1 00:00:00,500 --> 00:00:01,968 [SQUEAKING] 2 00:00:01,968 --> 00:00:03,444 [RUSTLING] 3 00:00:03,444 --> 00:00:04,920 [CLICKING] 4 00:00:10,350 --> 00:00:13,085 GARY GENSLER: Welcome back to fintech. 5 00:00:16,250 --> 00:00:19,400 We're getting together for our fourth class, 6 00:00:19,400 --> 00:00:21,530 and we're going to get going now. 7 00:00:21,530 --> 00:00:28,830 Thank you for tying in online via this modern technology. 8 00:00:28,830 --> 00:00:35,430 Today we're going to move to important customer interfaces-- 9 00:00:35,430 --> 00:00:38,340 user interfaces and user experiences that 10 00:00:38,340 --> 00:00:44,460 have been greatly enhanced by open application program 11 00:00:44,460 --> 00:00:46,250 interfaces. 12 00:00:46,250 --> 00:00:48,810 Now, application program interfaces 13 00:00:48,810 --> 00:00:53,910 is not new to finance, and it didn't start in M&A finance. 14 00:00:53,910 --> 00:00:57,240 It's a basic concept of how one computer is 15 00:00:57,240 --> 00:00:59,340 talking to another computer. 16 00:00:59,340 --> 00:01:01,560 And again, this is not a computer science class, 17 00:01:01,560 --> 00:01:03,870 but we're going to spend a little bit of time 18 00:01:03,870 --> 00:01:08,130 on what application program interfaces are 19 00:01:08,130 --> 00:01:09,750 and what this concept of open API 20 00:01:09,750 --> 00:01:13,800 is and some of the companies around it 21 00:01:13,800 --> 00:01:16,710 and how companies have organized themselves, 22 00:01:16,710 --> 00:01:20,160 both in terms of finance, and a little bit even 23 00:01:20,160 --> 00:01:24,030 outside of finance and some of the challenges with regard 24 00:01:24,030 --> 00:01:25,710 to that. 25 00:01:25,710 --> 00:01:29,210 What API has also led to is a movement of 26 00:01:29,210 --> 00:01:32,620 and a greater concentration around companies 27 00:01:32,620 --> 00:01:34,210 called data aggregators. 28 00:01:34,210 --> 00:01:37,720 And we'll get into that as well. 29 00:01:37,720 --> 00:01:42,130 As I've sort of worked through with the class already, 30 00:01:42,130 --> 00:01:44,230 it's my thought that certain technologies 31 00:01:44,230 --> 00:01:48,250 have been moving into the finance technology stack, 32 00:01:48,250 --> 00:01:52,810 and are really part and parcel to what finance is in 2020. 33 00:01:52,810 --> 00:01:56,107 And I think open API, with its movement around the globe-- 34 00:01:56,107 --> 00:01:57,940 and some countries have moved further in it, 35 00:01:57,940 --> 00:02:00,880 some countries have moved less, some companies have moved 36 00:02:00,880 --> 00:02:03,360 further, some companies have moved less-- 37 00:02:03,360 --> 00:02:05,230 that this is a trend that will be 38 00:02:05,230 --> 00:02:09,490 fully incorporated into the technology stack of finance 39 00:02:09,490 --> 00:02:13,590 country by country, if it hasn't already been so. 40 00:02:13,590 --> 00:02:17,510 But it's been a major disruptor in the last five, seven years. 41 00:02:17,510 --> 00:02:20,050 And it will continue to be a bit of a disruptor 42 00:02:20,050 --> 00:02:22,150 for the next handful of years. 43 00:02:22,150 --> 00:02:25,610 By the end of this decade, we'll look back, 44 00:02:25,610 --> 00:02:27,940 we'll say that that's probably already happened, 45 00:02:27,940 --> 00:02:29,620 and there'll be enhancements around it. 46 00:02:29,620 --> 00:02:32,800 But we'll be fully inside of it. 47 00:02:32,800 --> 00:02:36,490 Maybe not as dramatic as artificial intelligence 48 00:02:36,490 --> 00:02:38,380 and machine learning to our minds-- 49 00:02:38,380 --> 00:02:41,230 thinking of machines extracting correlations 50 00:02:41,230 --> 00:02:43,490 better than we can. 51 00:02:43,490 --> 00:02:45,760 But I think it's fairly dramatic, 52 00:02:45,760 --> 00:02:48,160 in terms of the advancement of finance 53 00:02:48,160 --> 00:02:51,200 and all that we have in front of us. 54 00:02:51,200 --> 00:02:54,910 So that's why I decided to take one class and one lecture 55 00:02:54,910 --> 00:02:56,080 on it. 56 00:02:56,080 --> 00:02:57,400 I also want to thank everybody. 57 00:02:57,400 --> 00:02:59,620 Romain put out a survey. 58 00:02:59,620 --> 00:03:01,630 About a third of you took it. 59 00:03:01,630 --> 00:03:03,100 It was a voluntary survey. 60 00:03:03,100 --> 00:03:04,730 And it's helpful. 61 00:03:04,730 --> 00:03:11,650 And what I took back from the desires of students were 62 00:03:11,650 --> 00:03:13,990 could I do a little bit more talking about the companies 63 00:03:13,990 --> 00:03:16,640 in the fintech space? 64 00:03:16,640 --> 00:03:21,940 And maybe add, if I might say, a little less, whether it's 65 00:03:21,940 --> 00:03:27,040 on the broader technology or broader policy side. 66 00:03:27,040 --> 00:03:28,425 I'll do my best. 67 00:03:28,425 --> 00:03:29,800 But I also want to share with you 68 00:03:29,800 --> 00:03:31,870 why I think there's a reason to have 69 00:03:31,870 --> 00:03:34,930 some foundation-- some foundation of what 70 00:03:34,930 --> 00:03:38,200 the technologies are, or even some foundation 71 00:03:38,200 --> 00:03:40,600 on the public policy side, as we move 72 00:03:40,600 --> 00:03:43,130 into the latter part of this class. 73 00:03:43,130 --> 00:03:46,540 The latter part of this class is really to talk sector by sector 74 00:03:46,540 --> 00:03:49,790 what the competitive landscape is, which companies are up, 75 00:03:49,790 --> 00:03:52,910 which companies are down, and so forth. 76 00:03:52,910 --> 00:03:57,390 But as we've already talked about a bit today, 77 00:03:57,390 --> 00:03:59,200 the companies really matter. 78 00:03:59,200 --> 00:04:02,680 And I'll sort of interject that more as we go forward 79 00:04:02,680 --> 00:04:04,370 during this semester. 80 00:04:04,370 --> 00:04:07,630 One other question is whether we could have some guest speakers. 81 00:04:07,630 --> 00:04:10,850 I'm kind of going to have to apologize in advance. 82 00:04:10,850 --> 00:04:13,120 This is kind of what you get. 83 00:04:13,120 --> 00:04:17,480 But over the next handful of sessions, 84 00:04:17,480 --> 00:04:20,500 if I come up with somebody to bring in-- it's 85 00:04:20,500 --> 00:04:24,880 a little hard at this juncture to bring in guest speakers. 86 00:04:24,880 --> 00:04:27,820 But I certainly appreciate the request. 87 00:04:27,820 --> 00:04:31,960 And with four weeks to go, you never know. 88 00:04:31,960 --> 00:04:37,870 But with that, I'm going to start to share some slides. 89 00:04:37,870 --> 00:04:39,610 I don't know, Romain if you have anything 90 00:04:39,610 --> 00:04:44,172 you took from the feedback, which was very helpful. 91 00:04:44,172 --> 00:04:45,130 PROFESSOR 2: Thank you. 92 00:04:45,130 --> 00:04:48,165 Thank you very much for filling out the survey. 93 00:04:48,165 --> 00:04:48,790 That was great. 94 00:04:48,790 --> 00:04:51,490 Just two small announcements from my side. 95 00:04:51,490 --> 00:04:53,980 One, a reminder for your team formation. 96 00:04:53,980 --> 00:04:55,390 Friday is the deadline. 97 00:04:55,390 --> 00:04:59,227 Please register your teams in Canvas in the People's tab 98 00:04:59,227 --> 00:05:00,310 using the Groups function. 99 00:05:00,310 --> 00:05:01,300 If you don't know how to do that, 100 00:05:01,300 --> 00:05:02,710 I'll be happy to show you. 101 00:05:02,710 --> 00:05:05,650 So please register your teams of three to four students 102 00:05:05,650 --> 00:05:07,360 for the group assignments. 103 00:05:07,360 --> 00:05:09,840 And then second reminder, as per the syllabus, 104 00:05:09,840 --> 00:05:11,590 have one of your team members just send me 105 00:05:11,590 --> 00:05:14,860 an email saying what is the topic that you wish to study 106 00:05:14,860 --> 00:05:16,088 for your group assignments. 107 00:05:16,088 --> 00:05:17,380 Please don't forget to do that. 108 00:05:17,380 --> 00:05:20,440 The deadline is also Friday, April the 10th. 109 00:05:20,440 --> 00:05:21,790 Thank you. 110 00:05:21,790 --> 00:05:22,790 GARY GENSLER: Thank you. 111 00:05:22,790 --> 00:05:24,980 And part of that is just to really help 112 00:05:24,980 --> 00:05:27,540 you engage in this subject. 113 00:05:27,540 --> 00:05:31,570 To Romain and myself, it doesn't matter which sector you pick 114 00:05:31,570 --> 00:05:33,430 or whether you choose to sort of write 115 00:05:33,430 --> 00:05:37,660 your paper from the perspective of incumbents, to big tech, 116 00:05:37,660 --> 00:05:39,700 or to a venture capitalist. 117 00:05:39,700 --> 00:05:41,860 But it's really to help you sort of engage 118 00:05:41,860 --> 00:05:45,550 in the world of fintech, that which we're studying, 119 00:05:45,550 --> 00:05:47,410 and to pick one sector. 120 00:05:47,410 --> 00:05:51,670 And are you taking it from the entrepreneurial strategy 121 00:05:51,670 --> 00:05:53,260 of the disruptor, or are you taking it 122 00:05:53,260 --> 00:05:56,830 from a big firm, either incumbent or a tech firm's 123 00:05:56,830 --> 00:06:00,950 perspective in this regard? 124 00:06:00,950 --> 00:06:03,370 So as I said, we're going to talk a little bit 125 00:06:03,370 --> 00:06:07,240 about application program interface, how 126 00:06:07,240 --> 00:06:11,110 this fits into the history of finance, 127 00:06:11,110 --> 00:06:15,040 deeper dive in open API and open banking. 128 00:06:15,040 --> 00:06:17,410 You were all asked to prepare and sort 129 00:06:17,410 --> 00:06:20,040 of look at the Bank of International Settlements 130 00:06:20,040 --> 00:06:20,710 report. 131 00:06:20,710 --> 00:06:22,000 So I'm not going to dive in. 132 00:06:22,000 --> 00:06:24,010 I'm going to hopefully assume that you've had 133 00:06:24,010 --> 00:06:27,520 a chance to engage with that. 134 00:06:27,520 --> 00:06:30,790 And then talk a little bit also about data aggregators 135 00:06:30,790 --> 00:06:34,420 and a movement amongst many in large financial firms 136 00:06:34,420 --> 00:06:38,070 to have a standard setting group-- 137 00:06:38,070 --> 00:06:42,880 a financial data exchange on open APIs. 138 00:06:42,880 --> 00:06:46,150 And then try to close out a little bit on robotic process 139 00:06:46,150 --> 00:06:46,820 automation. 140 00:06:46,820 --> 00:06:48,430 So a lot to cover. 141 00:06:48,430 --> 00:06:50,270 But I think we'll keep it going. 142 00:06:50,270 --> 00:06:53,860 And I ask again, use the Participate 143 00:06:53,860 --> 00:06:56,470 button to raise the blue hands. 144 00:06:56,470 --> 00:06:58,840 Use the chat function. 145 00:06:58,840 --> 00:07:01,330 Romain's going to keep interrupting me. 146 00:07:01,330 --> 00:07:05,830 And we'll have hopefully a good session together. 147 00:07:05,830 --> 00:07:09,410 So the readings, again, Bank of International Settlements, 148 00:07:09,410 --> 00:07:14,050 this consortium of 60 or so central banks around the globe, 149 00:07:14,050 --> 00:07:16,000 stands up reports and studies from time 150 00:07:16,000 --> 00:07:19,360 to time on new challenges they're seeing. 151 00:07:19,360 --> 00:07:22,330 So just that they pick this topic to write a report that 152 00:07:22,330 --> 00:07:25,300 came out this past November of 2019 153 00:07:25,300 --> 00:07:28,600 is indicative right there in and of itself 154 00:07:28,600 --> 00:07:30,610 that it's a movement around the globe that they 155 00:07:30,610 --> 00:07:35,620 thought they needed to study, and an important aspect. 156 00:07:35,620 --> 00:07:38,170 I picked one article out of the many articles 157 00:07:38,170 --> 00:07:41,560 that you can grab off the internet about fintech 158 00:07:41,560 --> 00:07:42,610 for the next decade. 159 00:07:42,610 --> 00:07:45,010 And this was one article that I thought was relevant, 160 00:07:45,010 --> 00:07:47,500 because it really did also include 161 00:07:47,500 --> 00:07:51,610 this concept of the interface-- the customer user interface 162 00:07:51,610 --> 00:07:56,620 and what was going on with regard to APIs. 163 00:07:56,620 --> 00:07:58,630 And then one article, again, just 164 00:07:58,630 --> 00:08:02,950 to give you sort of a primer on robotic process automation 165 00:08:02,950 --> 00:08:07,410 that I think is relevant as well. 166 00:08:07,410 --> 00:08:13,370 So again, just if anybody wants to give a little short brief as 167 00:08:13,370 --> 00:08:17,270 to what are the major trends on the marketing side 168 00:08:17,270 --> 00:08:19,860 that's affecting financial services right now. 169 00:08:19,860 --> 00:08:22,070 And we can go broader than API, of course. 170 00:08:22,070 --> 00:08:26,622 But Romain anybody want to sort of contribute now? 171 00:08:26,622 --> 00:08:28,080 PROFESSOR 2: Let's see, who will be 172 00:08:28,080 --> 00:08:29,440 our first volunteer of the day? 173 00:08:33,705 --> 00:08:34,205 Michael. 174 00:08:40,429 --> 00:08:42,890 AUDIENCE: Not sure if this is exactly marketing side, 175 00:08:42,890 --> 00:08:46,120 but I think customer trust is maybe more-- 176 00:08:46,120 --> 00:08:50,600 must have trust in connecting your personal data 177 00:08:50,600 --> 00:08:51,755 to a third party. 178 00:08:51,755 --> 00:08:54,920 I'm thinking something like mint.com, 179 00:08:54,920 --> 00:08:56,330 where I add my bank account. 180 00:08:56,330 --> 00:08:58,670 I can connect it and feel comfortable 181 00:08:58,670 --> 00:09:01,370 giving that information to a site that's not necessarily 182 00:09:01,370 --> 00:09:03,058 my bank or my credit card. 183 00:09:03,058 --> 00:09:04,100 GARY GENSLER: Absolutely. 184 00:09:04,100 --> 00:09:04,600 Absolutely. 185 00:09:04,600 --> 00:09:06,260 So one of the major changes-- 186 00:09:06,260 --> 00:09:09,290 and this has been a number of years, 187 00:09:09,290 --> 00:09:11,110 as companies like Mint and others, 188 00:09:11,110 --> 00:09:14,570 where you'll give your data to a company 189 00:09:14,570 --> 00:09:17,680 that you've personally never met a human being. 190 00:09:17,680 --> 00:09:20,570 You've downloaded an app and you're giving your data 191 00:09:20,570 --> 00:09:23,690 to that company, to give you advice 192 00:09:23,690 --> 00:09:25,610 or to give you a better rate. 193 00:09:25,610 --> 00:09:29,120 Have you used Mint, by the way? 194 00:09:29,120 --> 00:09:32,135 And do you mind just saying what product were you shopping for, 195 00:09:32,135 --> 00:09:35,987 or what service were you looking for? 196 00:09:35,987 --> 00:09:38,404 AUDIENCE: I was just trying to track my spending because I 197 00:09:38,404 --> 00:09:39,770 was spending way too much. 198 00:09:39,770 --> 00:09:41,520 GARY GENSLER: All right, so there are also 199 00:09:41,520 --> 00:09:44,970 sites where it can help you keep your personal budget, 200 00:09:44,970 --> 00:09:45,880 if you wish-- 201 00:09:45,880 --> 00:09:48,900 like personal accounting and so forth. 202 00:09:48,900 --> 00:09:52,170 And you give them permission to look into your bank account. 203 00:09:52,170 --> 00:09:54,460 But this trend of using-- 204 00:09:54,460 --> 00:09:56,700 and there many competitors of Mint 205 00:09:56,700 --> 00:09:59,470 that we'll dive into in the next couple weeks. 206 00:09:59,470 --> 00:10:00,930 But there are many competitors. 207 00:10:00,930 --> 00:10:02,340 It's basically saying, all right, 208 00:10:02,340 --> 00:10:06,250 I am going to permission a website-- 209 00:10:06,250 --> 00:10:09,500 a data service provider, fintech company-- 210 00:10:09,500 --> 00:10:14,450 I'm going to permission them to look inside my bank account. 211 00:10:14,450 --> 00:10:19,490 I have to give them my password and my username in some way 212 00:10:19,490 --> 00:10:22,290 and have them linked in. 213 00:10:22,290 --> 00:10:24,120 Big change. 214 00:10:24,120 --> 00:10:27,810 It takes away a little of the comparative advantage possibly 215 00:10:27,810 --> 00:10:31,230 of the big incumbents, because you and I can give somebody 216 00:10:31,230 --> 00:10:35,980 else permission to go in and look at that data. 217 00:10:35,980 --> 00:10:38,438 PROFESSOR 2: [? Iqbal? ?] 218 00:10:38,438 --> 00:10:38,980 AUDIENCE: Hi. 219 00:10:38,980 --> 00:10:39,890 Morning, professor. 220 00:10:39,890 --> 00:10:44,090 I just want to highlight that there's 221 00:10:44,090 --> 00:10:49,790 an avenue to market or to showcase financial services. 222 00:10:49,790 --> 00:10:52,410 It's starting to be on a lifestyle platform. 223 00:10:52,410 --> 00:10:57,470 So if you look at, for example, Alipay, it's all about access 224 00:10:57,470 --> 00:11:00,290 to [INAUDIBLE] online shopping. 225 00:11:00,290 --> 00:11:02,960 And WeChat's about access to its gaming. 226 00:11:02,960 --> 00:11:08,490 But there's a lot of these conflicting financial services 227 00:11:08,490 --> 00:11:11,060 and key services that people want. 228 00:11:11,060 --> 00:11:12,690 So it's nothing new. 229 00:11:12,690 --> 00:11:14,940 In fact, Walmart and even now [? Star Market ?] 230 00:11:14,940 --> 00:11:19,660 has banks opened up within its grocery stores. 231 00:11:19,660 --> 00:11:21,570 So I think there'll be a trend that 232 00:11:21,570 --> 00:11:24,230 will be for the [INAUDIBLE]. 233 00:11:24,230 --> 00:11:27,170 GARY GENSLER: And so what [? Iqbal ?] has raised is not 234 00:11:27,170 --> 00:11:29,270 new-- decades old-- 235 00:11:29,270 --> 00:11:32,660 that even big retailers, bricks and mortar retailers, 236 00:11:32,660 --> 00:11:34,700 thought, well, should we have a counter for you 237 00:11:34,700 --> 00:11:35,660 to do some finance? 238 00:11:35,660 --> 00:11:39,770 Should we embed finance into the retailing business 239 00:11:39,770 --> 00:11:43,280 with the trends with in-store and branded credit 240 00:11:43,280 --> 00:11:45,000 cards for decades? 241 00:11:45,000 --> 00:11:49,430 But now online embedded finance inside 242 00:11:49,430 --> 00:11:53,600 of Alipay, which was mentioned just then, and other companies, 243 00:11:53,600 --> 00:11:56,900 to embed the financial aspects right in their product 244 00:11:56,900 --> 00:11:57,920 offerings. 245 00:11:57,920 --> 00:11:59,960 And some of that embedded finance 246 00:11:59,960 --> 00:12:04,430 is an Amazon credit card or an Alipay payment mechanism. 247 00:12:04,430 --> 00:12:07,340 Some of the embedded finance can even 248 00:12:07,340 --> 00:12:11,240 be that if you're buying a mattress from Casper-- 249 00:12:11,240 --> 00:12:13,490 I think I may have referenced this already-- 250 00:12:13,490 --> 00:12:16,940 but if you're buying a mattress from cash online, 251 00:12:16,940 --> 00:12:21,390 that they might offer what's called point-of-sale financing. 252 00:12:21,390 --> 00:12:23,330 Now, you and I don't think of it that way. 253 00:12:23,330 --> 00:12:26,120 They just say, you can pay over four installments. 254 00:12:26,120 --> 00:12:28,700 But in fact, they're lending you money for those four 255 00:12:28,700 --> 00:12:31,880 installments, and it's embedded right in the product. 256 00:12:31,880 --> 00:12:36,680 And Uber and Lyft actually embed some finance right 257 00:12:36,680 --> 00:12:38,490 in their app. 258 00:12:38,490 --> 00:12:40,925 Now, they've built it on top of credit card companies. 259 00:12:40,925 --> 00:12:42,300 But they could choose to build it 260 00:12:42,300 --> 00:12:45,810 on top of their own sort of financing mechanism 261 00:12:45,810 --> 00:12:48,100 with their drivers and the like. 262 00:12:48,100 --> 00:12:51,620 So big change, embedded finance. 263 00:12:51,620 --> 00:12:53,540 Multi decades in coming, but now it's 264 00:12:53,540 --> 00:12:56,060 embedded finance in a technological way 265 00:12:56,060 --> 00:12:59,390 that it looks seamless to the user-- 266 00:12:59,390 --> 00:13:00,320 or can look seamless. 267 00:13:03,630 --> 00:13:05,713 PROFESSOR 2: Nikhil. 268 00:13:05,713 --> 00:13:07,380 AUDIENCE: The other piece that I thought 269 00:13:07,380 --> 00:13:11,880 was interesting was the article talk about first the uncoupling 270 00:13:11,880 --> 00:13:12,970 of financial services. 271 00:13:12,970 --> 00:13:16,720 So a lot of fintech incumbents decouple the product offerings. 272 00:13:16,720 --> 00:13:19,560 But then there's a resurgence of bundling products again. 273 00:13:19,560 --> 00:13:21,450 So when you look at Wealthfront, they 274 00:13:21,450 --> 00:13:23,070 started with just asset management. 275 00:13:23,070 --> 00:13:24,910 But now they're offering cash management. 276 00:13:24,910 --> 00:13:25,990 Robinhood's the same way. 277 00:13:25,990 --> 00:13:28,032 It came out trading, but again, they're 278 00:13:28,032 --> 00:13:29,490 trying to pivot to cash management. 279 00:13:29,490 --> 00:13:32,610 So they focus on very niche products to begin with, 280 00:13:32,610 --> 00:13:34,977 but there seems to be bundling happening again. 281 00:13:34,977 --> 00:13:35,810 GARY GENSLER: Right. 282 00:13:35,810 --> 00:13:40,320 And I think this is unbundling and rebundling 283 00:13:40,320 --> 00:13:44,050 happens in many industries. 284 00:13:44,050 --> 00:13:48,750 Some of the proponents in the most optimistic side 285 00:13:48,750 --> 00:13:51,600 of financial technology say, we can 286 00:13:51,600 --> 00:13:53,100 have a flattening of finance. 287 00:13:53,100 --> 00:13:55,590 We can have a full unbundling, where 288 00:13:55,590 --> 00:13:58,410 you can get each of your products separately. 289 00:13:58,410 --> 00:14:02,010 And while the technology can and does 290 00:14:02,010 --> 00:14:06,210 facilitate that, I think our tendencies 291 00:14:06,210 --> 00:14:10,530 in markets and as humans, and our behavioral science 292 00:14:10,530 --> 00:14:13,900 tells us that bundling happens. 293 00:14:13,900 --> 00:14:16,900 So even if this transitional period of time-- 294 00:14:16,900 --> 00:14:21,810 even if this late 20-teens and early 2020's allows 295 00:14:21,810 --> 00:14:26,400 for a lot of unbundling, I believe that a lot of this 296 00:14:26,400 --> 00:14:31,840 will stay bundled and rebundled within particular websites, 297 00:14:31,840 --> 00:14:33,990 particular product offerings. 298 00:14:33,990 --> 00:14:36,210 I don't know of Leonora is trying to speak, 299 00:14:36,210 --> 00:14:41,330 because your name's popping up here on the screen. 300 00:14:41,330 --> 00:14:46,410 But if not-- how would open API initiatives 301 00:14:46,410 --> 00:14:48,150 influence all of this? 302 00:14:48,150 --> 00:14:50,340 I've sort of said it a little bit myself, 303 00:14:50,340 --> 00:14:52,650 but anybody want to chime in? 304 00:14:59,380 --> 00:15:02,100 PROFESSOR 2: Geetha? 305 00:15:02,100 --> 00:15:06,100 AUDIENCE: Yeah, I think I speak from the incumbent side. 306 00:15:06,100 --> 00:15:10,180 I think open APIs have made it so easy from a banking 307 00:15:10,180 --> 00:15:17,670 perspective, in terms of ability to track what is being accessed 308 00:15:17,670 --> 00:15:19,390 of customer data. 309 00:15:19,390 --> 00:15:21,460 Because before APIs and before some 310 00:15:21,460 --> 00:15:24,730 of the standouts like OAuth and OIDC, 311 00:15:24,730 --> 00:15:29,770 they were pretty much dependent on screen scraping. 312 00:15:29,770 --> 00:15:33,420 And it was very difficult for us to track. 313 00:15:33,420 --> 00:15:36,640 And the customers pretty much gave their user ID and password 314 00:15:36,640 --> 00:15:38,230 to these vendors. 315 00:15:38,230 --> 00:15:41,200 And they could pretty much log in and screen scrape anything, 316 00:15:41,200 --> 00:15:43,400 and we won't even know. 317 00:15:43,400 --> 00:15:47,340 GARY GENSLER: So you're saying Gita, from the incumbent side-- 318 00:15:47,340 --> 00:15:52,580 from the big financial firm side, these new startups-- 319 00:15:52,580 --> 00:15:54,760 the Mints of the world, so to speak-- 320 00:15:54,760 --> 00:15:59,170 were able to get folks' usernames and passwords, 321 00:15:59,170 --> 00:16:02,440 and come into the incumbent systems, 322 00:16:02,440 --> 00:16:04,510 and grab the data through a method 323 00:16:04,510 --> 00:16:08,170 called screen scraping, or even through a robotic process 324 00:16:08,170 --> 00:16:10,870 automation, which we're going to chat about. 325 00:16:10,870 --> 00:16:14,200 And you're saying you feel from an incumbent side it's better 326 00:16:14,200 --> 00:16:17,080 to manage all of that-- 327 00:16:17,080 --> 00:16:21,550 those interfaces, which could be disruptive. 328 00:16:21,550 --> 00:16:24,970 You'd have more awareness and more security 329 00:16:24,970 --> 00:16:27,071 around that interface. 330 00:16:27,071 --> 00:16:29,502 AUDIENCE: Correct. 331 00:16:29,502 --> 00:16:31,460 GARY GENSLER: And then how does it influence it 332 00:16:31,460 --> 00:16:32,630 from the consumer side? 333 00:16:32,630 --> 00:16:35,660 Anybody want to throw an idea from the customer side? 334 00:16:35,660 --> 00:16:38,120 I agree with what Gita's saying. 335 00:16:38,120 --> 00:16:40,120 PROFESSOR 2: So I'm not sure it will be related, 336 00:16:40,120 --> 00:16:42,120 but Yi had his hand up. 337 00:16:42,120 --> 00:16:43,430 GARY GENSLER: Sure. 338 00:16:43,430 --> 00:16:44,930 AUDIENCE: Yeah, I think the benefits 339 00:16:44,930 --> 00:16:48,290 is more for the fintech company rather than the incumbent. 340 00:16:48,290 --> 00:16:53,330 Because it basically lowered the entry barriers. 341 00:16:53,330 --> 00:16:56,330 So I guess from the consumer perspective, 342 00:16:56,330 --> 00:16:58,820 now you have many more options. 343 00:16:58,820 --> 00:17:02,470 Previously, you can only work with those like large banks. 344 00:17:02,470 --> 00:17:06,800 But now you can work with many more companies that are there. 345 00:17:06,800 --> 00:17:10,530 So the market is more competitive, basically. 346 00:17:10,530 --> 00:17:13,430 GARY GENSLER: So if I permission a company like Mint, 347 00:17:13,430 --> 00:17:16,579 Mint can give me advice around my cash flow management. 348 00:17:16,579 --> 00:17:18,380 If I have permission a company like so 349 00:17:18,380 --> 00:17:21,920 SoFi, which is, frankly, competing with the banks-- 350 00:17:21,920 --> 00:17:24,200 competing with lending products. 351 00:17:24,200 --> 00:17:28,200 But if my SoFi to see my accounts, 352 00:17:28,200 --> 00:17:31,020 then I get more options as a consumer. 353 00:17:31,020 --> 00:17:32,940 SoFi might compete with the banks, 354 00:17:32,940 --> 00:17:37,200 but also give me a better user interface-- user experience 355 00:17:37,200 --> 00:17:39,180 as well. 356 00:17:39,180 --> 00:17:46,460 So it's far broadened out the world for the consumers. 357 00:17:46,460 --> 00:17:48,980 And then I put a couple of readings 358 00:17:48,980 --> 00:17:52,780 in about robotic process automation and screen scraping. 359 00:17:52,780 --> 00:17:57,410 Gita's already raised this a bit. 360 00:17:57,410 --> 00:17:59,410 And it's being used. 361 00:17:59,410 --> 00:18:01,890 It's also being used by the financial sector. 362 00:18:01,890 --> 00:18:06,440 So anybody want to add how you think about robotic process 363 00:18:06,440 --> 00:18:08,600 automation in a positive way? 364 00:18:08,600 --> 00:18:11,990 In a way from the incumbents and the financial firms, 365 00:18:11,990 --> 00:18:14,060 how they're using it. 366 00:18:14,060 --> 00:18:17,090 It's certainly also being used by the disruptors trying 367 00:18:17,090 --> 00:18:20,330 to get information out of the incumbents. 368 00:18:20,330 --> 00:18:23,300 They're using it as sort of an automated screen scraping 369 00:18:23,300 --> 00:18:25,880 to grab information from the incumbents. 370 00:18:25,880 --> 00:18:27,380 But does anybody want to sort of say 371 00:18:27,380 --> 00:18:29,230 how maybe the incumbents are using it 372 00:18:29,230 --> 00:18:30,600 in their business models? 373 00:18:36,027 --> 00:18:36,860 PROFESSOR 2: Hassan. 374 00:18:39,860 --> 00:18:40,860 AUDIENCE: Hi, professor. 375 00:18:40,860 --> 00:18:46,780 Yeah, I just have a question about the APIs. 376 00:18:46,780 --> 00:18:49,460 I don't actually totally understand what are these. 377 00:18:49,460 --> 00:18:51,830 I mean, is it like an operating system? 378 00:18:51,830 --> 00:18:53,960 I mean, what's an API exactly? 379 00:18:53,960 --> 00:18:56,570 I've tried to google it, but I've not quite understood. 380 00:18:56,570 --> 00:18:58,790 GARY GENSLER: So I'm going to dive into it. 381 00:18:58,790 --> 00:19:00,590 Can I hold that, Hassan, just for a minute? 382 00:19:00,590 --> 00:19:03,830 Because I'm going to dive into it in the next six minutes. 383 00:19:03,830 --> 00:19:10,800 But fundamentally, an application program interface 384 00:19:10,800 --> 00:19:14,010 is a method that two computers can speak to. 385 00:19:14,010 --> 00:19:16,330 And the traditional way, and the old way, 386 00:19:16,330 --> 00:19:22,820 it was just how one system can go in and get information, 387 00:19:22,820 --> 00:19:26,280 get objects out of the other system. 388 00:19:26,280 --> 00:19:30,470 And then we'll do a little cursory dive 389 00:19:30,470 --> 00:19:31,910 into what it looks like today. 390 00:19:31,910 --> 00:19:35,950 Because it's moved a lot from that traditional method. 391 00:19:35,950 --> 00:19:41,560 On a web-based API system, you can actually sort of almost-- 392 00:19:41,560 --> 00:19:45,860 it feels like on your computer, on your laptop, so to speak, 393 00:19:45,860 --> 00:19:50,760 that you're actually operating that system at a distance. 394 00:19:50,760 --> 00:19:53,270 So you sort of think of it like it's not only just grabbing 395 00:19:53,270 --> 00:19:56,690 some data or objects or records, but it's actually 396 00:19:56,690 --> 00:20:00,770 sort of grabbing some application, 397 00:20:00,770 --> 00:20:02,600 as if it feels like you're really 398 00:20:02,600 --> 00:20:05,090 operating it on your system. 399 00:20:05,090 --> 00:20:09,080 Open API is this conceptual framework 400 00:20:09,080 --> 00:20:12,020 that it's a public interface, rather than 401 00:20:12,020 --> 00:20:13,630 a private interface. 402 00:20:13,630 --> 00:20:15,890 And if you think that you could literally 403 00:20:15,890 --> 00:20:19,430 have just a restricted private interface that you give one 404 00:20:19,430 --> 00:20:22,820 other computer, whether it's an affiliate of yours 405 00:20:22,820 --> 00:20:27,170 or a customer of yours, a contractual private 406 00:20:27,170 --> 00:20:32,430 interface to do that which they wish to do, or is it 407 00:20:32,430 --> 00:20:37,350 a public interface, where broadly it can be interfaced? 408 00:20:37,350 --> 00:20:40,080 And so that's sort of a thumbnail. 409 00:20:40,080 --> 00:20:42,930 But let me sort of develop that a little bit more 410 00:20:42,930 --> 00:20:44,910 over the next half hour. 411 00:20:44,910 --> 00:20:47,300 Does that help? 412 00:20:47,300 --> 00:20:48,010 AUDIENCE: Yeah. 413 00:20:48,010 --> 00:20:48,873 Thank you very much. 414 00:20:48,873 --> 00:20:50,790 PROFESSOR 2: So to answer your question, Gary, 415 00:20:50,790 --> 00:20:52,962 we have Ivy and then Andrea. 416 00:20:52,962 --> 00:20:53,670 GARY GENSLER: OK. 417 00:20:53,670 --> 00:20:54,870 And then I'll move on. 418 00:20:54,870 --> 00:20:55,540 Thanks, Ivy. 419 00:20:58,430 --> 00:21:00,990 AUDIENCE: Yeah, I just wanted to answer the question. 420 00:21:00,990 --> 00:21:02,850 Because I thought the Medium article did 421 00:21:02,850 --> 00:21:05,910 a good job of talking about the different ways 422 00:21:05,910 --> 00:21:08,920 that are banking use cases. 423 00:21:08,920 --> 00:21:10,770 And I've actually gone through a lot 424 00:21:10,770 --> 00:21:12,690 in terms of just on the banking side 425 00:21:12,690 --> 00:21:16,270 like actually having to do customer onboarding and KYC. 426 00:21:16,270 --> 00:21:20,880 And like we actually have what's called a KYC analyst. 427 00:21:20,880 --> 00:21:24,270 So all this is actually very, very manual typically. 428 00:21:24,270 --> 00:21:27,000 And going through compliance checks, 429 00:21:27,000 --> 00:21:30,543 we'll be working on a deal, and at the last minute, 430 00:21:30,543 --> 00:21:32,210 you can't get compliance checks through. 431 00:21:32,210 --> 00:21:34,440 And it's usually a huge pain. 432 00:21:34,440 --> 00:21:36,900 So I was excited to see that they could potentially 433 00:21:36,900 --> 00:21:39,930 change the system, because it ends up 434 00:21:39,930 --> 00:21:42,510 being a huge headache for everyone. 435 00:21:42,510 --> 00:21:45,210 I think my question, though, is like, when I read about this, 436 00:21:45,210 --> 00:21:46,500 obviously it seems like-- 437 00:21:46,500 --> 00:21:48,120 I'm sure the technology is there. 438 00:21:48,120 --> 00:21:53,220 But I think about just like any bulge bracket banks-- 439 00:21:53,220 --> 00:21:56,370 their IT infrastructure has been in place 440 00:21:56,370 --> 00:21:57,750 for like 50-plus years. 441 00:21:57,750 --> 00:22:02,700 And changing that technology is like removing 442 00:22:02,700 --> 00:22:05,370 one piece of a really big complicated puzzle. 443 00:22:05,370 --> 00:22:11,700 So I almost just question how feasible that is. 444 00:22:11,700 --> 00:22:14,370 Like the article does a really good job 445 00:22:14,370 --> 00:22:17,250 of laying out what we can actually automate. 446 00:22:17,250 --> 00:22:20,130 But I guess I am a little bit skeptical as 447 00:22:20,130 --> 00:22:22,340 to like when that time will come. 448 00:22:22,340 --> 00:22:24,250 GARY GENSLER: And Ivy, may I ask which bank 449 00:22:24,250 --> 00:22:29,270 or what country you were working in that was doing this? 450 00:22:29,270 --> 00:22:32,450 AUDIENCE: I worked at Merrill and Barclays. 451 00:22:32,450 --> 00:22:35,030 GARY GENSLER: So Merrill-Lynch or Barclays, you're right, 452 00:22:35,030 --> 00:22:35,960 they're incumbents. 453 00:22:35,960 --> 00:22:38,000 And they can automate many things. 454 00:22:38,000 --> 00:22:41,840 The technology is in advance of what they are actually 455 00:22:41,840 --> 00:22:43,600 doing right now. 456 00:22:43,600 --> 00:22:49,010 But robotic process automation can provide both incumbents 457 00:22:49,010 --> 00:22:50,810 and disruptors. 458 00:22:50,810 --> 00:22:53,600 But the Merrills and Barclays are already 459 00:22:53,600 --> 00:22:56,510 using it to some extent to automate 460 00:22:56,510 --> 00:22:59,120 that which humans are doing. 461 00:22:59,120 --> 00:23:02,240 Some of it, as we'll dive into later on, 462 00:23:02,240 --> 00:23:04,190 is about reading documents. 463 00:23:04,190 --> 00:23:09,410 Just literally an object reading a document. 464 00:23:09,410 --> 00:23:12,440 You could say that one form of robotic process automation 465 00:23:12,440 --> 00:23:17,160 that we already have is when we can deposit a check by using 466 00:23:17,160 --> 00:23:18,550 our mobile phone. 467 00:23:18,550 --> 00:23:21,350 Now, that also uses natural language processing, 468 00:23:21,350 --> 00:23:26,130 and it sort of reads that into the computer, and so forth. 469 00:23:26,130 --> 00:23:30,920 But you're absolutely right that each in incumbent 470 00:23:30,920 --> 00:23:35,210 is dedicated to ensuring that if they're moving from a legacy 471 00:23:35,210 --> 00:23:41,670 system to a new system, that they do that in their own way, 472 00:23:41,670 --> 00:23:44,390 so that it doesn't break in the middle, 473 00:23:44,390 --> 00:23:46,370 and that they actually do save money. 474 00:23:46,370 --> 00:23:47,990 And they have to schedule it. 475 00:23:47,990 --> 00:23:50,810 And they might not get to that project for 18 months 476 00:23:50,810 --> 00:23:52,700 or even five or seven years. 477 00:23:52,700 --> 00:23:56,060 And sometimes that's what gives the fintech disruptor 478 00:23:56,060 --> 00:23:57,560 the opportunity-- 479 00:23:57,560 --> 00:24:02,150 that the incumbent can use a new technology. 480 00:24:02,150 --> 00:24:03,830 In this case, RPA. 481 00:24:03,830 --> 00:24:05,780 They can use that technology. 482 00:24:05,780 --> 00:24:09,680 But they can't schedule it for three years or even longer. 483 00:24:09,680 --> 00:24:11,240 Or if they're scheduling it, they 484 00:24:11,240 --> 00:24:14,490 have to do it in a very controlled way. 485 00:24:14,490 --> 00:24:16,370 And the disruptor then is coming along 486 00:24:16,370 --> 00:24:19,010 and having an opportunity in the meantime 487 00:24:19,010 --> 00:24:21,450 to attack that business model. 488 00:24:21,450 --> 00:24:23,310 So I thank you. 489 00:24:23,310 --> 00:24:25,730 Romain are there others, or should I move on? 490 00:24:25,730 --> 00:24:27,950 PROFESSOR 2: We finally had Andrea, who I believe 491 00:24:27,950 --> 00:24:29,210 has experience on this topic. 492 00:24:29,210 --> 00:24:32,150 So we could all benefit from it. 493 00:24:32,150 --> 00:24:36,050 AUDIENCE: Yeah, but I wanted to add a similar point that we 494 00:24:36,050 --> 00:24:37,430 just discussed with Ivy. 495 00:24:37,430 --> 00:24:39,890 So we can move on definitely. 496 00:24:39,890 --> 00:24:43,310 But what I appreciate about the RPA in the incumbents 497 00:24:43,310 --> 00:24:45,290 is that it helps them to increase 498 00:24:45,290 --> 00:24:47,540 the efficiency in operations. 499 00:24:47,540 --> 00:24:53,330 But at the same time, it also brings in the standardization. 500 00:24:53,330 --> 00:24:59,090 And people can focus-- or employees of the bank 501 00:24:59,090 --> 00:25:02,130 can focus on more value-adding services. 502 00:25:02,130 --> 00:25:06,350 For example, the advisory to the client when 503 00:25:06,350 --> 00:25:08,390 it comes to more difficult products, 504 00:25:08,390 --> 00:25:11,770 such as mortgages or consumer loans. 505 00:25:11,770 --> 00:25:13,910 But at the same time, as you mentioned, 506 00:25:13,910 --> 00:25:18,120 the time and the resources that you need, especially 507 00:25:18,120 --> 00:25:22,070 in huge incumbents like Merrill or Barclays, 508 00:25:22,070 --> 00:25:24,177 it takes a lot of time to get there. 509 00:25:24,177 --> 00:25:25,010 GARY GENSLER: Right. 510 00:25:25,010 --> 00:25:27,690 And so what's interesting about each of these technologies-- 511 00:25:27,690 --> 00:25:30,900 and this is true if we broaden it out-- 512 00:25:30,900 --> 00:25:35,010 open API, robotic process automation, 513 00:25:35,010 --> 00:25:38,550 they're both a tool for the disruptors 514 00:25:38,550 --> 00:25:40,980 to produce services and new business models 515 00:25:40,980 --> 00:25:44,950 to the consuming public, And, they're also 516 00:25:44,950 --> 00:25:47,440 tools to the incumbents, depending 517 00:25:47,440 --> 00:25:49,450 upon on how they can invest in those tools, how 518 00:25:49,450 --> 00:25:52,390 they can update their business models, and the like. 519 00:25:52,390 --> 00:25:55,360 And in particular in open API, we'll see there's a real 520 00:25:55,360 --> 00:25:58,020 challenge, that the large incumbents-- 521 00:25:58,020 --> 00:26:04,060 the JPMorgan Chases and the PNC Banks and so forth-- 522 00:26:04,060 --> 00:26:07,190 are sharing data now with a lot of the disruptors. 523 00:26:07,190 --> 00:26:10,300 They're providing greater services to their ultimate end 524 00:26:10,300 --> 00:26:12,040 users and customers. 525 00:26:12,040 --> 00:26:14,350 But then also they're concerned about, 526 00:26:14,350 --> 00:26:18,220 are they giving up some of their inherent competitive advantage 527 00:26:18,220 --> 00:26:20,440 of controlling that data? 528 00:26:20,440 --> 00:26:22,060 And so there's some real trade offs, 529 00:26:22,060 --> 00:26:24,670 particularly in the open API world, 530 00:26:24,670 --> 00:26:26,900 as to how banks look at this. 531 00:26:26,900 --> 00:26:30,430 So I wanted to go back to the question of like, what is API? 532 00:26:30,430 --> 00:26:34,390 I just was able to find a slide of TechTarget 533 00:26:34,390 --> 00:26:38,320 that I thought was helpful, just to think about this. 534 00:26:38,320 --> 00:26:41,050 An API, just think of two systems 535 00:26:41,050 --> 00:26:42,220 that are linking together. 536 00:26:42,220 --> 00:26:46,540 And sort of the original APIs was just a local application 537 00:26:46,540 --> 00:26:51,810 program interface that sort of could connect two computers 538 00:26:51,810 --> 00:26:52,810 together. 539 00:26:52,810 --> 00:26:56,560 Web-based API, which went further-- and web-based APIs 540 00:26:56,560 --> 00:26:58,070 have been around for 30 years. 541 00:26:58,070 --> 00:26:59,560 This is not a new technology. 542 00:26:59,560 --> 00:27:01,960 But web-based are designed really 543 00:27:01,960 --> 00:27:06,370 to represent such that they could be widely used. 544 00:27:06,370 --> 00:27:08,920 HTML now and later something called 545 00:27:08,920 --> 00:27:13,000 REST API, which is more recently, these are protocols. 546 00:27:13,000 --> 00:27:16,900 These are protocols that multiple parties on the web 547 00:27:16,900 --> 00:27:19,570 can come into your service in some way. 548 00:27:22,160 --> 00:27:24,910 And then we talk about public APIs, 549 00:27:24,910 --> 00:27:29,740 which really sort of broadens it out further. 550 00:27:29,740 --> 00:27:31,630 And you could even think of this as sort 551 00:27:31,630 --> 00:27:37,600 of private registered APIs to broadly publicly available 552 00:27:37,600 --> 00:27:39,400 APIs. 553 00:27:39,400 --> 00:27:41,830 And what many businesses are trying to use-- 554 00:27:41,830 --> 00:27:42,880 and they manage it. 555 00:27:42,880 --> 00:27:45,430 And it's a very active part of strategy, 556 00:27:45,430 --> 00:27:49,180 is to manage that interface to allow the developer 557 00:27:49,180 --> 00:27:53,110 community to interface with your platform. 558 00:27:53,110 --> 00:27:54,370 Facebook needs to do it. 559 00:27:54,370 --> 00:27:56,200 Instagram needs to do it. 560 00:27:56,200 --> 00:28:01,570 Any network-based or platform-based service actually 561 00:28:01,570 --> 00:28:05,140 wants to encourage developers to build applications 562 00:28:05,140 --> 00:28:07,535 on top of the platform. 563 00:28:07,535 --> 00:28:08,910 So now you bring this to finance. 564 00:28:08,910 --> 00:28:12,700 Does JPMorgan Chase, does Bank of America, does Barclays, does 565 00:28:12,700 --> 00:28:15,610 Merrill want to do the same-- 566 00:28:15,610 --> 00:28:20,490 encourage development of applications 567 00:28:20,490 --> 00:28:22,600 on top of their platform? 568 00:28:22,600 --> 00:28:25,630 Recognizing that they make a lot on their platform-- 569 00:28:25,630 --> 00:28:29,280 their balance sheet, their funding models, their data, 570 00:28:29,280 --> 00:28:30,550 their analytics. 571 00:28:30,550 --> 00:28:33,570 Do they want to encourage the development of applications 572 00:28:33,570 --> 00:28:34,830 on top? 573 00:28:34,830 --> 00:28:38,220 And it's a little bit of a push and pull, because they're also 574 00:28:38,220 --> 00:28:40,470 encouraging development of applications 575 00:28:40,470 --> 00:28:43,520 that might compete with them, but also 576 00:28:43,520 --> 00:28:44,870 enhance what they're doing. 577 00:28:48,410 --> 00:28:51,860 So I sort of found this other spaghetti slide 578 00:28:51,860 --> 00:28:53,920 that I thought was really helpful. 579 00:28:53,920 --> 00:28:56,330 Now, if you think about thousands 580 00:28:56,330 --> 00:28:59,960 of financial firms dealing with thousands of public, 581 00:28:59,960 --> 00:29:03,000 this spaghetti wire gets even worse. 582 00:29:03,000 --> 00:29:05,210 But application program interface 583 00:29:05,210 --> 00:29:08,390 is a classic problem of math. 584 00:29:08,390 --> 00:29:11,990 It's that if you only have two computers by two computers, 585 00:29:11,990 --> 00:29:14,690 you might need four lines of connections. 586 00:29:14,690 --> 00:29:16,550 But all of a sudden when you have thousands, 587 00:29:16,550 --> 00:29:20,550 you could have this spaghetti wire of interfaces. 588 00:29:20,550 --> 00:29:25,580 So what's happened in the midst of this in sector by sector 589 00:29:25,580 --> 00:29:27,020 is that there have been companies 590 00:29:27,020 --> 00:29:31,670 that really have an API management layer in between. 591 00:29:31,670 --> 00:29:35,180 And in many sectors, you've seen a consolidation 592 00:29:35,180 --> 00:29:39,140 around either one standard for APIs or companies 593 00:29:39,140 --> 00:29:41,870 that are really service providers in the middle. 594 00:29:41,870 --> 00:29:45,260 And here are some examples in non-finance. 595 00:29:45,260 --> 00:29:50,390 And whether you know underneath the hood 596 00:29:50,390 --> 00:29:52,820 about Twilio or others-- 597 00:29:52,820 --> 00:29:55,370 Twilio probably has, I'm told, something 598 00:29:55,370 --> 00:30:00,770 close to an 80% market share in voice and messaging. 599 00:30:00,770 --> 00:30:06,380 That they're sort of that layer in between the disruptors 600 00:30:06,380 --> 00:30:11,690 and the service providers and the classic companies, 601 00:30:11,690 --> 00:30:15,380 providing a layer in the middle. 602 00:30:15,380 --> 00:30:17,360 A layer in the middle. 603 00:30:17,360 --> 00:30:22,670 Checkr does it on employment checks when we want a check. 604 00:30:22,670 --> 00:30:25,410 Each one of these companies were startups over the last 20 605 00:30:25,410 --> 00:30:25,910 years. 606 00:30:25,910 --> 00:30:28,670 Salesforce is a little bit earlier. 607 00:30:28,670 --> 00:30:30,170 And they really took-- 608 00:30:30,170 --> 00:30:32,810 even if they didn't know it at the time, 609 00:30:32,810 --> 00:30:40,390 they took the concept of open API and API and they built it. 610 00:30:40,390 --> 00:30:45,510 And in fact, there's these words they call API first-- 611 00:30:45,510 --> 00:30:47,490 that you can build a strategy saying, 612 00:30:47,490 --> 00:30:51,550 we're going to be the best providing APIs. 613 00:30:51,550 --> 00:30:53,350 Now, who's done this in finance? 614 00:30:53,350 --> 00:30:54,475 Who's done this in finance? 615 00:30:54,475 --> 00:30:57,370 One company that you might think of as a payment service 616 00:30:57,370 --> 00:31:00,400 provider, Stripe, did it very well. 617 00:31:00,400 --> 00:31:05,920 Stripe basically said, we will be that API provider 618 00:31:05,920 --> 00:31:08,450 in the payment space. 619 00:31:08,450 --> 00:31:11,300 What company did it really well most recently that we talked 620 00:31:11,300 --> 00:31:14,360 about last class was Plaid. 621 00:31:14,360 --> 00:31:17,120 And we're going to come back into Plaid as well. 622 00:31:17,120 --> 00:31:19,200 But Plaid did this very well. 623 00:31:19,200 --> 00:31:20,600 Now, they didn't start by saying, 624 00:31:20,600 --> 00:31:24,760 we're going to be only an API company. 625 00:31:24,760 --> 00:31:27,970 But they did start with being an API company. 626 00:31:27,970 --> 00:31:30,280 It was two entrepreneurs, in Plaid's case, 627 00:31:30,280 --> 00:31:33,100 about six or seven years ago that said, 628 00:31:33,100 --> 00:31:35,940 there really isn't a good API in the payment space. 629 00:31:35,940 --> 00:31:38,890 Now, there was Stripe. 630 00:31:38,890 --> 00:31:41,530 And I'm not sure about their foundational stories, 631 00:31:41,530 --> 00:31:46,370 why they didn't feel that the Stripe APIs were good enough. 632 00:31:46,370 --> 00:31:48,070 But they created an API. 633 00:31:48,070 --> 00:31:51,400 And of course, Plaid just recently sold for $5 billion 634 00:31:51,400 --> 00:31:55,980 plus to Visa, or announced the sale. 635 00:31:55,980 --> 00:31:58,500 So we talked a little bit about finance 636 00:31:58,500 --> 00:32:00,240 and this history, the stack. 637 00:32:00,240 --> 00:32:03,870 And I'm going back to the customer interface stack. 638 00:32:03,870 --> 00:32:06,360 We don't need to go back to bricks and mortars 639 00:32:06,360 --> 00:32:08,280 and electronic payments. 640 00:32:08,280 --> 00:32:11,010 But again, sort of the conceptual framework 641 00:32:11,010 --> 00:32:13,140 of this class and how I think about it, 642 00:32:13,140 --> 00:32:15,570 the most recent things added to the stack-- 643 00:32:15,570 --> 00:32:19,620 the internet, mobile phones, contact cards, and the like-- 644 00:32:19,620 --> 00:32:24,870 those are all fully engaged and part of the finance technology 645 00:32:24,870 --> 00:32:26,020 stack. 646 00:32:26,020 --> 00:32:30,180 And now we're adding open API and chat 647 00:32:30,180 --> 00:32:32,460 bots, which we talked about last class, 648 00:32:32,460 --> 00:32:36,060 and natural language processing, robotic processing. 649 00:32:36,060 --> 00:32:39,210 We're adding that to the finance technology stack. 650 00:32:39,210 --> 00:32:41,610 And that's sort of the conceptual framework. 651 00:32:41,610 --> 00:32:44,900 And I think those are where the opportunities are. 652 00:32:44,900 --> 00:32:46,830 In five or eight or 10 years from now, 653 00:32:46,830 --> 00:32:48,570 it will be taken for granted. 654 00:32:48,570 --> 00:32:50,670 Of course, there'll be enhancements. 655 00:32:50,670 --> 00:32:52,560 And every change, there's an enhancement 656 00:32:52,560 --> 00:32:55,950 that creates some opportunities as well. 657 00:32:55,950 --> 00:32:58,300 So what do we mean by open API? 658 00:32:58,300 --> 00:32:59,990 To [? Iqbal's ?] earlier question. 659 00:32:59,990 --> 00:33:03,080 Open Application Program Interface 660 00:33:03,080 --> 00:33:07,590 provides developers access to and an ability 661 00:33:07,590 --> 00:33:09,450 to get permissioned customer data. 662 00:33:09,450 --> 00:33:11,610 That means you and I give a password, you 663 00:33:11,610 --> 00:33:16,350 and I give a username. 664 00:33:16,350 --> 00:33:19,560 Open banking initiatives-- open banking 665 00:33:19,560 --> 00:33:21,195 is unique to the financial world-- 666 00:33:21,195 --> 00:33:24,930 are initiatives usually promoted by the official sector. 667 00:33:24,930 --> 00:33:27,930 Usually promoted either by the central banks 668 00:33:27,930 --> 00:33:30,300 or by the banking regulators, and sometimes 669 00:33:30,300 --> 00:33:32,610 promoted by competition authority, 670 00:33:32,610 --> 00:33:35,940 to say you must share certain data 671 00:33:35,940 --> 00:33:40,640 or you must create some standards for these open APIs. 672 00:33:40,640 --> 00:33:43,200 And again, as I just mentioned, open APIs 673 00:33:43,200 --> 00:33:48,000 are in every field of online services, 674 00:33:48,000 --> 00:33:52,030 from voice to employment to retailing. 675 00:33:52,030 --> 00:33:54,960 But the competition authorities and the central banks 676 00:33:54,960 --> 00:33:57,180 around the globe said, well, maybe we 677 00:33:57,180 --> 00:33:59,820 should lean into this-- 678 00:33:59,820 --> 00:34:02,940 the official sector-- lean in and promote competition 679 00:34:02,940 --> 00:34:06,370 for this sharing of data. 680 00:34:06,370 --> 00:34:08,280 And these initiatives, which you've 681 00:34:08,280 --> 00:34:12,520 got a chance to read about, do have policy trade offs. 682 00:34:12,520 --> 00:34:15,600 And why I focus on these policy trade offs 683 00:34:15,600 --> 00:34:18,280 is they're also about competition. 684 00:34:18,280 --> 00:34:21,400 Incumbents are not going to really want 685 00:34:21,400 --> 00:34:26,000 to embrace competition like this. 686 00:34:26,000 --> 00:34:29,679 Having said that, many of the incumbents 687 00:34:29,679 --> 00:34:33,123 recognize that it might be that they're shifting 688 00:34:33,123 --> 00:34:35,290 right now-- that their business models are shifting, 689 00:34:35,290 --> 00:34:37,900 and they have to think like platform companies. 690 00:34:37,900 --> 00:34:40,840 And to the extent that the JPMorgans or the Barclays 691 00:34:40,840 --> 00:34:42,699 think more like platform companies, 692 00:34:42,699 --> 00:34:45,219 they want to encourage the developer community 693 00:34:45,219 --> 00:34:49,969 to develop applications on top of their platforms. 694 00:34:49,969 --> 00:34:54,980 If they think it's simply losing market share to the SoFis 695 00:34:54,980 --> 00:34:57,310 and to the new banks, then they're 696 00:34:57,310 --> 00:35:00,620 going to be more closed to open API. 697 00:35:00,620 --> 00:35:02,600 If they think more like platform companies, 698 00:35:02,600 --> 00:35:05,160 they're going to want to promote it. 699 00:35:05,160 --> 00:35:06,740 But there are also trade offs. 700 00:35:06,740 --> 00:35:09,050 There are also trade offs that the big incumbents 701 00:35:09,050 --> 00:35:11,600 would say about cybersecurity. 702 00:35:11,600 --> 00:35:14,720 If we just let anybody into our platform, 703 00:35:14,720 --> 00:35:18,490 they can also bring down that platform possibly. 704 00:35:18,490 --> 00:35:21,910 And so these are real public policy trade offs, 705 00:35:21,910 --> 00:35:26,345 but they're also commercial trade offs as well. 706 00:35:26,345 --> 00:35:28,220 [Romain I don't know if there was a question. 707 00:35:28,220 --> 00:35:30,200 I see some activity in the chat. 708 00:35:30,200 --> 00:35:32,080 PROFESSOR 2: We have a question from Geetha. 709 00:35:32,080 --> 00:35:32,955 GARY GENSLER: Please. 710 00:35:35,160 --> 00:35:39,720 PROFESSOR 2: Gary, I've been thinking a lot about incumbents 711 00:35:39,720 --> 00:35:42,280 becoming platforms. 712 00:35:42,280 --> 00:35:47,560 Other than an ability to allow access to their customer data 713 00:35:47,560 --> 00:35:53,610 to lenders, I'm not able to think 714 00:35:53,610 --> 00:35:58,680 of examples of what kind of platforms incumbents can be. 715 00:35:58,680 --> 00:36:00,930 Like, they are not a specific payment company. 716 00:36:00,930 --> 00:36:04,180 They're only servicing their own customers. 717 00:36:04,180 --> 00:36:08,130 So I wonder if you have any examples. 718 00:36:08,130 --> 00:36:10,350 GARY GENSLER: So I think many of the major banks 719 00:36:10,350 --> 00:36:13,050 are already platforms. 720 00:36:13,050 --> 00:36:16,200 They don't necessarily think of themselves this way, 721 00:36:16,200 --> 00:36:21,360 but they are bringing together-- 722 00:36:21,360 --> 00:36:25,300 and this is true for centuries-- they are bringing together 723 00:36:25,300 --> 00:36:27,970 borrowers and lenders, or savers. 724 00:36:27,970 --> 00:36:31,420 So they have a balance sheet that brings together, 725 00:36:31,420 --> 00:36:36,010 in some cases, millions of savers and tens of thousands 726 00:36:36,010 --> 00:36:38,170 or hundreds of thousands of borrowers. 727 00:36:38,170 --> 00:36:40,010 And they're bringing that together. 728 00:36:40,010 --> 00:36:42,850 And when you consider that sort of the modern large financial 729 00:36:42,850 --> 00:36:45,730 firm with $1 trillion or $2 trillion balance sheet, 730 00:36:45,730 --> 00:36:49,810 many of the largest [? were ?] [? all-- ?] and so forth. 731 00:36:49,810 --> 00:36:51,820 And so they're bringing that together, 732 00:36:51,820 --> 00:36:53,650 and they're managing risk. 733 00:36:53,650 --> 00:36:59,300 And they're very good at it, or some of them 734 00:36:59,300 --> 00:37:01,550 are very good at that risk management 735 00:37:01,550 --> 00:37:05,600 of those millions of depositors and tens 736 00:37:05,600 --> 00:37:07,970 or hundreds of thousands of borrowers. 737 00:37:07,970 --> 00:37:10,610 That in itself is a platform. 738 00:37:10,610 --> 00:37:12,830 That creates a data advantage that they 739 00:37:12,830 --> 00:37:14,870 have about risk management. 740 00:37:14,870 --> 00:37:16,610 And it could be a real one. 741 00:37:16,610 --> 00:37:21,620 And they can offer activities and products on top of that. 742 00:37:21,620 --> 00:37:24,350 The fintech startups can be, in a sense, 743 00:37:24,350 --> 00:37:28,160 a front end distribution channel to bring more in. 744 00:37:28,160 --> 00:37:30,500 Because a lot of the fintech startups 745 00:37:30,500 --> 00:37:34,190 don't have that same balance sheet capability-- 746 00:37:34,190 --> 00:37:38,660 that same ability to in essence earn a spread 747 00:37:38,660 --> 00:37:42,510 between savers and borrowers. 748 00:37:42,510 --> 00:37:44,670 Now, you might contend, well no, they're 749 00:37:44,670 --> 00:37:46,500 losing their competitive advantage, 750 00:37:46,500 --> 00:37:50,490 because those startups sort of have that customer 751 00:37:50,490 --> 00:37:52,780 interface, that direct-- 752 00:37:52,780 --> 00:37:58,140 one might consider it the last mile with that customer. 753 00:37:58,140 --> 00:38:00,780 But the technology companies have taken another approach 754 00:38:00,780 --> 00:38:04,200 in saying, we like developers developing applications 755 00:38:04,200 --> 00:38:05,520 on top of our platforms. 756 00:38:05,520 --> 00:38:07,320 We think it brings more activities. 757 00:38:07,320 --> 00:38:10,240 We think it brings more interface, more data. 758 00:38:10,240 --> 00:38:13,920 And so they're sort of a network that is bringing data 759 00:38:13,920 --> 00:38:17,505 and activity on the platform. 760 00:38:17,505 --> 00:38:18,380 But I agree with you. 761 00:38:18,380 --> 00:38:20,180 It's a real strategic trade off. 762 00:38:20,180 --> 00:38:22,230 It's a real strategic challenge. 763 00:38:22,230 --> 00:38:26,487 And the incumbents are still sorting it out. 764 00:38:26,487 --> 00:38:27,320 PROFESSOR 2: Camilo? 765 00:38:30,238 --> 00:38:31,030 AUDIENCE: Yeah, hi. 766 00:38:31,030 --> 00:38:37,060 I just wanted to ask, in one of the other lectures, 767 00:38:37,060 --> 00:38:40,440 they tried to classify the open banking between 768 00:38:40,440 --> 00:38:42,130 [? prescriptive, ?] to facilitate 769 00:38:42,130 --> 00:38:44,800 market-driven, and in-process. 770 00:38:44,800 --> 00:38:49,000 But there were many countries, as for example Colombia, who 771 00:38:49,000 --> 00:38:52,150 didn't have any classification. 772 00:38:52,150 --> 00:38:53,320 What's the meaning of that? 773 00:38:53,320 --> 00:38:56,530 It means that it's forbidden. 774 00:38:56,530 --> 00:38:58,660 I mean, is it impossible to scrape the data 775 00:38:58,660 --> 00:39:02,457 from that from the bank? 776 00:39:02,457 --> 00:39:04,040 GARY GENSLER: So it's a good question. 777 00:39:04,040 --> 00:39:06,190 The Bank of International Settlements paper I think 778 00:39:06,190 --> 00:39:08,090 is to which you refer. 779 00:39:08,090 --> 00:39:12,270 And what they did was country by country review where it is. 780 00:39:12,270 --> 00:39:15,530 And in fact, that's probably good 781 00:39:15,530 --> 00:39:21,980 leading to the map that was in that report that I'll show. 782 00:39:21,980 --> 00:39:26,530 In many countries, the central bank, bank regulators, 783 00:39:26,530 --> 00:39:30,050 and competition authorities have not said much. 784 00:39:30,050 --> 00:39:33,070 I'm not close enough to Colombia to know. 785 00:39:33,070 --> 00:39:36,490 But in those countries, you can still technologically 786 00:39:36,490 --> 00:39:37,750 go in and scrape. 787 00:39:37,750 --> 00:39:39,940 An alternative, as I lay out here, 788 00:39:39,940 --> 00:39:42,880 is actually scraping from the screen. 789 00:39:42,880 --> 00:39:44,900 And so I think actually it's the next-- 790 00:39:44,900 --> 00:39:48,910 I grab this slide from the Bank of International Settlements 791 00:39:48,910 --> 00:39:49,450 reports. 792 00:39:49,450 --> 00:39:52,390 And thank you, it's a great setup, Camilo. 793 00:39:52,390 --> 00:39:55,240 That Europe passed something called 794 00:39:55,240 --> 00:39:58,330 the Payments System Directive. 795 00:39:58,330 --> 00:40:00,010 And it was actually the second time 796 00:40:00,010 --> 00:40:04,210 that the European parliament came together and said, 797 00:40:04,210 --> 00:40:07,060 we need to promote in the payment space-- 798 00:40:07,060 --> 00:40:09,170 not necessarily in the lending space, 799 00:40:09,170 --> 00:40:11,080 but in the payment space-- 800 00:40:11,080 --> 00:40:15,380 that there's more availability from merchants 801 00:40:15,380 --> 00:40:20,600 and from fintech startups to access your data and move 802 00:40:20,600 --> 00:40:21,200 money. 803 00:40:21,200 --> 00:40:22,960 That's Europe. 804 00:40:22,960 --> 00:40:26,550 And as the Bank of International Settlements says, 805 00:40:26,550 --> 00:40:28,370 it's more prescriptive. 806 00:40:28,370 --> 00:40:32,360 In the countries that are gray on this map-- most of Africa, 807 00:40:32,360 --> 00:40:35,930 many parts of the Mideast, and so forth, 808 00:40:35,930 --> 00:40:38,700 parts of Latin America-- 809 00:40:38,700 --> 00:40:41,340 the Bank of International Settlements, I should say, 810 00:40:41,340 --> 00:40:43,150 only covers about 60 countries. 811 00:40:43,150 --> 00:40:44,820 So many of the parts that are gray 812 00:40:44,820 --> 00:40:47,220 are countries that Bank of International Settlements 813 00:40:47,220 --> 00:40:48,950 didn't survey. 814 00:40:48,950 --> 00:40:51,980 But those countries are probably, 815 00:40:51,980 --> 00:40:54,950 like the US, technically market-driven. 816 00:40:54,950 --> 00:40:57,560 There is no federal law in the US, 817 00:40:57,560 --> 00:40:59,570 and there's no banking rule, that 818 00:40:59,570 --> 00:41:04,070 prescribes that you must have open API in the US. 819 00:41:04,070 --> 00:41:07,130 So I think that in Colombia it might be a technically 820 00:41:07,130 --> 00:41:10,850 market-driven, where the startups are using 821 00:41:10,850 --> 00:41:14,660 some form of screen scraping and some form of robotic process 822 00:41:14,660 --> 00:41:15,950 automation. 823 00:41:15,950 --> 00:41:20,220 But what we found in the US is companies like Stripe, 824 00:41:20,220 --> 00:41:22,650 companies like Plaid, others that I'm 825 00:41:22,650 --> 00:41:27,780 going to list in a moment, like Yodlee and others, 826 00:41:27,780 --> 00:41:34,050 just built a business model out of creating public APIs. 827 00:41:34,050 --> 00:41:36,270 And the incumbents possibly in Colombia 828 00:41:36,270 --> 00:41:37,530 are dealing with the same. 829 00:41:37,530 --> 00:41:38,190 I don't know. 830 00:41:40,335 --> 00:41:40,960 AUDIENCE: Yeah. 831 00:41:40,960 --> 00:41:42,610 Well, I don't know. 832 00:41:42,610 --> 00:41:46,840 In Colombia, the banks are happy to not let create APIs, 833 00:41:46,840 --> 00:41:50,360 because they don't believe themselves as platforms. 834 00:41:50,360 --> 00:41:52,960 They prefer to charge any type of fees 835 00:41:52,960 --> 00:41:56,120 for whatever they can do it in the small print letter. 836 00:41:56,120 --> 00:41:59,830 So they get more the defensive approach, 837 00:41:59,830 --> 00:42:02,980 rather than what you're saying, seeing themselves as a platform 838 00:42:02,980 --> 00:42:07,790 and monetizing over many layers of application. 839 00:42:07,790 --> 00:42:11,560 GARY GENSLER: Yeah, I think that in many countries 840 00:42:11,560 --> 00:42:14,350 and in many incumbents, they're taking 841 00:42:14,350 --> 00:42:16,450 a sort of guarded approach, and sort of saying, 842 00:42:16,450 --> 00:42:21,262 no, no, no, this is dangerous for our business model. 843 00:42:21,262 --> 00:42:22,970 This is dangerous for our business model. 844 00:42:22,970 --> 00:42:28,730 But to that incumbent that does that, often they lose out. 845 00:42:28,730 --> 00:42:30,650 It sometimes takes years to lose out. 846 00:42:30,650 --> 00:42:33,540 But it's that somebody finds a way 847 00:42:33,540 --> 00:42:36,330 to encroach on that business model. 848 00:42:36,330 --> 00:42:39,710 And what we're finding around the globe 849 00:42:39,710 --> 00:42:41,870 is that more and more countries are promoting this. 850 00:42:41,870 --> 00:42:43,670 Brazil, for instance, is taking it 851 00:42:43,670 --> 00:42:45,170 to their national legislature. 852 00:42:45,170 --> 00:42:49,325 Now, with the corona crisis, I haven't stayed closely enough. 853 00:42:49,325 --> 00:42:51,360 It looked like it was going to pass this spring. 854 00:42:51,360 --> 00:42:54,470 And maybe one of the students knows whether they did. 855 00:42:54,470 --> 00:42:57,830 But the central bank was trying to, in essence, 856 00:42:57,830 --> 00:43:02,150 insert more competition, so that E2O and others in Brazil 857 00:43:02,150 --> 00:43:05,000 had to sort of take a little bit more competition 858 00:43:05,000 --> 00:43:08,010 in that environment. 859 00:43:08,010 --> 00:43:11,690 So my prediction would be, even in Colombia, it 860 00:43:11,690 --> 00:43:14,780 would be hard to guard against this for multiple years. 861 00:43:14,780 --> 00:43:16,500 Maybe a few more years. 862 00:43:16,500 --> 00:43:18,980 But they'll be guarding a business model that will 863 00:43:18,980 --> 00:43:21,500 have some encroachment on it. 864 00:43:21,500 --> 00:43:23,000 The Bank of International Settlement 865 00:43:23,000 --> 00:43:25,070 also sort of did this chart, which 866 00:43:25,070 --> 00:43:27,380 sort of said what's mandatory. 867 00:43:27,380 --> 00:43:30,110 Australia and Europe is fuller to the mandatory. 868 00:43:30,110 --> 00:43:32,060 That's why they're at the top of this chart. 869 00:43:32,060 --> 00:43:35,570 And you see the US, China, Argentina, probably 870 00:43:35,570 --> 00:43:39,320 Colombia, is technically market driven towards the bottom. 871 00:43:39,320 --> 00:43:41,000 It's also what products? 872 00:43:41,000 --> 00:43:43,550 Is it all about just funds transfer? 873 00:43:43,550 --> 00:43:45,890 Europe was more about payments. 874 00:43:45,890 --> 00:43:48,830 Was it also about the underlying product information? 875 00:43:48,830 --> 00:43:51,770 Because if it's forced upon the banks 876 00:43:51,770 --> 00:43:54,260 from the official sector-- 877 00:43:54,260 --> 00:43:55,770 in the US it's not. 878 00:43:55,770 --> 00:43:58,850 But in Europe, in Australia, Hong Kong, it has been. 879 00:43:58,850 --> 00:44:02,240 If it's sort of by the official sector, 880 00:44:02,240 --> 00:44:04,580 then it might well be that you have 881 00:44:04,580 --> 00:44:07,130 to include the product information, as well as 882 00:44:07,130 --> 00:44:11,130 the information underneath it. 883 00:44:11,130 --> 00:44:13,850 So back to the market-driven side. 884 00:44:13,850 --> 00:44:18,020 So the market-driven side is what's happened here in the US. 885 00:44:18,020 --> 00:44:20,960 And what's happened here in the US-- and I list-- 886 00:44:20,960 --> 00:44:23,750 this is not just the US, but also Europe, Australia, 887 00:44:23,750 --> 00:44:24,820 and some places. 888 00:44:24,820 --> 00:44:28,970 I list about a dozen firms that are very active in this place. 889 00:44:28,970 --> 00:44:32,000 This is not forced by the official sector. 890 00:44:32,000 --> 00:44:37,070 This is basically the private sector saying, 891 00:44:37,070 --> 00:44:39,140 let's avoid that spaghetti wire. 892 00:44:39,140 --> 00:44:40,870 These are companies that are going 893 00:44:40,870 --> 00:44:45,880 in between all the startups and the financial firms and saying, 894 00:44:45,880 --> 00:44:52,710 we can provide a standard way to do these interfaces. 895 00:44:52,710 --> 00:44:55,720 Plaid that I talked about just announced 896 00:44:55,720 --> 00:44:59,550 it's being sold to Visa for $5.3 billion. 897 00:44:59,550 --> 00:45:02,860 What do these firms compete on? 898 00:45:02,860 --> 00:45:07,370 They compete on a lot of things. 899 00:45:07,370 --> 00:45:08,860 First is coverage. 900 00:45:08,860 --> 00:45:12,420 If I'm Plaid, do I cover all the banks that you need? 901 00:45:12,420 --> 00:45:14,620 If I'm a new startup, and I'm thinking, 902 00:45:14,620 --> 00:45:16,360 should I use Plaid or Yodlee? 903 00:45:16,360 --> 00:45:19,780 And Yodlee and Plaid are deep competitors. 904 00:45:19,780 --> 00:45:22,810 Maybe one can stay just as competitive with each other 905 00:45:22,810 --> 00:45:25,720 as Goldman Sachs and Morgan Stanley. 906 00:45:25,720 --> 00:45:30,310 Yodlee was acquired years ago by a company called Envestnet. 907 00:45:30,310 --> 00:45:33,880 But Yodlee and Plaid, deep competitors to each other. 908 00:45:33,880 --> 00:45:36,610 Well first, what coverage do they have? 909 00:45:36,610 --> 00:45:39,560 As a startup, if I use Plaid instead of Yodlee, 910 00:45:39,560 --> 00:45:42,820 can I get to all of the banks? 911 00:45:42,820 --> 00:45:45,460 What services do they give me as a developer? 912 00:45:45,460 --> 00:45:46,840 Obviously, what price? 913 00:45:46,840 --> 00:45:47,740 Are they reliable? 914 00:45:47,740 --> 00:45:50,270 What's their security, and the like? 915 00:45:50,270 --> 00:45:51,800 So all of these are important. 916 00:45:51,800 --> 00:45:55,360 And one of the sort of incumbents in this space, 917 00:45:55,360 --> 00:45:58,030 Fiserv bought CashEdge. 918 00:45:58,030 --> 00:46:01,270 And Plaid was being bought by Visa. 919 00:46:01,270 --> 00:46:03,550 So some of these get caught up. 920 00:46:03,550 --> 00:46:07,680 SaltEdge is in 55 countries. 921 00:46:07,680 --> 00:46:10,650 And then I did a little research on a paper that 922 00:46:10,650 --> 00:46:12,390 was written three years ago-- 923 00:46:12,390 --> 00:46:16,050 or it's really a website that keeps a list. 924 00:46:16,050 --> 00:46:19,380 And I sort of put together-- this list is three years old. 925 00:46:19,380 --> 00:46:22,620 These are the account aggregators 926 00:46:22,620 --> 00:46:23,910 in all of these countries. 927 00:46:23,910 --> 00:46:27,027 And I'm sorry, Camilo, I don't see one for Colombia. 928 00:46:29,680 --> 00:46:33,730 But all of these companies in some way 929 00:46:33,730 --> 00:46:36,070 built up a business model to say, 930 00:46:36,070 --> 00:46:40,750 we will serve in these various countries 931 00:46:40,750 --> 00:46:45,370 to be between the incumbents and the startups. 932 00:46:45,370 --> 00:46:48,490 We will build some business model in between. 933 00:46:51,380 --> 00:46:54,410 Now, what can you do once you do that? 934 00:46:54,410 --> 00:46:56,370 You get a lot of data. 935 00:46:56,370 --> 00:46:59,670 And so what used to be just a software company-- 936 00:46:59,670 --> 00:47:03,920 Plaid, for instance, started as, in essence, we 937 00:47:03,920 --> 00:47:07,190 can provide this software. 938 00:47:07,190 --> 00:47:11,830 Started to have remarkable amounts of data. 939 00:47:11,830 --> 00:47:16,480 And that remarkable amounts of data in Plaid-- 940 00:47:16,480 --> 00:47:20,290 founded in 2013, sells for $5.3 billion to Visa, 941 00:47:20,290 --> 00:47:24,520 or announces the sale, only 450 employees. 942 00:47:24,520 --> 00:47:29,170 And Forbes estimates it's only had $109 million of revenues. 943 00:47:29,170 --> 00:47:31,710 So that's nearly 50 times revenues. 944 00:47:31,710 --> 00:47:33,810 Quite a multiple. 945 00:47:33,810 --> 00:47:36,720 You might wonder why Visa's doing this. 946 00:47:36,720 --> 00:47:40,170 This is from the Visa slides that Visa announced in January. 947 00:47:40,170 --> 00:47:44,880 This is how they sort of picture. 948 00:47:44,880 --> 00:47:45,840 They look at Plaid. 949 00:47:45,840 --> 00:47:53,540 They say, Plaid is connected to 2,600 fintech firms. 950 00:47:53,540 --> 00:47:58,310 That's 2,600 fintech firms connected to Plaid. 951 00:47:58,310 --> 00:48:00,950 And then Plaid on the other side provides an API 952 00:48:00,950 --> 00:48:04,580 into 11,000 data sources-- 953 00:48:04,580 --> 00:48:06,980 11,000 financial firms. 954 00:48:06,980 --> 00:48:10,790 2,600 fintech firms connected to Plaid, Plaid 955 00:48:10,790 --> 00:48:13,390 connects to 11,000. 956 00:48:13,390 --> 00:48:15,400 Again, that's the classic place you want to be. 957 00:48:15,400 --> 00:48:17,880 Plaid is at the bottleneck. 958 00:48:17,880 --> 00:48:21,250 Plaid is at the gateway, you would say. 959 00:48:21,250 --> 00:48:24,550 And they can collect data on 200 million users. 960 00:48:27,550 --> 00:48:29,440 Visa thinks there's a value in that. 961 00:48:29,440 --> 00:48:30,370 We'll have to see. 962 00:48:30,370 --> 00:48:32,320 We'll have to see whether that ends up 963 00:48:32,320 --> 00:48:35,200 being a good acquisition for Visa or not. 964 00:48:35,200 --> 00:48:38,350 But these companies in the middle-- 965 00:48:38,350 --> 00:48:43,450 the Plaids, the Yodlees, have right in the middle 966 00:48:43,450 --> 00:48:45,280 created value. 967 00:48:45,280 --> 00:48:47,050 Romain I'm going to pause there and see 968 00:48:47,050 --> 00:48:49,610 if there's some discussion that we can get going. 969 00:48:54,730 --> 00:48:58,280 PROFESSOR 2: [? Iqbal. ?] 970 00:48:58,280 --> 00:49:00,030 AUDIENCE: Yeah, professor. 971 00:49:00,030 --> 00:49:04,528 Can you shed some insight as to the relationship between Plaid 972 00:49:04,528 --> 00:49:05,820 and the financial institutions? 973 00:49:05,820 --> 00:49:07,612 What's in it for the financial institutions 974 00:49:07,612 --> 00:49:10,850 to face Plaid in reality? 975 00:49:10,850 --> 00:49:12,030 Were there fees involved? 976 00:49:14,580 --> 00:49:16,590 Because in the example we looked at some 977 00:49:16,590 --> 00:49:19,490 for some Latin American countries, such as Brazil, 978 00:49:19,490 --> 00:49:23,430 the incumbents are reluctant to participate in open API. 979 00:49:23,430 --> 00:49:26,670 And it would be [INAUDIBLE]. 980 00:49:26,670 --> 00:49:29,760 GARY GENSLER: It's a terrific question. 981 00:49:29,760 --> 00:49:31,520 And I'm going to apologize, because I'm 982 00:49:31,520 --> 00:49:35,726 going to slide back here on the slides a bit 983 00:49:35,726 --> 00:49:40,140 to go back to something. 984 00:49:40,140 --> 00:49:42,960 This is what they want to avoid. 985 00:49:42,960 --> 00:49:47,160 Even the incumbents don't want thousands of folks 986 00:49:47,160 --> 00:49:50,130 to enter into contractual arrangements, 987 00:49:50,130 --> 00:49:53,580 enter into secure private APIs. 988 00:49:53,580 --> 00:49:56,250 Even the incumbents want to sort of avoid 989 00:49:56,250 --> 00:49:59,610 this spaghetti wire of mess. 990 00:49:59,610 --> 00:50:06,150 Now, it might have been less strategic and more incremental 991 00:50:06,150 --> 00:50:13,110 that when you think about all these companies on this list-- 992 00:50:13,110 --> 00:50:16,890 these 10 or 12 that have built up these remarkable businesses, 993 00:50:16,890 --> 00:50:19,310 or around the globe all these-- 994 00:50:19,310 --> 00:50:23,130 that many of these started as small software providers. 995 00:50:23,130 --> 00:50:25,470 And they were saying to the big incumbents, 996 00:50:25,470 --> 00:50:28,960 we can handle that spaghetti wire. 997 00:50:28,960 --> 00:50:31,050 We can ensure that people aren't just screen 998 00:50:31,050 --> 00:50:34,220 scraping and using RPA. 999 00:50:34,220 --> 00:50:36,450 It's going to happen anyway. 1000 00:50:36,450 --> 00:50:39,180 You need somebody to help manage. 1001 00:50:39,180 --> 00:50:42,540 And so every modern online company 1002 00:50:42,540 --> 00:50:45,390 in finance and outside of finance 1003 00:50:45,390 --> 00:50:49,620 has to have an AP management strategy. 1004 00:50:49,620 --> 00:50:55,020 You don't normally think about something in the CEO suite, 1005 00:50:55,020 --> 00:50:57,930 this chief executive officer who's managing this. 1006 00:50:57,930 --> 00:51:00,660 But here we are at MIT, I'd like to bring this technology. 1007 00:51:00,660 --> 00:51:03,360 This is an important technology question 1008 00:51:03,360 --> 00:51:06,000 that you need to be asking is, how are we 1009 00:51:06,000 --> 00:51:09,510 managing our developer community, 1010 00:51:09,510 --> 00:51:11,700 our public interface? 1011 00:51:11,700 --> 00:51:14,250 Are we going to do it internally, 1012 00:51:14,250 --> 00:51:16,450 or are we going to hire a Plaid? 1013 00:51:16,450 --> 00:51:20,340 And I can tell you that the banks now-- this 1014 00:51:20,340 --> 00:51:23,790 has happened in the last four to seven years. 1015 00:51:23,790 --> 00:51:25,710 And many of the banks are sort of saying, 1016 00:51:25,710 --> 00:51:26,970 this has gotten out of hand. 1017 00:51:26,970 --> 00:51:28,680 This has gotten out of control. 1018 00:51:28,680 --> 00:51:30,450 Is this something that we really want 1019 00:51:30,450 --> 00:51:34,140 to allow the Plaids and the Yodlees and others 1020 00:51:34,140 --> 00:51:38,020 to commercialize our data? 1021 00:51:38,020 --> 00:51:42,510 But I think that in part, it was incrementally decisions 1022 00:51:42,510 --> 00:51:46,610 about how to manage broadly this interface. 1023 00:51:46,610 --> 00:51:50,960 And it's also possible that it wasn't far enough 1024 00:51:50,960 --> 00:51:54,110 up in the C-suite, and it was sort of managed 1025 00:51:54,110 --> 00:51:56,810 down in some regards. 1026 00:51:56,810 --> 00:52:01,370 But one reaction is this financial data exchange 1027 00:52:01,370 --> 00:52:02,390 that's happened here. 1028 00:52:02,390 --> 00:52:07,190 And I've listed the major founders of FDX. 1029 00:52:07,190 --> 00:52:09,890 But Financial Data Exchange-- and this is straight from 1030 00:52:09,890 --> 00:52:12,140 their own white paper-- 1031 00:52:12,140 --> 00:52:15,830 they want to create this group of remarkably large banks. 1032 00:52:15,830 --> 00:52:23,330 I think the members of FDX have $2 trillion of market value, 1033 00:52:23,330 --> 00:52:25,260 if I remember. 1034 00:52:25,260 --> 00:52:30,470 Because many of these actually have $2 trillion balance 1035 00:52:30,470 --> 00:52:31,610 sheets. 1036 00:52:31,610 --> 00:52:33,523 When you add up all of these companies, 1037 00:52:33,523 --> 00:52:34,940 they've come together and said, we 1038 00:52:34,940 --> 00:52:38,540 want to create standards for financial data sharing 1039 00:52:38,540 --> 00:52:41,090 and secure APIs. 1040 00:52:41,090 --> 00:52:45,950 We want to basically not sort of have the Plaids and the others 1041 00:52:45,950 --> 00:52:47,780 commercialize our data. 1042 00:52:47,780 --> 00:52:52,790 Even though that's not in FDX' mandate. 1043 00:52:56,180 --> 00:53:00,810 But in some regards, the cat's already out of the barn. 1044 00:53:00,810 --> 00:53:03,710 And now they're grappling with it. 1045 00:53:03,710 --> 00:53:08,670 But the JPMorgans and the PNC actually, I believe-- 1046 00:53:08,670 --> 00:53:13,560 which is a large mid-sized bank in the US-- 1047 00:53:13,560 --> 00:53:19,510 has stopped having Plaid have access, if I'm not mistaken. 1048 00:53:19,510 --> 00:53:23,810 So it's sort of an interesting challenge back and forth. 1049 00:53:23,810 --> 00:53:26,525 PROFESSOR 2: You now have Geetha and then Lin. 1050 00:53:26,525 --> 00:53:29,150 GARY GENSLER: Geetha's from the incumbent side going to tell us 1051 00:53:29,150 --> 00:53:32,170 why they're using the Plaids. 1052 00:53:32,170 --> 00:53:34,690 AUDIENCE: Yeah, I just wanted to make a comment on Plaid 1053 00:53:34,690 --> 00:53:36,310 specifically. 1054 00:53:36,310 --> 00:53:39,940 We allow access to a number of fintech lenders to our data, 1055 00:53:39,940 --> 00:53:41,680 because then we have no choice. 1056 00:53:41,680 --> 00:53:44,980 Because the customer is providing access to their data, 1057 00:53:44,980 --> 00:53:49,030 and we have to allow them to access it. 1058 00:53:49,030 --> 00:53:53,050 To somebody's question earlier, the biggest problem 1059 00:53:53,050 --> 00:53:53,680 with Plaid-- 1060 00:53:53,680 --> 00:53:55,420 and I'll probably send some articles 1061 00:53:55,420 --> 00:53:59,830 on kind of issues between Plaid and Capital One. 1062 00:53:59,830 --> 00:54:03,160 There was a period of time when we stopped giving access 1063 00:54:03,160 --> 00:54:06,070 to Plaid, because Plaid wouldn't comply 1064 00:54:06,070 --> 00:54:08,920 with any of the standards that we imposed. 1065 00:54:08,920 --> 00:54:11,710 For example, we needed fintech lenders 1066 00:54:11,710 --> 00:54:14,880 to come from certain IPs, because they looked like bots, 1067 00:54:14,880 --> 00:54:15,380 right? 1068 00:54:15,380 --> 00:54:17,350 Because they do all these batch accesses, 1069 00:54:17,350 --> 00:54:19,120 and they look like bots. 1070 00:54:19,120 --> 00:54:22,930 But Plaid wouldn't comply with those standards, 1071 00:54:22,930 --> 00:54:27,280 and it became very difficult for us to differentiate Plaid 1072 00:54:27,280 --> 00:54:30,220 from a bot. 1073 00:54:30,220 --> 00:54:33,790 And even though we had APIs, they would do screen scraping. 1074 00:54:33,790 --> 00:54:36,070 Sometimes they would come through our mobile channels. 1075 00:54:36,070 --> 00:54:38,195 Sometimes they would come through our web channels. 1076 00:54:38,195 --> 00:54:40,540 They would come through any API. 1077 00:54:40,540 --> 00:54:43,090 It became really hard for us to differentiate Plaid 1078 00:54:43,090 --> 00:54:46,300 from fraudsters. 1079 00:54:46,300 --> 00:54:48,700 So there was a period of time when we blocked Plaid. 1080 00:54:48,700 --> 00:54:51,420 And then they came to the table, we had conversations, 1081 00:54:51,420 --> 00:54:52,420 and then we let them in. 1082 00:54:54,950 --> 00:54:57,610 GARY GENSLER: And Geetha, you were at Cap One? 1083 00:54:57,610 --> 00:54:58,480 AUDIENCE: Yes. 1084 00:54:58,480 --> 00:54:58,980 Yeah. 1085 00:54:58,980 --> 00:55:01,188 GARY GENSLER: So for those who might not be familiar, 1086 00:55:01,188 --> 00:55:05,220 Cap One is one of the largest credit card banks in the US, 1087 00:55:05,220 --> 00:55:11,460 competing head to head with Bank of America and Citi and Chase, 1088 00:55:11,460 --> 00:55:15,180 Discover, and American Express and so forth. 1089 00:55:15,180 --> 00:55:18,090 I sort of call it the Big Seven. 1090 00:55:18,090 --> 00:55:24,120 And Cap One's the largest that isn't in and of itself 1091 00:55:24,120 --> 00:55:26,145 one of the big four banks in the US. 1092 00:55:29,240 --> 00:55:31,110 So it's a remarkable organization, 1093 00:55:31,110 --> 00:55:35,760 and it's cut out a big chunk of its business model 1094 00:55:35,760 --> 00:55:39,575 around serving customers in the credit card space. 1095 00:55:39,575 --> 00:55:40,950 And here you're hearing Gita say, 1096 00:55:40,950 --> 00:55:46,680 but we didn't really have a choice, is what you're saying. 1097 00:55:52,040 --> 00:55:53,670 Romane, others? 1098 00:55:53,670 --> 00:55:55,670 PROFESSOR 2: Let's go with Lin, and then Carlos, 1099 00:55:55,670 --> 00:55:56,670 if we have time. 1100 00:55:56,670 --> 00:55:57,645 GARY GENSLER: Please. 1101 00:55:57,645 --> 00:55:58,270 AUDIENCE: Yeah. 1102 00:55:58,270 --> 00:56:02,360 So in the example for Plaid, it seems like most of the data, 1103 00:56:02,360 --> 00:56:04,820 if not all of the data, is financial data coming 1104 00:56:04,820 --> 00:56:06,440 from financial institutions. 1105 00:56:06,440 --> 00:56:08,690 I'm just curious to hear your thoughts 1106 00:56:08,690 --> 00:56:12,270 on the market moving towards using non-financial data. 1107 00:56:12,270 --> 00:56:15,200 So customer data in the retail health 1108 00:56:15,200 --> 00:56:19,920 sector to inform or build financial services upon, 1109 00:56:19,920 --> 00:56:22,135 and whether there are any people doing that. 1110 00:56:22,135 --> 00:56:22,635 PROFESSOR 2: 1111 00:56:22,635 --> 00:56:26,450 GARY GENSLER: No, it's a really good observation 1112 00:56:26,450 --> 00:56:30,530 that the data aggregators-- 1113 00:56:30,530 --> 00:56:32,550 I'll go back a slide. 1114 00:56:32,550 --> 00:56:36,000 This whole cacophony from three years 1115 00:56:36,000 --> 00:56:37,980 ago around the globe, or the data 1116 00:56:37,980 --> 00:56:39,960 aggregators here in the US-- 1117 00:56:39,960 --> 00:56:43,620 these big eight or 10 in the US and Europe-- 1118 00:56:43,620 --> 00:56:45,930 these data aggregators are largely 1119 00:56:45,930 --> 00:56:50,750 about financial data for now. 1120 00:56:50,750 --> 00:56:54,010 So their business strategy-- these dozen 1121 00:56:54,010 --> 00:56:55,810 companies, their business strategies 1122 00:56:55,810 --> 00:56:59,080 was sort of API first. 1123 00:56:59,080 --> 00:57:03,370 They built an API upon which they got a network advantage 1124 00:57:03,370 --> 00:57:06,400 to get a lot of financial data. 1125 00:57:06,400 --> 00:57:09,950 But it is also true that they could add other data. 1126 00:57:09,950 --> 00:57:13,120 And in fact, Plaid gets bought by Visa. 1127 00:57:13,120 --> 00:57:15,640 Yodlee got bought by Envestnet. 1128 00:57:15,640 --> 00:57:21,730 And Envestnet is also in asset management and fund management 1129 00:57:21,730 --> 00:57:23,210 and so forth. 1130 00:57:23,210 --> 00:57:24,550 I believe you're right. 1131 00:57:24,550 --> 00:57:25,960 Your observation is right. 1132 00:57:25,960 --> 00:57:31,010 A lot of these "API first, data aggregator second" 1133 00:57:31,010 --> 00:57:35,300 are also going to be adding other non-financial data. 1134 00:57:35,300 --> 00:57:37,880 Alternative data, as we've talked about, 1135 00:57:37,880 --> 00:57:40,700 can come from finance, like whether you're paying 1136 00:57:40,700 --> 00:57:42,740 your utility bill or not. 1137 00:57:42,740 --> 00:57:46,370 It can come from finance and building full cash flow 1138 00:57:46,370 --> 00:57:47,990 underwriting models. 1139 00:57:47,990 --> 00:57:51,860 But it also can be your social media footprint. 1140 00:57:51,860 --> 00:57:53,120 And it often is. 1141 00:57:53,120 --> 00:57:55,490 Alternative data is about your social media 1142 00:57:55,490 --> 00:58:00,200 input and your footprint, I should say. 1143 00:58:00,200 --> 00:58:01,430 Your footprint. 1144 00:58:01,430 --> 00:58:04,070 So I think you're right. 1145 00:58:04,070 --> 00:58:09,750 Their comparative advantage is around this financial data. 1146 00:58:09,750 --> 00:58:16,910 Plaid, that's their advantage. 1147 00:58:16,910 --> 00:58:21,320 I want to note one company that's on this list here, 1148 00:58:21,320 --> 00:58:23,590 I think I included, was Galileo. 1149 00:58:23,590 --> 00:58:24,090 Did I? 1150 00:58:24,090 --> 00:58:25,880 Yeah, I included Galileo. 1151 00:58:25,880 --> 00:58:28,730 Galileo has done a recent fundraising round 1152 00:58:28,730 --> 00:58:31,670 in the middle of 2019. 1153 00:58:31,670 --> 00:58:32,960 It's still private. 1154 00:58:32,960 --> 00:58:36,007 I think their valuation was $1.2 billion. 1155 00:58:36,007 --> 00:58:37,340 Might have been a little higher. 1156 00:58:37,340 --> 00:58:39,260 I can't recall the figure. 1157 00:58:39,260 --> 00:58:40,460 So it's a unicorn. 1158 00:58:40,460 --> 00:58:42,650 Not many of us have heard of Galileo. 1159 00:58:42,650 --> 00:58:44,340 But two of their big customers-- 1160 00:58:44,340 --> 00:58:47,990 Chime is one. 1161 00:58:47,990 --> 00:58:51,170 I'll just say, Chime is one fintech company that's 1162 00:58:51,170 --> 00:58:52,790 one of their big customers. 1163 00:58:52,790 --> 00:58:55,610 Last year, Galileo had a glitch. 1164 00:58:55,610 --> 00:58:58,370 And Chime went down for 24 hours, 1165 00:58:58,370 --> 00:59:03,150 because Galileo's standards in API had a glitch. 1166 00:59:03,150 --> 00:59:06,170 So sometimes, you're that fintech startup 1167 00:59:06,170 --> 00:59:10,560 relying upon a Galileo, in that instance. 1168 00:59:10,560 --> 00:59:16,070 And if Galileo software has a problem, you have a problem. 1169 00:59:16,070 --> 00:59:21,170 So there's that challenge too in the midst of this. 1170 00:59:21,170 --> 00:59:23,630 Marqeta has done a fundraising round. 1171 00:59:23,630 --> 00:59:26,240 I think its valuation is about $2 billion. 1172 00:59:26,240 --> 00:59:32,450 Again, still private, still in the venture land, and so forth. 1173 00:59:32,450 --> 00:59:37,620 So there's significant value that has been built, 1174 00:59:37,620 --> 00:59:40,230 API first, data aggregation second. 1175 00:59:40,230 --> 00:59:42,600 And these companies are all somewhere 1176 00:59:42,600 --> 00:59:45,810 between the incumbents and the developer community 1177 00:59:45,810 --> 00:59:48,170 and the startups. 1178 00:59:48,170 --> 00:59:50,150 Back to the class, Romain? 1179 00:59:50,150 --> 00:59:52,160 PROFESSOR 2: Carlos is up. 1180 00:59:52,160 --> 00:59:53,440 AUDIENCE: Hi. 1181 00:59:53,440 --> 00:59:55,780 So I just want to give an example of a shout 1182 00:59:55,780 --> 00:59:57,340 out to Plaid, really, for something 1183 00:59:57,340 --> 00:59:58,507 they're doing at the moment. 1184 00:59:58,507 --> 01:00:01,990 So as part of the SBA PPP program for emergency funding 1185 01:00:01,990 --> 01:00:04,850 for small businesses with the whole corona crisis, 1186 01:00:04,850 --> 01:00:07,263 there's $200 billion being disbursed over the course 1187 01:00:07,263 --> 01:00:07,930 of three months. 1188 01:00:07,930 --> 01:00:09,463 And actually probably this Thursday 1189 01:00:09,463 --> 01:00:11,380 there's going to be an additional $200 billion 1190 01:00:11,380 --> 01:00:12,830 approved. 1191 01:00:12,830 --> 01:00:13,750 $250, sorry. 1192 01:00:13,750 --> 01:00:15,640 So what Plaid is doing is they're actually 1193 01:00:15,640 --> 01:00:18,640 helping, because of their role as a data aggregator, 1194 01:00:18,640 --> 01:00:22,630 to populate the payroll portion of these loan applications, 1195 01:00:22,630 --> 01:00:25,020 because a lot of banks are trying to do it manually, 1196 01:00:25,020 --> 01:00:27,220 which is completely infeasible, given 1197 01:00:27,220 --> 01:00:29,660 the massive size of the rollout that they're proposing. 1198 01:00:29,660 --> 01:00:32,230 So thanks to Plaid and what they're trying to do, 1199 01:00:32,230 --> 01:00:34,630 at least, a lot of loans can be approved, 1200 01:00:34,630 --> 01:00:37,937 or at least applications submitted exponentially faster. 1201 01:00:37,937 --> 01:00:40,020 And that can make a huge difference for businesses 1202 01:00:40,020 --> 01:00:41,860 to survive. 1203 01:00:41,860 --> 01:00:44,440 GARY GENSLER: And I thank you for bringing that information 1204 01:00:44,440 --> 01:00:44,940 in. 1205 01:00:44,940 --> 01:00:46,510 Do you work with Plaid? 1206 01:00:46,510 --> 01:00:47,770 Or have you worked there? 1207 01:00:47,770 --> 01:00:50,265 Or is this something you're reading about online? 1208 01:00:50,265 --> 01:00:51,640 AUDIENCE: So I've been following, 1209 01:00:51,640 --> 01:00:53,920 because I work for a fintech-focused VC that 1210 01:00:53,920 --> 01:00:56,700 actually invested in Quovo, which was acquired by Plaid. 1211 01:00:56,700 --> 01:00:58,840 So I know the model pretty well. 1212 01:00:58,840 --> 01:01:02,410 GARY GENSLER: These companies have brought a great deal 1213 01:01:02,410 --> 01:01:04,690 more financial inclusion. 1214 01:01:04,690 --> 01:01:07,510 These companies have opened up the banking sector, 1215 01:01:07,510 --> 01:01:08,410 to some extent. 1216 01:01:08,410 --> 01:01:10,330 And now the banking sector was probably 1217 01:01:10,330 --> 01:01:13,270 going to be opened up by competitive pressures 1218 01:01:13,270 --> 01:01:15,430 in any regard. 1219 01:01:15,430 --> 01:01:20,900 But with the Plaids, earlier Stripe, the Galileos 1220 01:01:20,900 --> 01:01:23,350 and the Yodlees and others, what they've done 1221 01:01:23,350 --> 01:01:25,690 is they've brought some standardization 1222 01:01:25,690 --> 01:01:28,840 to that spaghetti wire that I put earlier-- 1223 01:01:28,840 --> 01:01:31,470 these interfaces. 1224 01:01:31,470 --> 01:01:34,180 They're in the weeds. 1225 01:01:34,180 --> 01:01:39,423 To some of us, it might just be mundane connections. 1226 01:01:39,423 --> 01:01:40,090 But they're not. 1227 01:01:40,090 --> 01:01:43,640 They're really, really important connections. 1228 01:01:43,640 --> 01:01:46,480 And I would say if you take anything out of this lecture 1229 01:01:46,480 --> 01:01:49,210 and moving forward, if you're at an incumbent 1230 01:01:49,210 --> 01:01:52,870 or you're at a startup, that you sort of think 1231 01:01:52,870 --> 01:01:54,920 about your API strategy. 1232 01:01:54,920 --> 01:01:56,290 What is it? 1233 01:01:56,290 --> 01:01:59,600 You might not manage it, but it's good to know what it is. 1234 01:01:59,600 --> 01:02:03,800 You might know what your competitors view of it. 1235 01:02:03,800 --> 01:02:08,350 But I agree with Carlos that in this time of crisis-- 1236 01:02:08,350 --> 01:02:12,670 the corona crisis in the US, the program that he's referencing, 1237 01:02:12,670 --> 01:02:19,180 was of this $2 trillion package that Congress just passed, 1238 01:02:19,180 --> 01:02:23,570 there was $350 billion to fund small businesses. 1239 01:02:23,570 --> 01:02:25,590 They have to be 500 employees or less. 1240 01:02:25,590 --> 01:02:28,750 But small businesses. 1241 01:02:28,750 --> 01:02:34,233 It's called a Payroll Protection Program, PPP. 1242 01:02:34,233 --> 01:02:36,150 And through the Small Business Administration, 1243 01:02:36,150 --> 01:02:40,570 it guarantees the banking sector can 1244 01:02:40,570 --> 01:02:42,580 make loans that are forgiven. 1245 01:02:42,580 --> 01:02:47,430 So they're forgivable loans up to $10 million a piece. 1246 01:02:47,430 --> 01:02:49,680 They're a loan, but they're really 1247 01:02:49,680 --> 01:02:53,580 a grant if you can show in your application 1248 01:02:53,580 --> 01:02:58,030 that the money is going to support employment. 1249 01:02:58,030 --> 01:03:03,060 And there's various conditions and conditionality. 1250 01:03:03,060 --> 01:03:04,830 And so what Carlos is talking about 1251 01:03:04,830 --> 01:03:08,370 is in the loan applications, the loan applications 1252 01:03:08,370 --> 01:03:10,680 were first being filed I think last Friday, 1253 01:03:10,680 --> 01:03:12,780 but it might have been over the weekend. 1254 01:03:12,780 --> 01:03:16,020 And these loan applications had to go through your employment 1255 01:03:16,020 --> 01:03:19,260 and show who you had earning less than a certain amount 1256 01:03:19,260 --> 01:03:20,980 of dollars per employee. 1257 01:03:20,980 --> 01:03:22,367 So a lot of data. 1258 01:03:22,367 --> 01:03:22,950 A lot of data. 1259 01:03:22,950 --> 01:03:25,200 And it sounds like Plaid is actually 1260 01:03:25,200 --> 01:03:28,250 helping facilitate that, in some regard. 1261 01:03:28,250 --> 01:03:28,860 So thank you. 1262 01:03:34,010 --> 01:03:34,550 Romain. 1263 01:03:34,550 --> 01:03:37,820 I think I'm getting close to that 15 minute warning time. 1264 01:03:37,820 --> 01:03:39,860 But is there another question, or are we 1265 01:03:39,860 --> 01:03:41,780 just going to move on to robotic process 1266 01:03:41,780 --> 01:03:42,860 automation for a minute? 1267 01:03:42,860 --> 01:03:44,360 PROFESSOR 2: I think we can move on. 1268 01:03:44,360 --> 01:03:46,820 And indeed, we have 17 minutes left. 1269 01:03:46,820 --> 01:03:47,960 GARY GENSLER: All right. 1270 01:03:47,960 --> 01:03:49,910 So let me just pause for a minute 1271 01:03:49,910 --> 01:03:52,700 again on financial data exchange. 1272 01:03:52,700 --> 01:03:56,280 I'm not close enough to it to know 1273 01:03:56,280 --> 01:03:58,020 where they are in their development. 1274 01:03:58,020 --> 01:04:02,590 But basically, a large set of big incumbents-- and I've 1275 01:04:02,590 --> 01:04:05,860 just put a visual of their principal members. 1276 01:04:05,860 --> 01:04:09,370 But then they have behind this another 50 or 100 second 1277 01:04:09,370 --> 01:04:10,600 tier members. 1278 01:04:10,600 --> 01:04:13,830 And by second tier, this is the group 1279 01:04:13,830 --> 01:04:17,800 that's primarily running its board of directors and sort 1280 01:04:17,800 --> 01:04:19,150 of coordinating. 1281 01:04:19,150 --> 01:04:23,230 And it has all the big US banks on here, the largest-- 1282 01:04:23,230 --> 01:04:25,160 Quicken Loans, which is the largest 1283 01:04:25,160 --> 01:04:28,030 in mortgage lending in the US, and CIFBA, 1284 01:04:28,030 --> 01:04:30,633 which is the trade association. 1285 01:04:30,633 --> 01:04:32,050 They're coming together to say, we 1286 01:04:32,050 --> 01:04:35,110 need to standardize financial data sharing standards 1287 01:04:35,110 --> 01:04:41,200 for secure authentication and how this user experience works. 1288 01:04:41,200 --> 01:04:42,520 But part of it-- 1289 01:04:42,520 --> 01:04:44,530 well, part of it is the incumbents sort 1290 01:04:44,530 --> 01:04:49,403 of trying to reclaim a little bit of the territory. 1291 01:04:49,403 --> 01:04:51,820 I don't think the cat's going to be able to be pushed back 1292 01:04:51,820 --> 01:04:55,210 into the barn, so to speak. 1293 01:04:55,210 --> 01:04:57,320 Or you might say the horse is out of the barn, 1294 01:04:57,320 --> 01:04:59,170 and I won't use the cat. 1295 01:04:59,170 --> 01:05:03,310 But I think that this is an interesting dynamic that's 1296 01:05:03,310 --> 01:05:05,440 going on right now. 1297 01:05:05,440 --> 01:05:09,010 Now, you put that in the context of what we talked about, 1298 01:05:09,010 --> 01:05:10,930 robotic process automation. 1299 01:05:10,930 --> 01:05:16,200 Robotic process automation can be used in a positive way 1300 01:05:16,200 --> 01:05:19,830 by the incumbents and by the startups to do 1301 01:05:19,830 --> 01:05:24,480 account opening, onboarding, as we talked about earlier, 1302 01:05:24,480 --> 01:05:26,040 loan processing. 1303 01:05:26,040 --> 01:05:29,310 Anything that you can take and automate 1304 01:05:29,310 --> 01:05:32,810 that which humans are doing. 1305 01:05:32,810 --> 01:05:35,690 And I'm not sort of moving into chat bots 1306 01:05:35,690 --> 01:05:38,690 and the conversational interfaces 1307 01:05:38,690 --> 01:05:40,340 that we've already spoken about. 1308 01:05:40,340 --> 01:05:44,690 I'm talking about the more standardized documents, 1309 01:05:44,690 --> 01:05:49,550 reading the documents, and opening the accounts, 1310 01:05:49,550 --> 01:05:52,850 and, as was mentioned earlier and I maybe 1311 01:05:52,850 --> 01:05:54,380 should have put on this list, also 1312 01:05:54,380 --> 01:05:59,070 compliance for any money laundering 1313 01:05:59,070 --> 01:06:00,590 and Know Your Customer. 1314 01:06:00,590 --> 01:06:02,660 Those are very positive ways. 1315 01:06:02,660 --> 01:06:05,960 But it's also a way that challenger banks 1316 01:06:05,960 --> 01:06:10,020 and challengers can use it to scrape the data 1317 01:06:10,020 --> 01:06:12,960 and pull the data out of the incumbents, 1318 01:06:12,960 --> 01:06:16,350 as long as a customer permissions it. 1319 01:06:16,350 --> 01:06:23,290 So it's sort of both of those sides in a very positive way 1320 01:06:23,290 --> 01:06:26,440 across the entire field of finance-- 1321 01:06:26,440 --> 01:06:29,410 a very positive way that we can automate that which 1322 01:06:29,410 --> 01:06:31,740 humans are doing in the past. 1323 01:06:31,740 --> 01:06:36,910 This is frankly just the same as when the tractor came along 1324 01:06:36,910 --> 01:06:39,670 and we were automating what humans and the horse 1325 01:06:39,670 --> 01:06:42,580 and the cow and the oxen were doing in the field. 1326 01:06:42,580 --> 01:06:45,550 We can automate that which the human does. 1327 01:06:45,550 --> 01:06:48,180 But it changes the way that-- 1328 01:06:48,180 --> 01:06:52,020 the sort of dynamic between incumbents and startups. 1329 01:06:55,820 --> 01:06:59,660 And then I want to pause and see if there are questions. 1330 01:06:59,660 --> 01:07:02,905 And then I wanted to say bit on assignments, and then close. 1331 01:07:02,905 --> 01:07:05,030 PROFESSOR 2: We have a question in the chat for you 1332 01:07:05,030 --> 01:07:08,510 from Alex, a Sloan fellow, who tells us 1333 01:07:08,510 --> 01:07:11,210 that one of the readings one of the authors 1334 01:07:11,210 --> 01:07:14,130 ends with consumer will access financial services from one 1335 01:07:14,130 --> 01:07:15,590 central hub. 1336 01:07:15,590 --> 01:07:17,900 But then another author in another reading 1337 01:07:17,900 --> 01:07:21,230 speaks of another trend of embedded fintech. 1338 01:07:21,230 --> 01:07:22,820 So which ones of these perspectives 1339 01:07:22,820 --> 01:07:25,940 would you tend to agree with? 1340 01:07:25,940 --> 01:07:28,880 GARY GENSLER: Well Alex, I think that's the debate. 1341 01:07:28,880 --> 01:07:35,960 And I think you probably could guess that I'm at neither end. 1342 01:07:35,960 --> 01:07:46,200 I don't think that we'll end up with truly decentralized 1343 01:07:46,200 --> 01:07:48,720 and we can sort of disaggregate. 1344 01:07:48,720 --> 01:07:50,520 We had this discussion a little earlier. 1345 01:07:50,520 --> 01:07:52,950 Will it be completely unbundled? 1346 01:07:52,950 --> 01:07:53,920 I don't think so. 1347 01:07:53,920 --> 01:08:00,690 I think there's still going to be economic benefits to scale. 1348 01:08:00,690 --> 01:08:03,440 Sometimes that's driven by balance sheet. 1349 01:08:03,440 --> 01:08:08,690 The balance sheet financing side leads to scale, 1350 01:08:08,690 --> 01:08:12,530 though securitization and asset securitization as a way 1351 01:08:12,530 --> 01:08:14,620 to rent somebody else's balance sheet 1352 01:08:14,620 --> 01:08:17,870 and to sort of lay off that risk to somebody else. 1353 01:08:17,870 --> 01:08:21,260 But I think that there's a lot of reasons that scale 1354 01:08:21,260 --> 01:08:23,120 and network effects-- 1355 01:08:23,120 --> 01:08:25,910 balance sheet being one of the network effects, data 1356 01:08:25,910 --> 01:08:29,029 being another network effect, customer user 1357 01:08:29,029 --> 01:08:31,399 interface another network effect-- 1358 01:08:31,399 --> 01:08:36,290 that there's benefits to some scale, 1359 01:08:36,290 --> 01:08:38,810 that there will be some concentration. 1360 01:08:38,810 --> 01:08:40,520 Will there be one portal? 1361 01:08:40,520 --> 01:08:43,700 Will we go all the way to one portal? 1362 01:08:43,700 --> 01:08:47,120 Will it be that Amazon will sort of 1363 01:08:47,120 --> 01:08:53,000 be the one place I go to for all of my online financial needs? 1364 01:08:53,000 --> 01:08:55,220 Jeff Bezos might have a strategy that way. 1365 01:08:55,220 --> 01:08:57,439 You might write papers that way. 1366 01:08:57,439 --> 01:08:59,450 I would caution that I think that there 1367 01:08:59,450 --> 01:09:02,270 will be some competitive pressures 1368 01:09:02,270 --> 01:09:04,310 and economic pressures on the other side-- 1369 01:09:04,310 --> 01:09:07,220 that we'll still have some disaggregation. 1370 01:09:07,220 --> 01:09:10,460 For instance, in country after country, 1371 01:09:10,460 --> 01:09:13,189 for decades, there's been the debate 1372 01:09:13,189 --> 01:09:16,670 whether insurance products will be sold by banks and banking 1373 01:09:16,670 --> 01:09:18,920 products by insurance firms. 1374 01:09:18,920 --> 01:09:20,760 And for some reason around the globe 1375 01:09:20,760 --> 01:09:23,840 by and large, we still see this separation. 1376 01:09:23,840 --> 01:09:26,930 Some of it's just legacy and history. 1377 01:09:26,930 --> 01:09:29,180 Some of it is regulation. 1378 01:09:29,180 --> 01:09:31,220 But it's also how we think of the products, 1379 01:09:31,220 --> 01:09:35,630 that we think of buying our life insurance 1380 01:09:35,630 --> 01:09:38,323 or our household insurance or auto insurance 1381 01:09:38,323 --> 01:09:39,740 a little differently than we think 1382 01:09:39,740 --> 01:09:41,779 about funding the same things. 1383 01:09:41,779 --> 01:09:46,609 Funding our house is called a mortgage and funding our car 1384 01:09:46,609 --> 01:09:49,880 is what's called an auto loan. 1385 01:09:49,880 --> 01:09:53,359 Now, it's true in some countries, it's come together. 1386 01:09:53,359 --> 01:09:55,580 But by and large, it's not one portal. 1387 01:09:58,870 --> 01:10:01,170 But Alex, do you have a point of view? 1388 01:10:01,170 --> 01:10:03,810 Did you did you feel strongly one way or the other, 1389 01:10:03,810 --> 01:10:05,740 if you're there still? 1390 01:10:05,740 --> 01:10:13,590 AUDIENCE: I think the embedded fintech is quite insightful, 1391 01:10:13,590 --> 01:10:16,590 because nowadays that people talk 1392 01:10:16,590 --> 01:10:20,490 about customer-centric ideas even more than before, 1393 01:10:20,490 --> 01:10:24,390 because we have more data to provide relevant services. 1394 01:10:24,390 --> 01:10:29,580 But the thing is that when we talk about the jobs to be done, 1395 01:10:29,580 --> 01:10:32,580 financial services are not actually the job. 1396 01:10:32,580 --> 01:10:38,340 The ultimate job you want to get done, you want to buy house, 1397 01:10:38,340 --> 01:10:40,620 so you need a loan. 1398 01:10:40,620 --> 01:10:45,060 Which means that maybe it makes sense, some of the services 1399 01:10:45,060 --> 01:10:46,360 will be the second layer. 1400 01:10:46,360 --> 01:10:49,840 I mean, be embedded into those jobs that really need 1401 01:10:49,840 --> 01:10:51,952 to be done by the customers. 1402 01:10:51,952 --> 01:10:53,910 GARY GENSLER: I want to say this about embedded 1403 01:10:53,910 --> 01:10:57,840 finance and this concept that was in the article. 1404 01:10:57,840 --> 01:11:00,990 And I'm glad you reminded me of this, Alex. 1405 01:11:00,990 --> 01:11:03,690 I do think that we're going to see more 1406 01:11:03,690 --> 01:11:05,030 what's called embedded finance. 1407 01:11:05,030 --> 01:11:08,710 And by that term, what I take from it-- 1408 01:11:08,710 --> 01:11:10,390 and it's not a well-defined term. 1409 01:11:10,390 --> 01:11:12,550 You can't look at a Webster's dictionary 1410 01:11:12,550 --> 01:11:14,620 and say what's embedded finance. 1411 01:11:14,620 --> 01:11:20,560 But what I take from that term is that non-financial firms 1412 01:11:20,560 --> 01:11:23,350 will embed some aspect of finance 1413 01:11:23,350 --> 01:11:25,420 in their product offering. 1414 01:11:25,420 --> 01:11:26,890 Not a new concept. 1415 01:11:26,890 --> 01:11:30,040 Even the local grocery store in the 19th century 1416 01:11:30,040 --> 01:11:34,890 was saying you can buy your groceries and pay tomorrow. 1417 01:11:34,890 --> 01:11:36,670 You know, that local merchant. 1418 01:11:36,670 --> 01:11:40,300 And that led over 50 or 70 years to the modern credit 1419 01:11:40,300 --> 01:11:46,390 card being invented in the 1940s and taking off by the 1960s. 1420 01:11:46,390 --> 01:11:49,230 So we've had some concepts of embedded finance. 1421 01:11:49,230 --> 01:11:52,530 But I think that the modern technology 1422 01:11:52,530 --> 01:11:57,720 will provide more alternatives to embed a payment mechanism 1423 01:11:57,720 --> 01:11:59,380 or a lending mechanism. 1424 01:11:59,380 --> 01:12:01,800 And those will be the two principal things. 1425 01:12:01,800 --> 01:12:06,230 Some aspect of a loan or some aspect of payment right 1426 01:12:06,230 --> 01:12:07,890 in an app. 1427 01:12:07,890 --> 01:12:11,460 In a sense, like Uber and Lyft does. 1428 01:12:11,460 --> 01:12:16,850 In a sense, like of point-of-sale lending-- 1429 01:12:16,850 --> 01:12:21,140 that I can buy something online and pay 1430 01:12:21,140 --> 01:12:25,700 on installment over four or six payments, 1431 01:12:25,700 --> 01:12:28,870 maybe over multiple months. 1432 01:12:28,870 --> 01:12:32,170 I think embedded finance is a trend. 1433 01:12:32,170 --> 01:12:37,940 I also do think that because of this API movement-- 1434 01:12:37,940 --> 01:12:40,660 and this is a way to close on APIs-- 1435 01:12:40,660 --> 01:12:44,170 that startups can provide services in a way 1436 01:12:44,170 --> 01:12:46,270 they couldn't in the past. 1437 01:12:46,270 --> 01:12:48,460 That a Robinhood could come along-- 1438 01:12:48,460 --> 01:12:50,380 and we'll talk about Robinhood more when 1439 01:12:50,380 --> 01:12:51,820 we talk about capital markets. 1440 01:12:51,820 --> 01:12:54,400 But a Robinhood could come along and say, 1441 01:12:54,400 --> 01:12:58,210 we offer that you can open brokerage accounts 1442 01:12:58,210 --> 01:13:01,690 on this handheld device. 1443 01:13:01,690 --> 01:13:05,110 You can scan your documents using robotic process 1444 01:13:05,110 --> 01:13:07,180 automation and natural language processing. 1445 01:13:07,180 --> 01:13:11,250 You can scan your documents by taking a picture 1446 01:13:11,250 --> 01:13:15,420 from your phone, open a brokerage account, 1447 01:13:15,420 --> 01:13:19,920 and trade automatically, interfacing-- there's probably 1448 01:13:19,920 --> 01:13:23,730 dozens of APIs that stand between your Robinhood 1449 01:13:23,730 --> 01:13:26,970 app and the New York Stock Exchange. 1450 01:13:26,970 --> 01:13:28,890 And it's not just Robinhood directly 1451 01:13:28,890 --> 01:13:30,210 to the New York Stock Exchange. 1452 01:13:30,210 --> 01:13:34,780 There's multiple interfaces in between. 1453 01:13:34,780 --> 01:13:39,210 So I think it's facilitated a remarkable shift in finance, 1454 01:13:39,210 --> 01:13:43,230 that we can do those things right on our cell phones. 1455 01:13:43,230 --> 01:13:48,810 And embedded finance and these APIs relate to it. 1456 01:13:48,810 --> 01:13:50,850 I'm just not quite there that it will all 1457 01:13:50,850 --> 01:13:54,090 be bundled up into one service provider, 1458 01:13:54,090 --> 01:13:56,678 or that it will all be fragmented. 1459 01:13:56,678 --> 01:13:58,470 I think we'll have a little bit in between. 1460 01:14:02,080 --> 01:14:04,080 I just want to mention something on assignments, 1461 01:14:04,080 --> 01:14:06,060 because the April 8th. 1462 01:14:06,060 --> 01:14:09,870 Again, assignments-- particularly at this stage, 1463 01:14:09,870 --> 01:14:13,800 where we're pass emergency, no credit emergency, 1464 01:14:13,800 --> 01:14:16,770 or sort of colloquially pass/fail-- 1465 01:14:16,770 --> 01:14:19,560 assignments are a great way to engage in this subject. 1466 01:14:19,560 --> 01:14:23,490 I think of that even when we have A's and B's and things 1467 01:14:23,490 --> 01:14:24,780 like that. 1468 01:14:24,780 --> 01:14:27,570 I just really ask you to engage in this subject. 1469 01:14:27,570 --> 01:14:31,260 Pick some subject that interests you, that excites you, 1470 01:14:31,260 --> 01:14:33,690 and you want to write to an incumbent, 1471 01:14:33,690 --> 01:14:38,490 to a big tech firm like Amazon or Andreessen Horowitz. 1472 01:14:38,490 --> 01:14:42,270 Which Andreessen Horowitz owns a slice of a lot of the companies 1473 01:14:42,270 --> 01:14:44,850 we've been talking about already. 1474 01:14:44,850 --> 01:14:48,060 I mean, they're a remarkable venture capital firm. 1475 01:14:48,060 --> 01:14:51,900 And just think of some sort of strategy around it. 1476 01:14:51,900 --> 01:14:53,880 The individual papers are just a way 1477 01:14:53,880 --> 01:14:56,340 to break up the assignment in a way 1478 01:14:56,340 --> 01:14:58,410 that one person would write about it. 1479 01:14:58,410 --> 01:15:01,050 Overall market analysis, somebody 1480 01:15:01,050 --> 01:15:02,850 doing the technological trends. 1481 01:15:02,850 --> 01:15:10,120 If you pick that you're writing to Amazon about, let's say, 1482 01:15:10,120 --> 01:15:14,580 the consumer lending space, what are the technological trends 1483 01:15:14,580 --> 01:15:18,180 really doing to that space? 1484 01:15:18,180 --> 01:15:19,920 Who are the traditional competitors 1485 01:15:19,920 --> 01:15:21,990 and who are the big tech competitors? 1486 01:15:21,990 --> 01:15:26,850 And just breaking it up three or four pages, 900 to 1,200 words. 1487 01:15:26,850 --> 01:15:30,473 Again, we're just helping you kind of engage in a subject. 1488 01:15:30,473 --> 01:15:32,640 PROFESSOR 2: Perhaps you can answer another question 1489 01:15:32,640 --> 01:15:35,015 that I've received from a student who does not understand 1490 01:15:35,015 --> 01:15:36,135 how Plaid makes money. 1491 01:15:36,135 --> 01:15:37,260 What is the business model? 1492 01:15:40,560 --> 01:15:42,950 GARY GENSLER: With only $110 million in revenues-- now, 1493 01:15:42,950 --> 01:15:45,050 that's a Forbes estimate-- 1494 01:15:45,050 --> 01:15:46,300 you sort of go, wait a minute? 1495 01:15:46,300 --> 01:15:47,680 How does this work? 1496 01:15:47,680 --> 01:15:51,020 But how the space makes money-- 1497 01:15:51,020 --> 01:15:56,030 not just Plaid, but how API companies make 1498 01:15:56,030 --> 01:16:02,150 money in finance and outside of finance is they do charge fees. 1499 01:16:02,150 --> 01:16:05,750 I apologize, I don't know if it's a penny a transaction 1500 01:16:05,750 --> 01:16:06,560 or what it is. 1501 01:16:06,560 --> 01:16:08,770 And maybe somebody on this call actually knows-- 1502 01:16:08,770 --> 01:16:12,560 Gita or others know their actual revenue model. 1503 01:16:12,560 --> 01:16:18,560 But charging a very small fraction of money per trade. 1504 01:16:18,560 --> 01:16:21,650 And then what they do is data aggregation. 1505 01:16:21,650 --> 01:16:24,260 And with that data in other fields, 1506 01:16:24,260 --> 01:16:26,960 then you try to see how you can commercialize 1507 01:16:26,960 --> 01:16:34,430 that data by cross-selling either directly. 1508 01:16:34,430 --> 01:16:36,350 Or remember, a lot of these companies 1509 01:16:36,350 --> 01:16:40,640 like Plaid and Galileo and others are in between. 1510 01:16:40,640 --> 01:16:45,170 They're behind the scenes, between one side of the market 1511 01:16:45,170 --> 01:16:47,040 and another side of the market. 1512 01:16:47,040 --> 01:16:50,780 But when you create that like Twilio, 1513 01:16:50,780 --> 01:16:56,630 Twilio is an 80% market share between voice and messaging 1514 01:16:56,630 --> 01:16:59,300 and things like that-- in the US, at least. 1515 01:16:59,300 --> 01:17:01,190 So all of a sudden, then you can either 1516 01:17:01,190 --> 01:17:03,500 enhance your fee structure, because you're 1517 01:17:03,500 --> 01:17:07,090 such dominant oligopolies, sort of. 1518 01:17:07,090 --> 01:17:09,190 You can collect economic rent. 1519 01:17:09,190 --> 01:17:11,020 Or you commercialize the data. 1520 01:17:11,020 --> 01:17:13,720 And it's the data aggregation side-- 1521 01:17:13,720 --> 01:17:18,340 I think Visa's paying what might be 25 or 50 times revenues 1522 01:17:18,340 --> 01:17:21,190 for that data aggregation side.