have diverse perspectives on the risks and benefits of using alternative
data and modeling techniques in credit decisions. The Group is currently in learning mode, which
is why we’re so grateful, uh, for the opportunity to hear your feedback here today.
We’ve engaged in several listening sessions with external groups, including industry,
consumer advocates, other regulators, academics, and consultants, and as, uh, David earlier
mentioned, we released a request for information on February 16th. We released that RFI in
conjunction with a field hearing that we held in Charleston, West Virginia, on that same
date. You can find the RFI in our — on our website, and, uh, comments to that RFI are
due by May 19th. The RFI generally looks at, um, three basic
questions. The first is, uh: What type of data and model governance is out there, and
how are companies using it currently or planning to use it? This question is relevant to both
our attempt to gather information generally, and also to our attempt to flesh out what’s
really going out — going on in the real world as opposed to what’s being hyped up in the
press. So, for example, it may very well be that
companies have figured out how to underwrite someone based on their pictures on Facebook,
but our question, more specifically, is, is anyone out there actually using data for those
purposes in that way? And, uh, related to that, if they are using it, we are interested
in learning about what type of procedures they’re putting in place to monitor their
risk and ensure compliance. Uh, a second question involves the potential
benefits and potential risks to consumers and the market from using alternative data
and modeling techniques, and Albert and I will go into some of the potential risks and
benefits that the Bureau has identified in the RFI in a couple of slides.
And, finally, we’re interested in finding out how stakeholders view and understand the
application of specific statutes and regulations to the use of alternative data and modeling
techniques in the credit process. Oops. Okay. So, it’s important, in engaging
in this discussion, to establish a common understanding of the various terms that we
use in the RFI. When we talk about alternative data, we are not using a normative concept
but rather a relative one. So, we’re talking about alternative data as it can be contrasted
with data that’s traditionally used in the credit process.
So, with respect to traditional data, uh, we understand traditional data that’s used
in the credit process to include things that are commonly included in the consumer’s core
credit file and, in addition to that, any items that are regularly provided in the process
of applying for credit, such as the consumer’s reported income or assets. Alternative data
is, in essence, anything outside of that core traditional data, including data that isn’t
necessarily held by a consumer reporting company. And we want to, um, make sure that we distinguish
the — the term “alternative data,” in this context, from another commonly used term:
big data. We understand big data to be a reference to the volume or speed of the — the data,
and not necessarily the kind of data that’s being used.
With respect to modeling techniques, we understand traditional techniques used credit decisioning
include linear regression or logistic regression models, whereas alternative modeling techniques
can include other types of techniques, such as machine learning algorithms and such.
The request for information asked questions about the application of alternative data
and underwriting methods to the credit process, uh, and it’s important to understand that
that’s a very broad concept, in that it includes not just the credit decision itself but also
related processes, such as, uh, all underwriting, marketing, prescreening, fraud prevention,
servicing, and collections. That keeps happening. I mean — Yeah, it — it jumped to another
presentation. Sorry. Okay. Yep. Let me have the handout so we can
turn to that. I think I’m turning it over — back to you.
Great. Thanks, Shiri. So — — notwithstanding the, um — the definition
for alternative data that, uh, was, uh, described in the RFI and — and used by the Alternative
Data Working Group, we recognize that the line between traditional and alternative data
can sometimes be blurry, so, uh, slide 9, uh, tries to, uh — tries to, uh, walk through
some examples of, uh, alternative data. And — and these examples are meant to, sort of,
traverse the broad spectrum of what might constitute alternative data, uh, ranging from
data that’s more closely related to traditional data, to those forms of data that are less
closely, uh, tied to traditional data. And as the deputy director had mentioned in
his remarks, we recognize that many of these forms of data have already been used for quite
some time now, in particular by, uh, community banks, uh, in a judgmental way. So, we are
not, certainly, uh, sort of, suggesting that these are brand new types of data that are
being used in underwriting. So, on one side of the spectrum, uh, you have
what’s commonly referred to as trend data, which looks at patterns and trends in traditional
data over time. So, for example, that would be alternative data, for example, that looks
at whether or not credit, uh, card balances are — are rising or falling over time.
Uh, next, you have, um, for example, payment data. That’s not precisely credit related
but at the same time, uh, says something about, potentially, a consumer’s ability or willingness
to repay obligations, like, um, his or her cell phone bill, uh, uh, rent payments, insurance
premiums, or utility bills. You also have, uh, checking account or deposit
account information, which might look at the pattern of withdrawals and — and de — deposits,
uh, both — both from whom and — and to whom, which might also, uh, demonstrate a consumer’s
ability and willin — willingness to repay. Uh, moving further away from, sort of, core
financial product, uh, there’re a variety of stability measures that could be used,
like the number of changes in residence, the number of changes in employment, and the like.
Uh, there’s also other forms of data that might, uh, say something about future ability
to repay, including data about a consumer’s educational or occupational attainment, like
the degrees obtained by the applicant, the job positions held, or the schools attended.
Uh, and then, further afield, moving to the opposite side of the spectrum, you might have
behavioral data on where consumers shop, uh, where they travel, as well as how they might
interact with a website. And then, maybe, moving even further along
that spectrum, you have data fr — about a consumer’s friends and associates, including,
uh, their social media information. Question about that if I may. How does that
relate to creditworthiness? Well, uh, ECOA — ECOA does have some specific,
uh, provisions with respect to, uh, marketing and advertising, but, uh, certainly, you can
imagine that if you’re, um, not marketing or advertising to a particular group at all,
that that would, sort of, affect their ability to, sort of, be approved or have access to
credit, which would, uh, likely have some, uh, Fair Lending imp — implications on the
back end. Uh, and maybe — I don’t think that answered,
uh — answers the question. What — I understand how all of the other factors on the previous
slide relate to or could relate to — to, uh, ability to repay, to be used as a substitute
for a credit history or credit score, except for the final one. Ca you help me unders — What
— What — If I — If I don’t have credit history, but I — but I have friends on social
media, what does that have to do with my ability to repay?
Sure. And — and — and to be clear, you know, th — this is — it — it’s not entirely clear
what, necessarily, is the relationship, uh, with — with creditworthiness, but there are
some, uh, sort of, in — folks who have been reported in the media who suggest that, uh,
the creditworthiness of your associates might say something about, uh, your creditworthiness,
and — and that is, of course, uh, something that, uh, would need to be, sort of, stress-tested
or — or evaluated. I think some online lenders would suggest
that it’s useful, that — a fraud detection mechanism, that if you have somebody who doesn’t
exist and has no soci — no friends, no ath — uh, that maybe they’re — they’re phantom
people or that, maybe, actually — you know — and I’m not saying I agree with any of
this — or, maybe, that, uh, a consumer’s, at least, willingness to — If you know that
— uh, I know that you’re friends with these — these seven people, all of whom I know,
uh, that gives me more assurance that you’re likely to repay me and not — not stiff me.
Not saying — Well, uh —
— that’s true or not — I have friends —
— but that’s the hypothesis — I have friends —
That’s the hypothesis. — with good credit, and I have friends who
have bad credit — Right.
— but they’re still my friends. Jonathan —
Well, that’s the determination. That — that’s really the — The — the question that’s at
hand, I think, with all the data points in some regard is, what is the reliability of
these, I — I’ve got a feeling, because, at the end, uh, there might be some natural selection
there. But, um — but — but Jonathan, for —
— that — but that is the question. — for what it’s worth, you’ve hit upon a
source of considerable feedback to the Bureau when we published —
Absolutely. — this RFI.
Right. So, moving on to, uh, slide 10 — Slide 10,
um, uh, summarizes some of the potential pros and cons associated with alternative data
use, uh, at least from the perspective of the consumer, uh, and — and these pros and
cons are, uh, listed in our RFI. Uh, Shiri and I will — will discuss these in, sort
of, greater detail in subsequent slides. Uh, and of course, the — the goal, uh, of pointing
out these potential benefits and risks is to help facilitate a, uh, frank and full discussion
about, you know, how — how folks might mitigate the risks associated with alternative data,
while at the same time, uh, uh, realizing, uh, benefits.
Moving on to slide 11, which speaks to, uh, potential benefits — Uh, the key potential
benefit, uh, that’s described in the RFI is the one that I had alluded to earlier, which
is the prospect that alternative data, uh, could expand access to responsible credit
by giving those without a credit score or a traditional credit history a different way
to demonstrate creditworthiness. Another benefit could be in making credit
risk assessments, uh, more accurate or — or more predictive. So, for example, a lender
with a 680 credit score cutoff might, with alternative data, be able to distinguish,
uh, among those with a credit score lower than 680 as to who is more likely or less
likely to, uh, go delinquent or default. And this, in turn, could, uh, potentially, allow
that lender to relax, perhaps, their minimum credit score eligibility requirements and,
as such, uh, expand access. Um, another potential, uh, benefit could be
more timely information. So, to the extent that alternative data reduces the inherent
lag time between when consumer behavior occurs and when that behavior can be considered in
a creditworthiness assessment, uh — this might allow lenders to distinguish between
past financial problems and emerging ones. And then, there is the potential that alternative
data could lower costs if it allows for the automation of processes that replace or are
a substitute for, uh, more costly manual processes. And so, potentially, these, uh — these cost
savings could be passed along in the form of lower prices to consumers, or it could
enable lenders to make smaller loans more economically.
And I’ll turn it over to Shiri to speak about, uh, some of the risks.
So, first, it’s important to note that many of the risks that we’re going to be discussing
and that we note in the RFI also exist with respect to the use of traditional data or
modeling techniques. We’ve simply tried to highlight risks that may have unique implications
in the context of the use of alternative data and modeling techniques.
Uh, the first risk we point to here is privacy concerns, whether that’s because the data
is of a particularly sensitive nature or because consumers simply didn’t know that it was collected
or would be used for the purpose for which it — it was used.
Um, we also may have concerns about — Uh, there also may be a risk related to the quality
of the data, uh, especially to the extent that it’s kept by entities that may not comply
with consumer reporting obligations. There may also be concerns related to consumers’
ability to control the data, um, whether it’s because they don’t have access to the underlying
data or can’t correct any inaccuracies that they spot in the data, or because the data
is not actually about them — going back to an earlier conversation — but rather about
their peers or about groups that they may belong to.
There are, uh, also risks inherent in the complexity of the data that’s being used or
the modeling techniques that are being used. Obviously, the more complex and the more numerous
the factors that are used to evaluate someone, the harder it may be to explain to that individual
why a certain credit decision was made. Hmm. Okay. There may also be risks related
to barriers to mobility to the extent that the use of alternative data could potentially
penalize or reward certain groups for behaviors in ways that reinforce existing social barriers.
Uh, finally, there is the potential for, uh, discrimination. Using alternative data may,
in some ways, present a greater risk of unlawful discrimination on the basis of — of a protected
category, such as race or ethnicity, especially where we do not yet understand the potential
correlation between a piece of data and a protected class.
And, finally, there may be violations of other laws, such as the Fair Credit Reporting Act
or a violation of the prohibition against unfair and deceptive acts and practices. The
RFI generally seeks feedback on, uh, the potential for these risks, the extent of these risks,
and today we wanted to take the opportunity to seek your feedback on that same issue.
And so, with that, I think we’re going to turn it back, um, to David to tee off some
questions. So, with that presentation — Um, and thank
you, Albert. Thank you, Shiri. Um, as we talk about alternative data and community banks,
maybe it’s — I would throw it out to, um, the Council relative to, uh, think about what
types of alternative data do you use currently or have you used in the past, and maybe start
from that, and then — and work our way forward from there. It’s kind of a big topic to get
our — our arms around. So, I don’t know if there’s any — Angela?
You know — I can start out. Like, I know within, uh, the decisioning world, especially,
like, in mortgage lending or anything like that, a lot of times, that you are going to
have alternative data options when you go into, like, a government, um, Rural Development
Loan, FHA Loan, when you do a manual underwrite, you might need to grab — So, there’s no credit
report, and there’s no score, but they can still guarantee that, and you just need to
grab different sources. Uh, maybe you’re going to grab, uh — you
need a 1-year, with a 1 times 30 max late on it, from a phone bill or utility bill.
I know, uh, prior, I’ve done one from, like, a local oil company that somebody had gotten
their pill oils from for the last, you know, quarterly paids. Then, we kind of adjusted
that way. So, it’s just — I think there’s a lot of options out there that, uh, currently,
we are using as alternative data versus the — on a debt side, anyways, versus just the
traditional credit report style. And I would have to say, just kind of going
back to some very simplistic models, if we go back in — well, in a — in a very traditional
community bank, um, where the days of when you walked in with your parents, or you walked
in with a person from the community, and they quote/unquote “vouched for you.” There was
a character reference, if you will. Um, in that type of concept, from an alternative
data standpoint, um, goes back to ancient China, quite frankly, when you had savings
circles and people vouched for each other and pooled their money together to loan to
an individual for a certain event. And so, there are still, I think, potential profiles
or social, um, aspects that may come into play. Um, obviously, those are softer factors
than your traditional credit box, but I think those are types of practices that you saw
with banking, um, in the past. Um, as we move forward, obviously, there’s
— uh, Angela had mentioned, whether it’s utility bills or rent or cell phone or cable,
utility — all those regular type of payments, um, are items that you fill when there are
gaps. Um, I don’t know if there’s any other examples
of what people may use? Oh, Jon? We look at their deposit account, if they
— if they have one with us — Sure.
— how they’ve handled it. Many of them will have had savings accounts from when they were
kids, and we give them credit for that. So, the length of that relationship and then
— and how they’ve handled that relationship? I think you see the same thing, sometimes,
with small businesses and their merchant accounts, and their transaction volumes or deposit volumes
as a proxy for, um — as — as — uh, for income.
How about things like where people work or what their occupation is or — Is that stuff
that fig — that factors into your underwriting, uh — underwriting decisions? I would say, yes, it does. Uh, we talked earlier
in the group that, uh, you know, one time before a car dealership had sold out, we had
a fairly, uh, large population of immigrants moving in. And they’d be brought into us by
this car dealer, and — basically about four — um, and very few of them could speak English,
but, uh, they managed to get their transactions done, and that’s how we did business. Basically,
he was vouching for them. Uh, you know, on another source, uh, we do
marketing for credit cards, uh, on a large basis. Uh, I would call it a near prime, but
we reached to the — what I’d call the thin files. They have to have a file, but it’s
probably a thin file, and there’s probably 200 data points within a thin file, and we
are constantly researching which ones, uh, are likely to pay us back out of all those
data points. I mean, it’s a — it’s a lot of testing, it’s a lot of work, uh, but that’s
something that we work on, and I’d call it a near — near-prime-type product.
And I can speak, uh, kind of specifically to the employer channel. Um, we do — we have
a — an employer loan program, and I would say that the length of employment, um, seems
to be a factor of — of stability. I would say that there’s — when you dig into
that employer, what — what type of employment it is — is it part time, is it full time
— the — and that really — It speaks to the volatility of income, um, whether it’s
consistent or inconsistent, um, has some — Uh, I’ve seen some repayment characteristics to
that. Um, and so, those are things, I think, in that employment, uh, space, that we — we
do find or are seeing that are very credible. I think, also, on the employment — just to
add on to that, too — Sometimes you’re going to see that you want to, uh, uh, I’ve found,
especially in the agriculture world, when you have somebody who is, uh, farmhand style
or anything like that, they get paid in a different way. It’s a seasonal piece, and
their employment is different, so you are looking alternative ways than the traditional,
uh, employment verifications or anything like that for those.
And, generally, you know, the secondary market stuff isn’t going to accept it if it’s a handwritten
check that the farmer gave to the farmhand and then they get a free rent home as a payment
source. So, you have to have those alternative issues and help for those items, as well.
I would also say that there’s some other, uh, — I would, um — they’re not, maybe,
naturally occurring, but they’re regular, uh, payments. So, there’s an example of, if
you’re a — a cash farmer, there’s a cycle to your cash flow, right? It’s annual. So,
your mortgage payment may need to be annual as opposed to monthly or quarterly to — to
meet that. Um, it’s another thing that’s similar to,
let’s say, a tax refund, where you — you consistently see someone get a tax refund,
and they can use that either to repay debt or as a savings. You — you’re looking at
a specific event, but you’re — you’re familiar with that, but it doesn’t show anywhere, per
se. Um, but there is, again, some of that trust as to, you didn’t change your deductions;
you made the same. Okay, then there’s some reliability around that.
I can say, in, uh, our area, which is very small rural — and it’s agricultural, too
— we really don’t have the instance of people with no credit. Uh, maybe that’s more of an
inner city — I — I don’t know. Uh, the people that would come into our bank that would have
no credit score would be young people, and which we help them establish a credit score,
starting with minor loans or car loans, so forth, and so on.
A bigger issue, for us, with people, um, that we do all the time is, do figure out why their
credit is bad and get to the bottom of that and help them straighten that out. So, it’s
not — for us, it’s not an issue, so much, of alternative, uh, documents; it’s more figuring
out and helping them get their credit straightened out.
Uh, so, uh, we just don’t have a lot of people that come in without a credit score. We have
way more that come in that need help getting it cleaned up, and that’s just in my small
area, for instance. But it tie — And it ties when a consumer
will come in, if you look at the credit score and say, “No way, based on that credit score,
would I make a loan,” but you know something else about the consumer, or you do a process
which leads you think, “Yeah, this consumer is somebody I’m willing to make a loan to,”
even though the num — the three — the number doesn’t say — number — the number doesn’t
say that. Definitely, because the number could consist
of a cell phone, uh, collection issue. Now, the cell phone people do not, uh, report monthly
that this is a good customer paying their cell phone bill; however, they can turn it
over to a collection agency, and then it does end up on their credit report.
The same with medical bills. If there’s — uh, if we have a customer come in, and they have
issues with — and it’s because of medical, we throw that totally out. I don’t think it’s
right. Uh, these people, first of all, they may have issues with their insurance company,
with their doctor, with the hospital — has nothing to do, whether or not they’re paying
their other bi — uh, mortgage payments or car payments.
So, we throw that out, and, uh, uh, truthfully, I wish the industry would throw those out.
For one thing, they have enough problems already, you know, uh, with the medical issues. So,
we tear that apart. So, we don’t — So, for you — not so much bringing new information
into the system as — Correct.
— discounting stuff that — Just —
— you think is not — — summating what’s out there.
— it not probative but that — that the is — the — the traditional system thinks is
giving — is overweighing. Yes. Yes. At least, that’s my experience.
I’d, uh, love to hear from, uh, the Council members on — on how you feel that alternative
explanation is holding up in — in present day. We used to be able to, in our — used
to be able to mitigate a — a — or tell a story of a divorce or a — a life event of
some kind, or a medical issue, and write a little note in the file, and you’re good.
Now if — if the metrics in the box or the credit box isn’t — doesn’t quite fit, that
story doesn’t go as far as it used to. Um —
Uh, y — I don’t know if you want to talk about that.
I think that is an issue. I do think that’s an — Oh. From a Fair Lending
perspective — You know, as a Fair Lending officer, I’m fully aware of those type of
situations that come up. I was a former lender, also, and — and an underwriter, so I understand
that. But the way that the banks are being drilled
a — from a regulatory perspective, it — it’s scary. I mean, whenever you make an exception,
you almost have to write a book to explain why you’re making that exception. And you
better be consistent. That’s the bottom line is, you’ve got to be consistent, um, you know,
from borrower to borrower, because, you know, simi — similar situated type scenarios, they
better match as far as pricing and fees and — and all of that. It — it — it’s huge,
and it — it’s scary. We’ve actively been doing loans with alternative
credit for years and still continue to do them. Um, I have the same concern about fair
lending, I think, as we s — broaden this box, that the regulators will get ahold of
it, and — and we’ll have to continue to defend it. Um, to date, that’s not been an issue
for us. My concern is if we prescribe what you can
look at from an alternative credit source, that that will narrow our opportunity to help
consumers, because not every consumer is going to have the same opportunities for identifying
what their alternative credit might be. And so, I think we need to be very careful that
we don’t prescribe what it is we have to look at. We might be able to provide suggestions
of things we can look at but not provide things that we have to look at in every case.
One of the things this committee has talked about at length today is — is the degree
to which community banks rely on service providers. And, conceivably, as you think about alternative
data sources, many community banks are going to have to rely on companies that are producing
models, um, that they could use in some cases. Not everything can be a completely manual,
subjective decision, as much as — as much as we try.
So, I think that, um, inherently, uh, will raise concerns over cost, um, to have access
to that data in Reg B — what is it, empirically derived, statistically sound, I think, is
the correct terminology. Um, so, there’s — there’s — now there’s going to be a — a cost aspect.
I also worry, is — if the use of alternative data can help but not hurt, it makes a lot
more sense to me, then, if it becomes an under — then it becomes an — a — a — a straight-up
underwriting consideration, where, actually, alternative data may actually hurt you. So,
I think that’s a — that’s an important question to wrestle with, of, does it — is — can
it be used to put you over the bar, um, but at the same time, it shouldn’t actually hurt
you, nonetheless, because now all of a sudden, it raises all kinds of concerns about disputing
the accuracy of the source of the alternative data, and then it gets — it gets — seems
to get, uh, very ugly, quickly, there. I would — Oh, sorry, go ahead, Jack.
Just to dove-tail a little bit onto Trent’s comment: Uh, we just had a model that we wanted
a validation of, and I’m going to tell you, that was $135,000 for the validation. So,
when you develop these models to transpoint , they’re unbelievably expensive when you
start developing them and make them reg — you know, pass the regulatory muster.
Um, I would say, if — In — In one of our — our —
Oh, go ahead, Menzo. — areas, uh, we look at whether the individuals
that have applied have worked, um, successfully with another organization, such as Habitat
for Humanity. I mentioned the Workers — uh, Workmans Investment — Workers Investment
Board, or the, uh, youth program that we have through our county. So, if they’re showing
success in those programs, which are designed quite differently from what we’re able to
do, we’re more apt to sit down with those individuals and put them into a program that
is viable for them. One of my concerns, if I could — The data
that you — you quoted, I wonder — For those that — with, uh, no credit or inadequate
credit available, are these groups aggregated in regions that are underserved by true tradi
— true traditional community banks? Or has that a — analysis not been performed?
I’m not sure if that analysis has been, uh, performed. Uh, uh, uh, as I mentioned earlier,
I think, uh, what the Office of Research did find was that, uh, folks from lower-income
areas were disproportionately affected by this issue. And I’m not sure if there’s a
correlation between, uh, lower-income areas and — and, sort of, um, uh, whether or not
those areas are underserved by — by branches, for example, or — or banking.
Another key element in all of this is that for most of us with — with, sort of, non-traditional
approaches, it — it, sort of, starts with, what do we have to work with? Do you have
a job, and can we work with that record? Do you have a cell phone and can we check the
payment history? What’s incumbent in so many of the examination
processes we go through in Fair Lending is to do the same thing repetitively. So, if
11 different people come up with a — a — an alternative method that they did, it really,
sort of, drives that we each consider each of those, like, do you have a job? Do you
have a cell phone? The laundry list of what you have to go through
is — is part of where the debilitating decision process goes through. It used to be sort of
unique to the consumer, and now, every time somebody comes up with a new idea, it — it
really drives us together the data for — for all consumers in that set, and that — that
becomes cumbersome. I just have a very quick concern on the alternative
data usage and, uh, creating a module for that or a platform that you would think that
that would be, because I think that almost takes away from the ability to have it so
unique to the situation if you’re trying to drive it into one of those items. So — so,
that’s kind of a concern is, well, is it — that something that’s being more and more thought
through? You know, there’s the piece where you can
do that, and you’re going to enter that information, but at the end of the day, to keep the uniqueness
to what it is, the alternative data is going to be unique to regions, to areas, to persons,
to everything, to the bank. So, I don’t know that you can just put that into an AUS system
and then just to have an automatic system that does that.
Does it need to be either/or, or can it be just dependent upon the situation for that
particular bank and that particular area? I would think it would depend on the bank,
but you — I don’t know that you could have one or the other. You — I mean, you need
to keep the openness for the alternative data to be used within the community bank settings.
Is there information today that you would not use, even — even though you think it’s
relevant in making a credit assessment because you’re concerned that — too much Fair Lending
risk or, uh, too much privacy risk — whatever — whatever it is, that — where the law’s
constraining you from using information that you think would be releva — would — is actually
relevant and you’d actually like to use, or —
Yeah. — you just —
There’s a, uh, particular sect that we know, uh, will pay their debts because of their
cultural norms, and I can think of a few that I would never document it, and I would certainly
use it as a primary means of making the lending decision, which we have never done, but it’s
a — it’s a — I can see that being used, and quite effectively.
I think we’d probably never use, uh, place of employment, uh, type of job, degree, uh,
university graduated from or type of education. Maybe, to, um, uh, Jonathan’s point earlier,
the, uh, social media piece: I think those are areas that would concern me from a Fair
Lending perspective. Would you not use them because you don’t think
they provide valid information or because you would get in trouble with your safety
and soundness — Yes.
— regulator, or are there other issues? Fair Lending.
Fair Lending, yes. Well, one example is, if — if I, as a loan
officer, have a relationship with the applicant — a personal relationship, a friendship,
because we live in the same community, we see each other at church, for example — I
can’t use religion as a — as a factor. Um, maybe we’re in the same, kind of, social circles,
but — but that has poten — Like, if I treat people in my social circles differently than
people in other social circles, that could be dangerous.
Absolutely. But — but it could be relevant to — in — in
establishing, you know, a history, a relationship, with the applicant. It could be relevant in
their — assessing their likelihood of payment. But try — try documenting that in a file
and see what happens. Well — And — and — to Jonathon’s point,
I mean, we, traditionally, are relationship lenders. Uh, live in small communities, we
know their family, we know what house they lived in when they grew up, and these are
things that help us make credit decisions. But Fair Lending has pushed that out the door,
because every time we make an exception, we’re putting our organization at risk for, you
know, having to spend the money to answer it or be fined.
Uh, and the reality is that, you know, those relationships are important in evaluating
credit decisions. One of the best things we do is get to invest in people and — and help
them achieve their goals in life, whether it’s their first new car or their first new
home or whatever it is. And frankly, regulation’s taken that out of the — out of the, uh, equation.
I mean, because every time we make an exception, we’re at risk for a Fair Lending violation.
And yet, you know, in the case of Jonathan’s, uh, point, those people he knows from his
circles, he knows enough about them, he knows their character. Character is a huge factor
when you consider the C’s of lending. I mean, it’s just — It is.
Uh, and, you know, in my community, we have very conservative — We look a little bit
at what they’ve been able to amass without credit. I mean, because, frankly, we have
a lot of people who don’t borrow money; they just earn it and buy things with cash, period.
And so, one of the things we look at is, okay, what do they have now, without that credit,
uh, because, frankly, it tells us they worked for it.
In, uh — In prepping for this, I — I did some research with some microlending organizations
that operate around the world — Central America, primarily — and — and, uh, I mean, credit
scores don’t exist, and how do they make their credit decisions, and how do they operate
predictively and — and successfully for decades on end? And the key decision element that
de — describes the successfulness of a — of a — a borrower repaying is trust. It’s the
trust that exists between the — between the borrower and the lender, and it — it — and
it — it comes in two formats. One of them is, of course, that — that if
the borrower repla — repays the loan, the trust is instilled, and there’s the likelihood
of having the m — the money lent again. But most importantly, it’s the desire not to let
the lender down. It’s that — it’s that personal bond that exists. And — and where this is
relevant is on two fronts. First of all, this is what happens around
the world, conti — consistently, where things like FICO don’t exist. But — but also, it’s
r — it’s inherent in the root of everything that’s being said around the table. The trust
that we develop with the people that we get to know has an element of subjectivity to
it, and yet, it’s still rooted in trust. So, all the data in the world, um, based on
this microlending environment, suggests that whether you know that somebody’s only had
a 1 time 30 on their cell phone is — isn’t necessarily more important than the bond that
comes out of the discussion between the 2 people.
I — I — I —
Oh, go ahead, please. — I’d like to say something. I would — I’m
a prime example of, um, a single mom with a child, that never had a credit score because
I didn’t have a job because I got married straight out of high school, got divorced,
had a child. Um, I would have never had an opportunity to hold a job, because I wouldn’t
have been able to buy a car had the banker that loaned me the money to buy my first car
not loaned it to me because he knew my dad. And he knew if I didn’t pay, my dad would
kill me, so he made me the loan. We — we — we can’t do that anymore. We can’t
help people get started like that anymore because we know their family — just like
he said, we know, uh, the family, we know the culture, we know the character. We can’t
do that anymore. So, our hands are really tied to help people
get started, so the use of alternative data and means by which all we’re going to use
it for is to try and justify why we made them the loan — um, their rent — The landlord
can write us a letter. They can buy a house on a contract. They can bring in their receipts.
But we — we — we struggle, as community bankers, and try to help the people like we
were helped, um, to give them the opportunities that we had to get a loan that we just can’t
make them and justify without getting our bank in trouble anymore.
So, I think I’m really happy that they’re looking at this, trying to find a way to justify
us being able to use it, but I just hope they don’t tie our hands and list what only we
can use, and I hope that that doesn’t convert to hurt people rather than the ones we’re
able to help now and not be able to help them again because our hands are getting tied.
But — So. That’s all. Thank you. Well, um, Zixta, with that, I think
they answered your question to start with, that, uh, they will hold off from taking things
like credit bureau in light of Fair Lending, so there’s a risk/tradeoff between data, and
I think it was clear by the room that our communities are our third regulator, or fourth,
depending on how you do the math every day, um, in the community bank space.
So, um, Albert and Shiri, thank you very much for that informative presentation and an energetic
discussion. I think there’s more to come on this topic.
Um, during our next discussion, uh, we will hear from the Bureau about their inquiry into
challenges consumers face with accessing, using, and securely sharing their financial
records. Uh, today’s discussion will focus on what risks are for community banks, and
also what kinds of services or products do community banks use that potentially aggregate
consumer information. Uh, we will first hear from Bureau Staff Will
Wade-Grey — Gery, uh, Assistant Director, Card Payment and Deposit Markets, and from
Stephen Shin, uh, Managing Counsel, Office of Regulation, who serve as Bureau experts
on these consumer-access-to-information issues. Gentlemen?
Thanks very much. Thank you.
Um, uh, I’m Will . Um, we’re here to briefly describe the Bureau’s work on Section 1033,
uh, of the Dodd-Frank Act. Um, why briefly? Well, first, because, uh, David Silberman
already gave you some of the high-level, uh, summary there, and second, because as David
Reiling said, um, we’re really here to hear from you. Um, let me just quickly go forward
here. So, um, I know this is an issue, from reading the comments, that — that community
banks are — are far from in — indifferent on, so we would very much like to get to the
part where we hear from you. So, let me run through this briefly. I’ll
take, uh, a quick overview of background, and then Steve will cover the RFI that we
did, uh, at the end of last year, uh, and — and summarize, um, some of the information
that we learned, um, back from that, and then we’ll get into the discussion.
So, just starting with the first bullet here very briefly: The Dodd-Frank Act provides
for consumer rates to access financial account and account-related data in usable electronic
form. If we skip to the next slide, you can see
that in more detail. This is Section 1033. Um, uh, it — but in brief, it says: A covered
person shall make available to a consumer, upon request, information in the control or
possession of such person concerning the consumer financial product or service that the consumer
obtained from such covered person. Um, there are a number of exceptions written
into the statute. Uh, here, we have a — we have a few of them listed: Um, the information
must be in an electronic form that’s usable by the consumer. Uh, the information does
not need to be, uh — There is no additional obligation created by Section 1033 to maintain
data that otherwise would not be maintained. It is not a retention obligation. Um, uh,
and it applies only to data that the account holder or the, uh — sorry, the financial
account data holder, the provider of the account, can retrieve in the ordinary course of business.
There are other exceptions, too, that I think are fairly important. It doesn’t provi — apply
to proprietary information like credit-scoring algorithms. It doesn’t apply to fraud control
information or the like. If we just flip back a second, this gives
a little bit of a history of our work in the space. Uh, we launched our inquiry, uh, into,
sort of, sharing practices, issues that are implicated by Section 1033, at the end of
last year. Uh, we then launched the RFI. Um, I think, sort of, the core of the RFI is trying
to understand, essentially, what sharing is occurring, uh, what sharing may not be occurring
because of certain obstacles, what are the costs and benefits, risks, and so forth, to
different participants in that sharing ecosystem, consumers very much included.
Um, so then, if we move on a couple of slides, let’s just talk briefly about current market
practices. Obviously, what you have here is, you have a range of entities, from traditional
depositories of various kinds to newer fintech providers, using account data, in various
forms, obtained from, uh, uh, financial institutions to provide, um, sometimes new, sometimes existing,
financial services. You know, it’s financial services that are already being provided by
RDIs, but maybe these would compete with them. Sometimes, they’re newer services that aren’t
yet being provided. Um, many account data users do that through
aggregators, because account data users may go get account data from a — a wide range
of sources um, and, because of the transaction costs involved, they may choose to do that
through specialist aggregators rather than doing it themselves.
Um, we see in this ecosystem a wide range of different sharing practices. Obviously,
we all know about screen scraping. There are other methods of sharing data that exist here,
various forms of structured data feeds, APIs, and so forth.
We also see different authorization channels. So, obviously, we’re used to the notion of,
in the screen scraping world, the — the credential — you — you — consumers have shared their
credential, and that is, effectively, serving as the form of authorization, but there are
other forms of that that exist. You see here, for example, before the use of a structured
data feed or an API. Um, so, different channels both for authorization and for actual data
sharing. Um, and then, a recent trend that we have
seen, uh, is, sort of, there’s more players in this space, particularly larger banks and
larger aggregators, in negotiating bilateral deals, um, for sharing this. Uh, uh, at some
level, there have been a range of bilateral deals negotiated by aggregators with banks
for a long time. The scope of the deals appear to be broadening.
So, just quickly running through some of the consumer benefits — Um, uh, I think one that
we’re all probably relatively familiar with is, sort of, the — the BFM personal financial
management tools that exist in this space. This was, perhaps, one of the first uses that
were made of it — uh, made of data, and, often, it was actually done by traditional
DIs, not by newer players. Um, uh, other — other potential benefits:
sort of, automatic or motivational savings, um, apps of various kinds, which can provide
savings advice; they can actually, um, uh, provide, sort of, messaging at select moments.
They can provide algorithms that help determine the amount that you can save and — and give
you advice on that basis. Budgeting tools can be — can be very similar,
maybe — perhaps, with a broader focus, uh, on — on budgeting as a whole, not simply
for, um, savings. Um, uh, product recommendations are a fairly
big category. That comes in a variety of shapes and forms. There are some, uh, apps that use
data to, uh, advise you what to use in terms of products that are in your existing wallet.
Um, I’m — I’m familiar, in my card — with my card hat on, of — of one that tells you,
uh, which card in your wallet to use for maximizing your rewards earned. But there are other product
recommendation tools that tell you that, you know — how they analyze your transactions
then tell you that you’re using a debit card that has a high ATM fee, and you — maybe,
uh, there would be a better product available to you. There’s a — there’s a wide range
of things in that bucket. Account verification is, obviously, a very
important category. Um, uh, some of the, uh — some of the aggregation tools you may be
able to, sort of, move beyond what we’re doing things with things like micro payments and
so forth for account verification — account verification, obviously, being critical to
obtaining numerous, uh, consumer financial — add-on consumer financial products. Um,
loan application information verification is, sort of, a subset of that.
Credit decisioning is obviously a very large category and relates very much to what we’ve
been hearing in the first presentation. Um, cash flow management — again, I think a — a,
sort of, version of some of the budgetary-, um, uh, type tools we’ve already discussed.
Funds transfer and bill payment is a very big category. Most of the ones that we’ve
talked about, to this point, um, are more about data analysis per se. Uh, they may have
a messaging component, but they don’t actually then transact on your basis, but some of these
add a transactional component, as well. That might be for a variety of purposes. It
might be to make payments for your bills in a timely manner. It might be to make sure
that you, uh, transfer a certain amount to a retirement fund on a timely basis. Um, there
are different versions of that, too. It may tell you when you have available funds and
so forth. Um, fraud and identity theft protection is,
um, a — a — a — a fairly large category. We’re always familiar with that on the — on
the — on the supplier side, but — but some apps actually provide that information directly
to consumers by identifying, um — by running through their transactions and trying to spot
fraud patterns. Um, and then, obviously, there’s the investment version of the things that
we’ve been talking about from a consumer point of view.
Current market issues and risks, if we flip over — Um, uh, these are fairly broad buckets
that cover a wide range of issues. Um, you could disaggregate these further and generate
a longer list. This is a fairly high-level, um, aggregation.
Uh, data access issues — Obviously, there’s a range of access practices that we have seen.
In some cases, we see prohibited access. In other cases, we see access that is allowed
to the consumer but restricted in various ways. Maybe that’s de facto; maybe that’s
de jure. Um, we see certain conditions on access and so forth.
Data security — Obviously, this is a major concern, um, and it is one that we heard a
— a lot about from — from — from, um, commentators, uh, on all sides of these issues: Um, how
do aggregators and account data users get this data? Do they do so in a secure method?
Do they keep the data in secure ways? What regulatory — regulatory regimes apply to
them or should apply to them? Um, should they be treated differently from account data holders
now? If so, why, uh, uh, or why not? Um, related to that, sort of, these security
and privacy issues around retention and reuse: Do account data users obtain this data and
use it for — or hold it for longer than they need or use it for purposes other than those
that the consumer may understand that the data is going to be used for?
Um, another critical, uh, issue, obviously, concerns liability for unauthorized transactions,
uh, for example, from the result of a — a — of a data breach. This is a critical issue.
Um, one thing, obviously, that we hear a lot about in comments is, sort of, which party
in the chain may be responsible for this. We’re, obviously, very focused on, uh, the
issue of, is — you know, ways to enable sharing that don’t create liability for the consumer,
um, but, obviously, we’re also concerned with the allocation of risk as between parties
in the sharing system that aren’t the consumer. Um, uh, FCRA issues are, obviously, very significant
in this space, uh, similar with privacy: How does GLBA apply here to the aggregator or
to the user? How does the FCRA apply? Who is the CRA in this space? Are — are aggregators
CRAs? Are account data users CRAs? Under what circumstances are account data holders — Are
they furnishers of data and under what circumstances, and what are the implications of that?
Um, again, obviously, there are a large quantity of technical burdens. Um, uh, there are technical
burdens associated with being screen scraped. Um, there are technical burdens associated
with rolling out a system to put in place of screen scraping. Um, there are technical
burdens associated with different forms of authorization channel. Um, there may be technical
burdens associated with delivering to the consumer various services that may enable
them to control the flow of data, consumer dashboards that would let consumers, for example,
turn off data sharing at certain points. Uh, a wide range, uh, of technical burdens, um,
again, that we heard a lot about from the RFI.
And on that note, Steve? Thanks, Will. Um, so, I mean, it’s in that
context of rapid development in the marketplace for, um, products and services that have potential
— a lot of potential uses and benefits for consumers but also, along with that, uh, present
new or, um, uh, potential risks, as well. Uh, so we asked, uh, a — um, in the RFI,
a request of information, and solicited input on a — on — on a range of issues that we’ve
been, uh, kind of, discussing, um, and — and touched on a number of topics, uh, w — which
we have listed here. Um, but broadly speaking, you know, we asked,
um, about industry practices and current developments, uh, and the various use cases for products
and services that rely on access to consumer data and financial records, um, as well as,
you know, very much wanted to hear about what benefits consumers, um, receive and — and
their can — and their experience, uh, with these products.
Uh, in addition to practices and benefits, obviously, we are — we are very much interested
in hearing about, uh, the risks and challenges and concerns, uh, related to accessing consumer
data. And — and lastly, we, uh, requested feedback
on potential market developments: Uh, where, um, do responders see the market going? What
kind of developments? Are there industry standards that can come out from, um, uh — in the — in
the current, uh, ecosystem and the current players, um, uh, and — and along with that,
uh, asking for perspective and opinions on what the Bureau’s role, in particular, could
be, uh, to foster an environment where there is a development of responsible and consumer-friendly,
uh, innovation, um, uh, that benefits consumers. So, as was mentioned earlier, there were,
um, over 70 responders. Um, we, um, also, uh, receiving feedback, um, from, um, all
stakeholders. Uh, it was a wide range of, uh, responders. It was from financial institutions,
uh, big and small, uh, account aggregators, data aggregators, uh, information users, consumers,
and, uh, consumer interest groups, as well as trade associations. Um, in particular,
uh, we did receive comments from, uh, trades that represent the interests of con — uh,
co — community banks. Um, and, uh, while, uh, many of the commenters,
um, agree that responsible innovation benefits consumers and customers, um, alike, uh, as
was mentioned before, we did hear a number of, uh, concerns, uh, in — in — some that
we kind of listed out in terms of the risks that we were, kind of, aware of, but also,
you know, that list is, uh, uh, generated from, uh, the feedback that we were hearing
from financial institutions in particular, um, and, uh — And, again, uh, you know, on
a high level, those, uh, deal with, uh, data security and increase of potential fraud,
um, privacy concerns, as well as burdens, um, and — and, um — and increased costs,
uh, especially as they apply to community banks.
Um, at the same time, uh, we also heard that, uh, uh — whether anecdotal or from, um, various
feedback — that a commu — there are some community banks that are actively, um, partnering
or looking to partner with, um — with, uh — or — or develop innovative products and
services that — that rely on consumer data. I think, in recent developments, you know,
we do hear, kind of, market announcements about partnerships, um, uh, about using, uh,
data aggregation in order to, kind of, deliver products either more efficiently or faster
or — or, uh, you know, a smoother, uh, co — consumer experience.
Um, so, I — I feel like that — you know, with that, uh, probably turn the discussion
over to the Advisory Council, because it sounds like these are impressions that, um, uh — they’re
both, uh, uh, you know, the concerns but also the benefits, you know, are — are being highlighted
by — by community banks. Well, thank you, and thank you for that presentation.
Um, at this point in time, maybe we can gain some perspectives of the Advisory Council.
Please, Trent? I’ll —
Oh, sorry. I’ll go this — this time, Jack, and you — You go next.
Um, uh, we’re — we’re in the nationwide prepaid card issuing space, so we have customers,
sort of, all over the place. Um, and they, uh — they, uh, typically take advantage of
a lot of the providers that you listed on the — on the slide, the 12 or 13 types of
data sharing that — that’s going on. The one thing that is the greatest challenge
for us — and I’m sure many other banks — is that, right now, Reg E, in my mind, is, um,
completely ill-equipped to deal with this, um — this phenomena, um, and this desire
of a consumer to take information and move it to a — or have it moved to another provider,
and then, lo and behold, something bad happens. Um, right now, we start with the premise that,
uh, it’s the bank’s liability, in many cases, or they at least know us, and so we’re the
first place they call, because it’s easier to deal with us than it is to deal with, uh,
XYZ fintech, um, in Silicon Valley. And so, I think my concern is that I believe
the banks, right now, are sort of propping up this — that side of the industry, because
they’re not having to deal with the liability questions and taking best practices for securing
information and not sharing information, and then standing behind financially when something
doesn’t go right for the consumer. Um, so I think, just at its core, uh, Reg
E needs an overhaul to accommodate — to accommodate this particular, uh, issue, and — and specifically
needs a very clear line of demarcation between, banks’ liability starts here and stops here.
And no matter what happens post-that-line, the bank’s line is right here. And, uh, it’s
not there today. Just to touch on that — and I was going to
go into the Reg E, because that’s, obviously, a huge concern from a banker’s, uh, perspective
— but tapping on to that, you know, we have a duty to make the data available to the consumer.
The consumer doesn’t necessarily mean the third party in my mind. So, if we’re making
it available to the third party, that opens up a whole other, um, Pandora’s box. And it
requires third party oversight by us — huge with our regulatory bodies right now. Who
are we doing business with, who are we allowing to do business with, and are you doing your
due diligence on them? I’m not going to be flying out to the Silicon
Valley every other week for every new startup, to do due diligence. I can’t afford to. You
know, I spend, on an average year, at least a quarter of a million dollars on data security.
I don’t know what they spend. I haven’t got a clue. I mean, they might be running out
of their garage for all I know. Yeah.
So — And these partnerships you’re talking about with banks, to my opinion, this is another
one that I know that the OCC has been all over it. A lot of it is what I call rent-a-charter.
You know, they’re paid a fee for a transaction or per record that are coming through, so
they’re looking at it as a revenue source. So, I think there’s a fine line on where we’re
at here. They’re doing it for financial reasons. Are they doing it for any other reason? I
question that. So — And when you start looking at what they’re
doing and why we might be not wanting to share our data, some of those are our competitors.
You know, do we want to share that data with our direct competitor? And I would argue,
I don’t. At our cost. At our cost, yeah.
And our liability. That’s —
So, that’s — that’s where my — So, when you see the pushback, that’s just a handful
of the pushbacks. Can I see some of the positive? Yes, but we try to offer as many of those
products as we can within our bank. We have the budgeting. We’ll text message you when
your account’s getting so low. We’ll notify you of unusual transactions: Is this you?
We spend a lot of money on this type of stuff. So, why do we need to go out and outsource
it to another third party? I — I, you know, agree with that, and I think,
you know, with some of these third parties, we don’t even know who they are. Like you
said, they could be out of their garage. They could be two or three people that have a — you
know, a small, uh, business, and they’re not well cap — capitalized in some cases. So,
how are they financially responsible or — or reliable? I think the liability issue is huge
for financial institutions. We already have a liability right now. If
someone has, uh, their credit card on their cell phone, and they decide to take the password
off their cell phone so they don’t have to put in a password to use it to buy something,
and someone takes their phone and buys a lot of stuff, the bank eats it. It’s our — it’s
not their responsibility because they gave the password away or didn’t have one on there.
We’re responsible, and that’s the credit card industry putting us responsible, and we can’t
fight that if we want to have a relationship and offer a debit card at all.
So, we already have that huge liability, so of course we’re going to fight a customer
being able to give their information to anybody, because at the end of the day, it’s going
to be us, again, that eats it and that has the liability. So, if we can be guaranteed
the liability’s going to be on the consumer or whoever has the information, and have it
in writing, it’d be a lot easier for us to open something up.
There’s a catalog of regulations that, I think, has to be looked at along with this. Um, my
bank is — is a charter that’s 10 years old. It was built around the electronic delivery
of statements, uh, both on the loan side and on the consumer side, sort of built around
this whole idea. And the one that I scratch my head all day long is e-sign.
So, in order for me to open up an account for a customer who willingly comes wanting
electronic statements, opting in — no problem. Right? Two elements to eSign: The — the customer
has to voluntarily opt-in, but the thing that’s arcane is demonstrated consent. So, the consumer
has to — has to proactively show that they have the ability to open up a PDF. Now, the
reg was written when no one knew what a PDF was, and now it’s completely arcane.
So, I get the — the consumer set that we’re talking about here, who desires access to
this information, who are frustrated by the fact that I either have to send them home
to log onto the online banking to open a PDF to give me some form of demonstrated consent
that proves that they can do it, or if I’m lucky, I have a consumer who has a smart phone
in their possession right then and there, and then my staff can walk them through how
to log on while they’re in the ba — you know, to g — the — the process. You know, there’s
a series of these things that really need to be looked at.
There’s innovators, there’s entrepreneurs who say, “Hey, if the banking industry made
this more — this data more available, we could help consumers,” but, you know, part
of the ask here, with Reg E and things like e-sign — We could probably do a lot of things
to not hamstring ourselves. So, to — Oh. Uh, t —
, um, the potential asymmetry of the application that — Yes, that’s — I’m actually not sure
we’ve seen that comment in the comments, so. Yeah, and it’s — it — it makes — I mean,
it is what the regulation says, but there also is no, um — You know, there’s no latitude
in how it is examined. Right.
It is examined 100%. I mean, we have to — we have to demonstrate it for the consumer customer
on accounts end, on the loan side — Right.
— so we can’t send out electronic statements without going through this process. And it
— it really gets kind of tedious, because you not only have to take all the time, but
you have to document it. And that’s true even for consumers who originate
and service their account electronically. Right, so the — You know, so the — the — the
Millennial that’s just accustomed to doing all this information, they just don’t even
get it. Yep. That’s a really good point.
On a flip side to that, um, as community banks, we are dependent on, uh, data aggregators.
We use them all the time, right? We have credit bureaus. Uh, some cases, you’ll have a check
systems or some type of — Yep.
— system like that. There are legal background checks, uh, you know, identification like
a LexisNexis-type of system, and so we’re very much dependent upon those data aggregators.
Part of my worry, quite frankly, is, um, uh, can we still be in the game of — of data?
Um, or is it so vague and so expensive that the Googles, the Apples, the large institutions,
if you will, um, will own it? And is there access?
Um, and so, I guess my — my subsequent question to that, too, is: How do you get all these
bunnies back in the box? Um, you know, all this data is out there in — in various places.
Um, it’s not like I have the keys to my own financial — and that I can choose who to
give that key to. Um, maybe that type of scenario will work in a block chain or something, you
know, um, but today, I don’t know how you — you actually pull all that back.
But if it’s wrong, how do you correct it? That’s a great question. I don’t know.
Can I ask how many — how many folks here are offering their customers, you know, services
that require, sort of, consumer authorization to go get aggregated data services of one
kind or another? Uh —
We — — yeah.
We are. We have Yodlee in our online banking, but
what I find interesting about both these conversations is that the bankers are wrestling with regulatory
burden and struggle and, primarily, uh, overenforcement of those re — regulations. And we feel like
we can’t help the consumer and do the things they want, and yet, Silicon Valley can do
it, and they don’t have that regulatory burden; they don’t have the — the same constraints.
And yet, that’s what we sit here and talk about. Well, let’s go let them do it because
they’re able to, but the people that should be doing it can’t because of regulatory burden.
And that’s the part that’s so frustrating. I mean, you don’t think we want to provide
this stuff to them? Uh, we — we want to help consumers. We want to be able to make these
loans. We want to be able to — to provide these services. But the reality is, is that
we’re drowning in it, and, uh, we’re terrified to do anything, because everything we do is
wrong. And it’s not a — a — a fair, uh — you know,
as far as us and the large institutions, financial institutions. They have the resources, you
know, to implement all of this. We don’t as community banks. It’s not a fair playing field.
Yeah, they have legal staff to, uh — Right.
— argue with the regulators and — Exactly.
Well, I — I would say this: Rela — whether we like it or not, the — the digitization
of banking is in — in front of us, um, and one might even go so far to say as that we’re
over the chasm, or the tipping point, to where the adoption of mobile and technology is,
um — is now into the early adoption phase of innovation.
Um, and so, the access to data and the use of data, um — you know, obviously, a — a
blessing and a curse, right? Um, how do you use it? How do you use it safely? How do you
get access to it and provide it in a way to the consumer that is responsible and — and
secure? Um, I — I think when we look at — I was
kind of looking at the laundry list of — of all the regulations that we have to oversee
when we’re in that third party, um — You know, there’s — obviously, there’s privacy
regs. There’s the data security and BCP, uh, third-party risk, AML/BSA, and UDAP, to kind
of wrap it all together. I mean, it’s just kind of — I — I think, you know, there’s
a lot of regulation on there, some of which, as we discussed, may need to be modernized
to what’s happening, um, of which there may not be an end to the modernization, right?
Mm-hmm. We’ll — we’ll modernize it to today’s standards,
but 2 years from now, we’ll be saying, “Wow, is that antiquated,” in terms of its speed.
Um, I would ask one question, maybe, from the — the approach of — of looking at this.
Um, um, particularly when you look at the — the feedback that you got from the RFI,
are there — are there — is there anything — or nuggets in there when — Are there — Are
there people making progress in these phases, or is it s — Do you see a concentration in
the large institutions, bank and non-bank? Or not enough respondents?
I think, probably, on the community bank side. I mean, because you’re getting letters mostly
from trades, that that sort of tries to provide a holistic perspective and that, then — I
think then gets — it gets a little hard to sort of unpack that and go, you know, are
there players within this realm who are — are very far ahead of us? I mean, some of the
players who are further ahead than others, I think, while I wouldn’t call them community
banks, are at the smaller end of the large-bank spectrum —
Okay. — which is kind of interesting.
Yeah. Um, and they may see that as competitive play,
but if they’re trying to play to Millennials, maybe this is — you know, it’s not just that
we’re going to offer you the kind of services that you can get without an issue, but we’re
going to make it relatively easy for you to share your data so that you can get these
value-added services. Um, uh, I think, uh — but within community
banks, I think we’ve heard both demand for the services — I think something of a black
box, to me at least, at this point, reading the comments, is: To what extent do community
bankers feel that their customer base wants to enable sharing to get these services, be
it from them or be it from a third party? I certainly hear from community bankers that
their customers definitely have a demand for PFM-type tools, but to what extent do you
find, you know, online credentials being shared and a lot of screen scraping taking place?
Do you feel, sort of, the need to substitute something in that place, or is it not happening
at a scale that that is — that is so? I mean, I think — Yeah, I think the one thing
I would add is that, at least from — at least from, uh, some of the themes that are coming
up in the comments: Uh, it’s a tricky balance, you know, because I think, um, uh, you know,
there are — even from financial institutions, um, kind of saying, uh, different things because
there are different, uh, needs, like, at this time.
I think one — For example, I think, uh, um, many commenters said, CFPB — Whoa, like,
you know, don’t rush into a rulemaking, for example. Um, let industry kind of figure some
of these things out, whether we can figure out, uh, an industry standard. Uh, at the
same time, there’s also, kind of, calls for, um, uh, noting that, you know, there needs
to be a level playing field, that there are — uh, for all participants in the ecosystem,
uh, because, you know, there are a lot of regulations and requirements that banks have
which, um, may or may not, uh, translate over to non-banks or — uh, uh, uh, or, um, some
of the data aggregators. Uh, so I think — And — and — So, it’s one
of those things where it’s, uh, essentially, take your time, figure out and be aware of
these issues, but if you’re going to get into the — if you’re going to get into it, here
are the issues that really, kind of, matter. So, um, uh, you know — and — and I think
that there are, uh, you know, developmen — as we kind of mentioned earlier in our presentation,
there are developments in industry where, I think, there are some banks trying to figure
things out, whether it’s by bilateral agreements, things like that. I think, um, it’s an interesting
question, because it does turn a li — it may turn a little bit on whether you — the
interaction that you might have with an aggregator is via screen scraping, where you don’t even
know who the — the, um — who’s requesting the data, versus ones where you might actually
be thinking about, um, engaging in a — in — in some sort of agreement.
And, potentially, then presenting that information to the consumer. You are sharing with the
following entities. Do you want — Sorry. And, potentially, then sharing that information
with the consumer, so the kind of screen that would tell the consumer, your customer: Here’s
who you’re currently sharing with. Here — here’s who we’re sharing with, as we understand it,
because we have your authorization to do so, and here’s what we’re sharing. You know, “Turn
this off now if you want to” sort of thing, basically. Um, you can’t really do that if
all you have is screen scraping coming at you, and you don’t know who’s the other end.
And I — I think where — and I haven’t given this a ton of thought, but I think where that
— where the conversation is evolving is that we’re really talking about, um, two po — I’m
sure there’s lots of different arrangements, but the two primary arrangements: There’s
a direct relationship between bank and third-party entity, um, where I can, presumably, exercise
some sort of negotiation and requirements, contractually, around certain standards that
I feel comfortable with that sharing happening, and then you have screen scraping, which is,
sort of — to some degree, sort of, cowboy acquisition of — of information.
And I — I — uh, very soon, I’m — I’ll have to provide the Bureau with every cardholder
agreement and every fee schedule, um, that, uh, I offer in relation to my prepaid cards,
and that’s, presumably, that the — the consumers would find some sort of value in being able
to do comparative shopping. Um, if the third-party entities are within the realm of your regulatory
schema, um, perhaps it’s not too difficult for you to require some database reporting
of them, so that the public and we ha — can very easily go to a place and get some sort
of level of comfort that, in fact, that third-party entity has passed some level of competency
that you would be able to easily explain their sharing practices, so we have some sort of
comfort to be able to explain to a consumer, “Hey, before you do that, make sure you go
check this one out, because what we’re seeing, it could be troubling to you.”
Um, I think that’s really what’s — It’s — it’s impossible for us to educate the car — the
customer on all of the providers out there, so, I guess, if we’re in the database business,
and we’re in the — we’re in the comparative shopping business at the Bureau, um, this
— this could be as good a use as any. You know, I would tell you that my consumers
are very conservative, and they don’t want their data shared, yet they don’t know all
the places where they’re giving it away. And that’s the part that’s really maddening. And
yet, they often look back at us and say, “Are you — are you sharing it?”
And, uh, you know, I make the point because a lot of times, for instance, there’s the,
uh, tax prep websites that you’ll do it for free, but, uh, in that end-user agreement,
they’re saying, “We get to sell your data,” and it’s everything you file on your tax return.
That’s what you’re trading, and they’re getting, you know, a lot of money for that, and you
— you basically would have paid $35 or whatever it was to file. And yet, it’s in this big,
huge end-user agreement that nobody reads, and then that gets sold.
And — and that’s so true in so many places, and yet, as a bank, I have to conspicuously
tell them, “Here’s what we do in our data sharing practices.” In fact, I, you know,
up until, what, last year, I had to send it to them every year and say, “Here’s what we
do.” And — and it’s interesting to me. Like, I
see the ads on TV about the credit score cards: Well, you know, we don’t care if you’re our
customer or not. You call us up; we’ll give you your credit score.” Well, yeah, guess
what? And then you’re going to authorize that they turn around — They’re going to get enough
information about you, and they turn around and sell those to data warehouses.
So, there — there is a little bit of disparity in the marketplace that, uh, community banks
and banking in general has to be up front — UDAP — again, there’s another one of those
wonderful regulations that we talk about — uh, uh, very conspicuous about how we’re going
to use it, yet the rest of the marketplace doesn’t have to do that. And so, there’s an
unfair competitive advantage. Uh, it just — Uh, you know, and we sit there
and say, well, either make it — Make it equality. Either put them under the same requirements
we have, and then we can all talk, uh, again, or give us the same freedoms they have and
— and, you know, I’m sure we’d be glad to do all these things for them.
Yeah. It’s a generalization. I think it’s fair to say, in the — in the comments, aside
from the substantive comment of, “Don’t write a general term 33 rule,” the overwhelming
regulatory comments were about existing rules, um, and primarily focused on their asymmetric
application. If — If I can make a comment to the — maybe
to the techs deck that, um, I — I would say, banks in general, particularly community banks,
have been quite good, at the core, on a data security standpoint. The core, even though
it’s somewhat antiquated, and we have our complaints about that, um, but it’s been reliable,
and it’s been stable and secure. Um, the place where we lack is probably what
sits on top of that base, if you picture a cupcake. Um, that sweet spot on the stop — on
the — on the top of a cupcake is what I would consider some type of flexible middleware.
So, you — you can plug in different financial technologies yet still protect the core. Now,
you still may have some, um, relationships that you API direct into the core for business
intelligence or for your specific product lines, um, so I —
And I guess the only reason I — I really bring that up is, we’re going to have to access
different sources of data specific to the customers we serve. Um, generally, community
banks will find themselves in spaces of — of niches. Um, they have a niche or a particular
profile of a customer, whether it’s geographically based or demographically based or however
you want to slice that particular cake. So, in that customer experience, they’re going
to find the data and the technology to meet that. Um, they’re going to need some flexibility
in the middle to add and subtract as — as things change and protect the core at the
same time. So, I do think banks are particularly in a good spot as a data aggregator.
Um, uh, as — as we described, we — we fear some of the other data aggregators, because
— whether the standards are the — you know, the same or not. Um, and the — the question
is: Can we evolve fast enough, um, to be relevant to our customers and put those technologies
in place and access the data to do it? If you truly believe that banking’s being digitized
and you need to meet that. Are there questions, thoughts, comments relative
to the data aggregators? Got a feeling we will hear this topic, and
the previous one, again — — uh, in terms of the way banking is going.
Very good. Well, Will and Stephen, thank you very much
for your, uh, presentation, um, and, uh, thank you to the Council, uh, and everyone for that
great discussion. Um, if there’s no more comments, this concludes the meeting of the CFPB’s Community
Bank Advisory Council. Uh, thank you, again, for your time, and have a wonderful rest of
your day. Thank you. Thank you.

Leave a Reply

Your email address will not be published. Required fields are marked *