WEBVTT - This Is Why Credit Card Interest Rates Are So High 0:00:02.720 --> 0:00:15.840 Bloomberg Audio Studios, Podcasts, radio News. 0:00:18.640 --> 0:00:22.079 Hello and welcome to another episode of the Odd Lots podcast. 0:00:22.160 --> 0:00:24.400 I'm Jolle Wisenthal and I'm Tracy Alloway. 0:00:24.520 --> 0:00:28.000 Tracy, are you good about like frequent flyer miles and 0:00:28.040 --> 0:00:31.880 hotel rewards and cash back and using your credit card 0:00:32.040 --> 0:00:34.360 to get like good seats at the US Open like 0:00:34.400 --> 0:00:38.120 all the dining? Are you good about maximizing that stuff? 0:00:38.280 --> 0:00:39.120 Nope, I am not. 0:00:39.680 --> 0:00:42.520 I'm trying to be better, you know, I'm finally signing 0:00:42.560 --> 0:00:44.720 up to a bunch of frequent flyer programs and things 0:00:44.760 --> 0:00:48.960 like that. But in general, I am not a points strategy. 0:00:49.200 --> 0:00:50.880 Some people get really into it. 0:00:51.080 --> 0:00:51.239 Now. 0:00:51.280 --> 0:00:53.800 I do not have a very busy life. I do 0:00:53.840 --> 0:00:58.720 not have mental energy towards, you know, maximizing points or 0:00:58.760 --> 0:01:01.000 learning about the newest cars. It's like, oh's this card 0:01:01.000 --> 0:01:03.480 worth a four hundred dollars fee because I can get 0:01:03.640 --> 0:01:06.640 upgraded to Platinum faster this year. I do not want 0:01:06.680 --> 0:01:08.720 to think about that stuff. I'm not that interested. But 0:01:08.800 --> 0:01:12.080 I get the impression that means I'm probably paying for 0:01:12.120 --> 0:01:15.319 someone who is or something like that, or maybe you know, 0:01:15.319 --> 0:01:17.720 I'm paying these fees on my credit cards or these 0:01:17.720 --> 0:01:20.679 interchange fees, et cetera. Maybe I'm leaving money on the 0:01:20.720 --> 0:01:22.039 table by not doing it. 0:01:22.200 --> 0:01:22.440 I don't know. 0:01:22.480 --> 0:01:24.399 I find credit cards to be a weird business, Like 0:01:24.560 --> 0:01:27.240 I don't really know what visa does relative to say, 0:01:27.240 --> 0:01:30.120 the bank that issues a visa card, et cetera, to 0:01:30.120 --> 0:01:31.640 know how they slice them. I don't know any about 0:01:31.640 --> 0:01:32.200 credit cards. 0:01:32.520 --> 0:01:35.560 It's a very opaque business, for sure, and it's a 0:01:35.600 --> 0:01:38.680 weird business. I would say, like it's competitive, but also 0:01:38.760 --> 0:01:41.760 it's like not, you know, like everyone's kind of doing 0:01:41.800 --> 0:01:44.960 the same thing in many ways, so we should talk 0:01:44.959 --> 0:01:47.960 about it. It's also, I imagine, kind of sticky, in 0:01:48.000 --> 0:01:50.360 the same way that deposits at banks are sticky. 0:01:50.440 --> 0:01:52.960 We spoke with Joe Obote about that a while back. 0:01:53.000 --> 0:01:56.160 So I don't cycle through them a bunch and stuff 0:01:56.200 --> 0:01:59.040 like that. And there's so much credit card advertising. I 0:01:59.040 --> 0:02:00.760 don't know what's good, bad or whatever. 0:02:00.800 --> 0:02:01.000 I know. 0:02:01.120 --> 0:02:03.520 Yeah, Look, I use my credit card as a payments 0:02:03.520 --> 0:02:05.760 card because I don't really I don't carry a balance 0:02:05.840 --> 0:02:07.480 from month to month, so I don't I don't know, 0:02:07.480 --> 0:02:09.520 I think my interest rates or whatever. I pay it 0:02:09.560 --> 0:02:11.960 off at the end of every month because I just 0:02:11.960 --> 0:02:14.360 basically use it for payments et cetera. So I just 0:02:14.400 --> 0:02:17.280 don't know much about them, but they're a huge, major 0:02:17.320 --> 0:02:19.800 consumer financing source. Yeah, and every want to talk about 0:02:19.800 --> 0:02:22.440 fintech and BNPL and all these other things and stable 0:02:22.480 --> 0:02:24.880 coins and all this other stuff, and it's like, yeah, 0:02:24.919 --> 0:02:26.600 but the big one. Who's talking about the big one? 0:02:26.639 --> 0:02:27.200 Credit cards? 0:02:27.200 --> 0:02:27.919 Well that's the thing. 0:02:28.000 --> 0:02:31.960 So points have become a bigger attractant, I guess, to 0:02:32.040 --> 0:02:34.919 credit cards, and so people are spending more with their 0:02:34.960 --> 0:02:38.240 credit cards and carrying a bigger balance, which means that 0:02:38.360 --> 0:02:40.200 the rate that you're paying on the credit card is 0:02:40.200 --> 0:02:44.079 actually more important potentially than something like your mortgage rate. 0:02:44.360 --> 0:02:47.880 Totally well pleased to say. We do, in fact have 0:02:48.080 --> 0:02:52.000 the perfect guests. Someone we've had on the podcast before. 0:02:52.480 --> 0:02:54.120 I think the last time we were talking about red Q, 0:02:54.320 --> 0:02:57.280 which is main lending in the seventies. I like his 0:02:57.440 --> 0:02:59.720 work because he goes back to the simple things like 0:02:59.800 --> 0:03:01.799 let's talk about how this works. Let's talk about how 0:03:01.800 --> 0:03:04.000 this works, because I think we move on too quickly 0:03:04.040 --> 0:03:06.920 without sort of understanding the basics. Maybe there are stones 0:03:07.000 --> 0:03:07.800 left unturned. 0:03:08.200 --> 0:03:08.560 Literally. 0:03:08.560 --> 0:03:11.200 The perfect guest Itamar Drexler. He has a finance professor 0:03:11.240 --> 0:03:13.040 at Warden and he was the co author of a 0:03:13.080 --> 0:03:16.880 fairly recent paper sort of looking at the question of 0:03:17.120 --> 0:03:19.480 why are credit card rates so high? Because if you 0:03:19.520 --> 0:03:21.640 actually do borrow from them, sometimes the rates are like 0:03:21.639 --> 0:03:24.720 twenty something percent. Seems way higher than any other sort 0:03:24.720 --> 0:03:27.640 of lending. So itamar Thank you so much for coming 0:03:27.680 --> 0:03:28.959 back on odd lots. 0:03:29.160 --> 0:03:31.040 Thank you very much. It's really nice to be back. 0:03:31.040 --> 0:03:31.800 Thank you for having me. 0:03:31.919 --> 0:03:34.280 When I was doing some prep for this episode, there 0:03:34.400 --> 0:03:37.240 is not a ton of actually like fresh academic work 0:03:37.640 --> 0:03:40.560 on the credit card industry. There's not a ton of papers, 0:03:40.560 --> 0:03:42.000 but it's this huge space. 0:03:42.440 --> 0:03:43.080 Why did you. 0:03:43.200 --> 0:03:47.560 See a reason to go back and revisit the sort 0:03:47.560 --> 0:03:51.280 of basic simple question of looking at interest rates on 0:03:51.320 --> 0:03:51.960 credit cards? 0:03:52.120 --> 0:03:55.080 Yeah, so my interest is usually like we talked about 0:03:55.080 --> 0:03:57.680 when I was here last time's monetary policy, macro and 0:03:57.760 --> 0:04:00.840 out of banking and had some student to our co 0:04:00.920 --> 0:04:03.440 authors now on this paper a couple of years ago, 0:04:03.440 --> 0:04:05.800 I want to talk about fintech because fintech is a 0:04:05.920 --> 0:04:08.680 very popular topic. And then I was thinking, well, how 0:04:08.680 --> 0:04:12.120 do we analyze fintech and what's the potential room for 0:04:12.200 --> 0:04:15.840 fintech to grow if we don't really understand how the 0:04:16.200 --> 0:04:19.960 dominant incumbent players the credit card banks work. And then 0:04:19.960 --> 0:04:22.279 we look at this and I was very surprised to 0:04:22.320 --> 0:04:25.000 see something kind of simple, which is that the return 0:04:25.040 --> 0:04:27.839 on assets for credit card banks are just way higher 0:04:27.880 --> 0:04:32.200 than the average bank. So bank ROAs are typically you know, 0:04:32.279 --> 0:04:34.960 one one point two percent, they move a couple basis 0:04:34.960 --> 0:04:39.400 points are very exciting. Credit card banks ROAs and you know, 0:04:39.440 --> 0:04:41.760 most banks are not just credit cards, so it's actually 0:04:41.760 --> 0:04:43.560 even higher than this, are in the three and a 0:04:43.600 --> 0:04:46.520 half often four percent. So it's very shocked by this. 0:04:46.800 --> 0:04:48.680 How come it's so high when it's so hard to 0:04:48.680 --> 0:04:51.280 squeeze out a couple basis points. And then one of 0:04:51.320 --> 0:04:53.560 the reasons is just they you charge really high rates, 0:04:53.680 --> 0:04:55.240 Like okay, how did they get away with this? You 0:04:55.240 --> 0:04:58.120 know what is going on here? Just very simple question 0:04:58.200 --> 0:05:00.719 about how to decompose that rate into the pieces and 0:05:00.800 --> 0:05:02.160 kind of what's left over at the end. 0:05:02.400 --> 0:05:05.279 In the spirit of starting at basics, walk us through 0:05:05.920 --> 0:05:09.520 the revenue that credit card issuers or credit card banks 0:05:09.560 --> 0:05:12.960 are actually earning, the different types and who the players 0:05:12.960 --> 0:05:13.960 are in the system. 0:05:14.120 --> 0:05:18.120 Yeah, so let's separate first into two categories. One are 0:05:18.200 --> 0:05:21.640 people who revolve their balance, and that's what most of 0:05:21.680 --> 0:05:23.880 the papers about it I think that's the more interesting 0:05:24.080 --> 0:05:26.800 part and there's more details, and it's kind of the 0:05:26.839 --> 0:05:28.960 banking part of it. And there are actually a lot 0:05:28.960 --> 0:05:31.679 of people who revolve often. I find people are surprised 0:05:31.680 --> 0:05:34.279 to hear this, but about sixty percent of the credit 0:05:34.320 --> 0:05:37.400 card users actually revolve, so meaning that they don't pay 0:05:37.400 --> 0:05:38.839 in the grace period at the end of the month, 0:05:38.839 --> 0:05:42.440 and so they're hit with these very high usually interest charges. 0:05:42.680 --> 0:05:45.440 And then the other part are what people call transactors, 0:05:45.480 --> 0:05:48.800 so they're the kind that do pay during the grace periods, 0:05:48.800 --> 0:05:52.320 so they're not paying interest. Okay, So for the revolvers, 0:05:52.640 --> 0:05:55.760 there's again multiple parts, so you pay interest on the 0:05:55.760 --> 0:05:58.360 balance that you have. But before then there's the part 0:05:58.360 --> 0:06:01.200 which applies both to the transactor and revolvers. When you 0:06:01.240 --> 0:06:05.920 swipe the card, then there's immediately a percentage taken. People 0:06:05.960 --> 0:06:09.200 call it the swipe fee, and that is split up 0:06:09.200 --> 0:06:12.120 into a bunch of pieces. The ones that I used 0:06:12.120 --> 0:06:13.720 to be aware of that most people are aware of is 0:06:13.760 --> 0:06:18.400 the card network like Visa, MasterCard, Amex. There was Discover 0:06:18.600 --> 0:06:21.200 which is now part of Capital One, and that's there's 0:06:21.240 --> 0:06:25.479 a whole menu, but basically it's like fifteen twenty basis points. Okay, okay, 0:06:25.680 --> 0:06:27.800 it doesn't sound like a lot, but there's like ten 0:06:27.839 --> 0:06:31.400 trillion dollars of purchases between debit and credit cards. Turns 0:06:31.400 --> 0:06:34.040 out when you take twenty basis points of ten trillion 0:06:34.080 --> 0:06:35.120 dollars kind of adds. 0:06:34.960 --> 0:06:36.880 Up nice business if you can get it. 0:06:36.880 --> 0:06:39.480 It's really nice. Actually, be surprised that usually Visa and 0:06:39.560 --> 0:06:43.760 JP Morgan are the two most valuable financial services firms. 0:06:43.800 --> 0:06:47.160 They change who's number one, So Visa's, you know, been 0:06:47.200 --> 0:06:50.560 worth over six hundred billion dollars. It's a lot, yea 0:06:50.760 --> 0:06:54.400 and MasterCards is gigantic too, so there's that. Then the 0:06:54.480 --> 0:06:58.280 majority of that swipe fee, the majority of that remain there, 0:06:58.320 --> 0:07:01.000 actually goes to the bank that issued the card to 0:07:01.040 --> 0:07:05.000 the consumer. So that's called interchange fee. And again they 0:07:05.040 --> 0:07:07.240 don't make this like very easy to tell, but in 0:07:07.279 --> 0:07:10.040 our data it's a little over one point eight percent 0:07:10.080 --> 0:07:12.840 on average. I think it's largely been trending up over 0:07:12.880 --> 0:07:15.600 time slowly. So the bank gets that, it actually gets 0:07:15.720 --> 0:07:18.560 the vast majority of that, and then they pay your 0:07:18.600 --> 0:07:20.800 rewards and things. From that, a lot of that goes 0:07:20.800 --> 0:07:23.720 to just pass through to the rewards and things, and 0:07:23.800 --> 0:07:26.400 they keep a small portion of it for themselves. But 0:07:26.520 --> 0:07:29.720 the big part of their business where most of the 0:07:29.800 --> 0:07:32.720 money comes from that we analyze here is all these 0:07:32.720 --> 0:07:35.080 people that revolved. They pay an interest rate, and that 0:07:35.160 --> 0:07:38.640 interest rate now is on average twenty three percent, which 0:07:38.680 --> 0:07:40.600 is just like a shockingly high number. I mean, I 0:07:40.600 --> 0:07:42.560 guess I've seen that. It just when you work on 0:07:42.600 --> 0:07:44.440 assets and like, you know, think the kind of things 0:07:44.480 --> 0:07:47.720 you guys talk about bonds and bonds pay you know whatever, 0:07:47.760 --> 0:07:51.960 five percent investment grades spread is not even eighty basis points. Now, 0:07:52.000 --> 0:07:54.840 on top of it, high yield spreads under three percent, Like, 0:07:54.880 --> 0:07:56.840 how the hell do we get to twenty three percent? 0:07:57.720 --> 0:08:01.840 When you hear this number twenty three percent and you 0:08:01.920 --> 0:08:04.640 think about the fact that credit card users can be 0:08:04.760 --> 0:08:10.600 decomposed into transactors and revolvers, my first instinct would be, well, 0:08:11.080 --> 0:08:13.600 the transactors are very on the ball. They're like, not 0:08:13.720 --> 0:08:17.760 credit risks. I've always been just a transactor. I've never revolved. 0:08:18.600 --> 0:08:23.280 How much of that increase spread can just be explained 0:08:23.400 --> 0:08:28.400 by likelihood of default from the revolvers, which I presume 0:08:28.480 --> 0:08:31.760 are perhaps a little more you know, financially precurious and 0:08:31.800 --> 0:08:33.479 maybe less financially sophisticated. 0:08:33.800 --> 0:08:35.600 Right, So I think, if like me, you didn't know 0:08:35.679 --> 0:08:37.719 much about this, your assumption. If I think, if you 0:08:37.760 --> 0:08:40.680 ask most financial economists, the first thing they would think is, well, 0:08:40.800 --> 0:08:44.800 must be that most of the remainders is a charge offs, right, defaults, 0:08:45.360 --> 0:08:48.400 And that's not true. So you can find that pretty easily. 0:08:48.679 --> 0:08:52.280 So the average charge off rate on the revolvers, okay, 0:08:52.320 --> 0:08:53.640 so when you look at it, let's say you look 0:08:53.640 --> 0:08:56.440 it up online, you'll see kind of, you know, relative 0:08:56.440 --> 0:08:58.920 to the whole balance sheets includes both groups, and like 0:08:58.960 --> 0:09:02.320 you're saying, by definition, transactors don't borrow, so they can't default. 0:09:02.440 --> 0:09:04.319 That kind of makes it go down a little bit. 0:09:04.360 --> 0:09:06.880 But the majority our revolvers, So if we kind of 0:09:06.880 --> 0:09:10.320 clean that out, then on average in our sample it's 0:09:10.679 --> 0:09:14.440 five point seventy five percent of bounces are charged off, 0:09:14.720 --> 0:09:17.680 So it's not trivial by any means. That's a high number. 0:09:17.840 --> 0:09:20.200 But again we were talking about eighteen percent spread, so 0:09:20.240 --> 0:09:22.400 if you think, oh, it must be about eighteen percent 0:09:22.520 --> 0:09:25.280 charge us, it's not even close. And it's like never 0:09:25.320 --> 0:09:27.320 been that high, So you might think, well, maybe it's 0:09:27.440 --> 0:09:30.160 just that's on average, but sometimes it'll spike to be 0:09:30.280 --> 0:09:33.720 ridiculous numbers it does spike, but not for very long 0:09:33.760 --> 0:09:36.000 periods of time. So the bottom line is it's a 0:09:36.040 --> 0:09:38.120 substantial chunk of it, but not even close to a 0:09:38.160 --> 0:09:40.520 majority of it. So you know people default, but they 0:09:40.520 --> 0:09:42.120 don't default that much. 0:09:42.920 --> 0:09:45.440 Can I ask one more question on APR and the 0:09:45.520 --> 0:09:48.760 average there? Did you observe any trend over time? Like 0:09:48.880 --> 0:09:51.560 did the rate actually get higher as time went on? 0:09:51.920 --> 0:09:54.320 So that's something we haven't spent a lot of time 0:09:54.360 --> 0:09:56.360 on in this paper. But the answer to your question 0:09:56.480 --> 0:09:59.160 is this is obviously yes. So if you look at it. 0:09:59.200 --> 0:10:01.440 I think what can be found online is again it's 0:10:01.600 --> 0:10:04.079 I think there's something a little misleading there, but that 0:10:04.200 --> 0:10:06.600 has trended up pretty strongly. I think not as much 0:10:06.600 --> 0:10:08.800 as somebody you know goes to their computer and looks up. 0:10:09.040 --> 0:10:11.680 I'll say, Fred from the Call reports, what is the 0:10:11.679 --> 0:10:13.719 average APR? It looks crazy. It looks like it's gone 0:10:13.760 --> 0:10:16.839 up ten percent. It's gone up. We're gonna get to 0:10:16.880 --> 0:10:18.559 the bottom of like exactly how much I think it's 0:10:18.559 --> 0:10:21.880 gone up substantially since ten years ago. Let's say that 0:10:21.920 --> 0:10:23.600 trend is clear. I don't think it's as much as 0:10:23.640 --> 0:10:26.079 it looks like there, but yeah, it's been going up. 0:10:26.120 --> 0:10:30.560 Actually, okay, so you've established default rates for credit cards, 0:10:30.880 --> 0:10:34.200 and as you said, like this business is about volume, right, 0:10:34.240 --> 0:10:37.080 so is there an argument to be made that maybe 0:10:37.840 --> 0:10:41.240 if the world, you know, falls apart, then you have 0:10:41.640 --> 0:10:45.360 lots and lots and lots of consumers who are defaulting, 0:10:45.520 --> 0:10:47.800 but you know, potentially at a low rate of the total. 0:10:47.840 --> 0:10:50.280 But the volume makes it meaningful for banks. 0:10:50.320 --> 0:10:52.120 I mean, we are already looking at as a percentage 0:10:52.120 --> 0:10:55.480 of assets, so that kind of like valuates, it takes 0:10:55.520 --> 0:10:58.680 it all into consideration. I think the question, like in 0:10:58.720 --> 0:11:02.319 our minds was at first, you know, maybe in a crisis, 0:11:02.800 --> 0:11:06.079 something extreme happens. In a sense, it does, but this 0:11:06.120 --> 0:11:08.880 is already the average default rates so usually it's lower, 0:11:08.920 --> 0:11:10.679 and then you kind of include this in there, so 0:11:11.080 --> 0:11:13.200 we'll talk I guess a little bit about risk premium, 0:11:13.240 --> 0:11:16.079 which it turns out to be very clear here and important, 0:11:16.520 --> 0:11:18.560 but it's you know, just the average default rate is 0:11:18.600 --> 0:11:20.680 what it is, so it's not in expected defaults again 0:11:20.720 --> 0:11:23.320 they're it's surprising. But at the same time, if you 0:11:23.400 --> 0:11:26.520 just look at where banks actually suffer default losses in 0:11:26.559 --> 0:11:29.319 an average year, not a crisis year, something like fifty 0:11:29.360 --> 0:11:33.120 percent of banks default losses are actually coming from credit cards. 0:11:33.240 --> 0:11:35.320 The thing is is that they're not surprising, they're not 0:11:35.400 --> 0:11:38.960 unexpected losses. They're kind of the expected, but it's still 0:11:38.960 --> 0:11:41.439 really big. And the reason for that is even though 0:11:41.440 --> 0:11:44.000 credit cards only take up about five percent of banks 0:11:44.000 --> 0:11:47.760 balance sheet, the charge offs or the defaults on average 0:11:47.840 --> 0:11:49.800 bank assets is very low. I mean, you know, we 0:11:49.840 --> 0:11:52.840 have this impression of banks as being these like crazy, 0:11:53.080 --> 0:11:55.839 like risk taking lunatic to actually, I think the right 0:11:56.000 --> 0:11:58.080 way to look at them is that their average asset 0:11:58.120 --> 0:12:00.560 is extremely boring and low risk. They do take a 0:12:00.559 --> 0:12:03.400 lot of leverage, which is only possible because the average 0:12:03.400 --> 0:12:06.440 asset is extremely boring and low risk. But even after 0:12:06.520 --> 0:12:09.800 all that, their amount of defaults is not really that high. 0:12:09.840 --> 0:12:13.120 So if your average asset has about forty basis points 0:12:13.160 --> 0:12:16.000 average charge off and this thing has over five percent, 0:12:16.240 --> 0:12:18.320 then even if it's only five percent of the balance sheet, 0:12:18.400 --> 0:12:20.720 it can act like it's fifty percent of the charge of. 0:12:36.679 --> 0:12:40.520 Other forms of consumer borrowing. Right, Like, people are very 0:12:40.559 --> 0:12:44.040 assiduous about making their car payments making I don't know 0:12:44.080 --> 0:12:49.079 if up until recently, until recently, but historically the perception 0:12:49.280 --> 0:12:53.559 was people really prioritize car payments, right because that's essential 0:12:53.600 --> 0:12:56.360 to live and you can't have your car repossessed. Home 0:12:56.400 --> 0:13:01.080 mortgages obviously for obvious reason. I imagine that a stretched 0:13:01.200 --> 0:13:06.520 household will miss credit card payments are more inclined to 0:13:06.640 --> 0:13:08.200 if they're going to have to miss a payment, it's 0:13:08.200 --> 0:13:10.559 going to be there versus some of these other popular 0:13:10.559 --> 0:13:11.720 areas are borrowing. 0:13:11.600 --> 0:13:14.520 Totally, so the other ones are secured, and this is unsecured. 0:13:14.679 --> 0:13:18.400 So in that sense, this is an actual asset that's 0:13:18.600 --> 0:13:21.360 kind of risky and interesting for my vantage point, and 0:13:21.440 --> 0:13:26.200 that it's unsecured lending to normal people. All the rest 0:13:26.200 --> 0:13:29.200 of the stuff is secured, Like homes are obviously very important, 0:13:29.200 --> 0:13:32.320 collateral cars pretty much so the rates on those are 0:13:32.400 --> 0:13:35.720 much much lower. In these, they're nowhere near you know, 0:13:35.760 --> 0:13:37.920 the spread, there's nowhere near as juicy. I mean, when 0:13:37.920 --> 0:13:39.800 I teach students, we go through, like you know, the 0:13:39.880 --> 0:13:43.480 hierarchy of borrowing that the vast majority of borrowing is secured. 0:13:43.880 --> 0:13:45.880 You have to think it's crazy for a bank to 0:13:45.920 --> 0:13:49.160 come to somebody with the medium or lower credit score 0:13:49.200 --> 0:13:51.280 and say here have the line of credit of like 0:13:51.360 --> 0:13:54.760 five ten thousand dollars and you can default on it. 0:13:54.960 --> 0:13:56.880 You don't get shot for that. It's it's part of 0:13:56.880 --> 0:13:58.160 the law, it's part of the game. 0:13:58.360 --> 0:13:58.960 Yeah. 0:13:59.040 --> 0:14:02.040 I was reading an art from life in nineteen seventy 0:14:02.080 --> 0:14:04.520 where they were talking about how credit cards are becoming 0:14:04.559 --> 0:14:06.360 a big thing, and oh my god, these credit card 0:14:06.440 --> 0:14:10.240 companies are just mailing out applications to Americans. It's like 0:14:10.320 --> 0:14:13.840 giving sugar to diabetics. That was their analogy, speaking of 0:14:13.960 --> 0:14:18.640 unsecured versus secured. I'm also looking right now at a 0:14:18.679 --> 0:14:22.800 website that claims to have invented the first credit card 0:14:23.480 --> 0:14:27.080 that's based on your stock portfolio, So borrowing against your. 0:14:26.920 --> 0:14:29.360 Stock portfolio with the card. I gotta say, the card 0:14:29.440 --> 0:14:32.800 does look pretty nice. It's made of glass. Maybe that 0:14:32.920 --> 0:14:35.800 tells you something intentional metaphor yeah, exactly. 0:14:36.360 --> 0:14:38.880 Okay, So if it's not about risk premiums, if the 0:14:38.960 --> 0:14:42.680 rate isn't compensating for something like default, could it be 0:14:42.840 --> 0:14:48.040 compensating for all the points and benefits that customers are accruing. 0:14:48.160 --> 0:14:51.480 So it's not just compensation for expected default. I want 0:14:51.520 --> 0:14:53.120 I want to separate them from the risk premium. The 0:14:53.200 --> 0:14:56.280 risk premium is kind of the competition for unexpected default, 0:14:56.280 --> 0:14:58.640 which turns out to be pretty big here. But let's 0:14:58.680 --> 0:15:00.600 go back and talk about the point and stuff. So 0:15:00.880 --> 0:15:03.400 I find that people are more excited to talk about 0:15:03.400 --> 0:15:07.480 points than anything. The term rewards was really a marketing flourish. 0:15:07.920 --> 0:15:12.040 So yeah, So in total number of dollars this interchange was, 0:15:12.280 --> 0:15:14.800 I mean, again you have to look at fine exact numbers, 0:15:14.800 --> 0:15:16.720 but for credit cards alone, I think it was over 0:15:16.800 --> 0:15:19.520 one hundred and fifty billion dollars. So, like the GDP 0:15:19.680 --> 0:15:24.040 of a medium sized country gets transferred as interchange, and 0:15:24.160 --> 0:15:26.600 we find that about eighty five percent of that gets 0:15:26.640 --> 0:15:30.200 transferred through as rewards, you could wonder, I think it'd 0:15:30.240 --> 0:15:32.440 be natural to say, what is the point of this? 0:15:32.640 --> 0:15:36.160 Why charge people one point eight percent and then pass 0:15:36.240 --> 0:15:39.760 through one point five to seven percent as rewards? Where well, 0:15:39.760 --> 0:15:41.560 at least some people like I guess you and I 0:15:41.760 --> 0:15:44.320 Joe don't pay that much attention. I think I have 0:15:44.520 --> 0:15:46.960 enormous amounts of United Miles I'm never going to use 0:15:47.280 --> 0:15:49.960 because I have to actually travel to whatever place to 0:15:50.040 --> 0:15:52.720 use them. So why is that? I mean, I think 0:15:52.720 --> 0:15:55.080 a good economic question, and people have tackled this. I 0:15:55.080 --> 0:15:57.480 do think it creates a very strong network effect, so 0:15:57.600 --> 0:16:00.960 you are not actually seeing a charge for this. It's 0:16:00.960 --> 0:16:03.760 the retailer that has to eat it, and if you 0:16:03.840 --> 0:16:06.520 do not use a card that gives rewards, you're not 0:16:06.600 --> 0:16:08.960 going to get in most cases a lower price. So 0:16:09.080 --> 0:16:11.720 there's a whole series of litigation and fights over the 0:16:11.800 --> 0:16:17.000 years amazing about what retailers can do to discriminate prices 0:16:17.040 --> 0:16:19.320 based people who using cards and hurt And I thought 0:16:19.360 --> 0:16:21.200 a couple of years ago. The last couple years, I'm 0:16:21.200 --> 0:16:24.080 seeing more restaurants give you back a percentage or not 0:16:24.160 --> 0:16:26.520 charge you a percentage if you didn't do that. But 0:16:26.680 --> 0:16:29.080 it's a little bit beyond my legal expertise to sometimes 0:16:29.160 --> 0:16:31.560 understand these because for the longest time, I think you 0:16:31.600 --> 0:16:35.280 could give people a discount, but you couldn't do a searcharge. 0:16:35.640 --> 0:16:38.600 There was some like legal discrimination between those things. And 0:16:38.640 --> 0:16:41.560 as a result, people mostly don't pay attention to that 0:16:41.680 --> 0:16:44.160 kind of thing, and so you really want to stay 0:16:44.200 --> 0:16:46.560 inside the network and it kind of keeps you there. 0:16:46.600 --> 0:16:48.200 Even if at the end it would be a total 0:16:48.240 --> 0:16:51.680 pass through, it still helps for them to keep this business. 0:16:52.160 --> 0:16:55.200 You know, it's interesting. There's this crypto company. Have you 0:16:55.240 --> 0:16:57.960 heard of a Blackbird. Yeah, it's a crypto thing and 0:16:57.960 --> 0:16:59.640 they have a bunch of restaurants. You sign up and 0:16:59.640 --> 0:17:01.400 you're like pay in a coin. I don't know exactly 0:17:01.480 --> 0:17:05.040 how it works, but I think that they have to 0:17:05.160 --> 0:17:08.040 in some way because in theory would be nice, like 0:17:08.160 --> 0:17:10.199 maybe we'll get a little bit in the stable coins. 0:17:10.240 --> 0:17:11.960 It's like a payments rail in the future or in 0:17:12.000 --> 0:17:13.840 this conversation, and think it would be a nice way 0:17:13.880 --> 0:17:17.760 to like circumvent this. But even they, I think implicitly 0:17:17.880 --> 0:17:21.000 have to reinvent the rewards model to do it. Maybe 0:17:21.000 --> 0:17:24.119 you get premium seeds or you get reservations, et cetera. 0:17:24.600 --> 0:17:27.120 In order to sort of like bootstrap a new network, 0:17:27.920 --> 0:17:30.639 you start end up having to reinvent a lot of 0:17:30.640 --> 0:17:33.280 the rebates and the benefits, et cetera that come up 0:17:33.320 --> 0:17:35.560 the old network. Maybe we'll get into crypto a little 0:17:35.560 --> 0:17:37.480 bit more, but talk to us a little bit more 0:17:37.520 --> 0:17:42.320 then about like the persistence of this spread that can't 0:17:42.320 --> 0:17:43.960 fully be explained by defaults. 0:17:44.040 --> 0:17:46.280 Yes, so the default, like I said, is like a 0:17:46.320 --> 0:17:49.600 little under six percent, then I'll just I'll mention it. 0:17:49.640 --> 0:17:52.600 So defaults that do spike in bad times. So we estimate, 0:17:52.720 --> 0:17:55.360 using kind of the cross section of different Fyco scores, 0:17:55.720 --> 0:17:58.399 how much extra compensation you get as you go to 0:17:58.680 --> 0:18:01.280 lower and lower Fyco scores in terms of extra APR 0:18:01.920 --> 0:18:04.760 net of the defaults. So we estimate that the risk 0:18:04.800 --> 0:18:08.439 premium there is accounting for about similar sized piece. So 0:18:08.800 --> 0:18:11.760 there's a risk preum about five percent on average, which 0:18:11.840 --> 0:18:14.040 is much smaller. For let's say you're an eight hundred 0:18:14.080 --> 0:18:16.840 FICHO borrower, there's not that much risk premium. But if 0:18:16.880 --> 0:18:19.479 you're a six hundred Fyco borrower, the risk preum goes 0:18:19.560 --> 0:18:22.240 up to, like, you know, nine percent. So I think 0:18:22.240 --> 0:18:24.160 it's means something very important. I think the person who's 0:18:24.160 --> 0:18:27.439 borrowing there may not realize that they are paying a 0:18:27.560 --> 0:18:30.800 very large risk bemum. So if you're a low Fyco 0:18:30.920 --> 0:18:34.240 borrower and you aren't going to default like you know 0:18:34.280 --> 0:18:36.919 you're not, you're paying a very high risk premium. And 0:18:37.000 --> 0:18:40.240 that is because other people default in bad times. Even 0:18:40.280 --> 0:18:42.280 if you do think you're going to default. Sometimes I 0:18:42.280 --> 0:18:44.520 think one should realize how much of a risk premium 0:18:44.520 --> 0:18:46.800 you're actually paying for this. So but now let's go 0:18:46.840 --> 0:18:49.000 back to something else before we maybe talk more about 0:18:49.040 --> 0:18:51.840 that is the other pieces of this. So we talked 0:18:51.880 --> 0:18:54.800 about interchange and rewards. It's not zero. They do earn 0:18:54.800 --> 0:18:57.159 a little bit from it. Most of the transactors what 0:18:57.160 --> 0:18:59.920 they make off transactors. Is that difference because transactors spend, 0:19:00.400 --> 0:19:03.040 you know, recurringly a lot. Borrowers tend to kind of 0:19:03.040 --> 0:19:04.960 accumulate and they don't have that much more room to 0:19:05.000 --> 0:19:07.320 spend because they've borrowed. So that's not a big portion 0:19:07.440 --> 0:19:10.240 of the revenues there. Then there's fees that's another couple 0:19:10.240 --> 0:19:13.439 percent is actually making the puzzle worse. And then the 0:19:13.480 --> 0:19:15.800 part that turned out to be really big that surprised 0:19:15.880 --> 0:19:20.159 us is operating expenses, of which marketing you mentioned, this 0:19:20.480 --> 0:19:22.600 turns out to be really big, and. 0:19:23.160 --> 0:19:24.960 Yeah, this is the thing that I don't get. 0:19:25.080 --> 0:19:27.240