Nov. 27, 2024

From Polls to Partnerships: Election Season and the NCUA’s Role in Modern Lending

Apple Podcasts podcast player iconSpotify podcast player icon
Apple Podcasts podcast player iconSpotify podcast player icon

In a world where technology rapidly reshapes financial systems, credit unions must innovate to stay ahead.

In this episode, host Matt Snow chats with Nat Hoopes, Head of Public Policy at Upstart, as they navigate the fintech landscape and highlight the importance of actionable insights from industry professionals. They discuss the evolving role of the NCUA, the need for diversified products in credit unions, and share compelling anecdotes about Synapse's bankruptcy and Upstart's partnerships with over 100 institutions. Together, they explore the delicate balance between partnership and regulation, the transformative power of AI in lending, and the challenges and opportunities in this dynamic sector.

Join us as we discuss:

  • Why insights from real industry professionals outweigh the advice of self-proclaimed "gurus".
  • How AI enhances credit access and performance compared to traditional methods.
  • The importance of partnerships between credit unions and fintechs to offer diverse products and maintain competitiveness amid regulatory scrutiny and technological changes.
WEBVTT

1
00:00:02.879 --> 00:00:05.639
You are listening to Leaders in Lending from Upstart, a

2
00:00:05.679 --> 00:00:09.560
podcast dedicated to helping consumer lenders grow their programs and

3
00:00:09.640 --> 00:00:13.919
improve their product offerings. Each week, here decision makers in

4
00:00:13.960 --> 00:00:16.760
the finance industry offer insights into the future of the

5
00:00:16.800 --> 00:00:22.039
lending industry, best practices around digital transformation, and more. Let's

6
00:00:22.039 --> 00:00:22.800
get into the show.

7
00:00:26.399 --> 00:00:28.920
Welcome to another episode of Leaders in Lending. I'm your host,

8
00:00:28.960 --> 00:00:32.000
Matt Snow this week joined by Nat Hoops upstarts very

9
00:00:32.039 --> 00:00:34.520
own head of public policy and regulatory affairs.

10
00:00:34.600 --> 00:00:36.320
Natt Welcome, Thanks Matt.

11
00:00:36.399 --> 00:00:37.880
It's good to be back on the pod.

12
00:00:38.520 --> 00:00:41.200
Yeah, return guests, but first time I get the pleasure

13
00:00:41.240 --> 00:00:42.679
of chatting with you in this format.

14
00:00:42.679 --> 00:00:43.640
So I've been looking forward to this.

15
00:00:44.280 --> 00:00:46.560
Yeah, it's going to be It's a good month.

16
00:00:46.600 --> 00:00:49.679
I guess we're almost in November, a big, big election

17
00:00:49.840 --> 00:00:53.399
in Washington and around the country, so good time to

18
00:00:53.479 --> 00:00:55.119
talk politics and policy.

19
00:00:55.600 --> 00:00:58.240
Yeah, And I was wondering Mat, like, this close to

20
00:00:58.280 --> 00:01:02.000
an election a few weeks away, do you see increased activity?

21
00:01:02.079 --> 00:01:04.920
Are people distracted or you know, what's what's it like

22
00:01:04.920 --> 00:01:06.760
in Washington right now? What are people thinking about?

23
00:01:07.400 --> 00:01:11.040
Yeah, ironically, it's kind of a time when Washington slows down,

24
00:01:11.319 --> 00:01:14.200
and all of the operatives and all of the energy

25
00:01:14.319 --> 00:01:20.799
is in Pennsylvania or or Arizona, or Georgia or North Carolina.

26
00:01:21.879 --> 00:01:25.000
And so it's definitely a time when a lot of

27
00:01:25.159 --> 00:01:29.040
people are anxiously looking at the polls. They're looking at

28
00:01:29.040 --> 00:01:33.200
the weather forecast for election day. They're really trying to

29
00:01:33.280 --> 00:01:35.439
kind of get a sense of you know, who's going

30
00:01:35.519 --> 00:01:37.239
to come out. And I think the campaigns tend to

31
00:01:37.280 --> 00:01:39.840
focus on their core you're turning out their core voters.

32
00:01:40.840 --> 00:01:44.159
And it's really a time when policy takes a back seat.

33
00:01:44.239 --> 00:01:47.400
So where I spend the vast majority of my time,

34
00:01:47.519 --> 00:01:52.879
which is you know, regulatory work and and the the

35
00:01:52.920 --> 00:01:56.879
evolving policy landscape for for upstart in the financial technology

36
00:01:56.920 --> 00:02:01.519
industry really isn't the folks, is during an election season.

37
00:02:02.120 --> 00:02:05.879
Yeah, a lot of variables that go into the outcomes

38
00:02:05.879 --> 00:02:08.039
of an election. Maybe there's an AI model out there

39
00:02:08.080 --> 00:02:11.400
being worked on to get more precise on what's really

40
00:02:11.439 --> 00:02:11.960
going to happen.

41
00:02:12.039 --> 00:02:15.400
I can guarantee that AI is involved in the election,

42
00:02:15.479 --> 00:02:18.319
hopefully not in ways that we don't want, So we'll see.

43
00:02:18.479 --> 00:02:19.360
Yeah, exactly.

44
00:02:19.800 --> 00:02:23.520
Well, Hey, you mentioned you were in Phoenix recently meeting

45
00:02:23.520 --> 00:02:25.319
with a bunch of examiners from the nc UA.

46
00:02:26.199 --> 00:02:26.840
What was that like?

47
00:02:27.280 --> 00:02:29.199
What's it like when a bunch of examiners get together

48
00:02:29.240 --> 00:02:32.199
and invite someone from a fintech in there to speak

49
00:02:32.199 --> 00:02:32.840
with them.

50
00:02:32.879 --> 00:02:35.560
Sure, well, you know, we've just seen such tremendous growth

51
00:02:35.639 --> 00:02:39.039
of credit unions, right so of our you know, one

52
00:02:39.120 --> 00:02:43.080
hundred plus institutions, more than half our credit unions. It's

53
00:02:43.080 --> 00:02:46.800
a segment that has historically done a lot, you know,

54
00:02:46.879 --> 00:02:50.120
of what they would call signature loans, but unsecured personal

55
00:02:50.120 --> 00:02:54.439
loans are core product in Upstarts product suite, and so

56
00:02:54.639 --> 00:02:56.960
that means we have a lot of experience of working

57
00:02:56.960 --> 00:02:59.280
with credit unions, but we're also growing very quickly. And

58
00:02:59.319 --> 00:03:04.039
so the regional examiners, the examiners that are on the ground,

59
00:03:05.120 --> 00:03:08.560
they may be going through the first cycle where they're

60
00:03:08.560 --> 00:03:10.680
working with a credit union who signed up with upstarted.

61
00:03:10.680 --> 00:03:11.479
So there can just be.

62
00:03:11.439 --> 00:03:17.560
A lot of questions about AI, about unsecured personal lending,

63
00:03:17.599 --> 00:03:21.080
about the current macro environment, you name it. And so

64
00:03:21.879 --> 00:03:23.840
it was really good to just get in front of

65
00:03:23.879 --> 00:03:28.520
the audience where they could all ask any question they wanted,

66
00:03:29.479 --> 00:03:32.919
try to demystify some of the partnership, try to give

67
00:03:32.960 --> 00:03:37.080
them an understanding of how what we do is both

68
00:03:37.120 --> 00:03:39.000
similar to what has been done before if you think

69
00:03:39.000 --> 00:03:41.680
about credit scoring and all the machine learning that's always

70
00:03:41.680 --> 00:03:44.000
gone into an AI that's gone into the you know,

71
00:03:44.159 --> 00:03:46.560
the development of traditional credit scores, and then kind of

72
00:03:46.560 --> 00:03:49.240
taking it to the next level in developing this full

73
00:03:49.439 --> 00:03:53.360
turnkey solution so that credit unions can can really find customers,

74
00:03:53.919 --> 00:03:57.280
help help them price customers, help them add that extra

75
00:03:57.439 --> 00:04:01.479
layer of a tool to be more accurate in in

76
00:04:01.759 --> 00:04:03.680
how how they do what they do, and then also

77
00:04:03.719 --> 00:04:06.800
servicing the customers. And so I think just giving the

78
00:04:07.159 --> 00:04:10.520
examine our audience, from our perspective what a partnership looks like,

79
00:04:11.240 --> 00:04:14.479
you know how how the how the partnership is overseen

80
00:04:14.520 --> 00:04:18.920
by credit unions, that that was just a really good opportunity.

81
00:04:18.959 --> 00:04:21.079
And the session went I think I want to say

82
00:04:21.160 --> 00:04:24.360
it probably went a half an hour over just with

83
00:04:24.600 --> 00:04:28.040
the numbers of you know, really good back and forth

84
00:04:28.079 --> 00:04:29.160
and dialogue that we had.

85
00:04:29.639 --> 00:04:31.839
Yeah, that's great that there was so much engagement. Do

86
00:04:32.160 --> 00:04:34.920
you think they even understand or grasp like the need

87
00:04:35.000 --> 00:04:38.839
for outside partnership with fintech companies like Upstart, Like do

88
00:04:38.879 --> 00:04:41.959
they understand the premise of why maybe some of these

89
00:04:42.000 --> 00:04:44.120
smaller institutions can't even do this themselves.

90
00:04:44.240 --> 00:04:46.920
I do think that there's an education that needs to

91
00:04:46.920 --> 00:04:51.879
be done about the long term viability of their customer

92
00:04:51.959 --> 00:04:55.600
base wanting anything other than a deposit accountan we know

93
00:04:55.680 --> 00:04:57.839
deposits are pretty sticky. You know, if you have been

94
00:04:57.839 --> 00:04:59.360
with the credit union for a long time, you may

95
00:04:59.360 --> 00:05:02.519
not be planning to leave on a deposit basis. But

96
00:05:03.079 --> 00:05:06.439
just doing the education that people are really searching for credit,

97
00:05:06.560 --> 00:05:09.959
not with their primary institution, but they're out there at

98
00:05:10.000 --> 00:05:12.839
the comparison sites. They're trying to get the best possible rate,

99
00:05:13.360 --> 00:05:15.079
just the way we would buy you know, an airline

100
00:05:15.079 --> 00:05:16.600
ticket or anything else. Like, we're going to be out

101
00:05:16.600 --> 00:05:18.600
there trying to find the best option, and that means

102
00:05:18.720 --> 00:05:22.160
you know, finding the Internet, and so in doing that,

103
00:05:23.120 --> 00:05:25.560
you know, really explaining to them the whole order of

104
00:05:25.600 --> 00:05:28.000
operations of how somebody comes upstart dot com, how they

105
00:05:28.000 --> 00:05:31.079
get matched with the credit union in their region, you

106
00:05:31.079 --> 00:05:34.600
know why. You know, if the customer isn't walking in

107
00:05:34.639 --> 00:05:37.079
with a bunch of paperwork under their arm physically, it

108
00:05:37.120 --> 00:05:39.680
doesn't mean that the credit union can't have a viable

109
00:05:39.759 --> 00:05:41.959
lending program. But they're really going to need a partner,

110
00:05:42.000 --> 00:05:43.680
So we did spend a lot of time on that.

111
00:05:43.800 --> 00:05:48.240
I think I think the examiners, you know, they can

112
00:05:48.279 --> 00:05:50.720
be suspicious, that they can be skeptical.

113
00:05:50.759 --> 00:05:52.480
I think it's their job to be skeptical.

114
00:05:52.519 --> 00:05:56.040
There have been a lot of you know, snake oil

115
00:05:56.079 --> 00:05:58.680
salesman through the year saying this is going to be transformative,

116
00:05:58.680 --> 00:06:00.279
this is going to be the greatest thing, and then

117
00:06:00.439 --> 00:06:02.959
and then it doesn't always turn out, And so you

118
00:06:03.040 --> 00:06:05.560
have to kind of be willing to meet that skepticism

119
00:06:05.600 --> 00:06:07.800
head on and then show why, you know, we're the

120
00:06:07.879 --> 00:06:10.120
kind of company, as a public company that's really built

121
00:06:10.160 --> 00:06:12.439
to last and it's going to be here for Credit

122
00:06:12.560 --> 00:06:13.600
Unised for a really long time.

123
00:06:14.839 --> 00:06:15.079
Yeah.

124
00:06:15.160 --> 00:06:16.879
You know, the one thing that's helped me put some

125
00:06:16.920 --> 00:06:18.800
of that into context that is a memory that not

126
00:06:18.879 --> 00:06:21.560
only are they a regulator, they're the ensure, right, so

127
00:06:21.600 --> 00:06:23.839
they have a bit of a right and a reason

128
00:06:23.879 --> 00:06:25.759
to be skeptical. It's part of their job right to

129
00:06:25.800 --> 00:06:29.680
make sure that these credit unions are operating safely and

130
00:06:29.720 --> 00:06:32.399
soundly to protect their their customer's deposits.

131
00:06:32.639 --> 00:06:34.600
That's that they're not that's absolutely right.

132
00:06:34.879 --> 00:06:39.759
They're an insure and and there's a certain amount of

133
00:06:41.319 --> 00:06:45.079
you know, truth to the possibility that even if something

134
00:06:45.600 --> 00:06:48.319
is only a you know, reputational harm to the bank

135
00:06:48.600 --> 00:06:51.600
or the credit union like that, that that could.

136
00:06:51.360 --> 00:06:53.399
Then trigger sort of more problems.

137
00:06:53.519 --> 00:06:53.680
Right.

138
00:06:53.759 --> 00:06:58.199
So, they they tend to be very interested in understanding

139
00:06:58.199 --> 00:07:00.639
at the front end of a partnership, you know, when

140
00:07:00.680 --> 00:07:02.879
something is launching, because they know they may not be

141
00:07:02.959 --> 00:07:06.439
back for another eighteen months or two years, and so

142
00:07:06.800 --> 00:07:09.000
you know, they want to make sure that it's you know,

143
00:07:09.040 --> 00:07:11.120
as they like to say, if something's right at the beginning,

144
00:07:11.399 --> 00:07:12.600
it can be right in the middle, and then it

145
00:07:12.600 --> 00:07:14.439
can be right at the end. So I think I

146
00:07:14.480 --> 00:07:17.120
think we did a pretty good job of walking them

147
00:07:17.120 --> 00:07:20.279
through the full life cycle and showing them how the

148
00:07:20.319 --> 00:07:24.560
credit the credit union is really in charge of the size,

149
00:07:25.319 --> 00:07:29.959
the cadence of the partnership, the diligence process, the on

150
00:07:30.399 --> 00:07:33.199
going you know, the ongoing monitoring, the performance monitoring, the

151
00:07:33.199 --> 00:07:37.360
fair lending monitoring. And I think that really got them

152
00:07:37.839 --> 00:07:40.920
more certainly more comfortable than they were walking in. I

153
00:07:40.920 --> 00:07:43.800
think that you know, they they there were concerns not

154
00:07:43.839 --> 00:07:47.000
with Upstart, but just in general with financial technology companies,

155
00:07:47.399 --> 00:07:49.120
you know, maybe sort of trying to jump in and

156
00:07:49.160 --> 00:07:52.079
take advantage of small credit unions. And by the end,

157
00:07:52.120 --> 00:07:55.439
I think that they saw that upstarts on that list

158
00:07:55.480 --> 00:07:57.920
of companies, it's going to be around for credit unions

159
00:07:57.959 --> 00:07:58.680
for a long time.

160
00:07:59.759 --> 00:08:02.040
Yeah, you made a good point about them, you know,

161
00:08:02.279 --> 00:08:05.120
in their field, exams being every twelve to eighteen months,

162
00:08:05.120 --> 00:08:08.759
and the rate of pain, pace of change in technology

163
00:08:08.839 --> 00:08:12.399
now seems to be exponentially growing, you know, whether that's

164
00:08:12.439 --> 00:08:16.680
the availability of cloud computing, the different you know, algorithms

165
00:08:16.680 --> 00:08:18.600
and techniques you can use at that space, the number

166
00:08:18.600 --> 00:08:22.079
of third parties that are coming into play. So I

167
00:08:22.079 --> 00:08:24.839
would imagine they would want to be keeping up to

168
00:08:24.920 --> 00:08:26.639
date and learning as much as they could in between

169
00:08:26.680 --> 00:08:30.639
those exams and staying in sync with those things. Did

170
00:08:31.120 --> 00:08:33.919
you find there were any specific areas they were really

171
00:08:34.960 --> 00:08:37.559
you know, seeking more information on, or you found gaps

172
00:08:37.600 --> 00:08:40.799
and that they were trying to rectify.

173
00:08:41.200 --> 00:08:43.720
No, I think you know they were interested in, you know,

174
00:08:44.399 --> 00:08:47.159
when in the process, you know, they can get a

175
00:08:47.159 --> 00:08:50.200
better understanding of the upstart model. We talked a lot

176
00:08:50.200 --> 00:08:53.720
about what's appropriate to ask, you know, a credit union

177
00:08:53.759 --> 00:08:57.919
executive who may not be an AI machine learning expert

178
00:08:57.919 --> 00:09:00.759
with a computer science degree, versus where it's fair for

179
00:09:01.440 --> 00:09:03.960
you know, upstart to answer a straightforward question that might

180
00:09:03.960 --> 00:09:05.919
be posed in writing. And so we've sort of talked

181
00:09:05.919 --> 00:09:08.759
a little bit about that and how to have the

182
00:09:08.799 --> 00:09:12.399
exam process feel fair to somebody who isn't steeped in

183
00:09:12.440 --> 00:09:13.039
it the way.

184
00:09:12.840 --> 00:09:16.279
Our team is. And then we also spend a lot

185
00:09:16.279 --> 00:09:17.960
of time on open banking.

186
00:09:17.600 --> 00:09:22.360
Interestingly, because I think there's a lot of skepticism initially

187
00:09:22.639 --> 00:09:26.000
in this idea of like, well, would why would Congress

188
00:09:26.000 --> 00:09:28.000
in the DoD Frank Act, which you know I worked

189
00:09:28.000 --> 00:09:32.399
on back in twenty ten, you know, why would we

190
00:09:32.440 --> 00:09:34.840
want to empower and say that the customer owns their

191
00:09:34.879 --> 00:09:38.240
own data, And then you walk through just the importance

192
00:09:38.240 --> 00:09:40.120
of being able to shop around for the best possible

193
00:09:40.159 --> 00:09:43.120
product and relating it to our own lives of how

194
00:09:43.159 --> 00:09:45.240
we do that all the time, and that that applies

195
00:09:45.279 --> 00:09:47.840
in banking as well, and then making it come for

196
00:09:47.919 --> 00:09:49.919
all circle and say, well, but now credit unions can

197
00:09:49.960 --> 00:09:52.240
actually take advantage of that too with partnerships like I'll start,

198
00:09:52.279 --> 00:09:55.120
like you know it's it's the credit union upstart partnership.

199
00:09:55.120 --> 00:09:57.600
If it's enabling somebody to refinance their credit card that

200
00:09:57.679 --> 00:09:59.559
might be with one of the biggest banks in the country,

201
00:09:59.600 --> 00:10:01.159
and then all of a sudden, that's now become a

202
00:10:01.679 --> 00:10:04.399
credit union customer that could be become a member, right,

203
00:10:04.440 --> 00:10:06.879
that's going to be you know potentially like you know,

204
00:10:06.919 --> 00:10:09.159
a depositor with that credit union, or get other services

205
00:10:09.159 --> 00:10:10.919
from that credit union. We talked a little bit about

206
00:10:10.960 --> 00:10:13.639
that cross selling, and so I brought up an example

207
00:10:13.679 --> 00:10:15.679
where you know, a credit union that worked with Upstart

208
00:10:15.720 --> 00:10:17.960
was able to cross sell you know, something like ten

209
00:10:18.000 --> 00:10:20.480
percent of their customers into other products and so forth,

210
00:10:20.559 --> 00:10:22.879
or their member their new members through through through the

211
00:10:22.960 --> 00:10:24.720
Upstart partnership into into you.

212
00:10:24.679 --> 00:10:25.679
Know, other products.

213
00:10:25.679 --> 00:10:30.360
And so I think it just positioning everything as on

214
00:10:30.440 --> 00:10:32.240
the long term side of the health.

215
00:10:32.000 --> 00:10:34.519
Of the credit union, which is ultimately what they care about.

216
00:10:34.559 --> 00:10:34.679
Right.

217
00:10:34.720 --> 00:10:36.480
They are just like you said, they are the insurer.

218
00:10:37.679 --> 00:10:41.559
They want these institutions to survive, thrive. They're not so

219
00:10:41.639 --> 00:10:44.080
much interested whether they grow quickly, but they at least

220
00:10:44.159 --> 00:10:47.799
don't want them to wither and fail. And so I

221
00:10:47.840 --> 00:10:51.480
think that that was probably a really healthy kind of

222
00:10:52.559 --> 00:10:54.240
I think a little bit of an AHA moment of

223
00:10:54.279 --> 00:10:57.200
like open banking is meaning that our credit unions have

224
00:10:57.240 --> 00:10:59.240
a chance to compete for customers that they probably never

225
00:10:59.279 --> 00:11:00.679
had a shot at before.

226
00:11:02.240 --> 00:11:04.679
Yeah, that's interesting. I wasn't expecting you to bring up

227
00:11:04.720 --> 00:11:08.000
open banking. Maybe. Spoiler alert, there's a mutual friend of

228
00:11:08.000 --> 00:11:10.320
ours who's been on the podcast before as well. We've

229
00:11:10.320 --> 00:11:13.320
been having some conversations about this offline and are planning

230
00:11:13.320 --> 00:11:15.639
maybe a future conversation. So I might have to get

231
00:11:15.639 --> 00:11:17.759
some more talking points from you, some more insight in

232
00:11:17.799 --> 00:11:21.799
your perspective on that, because I think that the interesting

233
00:11:21.840 --> 00:11:24.159
correlation you mentioned like it's probably going to open up

234
00:11:24.279 --> 00:11:27.399
the opportunity for more partnership opportunities for credit unions. I

235
00:11:27.399 --> 00:11:30.799
would imagine in a world where data is more portable,

236
00:11:30.799 --> 00:11:32.120
our customers have more choice.

237
00:11:32.240 --> 00:11:33.879
Yeah, I think so one hundred percent.

238
00:11:34.159 --> 00:11:37.320
And I think the challenge for the smaller institutions, which

239
00:11:37.360 --> 00:11:40.240
we've seen across a lot of products, you know, small

240
00:11:40.320 --> 00:11:44.639
dollar product being you know a recent most recently there's

241
00:11:44.840 --> 00:11:47.840
you know, some studies of how many of the very

242
00:11:47.919 --> 00:11:50.320
large five or six banks of the country have built

243
00:11:50.679 --> 00:11:54.000
pretty viable you know, small dollar products for loans under

244
00:11:54.000 --> 00:11:57.320
a thousand dollars. But it's it's a difficult product, difficult

245
00:11:57.320 --> 00:12:02.759
customer to underwrite on, you know, in you know, online interface,

246
00:12:02.759 --> 00:12:05.559
people want their money quickly. You know, how are you

247
00:12:05.639 --> 00:12:08.360
going to compete in an area like that if you

248
00:12:08.360 --> 00:12:11.440
don't have a partnership. And so I think that there's

249
00:12:11.480 --> 00:12:13.879
a you know, there's a host of ways in which

250
00:12:14.080 --> 00:12:16.399
these credit unions can use the right type of partners

251
00:12:16.440 --> 00:12:19.480
to get them, you know, the type of members that

252
00:12:19.519 --> 00:12:22.279
they want to find in their regions and offer them

253
00:12:22.279 --> 00:12:25.039
a viable you know, such as the product.

254
00:12:25.320 --> 00:12:27.480
As you know, it's like the user experience is everything.

255
00:12:27.519 --> 00:12:29.159
I mean, if if it's going to take days and

256
00:12:29.240 --> 00:12:30.840
days to get through the process, they're going to just

257
00:12:31.440 --> 00:12:33.360
they're going to just drop out of the of the

258
00:12:33.399 --> 00:12:36.039
application and go go find somebody else who can get

259
00:12:36.080 --> 00:12:37.840
it done for them in an efficient manner.

260
00:12:38.960 --> 00:12:39.159
Yeah.

261
00:12:39.240 --> 00:12:41.759
Absolutely, I think we don't talk about that enough.

262
00:12:42.000 --> 00:12:43.919
You know. Obviously a lot of attention rightfully, so it

263
00:12:43.960 --> 00:12:47.679
goes to the underwriting, the verification, all the data and

264
00:12:47.759 --> 00:12:50.440
insights that drive that, but you know, the user experience

265
00:12:50.480 --> 00:12:53.120
has to be seamless and continually evolving as well. So

266
00:12:53.639 --> 00:12:56.559
definitely a huge benefit to partnerships there.

267
00:12:56.679 --> 00:12:57.159
Yeah, and we.

268
00:12:57.120 --> 00:13:00.320
Talked a lot with the examiners you know at nc

269
00:13:00.440 --> 00:13:04.600
u A about the fair lending testing as well, how

270
00:13:04.639 --> 00:13:08.720
to keep how to keep a rigorous set of tests

271
00:13:09.240 --> 00:13:13.440
on the upstart you know models, you know, doing some

272
00:13:13.600 --> 00:13:17.360
education on the upstart is just one piece of the decision.

273
00:13:17.519 --> 00:13:22.200
So so the person's you know, credit criteria as determined

274
00:13:22.240 --> 00:13:25.480
by other companies that are not upstart, Fyco being a

275
00:13:25.559 --> 00:13:28.720
great example, their their credit score, things like their debt

276
00:13:28.720 --> 00:13:31.679
to income ratio. Things that that tend to create like

277
00:13:31.720 --> 00:13:34.639
who is actually going to be even eligible can drive

278
00:13:34.720 --> 00:13:38.840
a lot of who gets declined, approved, et cetera. And

279
00:13:38.879 --> 00:13:42.080
then upstart is sort of this this piece that comes

280
00:13:42.120 --> 00:13:45.679
in afterwards. And so once they understood that, there's a

281
00:13:45.799 --> 00:13:49.480
very big difference between using AI and all the alternative

282
00:13:49.559 --> 00:13:54.600
data as this really helpful additional layer as opposed to, oh, no,

283
00:13:55.039 --> 00:13:57.559
you're just turning over your entire underwriting to some black

284
00:13:57.600 --> 00:14:00.679
box that's not you Like, that's a different story. And

285
00:14:00.720 --> 00:14:04.399
so I think people also, just examiners don't always appreciate

286
00:14:04.399 --> 00:14:07.440
that when they come into the first time to examine

287
00:14:08.000 --> 00:14:09.639
a credit union that is signed up with a financial

288
00:14:09.639 --> 00:14:12.919
technology provider, they think, oh, like, now this is out

289
00:14:12.960 --> 00:14:16.000
of their control and like showing no, no, this is within

290
00:14:16.039 --> 00:14:19.279
their control. And not only that, but it's really these

291
00:14:19.320 --> 00:14:22.080
other tools that are what is driving the vast majority

292
00:14:22.080 --> 00:14:25.559
of who gets credit and who doesn't, and then the

293
00:14:25.639 --> 00:14:28.559
upstart that's helping them really get a more refined look

294
00:14:28.600 --> 00:14:33.799
at those who would qualify within their desired region or

295
00:14:33.919 --> 00:14:35.519
other eligibility requirements.

296
00:14:36.559 --> 00:14:39.159
Hey there, former host Jeff kalviner here to let you

297
00:14:39.159 --> 00:14:42.440
know about an exciting opportunity to strengthen your understanding of

298
00:14:42.600 --> 00:14:46.039
artificial intelligence and financial services. If you want to better

299
00:14:46.120 --> 00:14:48.879
understand the fundamentals of how AI models work and how

300
00:14:48.879 --> 00:14:52.720
they will impact the financial services industry, your luck. Upstart

301
00:14:52.759 --> 00:14:55.399
is launched an online course and certification designed to do

302
00:14:55.759 --> 00:14:58.240
just that. This course will give you the knowledge you

303
00:14:58.279 --> 00:15:01.320
need to lead the conversation about how your organization can

304
00:15:01.360 --> 00:15:05.840
best leverage and manage this revolutionary technology. But that sounds interesting,

305
00:15:06.200 --> 00:15:08.360
check out the course as it's currently free for a

306
00:15:08.399 --> 00:15:11.279
limited time. You can learn more at upstart dot com

307
00:15:11.320 --> 00:15:16.320
slash AI Certification. Again, that's upstart dot com slash AI Certification.

308
00:15:16.840 --> 00:15:18.639
Thanks, and now back to the show.

309
00:15:20.679 --> 00:15:22.720
Did you find other than fairness? Were there other things

310
00:15:22.759 --> 00:15:25.960
they really leaned into and kind of that compare contrast,

311
00:15:26.000 --> 00:15:27.679
Like you mentioned, there's a lot of things that happen

312
00:15:27.720 --> 00:15:30.799
in the process that are the same as they've always been.

313
00:15:30.879 --> 00:15:35.000
You know, we just maybe newer techniques, faster, more more

314
00:15:35.080 --> 00:15:38.200
data points. Like as they went through that exercise of

315
00:15:38.240 --> 00:15:40.200
trying to think about, you know, how is this similar

316
00:15:40.279 --> 00:15:41.799
or dissimilar from what they're used to?

317
00:15:42.120 --> 00:15:46.120
Were there other things that they tended to investigate harder.

318
00:15:47.600 --> 00:15:51.200
I think that they just they had some prior assumptions.

319
00:15:50.559 --> 00:15:55.039
That where we had a really robust discussion. I mean,

320
00:15:55.080 --> 00:15:58.200
there was one comment in one of the earlier forums,

321
00:15:58.240 --> 00:16:02.200
not the main forum, where you know, why would you

322
00:16:02.279 --> 00:16:04.600
need The question was posed by an example, why would

323
00:16:04.639 --> 00:16:06.600
you need a lot of data points for a for

324
00:16:06.639 --> 00:16:09.720
a seven hundred fifty dollars loan? Right? And so I

325
00:16:09.759 --> 00:16:11.440
had to say, like, no, no, that's exactly where you

326
00:16:11.440 --> 00:16:12.200
probably need the.

327
00:16:12.080 --> 00:16:14.039
Most data points, do you know what I mean?

328
00:16:14.320 --> 00:16:16.559
In a certain sense that like it seems like a

329
00:16:16.639 --> 00:16:19.639
low stakes, but if you're trying to scale a program

330
00:16:19.679 --> 00:16:22.360
and you're going to have a whole lot of them.

331
00:16:22.559 --> 00:16:28.399
You know, people who are you know, at at at

332
00:16:28.440 --> 00:16:30.720
a point in their life where they need to borrow

333
00:16:30.799 --> 00:16:33.159
five hundred bucks or seven hundred fifty bucks or a

334
00:16:33.200 --> 00:16:35.720
thousand bucks from a financial services provider, they can't come

335
00:16:35.799 --> 00:16:38.279
up with that kind of money on their own. You

336
00:16:38.440 --> 00:16:43.840
really want to understand, you know, details about their you know,

337
00:16:43.919 --> 00:16:45.519
for instance, like it might be their cash flow in

338
00:16:45.559 --> 00:16:47.960
a writing speaking of open banking and through plaid, or

339
00:16:48.320 --> 00:16:52.279
you know, details about their their paycheck history to really

340
00:16:52.399 --> 00:16:54.639
understand that they're going to be able to repay that loan,

341
00:16:54.679 --> 00:16:57.159
because it's not as though you're lending into some cushion

342
00:16:57.240 --> 00:17:01.000
where you know, there's enormous there's enormous wealth there and

343
00:17:01.039 --> 00:17:03.399
you're sort of saying, well, you know that we know

344
00:17:03.480 --> 00:17:06.839
that this person is you know, both wealthy and has

345
00:17:06.880 --> 00:17:09.279
this huge high income. Like many times these you know,

346
00:17:09.400 --> 00:17:12.640
these smaller dollar loans, like you need more data. And

347
00:17:12.680 --> 00:17:14.799
so there again, we had a lot of a lot

348
00:17:14.839 --> 00:17:21.680
of these I think, really interesting prior assumptions that had

349
00:17:21.720 --> 00:17:24.960
to kind of be like addressed in a really direct way.

350
00:17:25.160 --> 00:17:27.039
And that doesn't mean it was uncomfortable.

351
00:17:27.039 --> 00:17:29.519
It's just it's a really important sort of evolution and

352
00:17:29.640 --> 00:17:30.880
conversation that we're having.

353
00:17:32.039 --> 00:17:35.039
Yeah, and I wondered that, like, how I don't regulators

354
00:17:35.079 --> 00:17:39.880
examiners think about that trade off of data versus human interaction,

355
00:17:40.400 --> 00:17:43.720
the cost, the ability to probe those decisions, you know,

356
00:17:43.759 --> 00:17:46.680
one by one or at scale, you know, the cost

357
00:17:46.759 --> 00:17:48.960
of those like did you guys get into those, whether

358
00:17:49.000 --> 00:17:52.680
it's you know, in terms of making the underwriting decision,

359
00:17:52.720 --> 00:17:56.000
reviewing fairness, like just the difference like you were saying

360
00:17:56.000 --> 00:17:59.400
with especially with a small dollar loan, every time you

361
00:17:59.400 --> 00:18:01.640
have to involve a human in any of that really

362
00:18:02.079 --> 00:18:05.440
makes that kind of a loan or a decision much

363
00:18:05.480 --> 00:18:08.559
more expensive. So you're you're not as able to expend

364
00:18:08.799 --> 00:18:11.119
extend credit to as many customers as you might like.

365
00:18:11.240 --> 00:18:12.400
Yeah, we talked a lot about that.

366
00:18:12.480 --> 00:18:16.039
We talked about the value in automation and again in

367
00:18:17.200 --> 00:18:20.240
the value also in scale from a you know, even

368
00:18:20.279 --> 00:18:24.559
from a performance standpoint, the importance of diversification, and I

369
00:18:24.599 --> 00:18:28.599
think that, uh, you know again, in some ways we're

370
00:18:28.640 --> 00:18:30.920
fortunate with the n c U A and credit unions

371
00:18:30.960 --> 00:18:34.839
because they have offered this product area a long time.

372
00:18:34.880 --> 00:18:37.799
As you know, there's like the pal loan program in

373
00:18:37.880 --> 00:18:40.960
credit unions for small dollar loans. They've won't long offered

374
00:18:41.519 --> 00:18:44.880
you know, signature loans and so, you know, as opposed

375
00:18:44.920 --> 00:18:47.279
to some in the banking greune that mostly do commercial

376
00:18:47.400 --> 00:18:49.920
lending and might just be for the first time dipping

377
00:18:49.920 --> 00:18:53.559
their toe into consumer you know, uh, closed in installment

378
00:18:53.799 --> 00:18:55.839
like loans. That this is an area that is sort

379
00:18:55.839 --> 00:18:58.000
of bread and butter for a long time for credit unions,

380
00:18:58.279 --> 00:19:01.599
I think, and so that gives there is a little

381
00:19:01.640 --> 00:19:06.559
bit more expertise. I think one of the things that

382
00:19:06.680 --> 00:19:10.119
was probably compelling to hear were just the anecdotal examples

383
00:19:10.119 --> 00:19:12.839
where I could say that those that have kept a.

384
00:19:12.799 --> 00:19:15.599
Traditional program running alongside upstart have.

385
00:19:15.680 --> 00:19:20.799
Reported uniformly that that the AI version has outperformed right

386
00:19:21.000 --> 00:19:23.680
or that products where they've made this conversion to AI,

387
00:19:23.880 --> 00:19:26.960
or if they went with a prior vendor that wasn't

388
00:19:27.039 --> 00:19:29.799
using as much data or AI in a similar product

389
00:19:30.720 --> 00:19:33.000
like that, that that the AI version may not have

390
00:19:33.000 --> 00:19:35.400
been perfect, credit performance may not be perfect, you know,

391
00:19:35.480 --> 00:19:38.519
do the macro or whatever, but that it outperformed what

392
00:19:38.599 --> 00:19:42.160
they had been doing or what they were doing side

393
00:19:42.200 --> 00:19:44.279
by side with a different program. And so I think

394
00:19:45.559 --> 00:19:48.279
that was I think helpful for people to understand that,

395
00:19:48.680 --> 00:19:50.839
you know, there's a context to what they have to

396
00:19:50.920 --> 00:19:52.799
view if they if they go into a credit union,

397
00:19:53.440 --> 00:19:55.880
you know, they they need to understand like, hey, why

398
00:19:55.920 --> 00:19:58.559
don't I take a look at whether or not you know,

399
00:19:58.640 --> 00:20:02.440
a traditional portfolio is a scalable is it actually going

400
00:20:02.480 --> 00:20:05.039
to become anything five years from now that can produce

401
00:20:05.519 --> 00:20:08.240
you know, some some sort of healthy you know income,

402
00:20:08.599 --> 00:20:12.200
you know for the credit union or so that's one question.

403
00:20:12.319 --> 00:20:15.680
And and then also is it is it actually performing

404
00:20:15.759 --> 00:20:18.440
in a decent way from a credit performance stample.

405
00:20:19.200 --> 00:20:20.119
Yeah, that's a good point.

406
00:20:20.160 --> 00:20:23.720
That's one thing that typically helps skepticism, I think is

407
00:20:23.799 --> 00:20:26.359
data and experience. So we're starting to get more and

408
00:20:26.400 --> 00:20:29.799
more proof points out there, and good point about the

409
00:20:29.799 --> 00:20:32.960
side by side comparison as well, So hopefully that helps you.

410
00:20:33.279 --> 00:20:35.119
And there's data on that, as you know, Matt, we

411
00:20:35.160 --> 00:20:36.880
talked about the data from d v O one a

412
00:20:36.880 --> 00:20:39.279
little bit, so we were able to kind of share

413
00:20:39.319 --> 00:20:43.799
some data on upstarts performance and you know, while acknowledging

414
00:20:43.839 --> 00:20:46.559
that that every credit union, based on their region, may have,

415
00:20:46.680 --> 00:20:48.440
you know, variations are based on their.

416
00:20:48.599 --> 00:20:49.680
Hard credit criteria.

417
00:20:49.680 --> 00:20:51.559
That doesn't have to do with upstart, like they could

418
00:20:51.599 --> 00:20:55.079
have different different outcomes and experiences. But but the you know,

419
00:20:55.160 --> 00:20:57.440
the bottom line, if you look in the aggregate at

420
00:20:57.440 --> 00:21:01.279
the performance and the risk adjusted returns and so forth

421
00:21:01.319 --> 00:21:05.000
of the upstart portfolio, you know it's been quite strong

422
00:21:05.359 --> 00:21:07.119
through the cycle. And so we spend a little bit

423
00:21:07.119 --> 00:21:08.880
of time on the macro cycle and sort of the

424
00:21:08.880 --> 00:21:16.640
importance of allowing credit unions the opportunity to see programs

425
00:21:16.799 --> 00:21:19.079
perform on all sides of the cycle, right, and and

426
00:21:19.119 --> 00:21:21.640
the way that we encourage you know, institutions. Obviously we

427
00:21:21.680 --> 00:21:23.440
want them to do as much as we they can

428
00:21:23.519 --> 00:21:26.000
with us, but we're not pressuring them to do so

429
00:21:26.079 --> 00:21:28.720
this is their program. They can decide to scale it

430
00:21:28.799 --> 00:21:31.000
up and scale it down at whatever times. You know,

431
00:21:31.079 --> 00:21:33.160
we have our own macro tools. We talked a little

432
00:21:33.160 --> 00:21:36.640
bit about you MI and so forth, and so how

433
00:21:37.079 --> 00:21:40.920
how does the macro index help inform the way people

434
00:21:40.960 --> 00:21:44.599
can be thinking about about their their risk appetite at

435
00:21:44.640 --> 00:21:45.680
any given time.

436
00:21:45.720 --> 00:21:48.599
But you know, just as we you know, invest in the.

437
00:21:48.519 --> 00:21:51.720
Stock market, you dollar cost average, it's it's not always

438
00:21:51.759 --> 00:21:53.440
good to be stopping and starting.

439
00:21:53.480 --> 00:21:55.559
In general, it's better to see things through the cycle.

440
00:21:55.599 --> 00:21:57.799
And so we talked a little bit about about that

441
00:21:57.880 --> 00:22:01.759
and how ultimately, you know, being as inclusive as possible

442
00:22:01.799 --> 00:22:03.799
and seeing through the cycle is how we think people

443
00:22:03.839 --> 00:22:05.240
will probably do the best.

444
00:22:05.960 --> 00:22:08.599
Yeah, I totally agree. That's typically when people need it

445
00:22:08.640 --> 00:22:11.799
the most. Right access to these types of credit products

446
00:22:11.799 --> 00:22:16.240
as well, yeap, you know, with the credit unions, a

447
00:22:16.319 --> 00:22:18.759
lot of them being a regionally focused or bound within

448
00:22:18.839 --> 00:22:21.200
certain states or field and membership. Did you guys discuss

449
00:22:21.279 --> 00:22:22.759
or what's your point of view on some of the

450
00:22:22.799 --> 00:22:25.240
state regulation that's happening out there, or do you see

451
00:22:25.559 --> 00:22:28.960
like how out of things modulate between states versus a

452
00:22:29.079 --> 00:22:32.920
national federal level and thinking about lending from that lens.

453
00:22:33.119 --> 00:22:36.880
Yeah, I mean, like, and this is an interesting dynamic

454
00:22:36.880 --> 00:22:40.559
because we know in DC there's this big, you know,

455
00:22:40.839 --> 00:22:42.720
long term battle.

456
00:22:42.400 --> 00:22:44.119
Between banks and credit unions.

457
00:22:44.160 --> 00:22:48.519
And most recently you probably saw the CEO of the

458
00:22:48.519 --> 00:22:51.599
American Bankers Association, Rob Nichols, wrote to the head of

459
00:22:51.880 --> 00:22:55.160
the NCUA, Todd Harburn's, you know about acquisitions and the

460
00:22:55.200 --> 00:22:58.680
transparency because there's now so many credit unions buying small

461
00:22:58.720 --> 00:23:02.960
community banks, and you know, that's just the latest iteration

462
00:23:03.079 --> 00:23:05.079
of this. There there is a fight when I was

463
00:23:05.440 --> 00:23:07.640
on the hill about business lending, how much business what

464
00:23:07.720 --> 00:23:10.440
percentage of the balance sheet credit unions could use for

465
00:23:10.440 --> 00:23:13.240
commercial lending, which kind of impedes in their turf and

466
00:23:13.240 --> 00:23:15.680
and then of course the tax status of credit unions,

467
00:23:15.720 --> 00:23:18.799
and so there is this long term battle that's always

468
00:23:18.839 --> 00:23:23.240
gone on and and questions about the the the the

469
00:23:23.240 --> 00:23:26.480
the overall role of credit unions in their communities and

470
00:23:26.519 --> 00:23:31.200
for specific communities as opposed to broadly broader uh, you know,

471
00:23:31.359 --> 00:23:34.039
serving broader constituencies. We have a lot of national credit us.

472
00:23:34.079 --> 00:23:37.160
I I just made the point of diversification that credit

473
00:23:37.279 --> 00:23:41.200
unions in general will do better if they're able to

474
00:23:41.720 --> 00:23:45.160
draw you know, members from as broad an area as

475
00:23:45.200 --> 00:23:47.839
they possibly can, and then they'll do better if they

476
00:23:47.839 --> 00:23:53.119
can you know, have you know, non concentrated portfolios again

477
00:23:53.200 --> 00:23:55.680
broadly speaking, as much as they can. So those that

478
00:23:55.759 --> 00:23:58.319
are in certain states and operating within those state guidelines

479
00:23:58.359 --> 00:24:00.480
and and are are you know going to.

480
00:24:00.400 --> 00:24:02.519
Stay there for a variety of reasons, that can be fine.

481
00:24:02.559 --> 00:24:05.359
And then those that are national there's going to be

482
00:24:05.759 --> 00:24:09.319
you know, kind of a constant, I believe push towards

483
00:24:09.359 --> 00:24:12.039
the national model because of the Internet, right, I mean

484
00:24:12.119 --> 00:24:14.880
this is just if you know, if you can find

485
00:24:15.160 --> 00:24:18.440
new members, if you can find you know, new members

486
00:24:18.440 --> 00:24:23.440
through you know, a credit channel, and then somebody signs

487
00:24:23.519 --> 00:24:25.359
up for a deposit account, like you're going to want

488
00:24:25.400 --> 00:24:28.039
to do that if you can. And so I think

489
00:24:28.079 --> 00:24:31.200
it's going to become a long term sort of trend

490
00:24:31.200 --> 00:24:33.400
that you're going to see of more and more of

491
00:24:33.599 --> 00:24:36.440
you know, credit unions that are sophisticated enough to handle

492
00:24:36.480 --> 00:24:37.920
it or going to want to be able to have

493
00:24:37.960 --> 00:24:39.839
as wide a footprint as they possibly can.

494
00:24:41.160 --> 00:24:44.039
Yeah. That leads me to think of another topic I've

495
00:24:44.400 --> 00:24:45.920
been debating with a few people. I don't if I'll

496
00:24:45.960 --> 00:24:47.599
have enough time in this episode to get into it,

497
00:24:47.599 --> 00:24:50.519
but like the idea of a primary financial institution, like

498
00:24:50.559 --> 00:24:52.640
does that exist anymore? What you know, the role of

499
00:24:52.680 --> 00:24:56.519
these institutions, whether the banks or credit unions or community banks,

500
00:24:56.759 --> 00:24:59.640
you know, do they serve a customer or is there

501
00:24:59.640 --> 00:25:02.240
specialization and other products? And you know, no one really

502
00:25:02.279 --> 00:25:04.759
owns a customer, but you're you're kind of fighting for

503
00:25:04.799 --> 00:25:05.680
that next product.

504
00:25:06.799 --> 00:25:07.000
Yeah.

505
00:25:07.039 --> 00:25:09.359
Well, if I mean think, I think from a regulatory standpoint,

506
00:25:10.720 --> 00:25:15.400
you know, you want you want the credit union to

507
00:25:15.480 --> 00:25:19.599
be able to offer the types of products and experiences

508
00:25:19.640 --> 00:25:23.200
that can keep it healthy over the long haul. And

509
00:25:23.599 --> 00:25:27.799
if it can't do that, then it's gonna wither and

510
00:25:27.880 --> 00:25:31.400
you're gonna see, you know, the consolidation.

511
00:25:31.079 --> 00:25:33.799
And and all the challenges around that.

512
00:25:35.440 --> 00:25:37.880
You know. So I think that there is interest in

513
00:25:38.200 --> 00:25:40.759
a lot of this innovation from that standpoint, It's just

514
00:25:40.799 --> 00:25:44.400
that there's also that healthy skepticism of is this innovation

515
00:25:45.400 --> 00:25:47.440
serving the credit union and the consumer?

516
00:25:47.680 --> 00:25:47.880
Right?

517
00:25:47.960 --> 00:25:51.240
And so if you can show that it does both,

518
00:25:51.279 --> 00:25:52.720
and I think, you know, we have a lot of

519
00:25:52.720 --> 00:25:55.839
great data and studies to show that that Upstart, you know,

520
00:25:56.200 --> 00:25:58.240
is one of those companies that does both, but you.

521
00:25:58.240 --> 00:26:00.240
I think you need to be able to be prayer

522
00:26:00.319 --> 00:26:01.359
to show that, you know.

523
00:26:02.200 --> 00:26:08.000
Moving back, I think we're at an interesting moment in Washington,

524
00:26:08.079 --> 00:26:10.640
not only because of the election, right where we're at

525
00:26:10.640 --> 00:26:12.799
a moment where you know, we're going to have a

526
00:26:12.839 --> 00:26:16.480
new administration either way, which doesn't happen very often. Right,

527
00:26:16.519 --> 00:26:18.960
We're not you know, Joe Biden is not going to

528
00:26:19.000 --> 00:26:22.400
be president in January, so like that's a new administration.

529
00:26:22.519 --> 00:26:24.000
Is that that's a that's a big deal.

530
00:26:24.039 --> 00:26:26.480
So we're gonna have a new a new administration either

531
00:26:26.519 --> 00:26:32.359
way to spend time with talking about about financial technology,

532
00:26:32.480 --> 00:26:39.200
about AI and underwriting, about partnerships between supervised institutions and

533
00:26:39.279 --> 00:26:42.519
institutions that are regulated but aren't directly supervised in the

534
00:26:42.519 --> 00:26:47.240
same way. And and we're doing that against the backdrop

535
00:26:47.279 --> 00:26:51.519
where as we mentioned, they are all of these compelling

536
00:26:51.599 --> 00:26:55.079
reasons why, especially for any institution but the very biggest

537
00:26:55.119 --> 00:26:58.319
of the of the you know, the largest institution of

538
00:26:58.319 --> 00:27:00.920
the country that build a lot of amazing products in

539
00:27:01.000 --> 00:27:04.000
house for all, but that you know, handful of companies.

540
00:27:04.039 --> 00:27:05.799
Partnerships are going to be the way of various kinds

541
00:27:05.799 --> 00:27:09.119
of way that that a lot of things move forward.

542
00:27:10.519 --> 00:27:13.880
And yet you have, uh, you know, there's.

543
00:27:13.720 --> 00:27:16.240
Negative there's been negative headlines, not in the lending area

544
00:27:16.319 --> 00:27:18.920
so much, but on you know, things like synapse bankruptcy

545
00:27:18.960 --> 00:27:22.839
and others. Right now, there's a request for information from

546
00:27:23.000 --> 00:27:27.359
the federal regulators about third party risk management at partnerships.

547
00:27:27.599 --> 00:27:30.359
We at Upstart will be responding to that. American Fintic

548
00:27:30.400 --> 00:27:31.559
Council will be responding to that.

549
00:27:32.160 --> 00:27:35.599
And so you know, we're also making the point all

550
00:27:35.640 --> 00:27:37.799
of these forums, whether it's the NCAA Examiner's Forum or

551
00:27:37.839 --> 00:27:39.680
when we meet with the OCC or the fd i

552
00:27:39.720 --> 00:27:41.359
C or the ten of these conferences, we're spending a

553
00:27:41.400 --> 00:27:45.079
lot of time talking about the value. And then we

554
00:27:45.200 --> 00:27:47.839
also have to bring the examples where the rovert meet

555
00:27:47.839 --> 00:27:52.440
throat of where we think a regulation or implementation or

556
00:27:52.519 --> 00:27:57.279
regulation or supervisory question or expectation may be so onerous

557
00:27:57.319 --> 00:27:59.799
that it is going to disrupt a partnership, right, And

558
00:27:59.839 --> 00:28:03.400
so so that's not where anybody wants to be. You know,

559
00:28:03.440 --> 00:28:06.559
nobody wants to be complaining, but you have to be

560
00:28:06.640 --> 00:28:10.559
willing to advocate on behalf of partnerships that have the

561
00:28:10.640 --> 00:28:13.559
chance to really move the whole financial system forward in

562
00:28:13.599 --> 00:28:16.160
a productive way, both for the institutions for the consumer.

563
00:28:17.680 --> 00:28:18.759
Yeah, totally agree.

564
00:28:18.880 --> 00:28:21.279
And I wonder not how did you leave Phoenix in

565
00:28:21.359 --> 00:28:23.440
that meeting with the examiners. Do you feel like they

566
00:28:23.480 --> 00:28:26.160
were bullish on the idea of partnerships and you know,

567
00:28:26.559 --> 00:28:30.079
companies being able to help credit unions in that sustainability

568
00:28:30.160 --> 00:28:32.519
like you mentioned, I think they were bullish.

569
00:28:32.640 --> 00:28:36.160
I think that they though want to think about how

570
00:28:36.200 --> 00:28:40.359
to how to tease out whether an institution like up

571
00:28:40.359 --> 00:28:43.480
Start with a lot of resources, is the partner and

572
00:28:43.559 --> 00:28:46.720
able you know, thousands of employees able to able to

573
00:28:46.759 --> 00:28:49.880
support a partnership in a viable long term way, versus

574
00:28:50.640 --> 00:28:53.000
a vendor that might come with a lot of promises

575
00:28:53.039 --> 00:28:56.319
and not the ability to like really deliver, right. And

576
00:28:56.400 --> 00:29:01.359
so I think they are going to be continuing to

577
00:29:01.480 --> 00:29:04.720
bring scrutiny, but I think it's going to be with

578
00:29:04.880 --> 00:29:08.279
more knowledge of Upstart, more and more understanding of how

579
00:29:08.279 --> 00:29:12.480
the partnership works, and more understanding of the whole reason why.

580
00:29:13.000 --> 00:29:15.680
I think I mentioned just in the context of open banking,

581
00:29:15.720 --> 00:29:18.359
like why why even why you know, why did you

582
00:29:18.440 --> 00:29:20.759
take their call? Like why are you spending time? Why

583
00:29:20.799 --> 00:29:23.960
are you thinking about it? And understanding the many problems

584
00:29:23.960 --> 00:29:28.279
that we're solving for a smaller institution, from finding the member,

585
00:29:29.160 --> 00:29:31.400
helping do it with the underwriting, helping with the pricing,

586
00:29:31.400 --> 00:29:36.279
helping with the servicing, help you know, supporting fair lending, testing, uh,

587
00:29:36.440 --> 00:29:41.759
all of that rigor, and ultimately you know, keeping the

588
00:29:41.880 --> 00:29:44.960
viability and health of the institutions that side of the

589
00:29:45.000 --> 00:29:46.839
balance sheet, you know, not the deposit so much, but

590
00:29:46.880 --> 00:29:48.920
the credit side of the balance that you know that

591
00:29:49.960 --> 00:29:53.839
in a in a in a good posture. So that

592
00:29:53.839 --> 00:29:56.920
that I think was the time well spent. Let alone,

593
00:29:56.920 --> 00:29:58.759
a lot of the time spent on third party risk

594
00:29:58.799 --> 00:30:03.000
management and don more of the in the weeds regulatory conversation.

595
00:30:04.480 --> 00:30:05.240
Yeah, totally.

596
00:30:06.079 --> 00:30:08.480
Well that it's always great catching up. I'm always impressed

597
00:30:08.480 --> 00:30:10.519
to you. You have to know so much about so

598
00:30:10.599 --> 00:30:13.279
many different things and how upstart operates in the space

599
00:30:13.720 --> 00:30:16.240
and the regulations that impact it. I wonder like, is

600
00:30:16.279 --> 00:30:18.599
there a source of information you can share with us,

601
00:30:18.640 --> 00:30:20.559
like that helps you keep up to date, whether it's

602
00:30:20.920 --> 00:30:23.359
you know, the points of view of some of these agencies,

603
00:30:23.960 --> 00:30:27.240
the news, the feelings of politicians. Like again, like so

604
00:30:27.279 --> 00:30:30.279
many variables can go into impacting how these things turn out,

605
00:30:30.319 --> 00:30:31.960
Like how do you ye're a source?

606
00:30:32.200 --> 00:30:32.960
I think it's interesting.

607
00:30:33.000 --> 00:30:33.880
I think it's interesting.

608
00:30:34.240 --> 00:30:36.680
I get that question a lot when I when I speak,

609
00:30:37.000 --> 00:30:37.960
Where do you get your news?

610
00:30:37.960 --> 00:30:40.960
How do you how do you keep on top of

611
00:30:40.960 --> 00:30:41.319
all of it?

612
00:30:42.920 --> 00:30:46.680
I think a lot of it is accumulated, and I think,

613
00:30:46.880 --> 00:30:48.960
you know, LinkedIn is a great example, you know, when

614
00:30:49.000 --> 00:30:52.000
you've developed a career's worth of relationships and then you're

615
00:30:52.000 --> 00:30:55.519
able to follow your network and curate your network.

616
00:30:55.440 --> 00:30:58.880
Based on what your you know, interests are for upstart,

617
00:30:59.079 --> 00:31:01.359
and then you're able of the track a lot of

618
00:31:01.359 --> 00:31:04.960
what's keeping your network either up at night or what

619
00:31:05.000 --> 00:31:06.759
they're spending that their time on. That can be an

620
00:31:06.759 --> 00:31:10.759
incredible source of news. So rather than what some editor

621
00:31:11.200 --> 00:31:15.079
it's almost like the ultimate crowdsourcing of of information, right,

622
00:31:15.160 --> 00:31:17.519
rather than what some editor at the New York Times

623
00:31:17.599 --> 00:31:19.640
or the Wall Street Journal, or the Financial Times or

624
00:31:20.039 --> 00:31:23.599
or Fintech Daily decided they thought was important you know,

625
00:31:23.640 --> 00:31:26.440
you read all those but then you also have you know,

626
00:31:26.599 --> 00:31:29.799
what those who are in the supervised community, like, you

627
00:31:29.839 --> 00:31:32.559
know your potential partners, like what they're posting about AI.

628
00:31:33.279 --> 00:31:34.880
You have the advocates, what.

629
00:31:34.720 --> 00:31:37.039
They're posting, the you know, the civil rights community, and

630
00:31:37.079 --> 00:31:39.400
you're able to kind of track, you know, in a

631
00:31:39.720 --> 00:31:42.440
real time what is bubbling and try to see a

632
00:31:42.440 --> 00:31:45.640
little bit around the corner of what's coming in a

633
00:31:45.680 --> 00:31:48.759
way that that wasn't probably possible before that platform. So

634
00:31:48.799 --> 00:31:51.960
I do think that that's that's an incredibly powerful one

635
00:31:52.079 --> 00:31:54.799
for the type of tracking that you're that you're talking about.

636
00:31:56.240 --> 00:31:59.480
Yeah, that's a good, good point. Yeah, LinkedIn with all

637
00:31:59.480 --> 00:32:02.920
its worth, and like self appointed gurus like you filter

638
00:32:03.000 --> 00:32:05.160
through that to your network of people and see what

639
00:32:05.200 --> 00:32:07.759
they care about and where they're finding information that's true.

640
00:32:08.599 --> 00:32:11.119
I mean, I just I just block the gurus like

641
00:32:11.200 --> 00:32:14.039
I It's it's like people that I know that are

642
00:32:14.480 --> 00:32:17.079
you know, they were just at Treasury or they were

643
00:32:17.119 --> 00:32:19.000
just at one of these regulatory bodies, and then they're

644
00:32:19.039 --> 00:32:21.440
posting something about something that they were wrestling with. Like

645
00:32:21.440 --> 00:32:23.200
that's where the magic is that people that are who

646
00:32:23.240 --> 00:32:26.920
are who are really employed in the in the workforce

647
00:32:27.160 --> 00:32:30.400
in the field that you're in and and really understanding

648
00:32:30.440 --> 00:32:31.119
what they're tracking.

649
00:32:31.160 --> 00:32:34.680
I think is is powerful. So thanks for the conversation, Matt.

650
00:32:34.720 --> 00:32:36.680
It's great and I think we'll be back at it

651
00:32:36.720 --> 00:32:39.119
maybe later this fall with a with the new sets,

652
00:32:39.160 --> 00:32:40.400
but it's fun to talk a little bit about the

653
00:32:40.519 --> 00:32:43.039
NCUA and the event in Phoenix.

654
00:32:43.559 --> 00:32:45.039
Definitely thanks again for joining that.

655
00:32:45.839 --> 00:32:48.359
I'll start partners with banks and credit unions to grow

656
00:32:48.400 --> 00:32:52.319
households and expand consumer lending through its leading AI lending platform.

657
00:32:53.000 --> 00:32:56.079
I'll start powered banks and credit unions leverage AI to

658
00:32:56.160 --> 00:32:59.440
offer higher approval rates and experience lower loss rates, while

659
00:32:59.480 --> 00:33:04.319
simultaneule delivering the exceptional digital first lending experience that consumers demand.

660
00:33:04.839 --> 00:33:07.640
Whether you're looking to grow and enhance your existing personal

661
00:33:07.640 --> 00:33:11.440
and auto lending programs or you're just getting started, Upstart

662
00:33:11.480 --> 00:33:14.759
can help. Upstart offers an end to end solution that

663
00:33:14.799 --> 00:33:17.640
can help you find more credit worthy borrowers within your

664
00:33:17.720 --> 00:33:23.119
risk profile. With all digital underwriting, verification, loan closing, and servicing,

665
00:33:23.559 --> 00:33:26.519
It's all possible with Upstart in your corner. Learn more

666
00:33:26.559 --> 00:33:30.359
about finding new borrowers, enhancing your credit decisioning process, and

667
00:33:30.440 --> 00:33:34.200
growing your business by visiting upstart dot com slash lenders.

668
00:33:34.599 --> 00:33:37.279
That's upstart dot com slash lenders.