WEBVTT
1
00:00:00.120 --> 00:00:02.839
I wish we had three percent thirty year mortgages again.
2
00:00:02.919 --> 00:00:05.040
We may never see those again, Like this is a
3
00:00:05.080 --> 00:00:06.719
new Normally, this type.
4
00:00:06.480 --> 00:00:09.000
Of environment could also last for an extended period. And
5
00:00:09.080 --> 00:00:13.000
this is what it looks like for five eight years.
6
00:00:13.199 --> 00:00:15.560
I almost don't see that changing all that much in
7
00:00:15.640 --> 00:00:17.679
twenty six, and then we're going to get into twenty seven,
8
00:00:17.679 --> 00:00:19.519
and there's gonna be election in twenty eight and who knows,
9
00:00:19.600 --> 00:00:20.239
right will.
10
00:00:20.039 --> 00:00:21.879
AI take jobs? It will take some, but it will
11
00:00:21.920 --> 00:00:24.359
create others. How does that impact the economy?
12
00:00:28.600 --> 00:00:31.000
All right, welcome to another episode of leaders in Lending. Today,
13
00:00:31.000 --> 00:00:34.119
we're going to talk about the twenty twenty five credit crunch.
14
00:00:34.200 --> 00:00:38.520
As we all know, most institutions are tightening credit, probably
15
00:00:38.560 --> 00:00:41.000
not for the best, but it's been something that's been
16
00:00:41.039 --> 00:00:44.439
ongoing now for the remitt for the entire year. Barry
17
00:00:44.439 --> 00:00:46.679
would love to get your perspective on kind of what's
18
00:00:47.039 --> 00:00:50.000
what's driving the titan credit in twenty twenty five.
19
00:00:50.560 --> 00:00:55.840
Overall, I think just uncertainty. Right since we've had a
20
00:00:55.840 --> 00:00:59.520
new administration come in in January, there were threats of
21
00:01:00.079 --> 00:01:02.600
reductions in government workforce, which some have come to pass
22
00:01:02.600 --> 00:01:05.519
and some haven't. The regulatory environment has sort of been
23
00:01:05.560 --> 00:01:10.000
turned a little bit upside down. The CFPB doesn't seem
24
00:01:10.040 --> 00:01:14.879
to have as much maybe forward influence as an expectation
25
00:01:14.959 --> 00:01:17.439
than maybe what they had in the past. But then,
26
00:01:17.599 --> 00:01:20.519
just in the economy, we've been waiting for a recession
27
00:01:20.599 --> 00:01:23.239
hasn't come, been waiting for rates to reduce, hasn't come,
28
00:01:23.400 --> 00:01:27.239
been waiting for unemployment to climb, hasn't necessarily I mean,
29
00:01:27.280 --> 00:01:28.879
it's come up a little bit, but the labor market
30
00:01:28.920 --> 00:01:32.159
is still really strong. So all the sort of fundamental
31
00:01:32.200 --> 00:01:39.640
economic conditions that would normally predate a credit crunch haven't
32
00:01:39.640 --> 00:01:42.959
necessarily come to pass. So it's been more about our
33
00:01:43.000 --> 00:01:47.000
expectation or worry that's coming. And that's actually a prudent
34
00:01:47.040 --> 00:01:51.159
approach in the face of uncertainty. May not necessarily be
35
00:01:51.359 --> 00:01:53.120
the time to take on a whole lot of risk,
36
00:01:53.519 --> 00:01:56.159
Although I'll argue with myself a little bit on this.
37
00:01:56.200 --> 00:01:59.280
The counterpoint is, if everyone else is pulling back, isn't
38
00:01:59.319 --> 00:02:01.159
that a really good time for you to pull forward
39
00:02:01.400 --> 00:02:05.959
in a in an aggressive but in somewhat measured fashion.
40
00:02:06.239 --> 00:02:09.199
When I say measured meaning more about the quality of risk.
41
00:02:10.639 --> 00:02:13.599
You know, maybe that's something that that where some credit
42
00:02:13.639 --> 00:02:15.240
unions and banks could win on.
43
00:02:15.479 --> 00:02:19.199
Sure, and I definitely think you know, geopolitical uncertainty as well.
44
00:02:19.240 --> 00:02:22.000
I mean you touched on the regulatory and the economic uncertainty,
45
00:02:22.120 --> 00:02:25.280
but you know, we continue to see turmoil in different
46
00:02:25.280 --> 00:02:28.159
areas of the world, and you know, particularly in the
47
00:02:28.159 --> 00:02:31.439
Middle East. And then you know tariffs on off do
48
00:02:31.479 --> 00:02:33.039
we have tariff do we not? On what products?
49
00:02:33.560 --> 00:02:33.680
Uh?
50
00:02:33.960 --> 00:02:37.080
And those rollbacks and the impact of that just continues
51
00:02:37.120 --> 00:02:41.039
to We've got uncertainty almost in every area of things
52
00:02:41.039 --> 00:02:43.080
that normally would be kind of part of the fundamentals
53
00:02:43.120 --> 00:02:44.639
of driving those decisions.
54
00:02:44.960 --> 00:02:47.879
Everyone's been waiting for a rate cut, right stock market up,
55
00:02:48.439 --> 00:02:51.159
labor market's still pretty good. Inflation is running around three
56
00:02:51.199 --> 00:02:53.960
percent right now. I know the Fed has won it
57
00:02:54.039 --> 00:02:55.879
at two for a while, but we haven't been there
58
00:02:55.919 --> 00:02:58.360
for a long time. Like, there's a lot of conditions
59
00:02:58.800 --> 00:03:01.840
that traditionally wouldn't call for a rate cut. And I
60
00:03:01.879 --> 00:03:05.879
think as as consumers, we all want to see a
61
00:03:05.919 --> 00:03:09.000
rate cut. Even in our business, certainly a rate cut
62
00:03:09.039 --> 00:03:13.599
would help to create maybe more in demand. And I
63
00:03:13.639 --> 00:03:16.759
think as a as someone who manages the balance sheet,
64
00:03:16.759 --> 00:03:20.120
if you're a CFO today, you're thinking about, Okay, can
65
00:03:20.199 --> 00:03:22.879
I put more of the higher rate business on my
66
00:03:22.879 --> 00:03:26.240
books now in the face of perhaps seeing the lower
67
00:03:26.319 --> 00:03:29.120
rates going forward, you know, maybe that's that's suprett approach to.
68
00:03:29.039 --> 00:03:32.879
Take everything's not ebbing and flowing like it should, right,
69
00:03:33.400 --> 00:03:37.719
So you think about there's many different reasons to tighten credit.
70
00:03:37.759 --> 00:03:41.919
You talked about geopolitical, We're talking you know, there's delinquencies
71
00:03:41.919 --> 00:03:44.319
that are rising. Do you think there's one honed in
72
00:03:44.520 --> 00:03:47.479
area that is driving the crunch or is it just
73
00:03:47.520 --> 00:03:48.840
a holistic view of all of them.
74
00:03:48.879 --> 00:03:52.159
Do you think about the you know, financial crisis in
75
00:03:52.240 --> 00:03:58.599
LA the GFC. Everything felt unprecedented right after, and it
76
00:03:58.639 --> 00:04:00.599
was this idea, you know, of the new normal, like
77
00:04:01.000 --> 00:04:04.159
the world has changed, and I think we're there's a
78
00:04:04.199 --> 00:04:06.560
lot of conversations like this where people are saying, how
79
00:04:06.560 --> 00:04:08.039
are we going to get back to that? Like how
80
00:04:08.039 --> 00:04:10.840
are we going to get to less uncertainty into more
81
00:04:10.879 --> 00:04:14.199
stability and to being able to see those you know,
82
00:04:14.560 --> 00:04:17.879
leading and trailing indicators and use them to make decisions.
83
00:04:17.959 --> 00:04:20.240
And it could be that we don't. And it could
84
00:04:20.279 --> 00:04:22.600
be that this is similar to how zero interest rates
85
00:04:22.600 --> 00:04:24.319
we thought that was a short term thing and it
86
00:04:24.399 --> 00:04:28.639
lasted for an extended period. This type of environment could
87
00:04:28.680 --> 00:04:31.240
also last for an extended period, and maybe this is
88
00:04:31.360 --> 00:04:36.240
what it looks like for five eight years, and there's
89
00:04:36.319 --> 00:04:39.079
not going to be a sudden return to the same
90
00:04:39.160 --> 00:04:41.639
fundamentals of what we thought were fundamentals before. There are
91
00:04:41.680 --> 00:04:43.360
now new fundamentals.
92
00:04:43.519 --> 00:04:46.720
Right, this is a new normal basically, right. Yeah, there's
93
00:04:46.759 --> 00:04:50.800
not been an event like COVID and the pandemic that
94
00:04:51.079 --> 00:04:53.360
sort of caused that really brief recession we add in
95
00:04:53.399 --> 00:04:57.120
twenty twenty. There's not been an event like the mortgage
96
00:04:57.120 --> 00:05:00.959
backed securities and the housie market kind of floating back
97
00:05:01.000 --> 00:05:04.079
in two thousand and eight. I laughed, because of the
98
00:05:04.120 --> 00:05:06.519
conditions at the time, it was very apparent to while
99
00:05:06.519 --> 00:05:08.759
you know, we were all in the business at that time,
100
00:05:09.240 --> 00:05:14.040
and you know, certainly in California we could see if
101
00:05:14.639 --> 00:05:17.439
as as a money mooning quarterback looking back at everything
102
00:05:17.439 --> 00:05:19.399
that was happening around the housing markets, like, yeah, of
103
00:05:19.439 --> 00:05:22.600
course this was unsustainable, right, But there's nothing like that
104
00:05:22.720 --> 00:05:24.199
happening right now.
105
00:05:24.240 --> 00:05:26.000
There's no one thing to see.
106
00:05:26.199 --> 00:05:27.879
And so to your point, this is somewhat of the
107
00:05:27.959 --> 00:05:31.759
new normal. You know, I hear from people saying, oh boy,
108
00:05:32.079 --> 00:05:35.040
I wish we had three percent thirty year mortgages again.
109
00:05:35.279 --> 00:05:37.680
We may never see those again, Like this is the
110
00:05:37.759 --> 00:05:39.920
new normal. If you're paying six or seven percent for
111
00:05:40.079 --> 00:05:44.120
thirty year mortgage, that's probably all as good as you're
112
00:05:44.120 --> 00:05:44.519
going to get.
113
00:05:44.639 --> 00:05:46.560
It's not like early eighties numbers. I mean, we're not
114
00:05:46.680 --> 00:05:48.920
you know, we could be paying fifteen percent seventeen.
115
00:05:48.959 --> 00:05:50.639
It's more like the nineties. You know, when I bought
116
00:05:50.639 --> 00:05:52.319
my first house, I think, you know, at my rate
117
00:05:52.360 --> 00:05:54.240
was somewhere around in the eighth and I was happy
118
00:05:54.240 --> 00:05:56.720
with it and that was just just the way it was.
119
00:05:56.800 --> 00:05:59.000
So do you think the ongoing uncertainty though, Like you
120
00:05:59.040 --> 00:06:01.959
think about talk to lending professionals on a daily basis,
121
00:06:02.000 --> 00:06:04.199
and they've been telling us that inflation is going to arise,
122
00:06:04.279 --> 00:06:06.720
We're tightening credit. They've been saying it now for almost
123
00:06:06.800 --> 00:06:09.160
three years, and we haven't really got to that that
124
00:06:09.279 --> 00:06:12.600
it factor similar to what the GFC was. Do you
125
00:06:12.600 --> 00:06:14.480
think that that's going to be the new Norman because
126
00:06:14.480 --> 00:06:17.800
there's so much tightening, it's just going to keep translating
127
00:06:17.839 --> 00:06:20.240
as is. Or do you think if we see a
128
00:06:20.279 --> 00:06:22.879
surge of untightened that things would pop or what are
129
00:06:22.879 --> 00:06:23.480
your thoughts there?
130
00:06:23.600 --> 00:06:26.199
I mean, you just sort of introduced maybe a concept
131
00:06:26.360 --> 00:06:30.319
of this arbitrary tightening across the industry, maybe that is
132
00:06:30.360 --> 00:06:34.399
somewhat governing things today, that that's why we're not seeing
133
00:06:34.800 --> 00:06:38.800
abundant losses. And I'll go back again to the PREGFC
134
00:06:38.879 --> 00:06:43.600
with housing, where it was stated income applications and no
135
00:06:43.720 --> 00:06:46.480
down payments and things like this, like all those sort
136
00:06:46.519 --> 00:06:51.639
of risky conditions somewhat cavalier underwrite standards that went into
137
00:06:51.680 --> 00:06:54.879
that that anyone can get a mortgage right And perhaps
138
00:06:54.959 --> 00:07:01.360
now to your point, drew that by lenders tightening more
139
00:07:01.360 --> 00:07:04.800
and more, that there's just less of a less of
140
00:07:04.959 --> 00:07:09.439
a risk out there of this becoming something that would
141
00:07:09.480 --> 00:07:13.000
be beyond the risk of what might other rest be expected.
142
00:07:13.199 --> 00:07:15.160
Well, there's such a reaction too, I mean, there was
143
00:07:15.199 --> 00:07:18.920
such a I guess penalization, and I mean people could
144
00:07:18.920 --> 00:07:21.600
people have different opinions of whether there was enough penalization
145
00:07:21.639 --> 00:07:24.519
of the financial industry after or you know, and how
146
00:07:24.600 --> 00:07:29.800
much the ownership the consumer versus the institutions had. But
147
00:07:30.240 --> 00:07:32.879
like there was there were certainly a lot of repercussions
148
00:07:32.920 --> 00:07:35.600
and consequences, and like DoD frank and and a lot
149
00:07:35.600 --> 00:07:38.959
of other regulation Federal Reserve, Triprairie repro reform came out
150
00:07:39.000 --> 00:07:42.920
of came out of the GFC, and so it is
151
00:07:42.959 --> 00:07:44.959
possible that a lot of those things are actually working
152
00:07:45.240 --> 00:07:49.040
well and better than maybe we think, because maybe we're
153
00:07:49.079 --> 00:07:50.839
all kind of waiting for that other shoot to drop
154
00:07:50.839 --> 00:07:51.519
and maybe it's.
155
00:07:51.439 --> 00:07:53.560
It's not Maybe there isn't another stree, it will.
156
00:07:53.399 --> 00:07:55.600
Not drop right, or maybe it'll drop in a different
157
00:07:55.600 --> 00:07:58.399
way next time that hasn't happened before.
158
00:07:58.560 --> 00:08:01.120
So then maybe we put our foot on the gas again,
159
00:08:01.720 --> 00:08:04.319
not all the way to the floor, right, but take
160
00:08:04.360 --> 00:08:05.720
it a little bit less off the brake, a little
161
00:08:05.759 --> 00:08:08.360
more on the gas than maybe that would be That
162
00:08:08.399 --> 00:08:09.000
would be a.
163
00:08:08.920 --> 00:08:10.920
Good approach right now, Yeah, I know we've talked a
164
00:08:10.920 --> 00:08:14.160
lot about this and prior podcast of leveraging the use
165
00:08:14.160 --> 00:08:17.120
of AI models to predict risk more effectively. Right, so
166
00:08:17.120 --> 00:08:18.759
you think about you wanted to take your foot off
167
00:08:18.759 --> 00:08:20.839
the brake and press the gas a little bit again,
168
00:08:21.720 --> 00:08:23.800
And going back to a prior podcast, we're giving more
169
00:08:23.800 --> 00:08:26.519
of a description of how to have the knowledge base
170
00:08:26.600 --> 00:08:29.720
to explain having these types of partnerships. But in actuality,
171
00:08:29.759 --> 00:08:33.399
about seventy percent of financial institutions do use some form
172
00:08:33.399 --> 00:08:37.320
of AI driven modeling for decisioning credit right, So how
173
00:08:37.360 --> 00:08:41.759
do you think financial institutions are adapting those models to
174
00:08:41.840 --> 00:08:44.039
have less risk or even to open things up and
175
00:08:44.080 --> 00:08:46.879
be more predictive and bringing in the types of borrowers
176
00:08:47.159 --> 00:08:48.440
that doesn't add the risk there.
177
00:08:48.679 --> 00:08:51.279
So models are only good as good as a data
178
00:08:51.320 --> 00:08:54.360
that's that's in them. And so what sort of data
179
00:08:54.639 --> 00:08:58.639
is being used to train those models? Is lending data
180
00:08:58.639 --> 00:09:01.360
from the past, call it ten years, twelve years? Now,
181
00:09:01.360 --> 00:09:03.039
what's been happening in the last ten to twelve years?
182
00:09:03.039 --> 00:09:07.399
Who went from zero from ZERP into you know, a
183
00:09:07.440 --> 00:09:10.120
little recession around COVID and you could almost sort of
184
00:09:10.120 --> 00:09:12.720
throw out that little eighteen month period because it just
185
00:09:12.879 --> 00:09:17.639
was unprecedented in terms of what happened with deposits and
186
00:09:17.639 --> 00:09:20.759
then you know it's kind of dropping again. So these
187
00:09:20.799 --> 00:09:23.519
models are trained on that data over that time period,
188
00:09:24.600 --> 00:09:28.279
which if we are in a new normal, then wouldn't
189
00:09:28.320 --> 00:09:32.519
that mean that the data that's in those models is
190
00:09:32.559 --> 00:09:37.320
actually going to be maybe a better predictive value than
191
00:09:37.480 --> 00:09:40.440
maybe data that was happening before the GFC for example.
192
00:09:40.519 --> 00:09:42.600
Sure that, yeah, like you need to need to actually
193
00:09:42.679 --> 00:09:44.480
look at them more like have a little bit more
194
00:09:44.519 --> 00:09:48.200
recency bias, like intentionally in how you're thinking about training
195
00:09:48.240 --> 00:09:50.039
a model, because I think there's a lot of times,
196
00:09:50.360 --> 00:09:52.639
you know, like COVID's a great example that people looked
197
00:09:52.679 --> 00:09:53.919
back at that and said, oh, well that was an
198
00:09:53.960 --> 00:09:57.000
unusual environment. Well maybe it was, but it also happened
199
00:09:57.039 --> 00:10:00.960
in real life. So things that are unusual environments actually occurred.
200
00:10:01.320 --> 00:10:04.080
So you can't skip it right in your model.
201
00:10:04.120 --> 00:10:05.639
Can't just toss that data.
202
00:10:05.720 --> 00:10:08.320
Yeah, Like you have to understand that there were you know,
203
00:10:08.559 --> 00:10:13.720
there were liquidity injections from the federal government and deposit
204
00:10:13.799 --> 00:10:16.279
increases and also with a zero interest rate. So it
205
00:10:16.320 --> 00:10:20.200
was unusual, but it wasn't but it actually occurred, And
206
00:10:20.279 --> 00:10:23.799
so like, how do you incorporate that in thinking about
207
00:10:24.039 --> 00:10:26.399
what could happen next and how to plan for it.
208
00:10:26.480 --> 00:10:28.360
I think most often, too, like we've seen it over
209
00:10:28.399 --> 00:10:31.200
the past few years, the resiliency of the American consumers
210
00:10:31.480 --> 00:10:34.200
is there. They figure it out whether a piece of
211
00:10:34.200 --> 00:10:37.480
paper tells us inflation has increased x percent, like, things
212
00:10:37.519 --> 00:10:40.759
have stayed fairly stable, at least to my knowledge base,
213
00:10:41.200 --> 00:10:43.159
to be able to be able to open up the
214
00:10:43.240 --> 00:10:45.919
risk the risk door a little bit more so thinking
215
00:10:45.919 --> 00:10:49.480
about you know, most institutions in some face that are
216
00:10:49.559 --> 00:10:53.720
leveraging an AI driven model for predictive risk. Is there
217
00:10:53.759 --> 00:10:57.360
a way that credit unions can expand their approval without
218
00:10:57.399 --> 00:11:00.919
adding additional risk outside of leverage the use of models.
219
00:11:01.399 --> 00:11:03.840
So outside of using models, like, how else can they
220
00:11:04.039 --> 00:11:06.320
kind of expand their approval? I mean, I certainly think
221
00:11:06.879 --> 00:11:10.200
you know, a key theme automation and it doesn't have
222
00:11:10.240 --> 00:11:12.399
to be predictive AI. Right, Like we talk a lot
223
00:11:12.440 --> 00:11:15.480
about underwriting, which is more of a predictive AI, but
224
00:11:16.039 --> 00:11:20.440
you know, thinking about using generative AI agent AI chat
225
00:11:20.440 --> 00:11:27.080
agents to improve their customer member experience, improve the digital process.
226
00:11:27.919 --> 00:11:29.679
I think a lot of those things can also drive
227
00:11:29.720 --> 00:11:32.360
improval because they reduce friction in the flow, They increase
228
00:11:32.440 --> 00:11:36.440
the chance of that new member being you know, satisfied
229
00:11:36.440 --> 00:11:39.320
and getting the product that they need. And that's not
230
00:11:39.480 --> 00:11:43.519
taking on any additional risk. It's actually reducing operational risk
231
00:11:44.440 --> 00:11:46.720
while increasing the likelihood of approvals.
232
00:11:47.279 --> 00:11:49.399
All right, So going back to the whole theme of
233
00:11:49.919 --> 00:11:52.559
different levers, So we talked about non model type of
234
00:11:52.600 --> 00:11:55.399
approvals leveraging models, which seems to be the kind of
235
00:11:55.399 --> 00:11:58.600
the new norm. Credit unions have their fingertips at a
236
00:11:58.679 --> 00:12:02.840
bulk of different ways of leveraging risk through participations through
237
00:12:02.840 --> 00:12:06.440
whole lowned purchases, through many different areas. So what are
238
00:12:06.440 --> 00:12:10.000
the smartest levers you think current state that credit unions
239
00:12:10.000 --> 00:12:13.120
can kind of pull to expand their access without adding