WEBVTT - A Look Behind The Celsius Curtain 0:00:03.120 --> 0:00:06.320 This is Bloomberg Crypto, a daily Bloomberg Ihad podcast, and 0:00:06.400 --> 0:00:09.920 I'm Stacy Marie Ishmael, Managing editor of Crypto for Bloomberg News. 0:00:10.480 --> 0:00:27.720 It's Tuesday, February twenty eight. Before it filed for bankruptcy 0:00:27.840 --> 0:00:31.920 last summer, crypto lenders Celsius had a devoted following among 0:00:32.000 --> 0:00:36.639 retail investors. These investors believed in the repeated affirmations from 0:00:36.640 --> 0:00:42.040 Celsius's charismatic CEO, Alex Mushinsky. Mushinski would say that these 0:00:42.080 --> 0:00:46.680 investors were making smart, low risk financial decisions by trusting 0:00:46.720 --> 0:00:50.440 their digital tokens to the company, But the reality of 0:00:50.479 --> 0:00:53.400 what was happening behind the scenes, according to a recent 0:00:53.440 --> 0:00:57.840 report from an independence examiner, sounds like it was very different. 0:00:58.400 --> 0:01:01.840 According to the allegations and report, the risk controls at 0:01:01.840 --> 0:01:06.319 Celsius were much less thorough and much less sophisticated than 0:01:06.400 --> 0:01:09.400 what they were telling their customers at the time. That 0:01:09.480 --> 0:01:12.759 report clocks in at six hundred and eighty nine pages, 0:01:13.040 --> 0:01:16.759 so there's a lot in there, But fundamentally, it alleges 0:01:16.800 --> 0:01:20.640 that Celsius misrepresented the true financial health of the company 0:01:22.160 --> 0:01:25.480 in ways that turned out to have devastating consequences for 0:01:25.520 --> 0:01:29.520 those customers who trusted them. Bloomberg's Olga Kreife has read 0:01:29.600 --> 0:01:32.280 many hundreds of those pages, and she joins me now 0:01:32.319 --> 0:01:42.200 with the details. Olga, welcome back to the show. Thank 0:01:42.240 --> 0:01:44.720 you for having me today. We are not going to 0:01:44.760 --> 0:01:49.360 talk about that's a big change, that's huge. I know 0:01:49.440 --> 0:01:51.960 it's a surprise, But we are going to talk about 0:01:52.400 --> 0:01:54.840 a company that I think, in a lot of ways 0:01:54.960 --> 0:02:00.360 is equally important to understanding what's been happening in crypto, 0:02:00.600 --> 0:02:05.000 and that company is Celsius. I and again reminding all 0:02:05.040 --> 0:02:07.680 others nerves, this is a theoretical example. I do not 0:02:07.880 --> 0:02:10.240 personally own a bunch of crypto. I do not own 0:02:10.280 --> 0:02:14.160 any crypto. But if I did own a bunch of crypto, 0:02:14.360 --> 0:02:17.280 I could have gone to Celsius, would say ten thousand 0:02:17.360 --> 0:02:23.560 dollars worth of bitcoin, for example, and then said, okay, Celsius, 0:02:23.720 --> 0:02:28.440 I'm gonna let you take this money and lend it 0:02:28.480 --> 0:02:31.760 out to other people, and you will pay me some 0:02:31.840 --> 0:02:35.040 kind of really high interest rate in return. Is that correct? 0:02:35.280 --> 0:02:40.840 That's exactly right. And you know, for years Alexandmaschinsky, CEO 0:02:40.919 --> 0:02:44.960 of Celsius, he basically said that, you know, if you 0:02:45.080 --> 0:02:48.120 put your money in a bank. You're basically going to 0:02:48.200 --> 0:02:52.360 get this tiny little interest if any. But you know, 0:02:52.400 --> 0:02:55.720 it's a smart financial move for you to come to 0:02:56.040 --> 0:03:00.480 Celsius and actually earn and much higher yield with you know, 0:03:01.000 --> 0:03:05.240 he said low risk. If I wanted to borrow money 0:03:05.280 --> 0:03:08.880 from Celsius, what would I do? So you would come 0:03:08.919 --> 0:03:13.400 to Celsius, And it depended on who you were. Basically, 0:03:13.760 --> 0:03:19.440 sometimes people had to put up collateral to borrow from Celsius, 0:03:19.480 --> 0:03:24.200 but sometimes, as we found out more recently, they didn't 0:03:24.200 --> 0:03:27.600 have to put up collateral, or not as much collateral. 0:03:27.760 --> 0:03:30.920 As you know, some customers who kept their funds on 0:03:31.000 --> 0:03:35.920 Celsius thought they needed to put up And so that 0:03:35.960 --> 0:03:40.080 brings us to this Examiner report that came out in 0:03:40.600 --> 0:03:44.040 January that sort of uncovered that there were a lot 0:03:44.080 --> 0:03:49.520 of discrepancies in between institutional for instance, and retail borers 0:03:49.640 --> 0:03:54.560 at Celsius. So when we talk about collateral, because I 0:03:54.560 --> 0:03:57.320 think this is a really important point and you know, 0:03:57.680 --> 0:04:00.480 it gets to the heart of as you say, this 0:04:00.680 --> 0:04:05.400 report that was published just recently, when you talk about 0:04:05.400 --> 0:04:07.960 collateral in a traditional loan context, that's like, if I 0:04:07.960 --> 0:04:10.080 want to borrow money to buy a house, I have 0:04:10.200 --> 0:04:15.320 to pledge some kind of asset in return. That might 0:04:15.360 --> 0:04:18.560 be a down payment, that might be some other asset 0:04:18.600 --> 0:04:20.839 that I have. But if I were very wealthy and 0:04:20.880 --> 0:04:24.680 I owned, say, you know, millions and millions of dollars 0:04:24.680 --> 0:04:27.920 worth of shares, I could pledge those shares as collateral 0:04:27.960 --> 0:04:29.880 if I wanted to borrow money to do something else 0:04:29.920 --> 0:04:31.839 like build a spaceship or whatever it is that really 0:04:31.839 --> 0:04:34.760 wealthy people do with their billions of dollars. But the 0:04:34.839 --> 0:04:37.279 idea is that in the case that I didn't pay 0:04:37.360 --> 0:04:41.720 my loan back, the bank or the person who let 0:04:41.800 --> 0:04:43.919 me money would have a way to at least cover 0:04:44.000 --> 0:04:48.120 the losses. Is that roughly how that works. Absolutely, and 0:04:48.279 --> 0:04:52.080 it worked very differently on Celsius. Let me give you 0:04:52.120 --> 0:04:55.800 one example. So, for instance, FTX, they change that when 0:04:55.920 --> 0:05:01.400 bankrupt last November, would borrow coins from Celsius and it 0:05:01.440 --> 0:05:05.840 would put up its own FTT token, which FTX is 0:05:05.880 --> 0:05:10.560 should as collateral. Now we know because FTX collapsed in 0:05:10.600 --> 0:05:15.120 November that you know, the value FTT tokens was very questionable, 0:05:15.800 --> 0:05:19.440 and actually people inside of Celsius knew that as well, 0:05:19.480 --> 0:05:23.520 that it was questionable and probably you know, should not 0:05:23.600 --> 0:05:27.039 be accepted as collateral, but it was still accepted as collateral. 0:05:27.160 --> 0:05:32.480 So those are the type of transactions that Celsius had facilitated. 0:05:32.680 --> 0:05:34.800 When I was reading the Examiner report, like one of 0:05:34.800 --> 0:05:37.719 the things that really jumped out to me was this 0:05:37.800 --> 0:05:40.960 idea that this wasn't something that happened recently. It's not 0:05:41.000 --> 0:05:45.240 that these standards only eroded in twenty twenty one or 0:05:45.279 --> 0:05:48.560 twenty twenty two before they filed for bankruptcy. It sounded 0:05:48.600 --> 0:05:52.479 like there had been a real pattern of having less 0:05:52.520 --> 0:05:57.480 than robust risk management for a much longer time. Absolutely 0:05:57.600 --> 0:06:01.440 so machine Ski and his co founder they came up 0:06:01.440 --> 0:06:05.520 with the idea for Celsius in twenty seventeen, and this 0:06:05.600 --> 0:06:09.680 company was the service was launched in twenty eighteen. And 0:06:09.880 --> 0:06:15.680 actually some risk control policies well they were not even imposed, 0:06:15.720 --> 0:06:18.440 they were just there was I think more of a 0:06:18.480 --> 0:06:23.520 pretense of imposing risk controls. But that only started in 0:06:23.600 --> 0:06:28.440 twenty twenty one, according to the Examiner report. And so 0:06:29.240 --> 0:06:33.720 one thing that they examined a highlighted throughout this, you know, 0:06:33.800 --> 0:06:38.560 almost seven hundred page report, is that this company they 0:06:38.600 --> 0:06:42.640 had trouble with accounting, they had trouble with pain taxes, 0:06:42.760 --> 0:06:45.960 they had trouble with risk controls. I mean they used 0:06:46.040 --> 0:06:49.400 quick books and Google spreadsheets to keep track of the 0:06:49.520 --> 0:06:54.279 coins that were you know, lent and borrowed. And obviously 0:06:54.360 --> 0:06:58.600 quick books is for small to mid sized businesses, and 0:06:59.279 --> 0:07:05.040 you know, right, everyone's apparently exactly and the examiner couldn't 0:07:05.040 --> 0:07:09.560 even find a lot of the older record keeping. It 0:07:09.600 --> 0:07:15.680 does not exist, apparently. And the risk controlled policies, you know, 0:07:15.720 --> 0:07:20.400 when when internal people were suggesting some of this risk 0:07:20.440 --> 0:07:24.160 control measures which are pretty standard that were not implemented. 0:07:24.720 --> 0:07:28.119 A lot of them were sort of not implemented even 0:07:28.200 --> 0:07:33.400 after the suggestions have been made and issues have been raised. 0:07:33.640 --> 0:07:37.720 You know, Alex Maschinski, according to the examiner's report, was 0:07:37.920 --> 0:07:42.480 very focused on pretty much just growing the user base 0:07:43.160 --> 0:07:48.400 of Celsius and not on anything. Well, how does that 0:07:48.480 --> 0:07:53.400 contrast with what Maschinsky was telling customers at the time. 0:07:54.840 --> 0:07:58.960 So at the time and actually, you know, up until 0:07:59.600 --> 0:08:05.360 self is froze withdrawals on June twelfth of last year, 0:08:06.200 --> 0:08:10.920 essentially machine Ski was telling everybody we have enough funds, 0:08:11.360 --> 0:08:15.320 we have all the funds we need to reimburse everybody. 0:08:15.360 --> 0:08:20.440 You know, everything is fine. And for actually several months 0:08:20.520 --> 0:08:24.840 before then, things actually for many months before then, things 0:08:24.840 --> 0:08:30.960 were not fine. Celsius was recording massive losses. It was 0:08:31.040 --> 0:08:35.400 spending customer and investor money to prop up its own 0:08:35.880 --> 0:08:41.800 cell token. And when things went south last May, when 0:08:41.920 --> 0:08:45.400 the Terra blockchain collapsed, this was sort of just the 0:08:45.520 --> 0:08:49.320 last nail in the coffin. But Celsius filed for bankruptcy 0:08:49.400 --> 0:08:52.520 with a hole in its balance sheet of more than 0:08:52.559 --> 0:08:55.000 a billion dollars. But you know, a lot of this 0:08:55.080 --> 0:08:59.560 whole actually happened way earlier than May. Everything that could 0:08:59.600 --> 0:09:02.840 have gone wrong with this business pretty much. Did you know, 0:09:02.920 --> 0:09:07.000 they borrowed money from a lender and then the lender 0:09:07.440 --> 0:09:11.920 refused to give them their collateral back, for example, or 0:09:12.320 --> 0:09:18.800 they entrusted their you know, thousands of ether too, somebody 0:09:18.920 --> 0:09:23.640 who lost access to that ether. You know, all kinds 0:09:23.640 --> 0:09:27.600 of strange things happened in this business. And I think 0:09:27.679 --> 0:09:29.960 partially this goes back to what you said, you know, 0:09:30.440 --> 0:09:35.000 the lack of risk controls and the business wasn't managed well. 0:09:35.440 --> 0:09:38.680 Now you've been a reporter for a long time, I 0:09:38.720 --> 0:09:42.000 feel like you've heard people say this before. But yeah, 0:09:42.000 --> 0:09:44.240 this time, for sure, we're going to pay attention to 0:09:44.280 --> 0:09:47.439 the fundamentals. We're not going to just try to chase 0:09:47.520 --> 0:09:49.680 after the thing that everybody's chasing at. We're not going 0:09:49.720 --> 0:09:52.720 to overpay for deals. We're going to be really thoughtful 0:09:52.760 --> 0:09:55.520 and deliberate and really care about risk. And then like 0:09:55.640 --> 0:09:59.760 three years later everything blows up again. Oh absolutely. I 0:09:59.760 --> 0:10:04.079 mean we saw this with initial coin offerings, where vcs 0:10:04.120 --> 0:10:07.120 were front and center in a lot of them, you know, 0:10:07.520 --> 0:10:11.600 and made a lot of money on icos before that 0:10:11.760 --> 0:10:16.880 whole space sort of collapsed because of enforcement and regulation, 0:10:17.160 --> 0:10:21.400 and actually many vcs ended up just escaping with the 0:10:21.440 --> 0:10:25.400 money they made, and a lot of the retail users 0:10:25.480 --> 0:10:28.520 ended up losing all of their money and getting hurt. 0:10:28.559 --> 0:10:31.920 And I think that's what happened last year again with 0:10:32.640 --> 0:10:35.960 the collapse of a lot of this crypto lenders, where 0:10:36.559 --> 0:10:40.000 you know, just regular a lot of regular people last 0:10:40.320 --> 0:10:44.280 everything they had. So I totally agree with you that 0:10:44.320 --> 0:10:48.920 I think is just memories are short up. Next, more 0:10:48.960 --> 0:10:52.199 from Bloomberg reported Ubercarif on why Celsius mathos in the 0:10:52.240 --> 0:11:03.880 crypto ecosystem. We'll be right back. I want to go 0:11:03.920 --> 0:11:09.000 back to something else. This idea of Celsius new Or 0:11:09.080 --> 0:11:14.679 believed that they owned customer tokens even while they were 0:11:14.720 --> 0:11:17.040 seeing the customers. No, no, no, these are always going 0:11:17.120 --> 0:11:21.360 to be your assets. Why has this become important in 0:11:21.520 --> 0:11:27.560 the bankruptcy fight? During sort of YouTube videos and all 0:11:27.640 --> 0:11:33.000 kinds of weekly you know, talks that Machine Ski did 0:11:33.480 --> 0:11:37.680 with users, he very frequently said that the sayo tokens, 0:11:38.120 --> 0:11:42.520 you know, we don't own them. You or the users 0:11:42.520 --> 0:11:47.520 own them. But many users, turns out, hadn't read Celsius's 0:11:47.760 --> 0:11:51.319 terms of service, at least in the state that they 0:11:51.360 --> 0:11:54.600 were in in the last couple of years. And by 0:11:54.640 --> 0:11:57.360 the way, the terms of service, there were so many 0:11:57.480 --> 0:12:01.760 versions of those it actually was part of the big 0:12:01.800 --> 0:12:05.080 part of the proceedings initially, you know, just trying to 0:12:05.120 --> 0:12:08.440 find all these different versions of terms of service. But 0:12:08.600 --> 0:12:13.560 in the terms of service it specifically stated that the 0:12:13.720 --> 0:12:18.920 users tokens if they sit in certain of the more 0:12:19.040 --> 0:12:26.160 popular Celsius accounts called Urn. In the scenario, Celsius basically 0:12:26.160 --> 0:12:30.520 owns the tokens. And when Celsius when bankrupt, what that 0:12:30.559 --> 0:12:33.800 meant is that all of the users whose funds were 0:12:33.840 --> 0:12:39.840 in the earn accounts, they became unsecured creditors. And obviously, 0:12:40.600 --> 0:12:44.439 and the judge actually ruled that the tokens do belong 0:12:45.040 --> 0:12:50.440 to selsius the company versus the users, and so that 0:12:51.120 --> 0:12:57.720 essentially dramatically reduced the amount of recovery available to those users, 0:12:58.240 --> 0:13:01.840 right because unsecured creditors, which is, you know, just like 0:13:01.840 --> 0:13:08.120 a jargon for your very low down in the repayment order, Like, 0:13:08.200 --> 0:13:11.079 if there's enough assets after we've paid everybody else in 0:13:11.160 --> 0:13:15.160 this long list above, you great, But otherwise you're probably 0:13:15.160 --> 0:13:16.840 not going to get very much, if you get anything 0:13:16.840 --> 0:13:23.760 at all. Now, most people do not read terms of service, 0:13:25.080 --> 0:13:27.760 very true. You sign up for something, you open a 0:13:27.800 --> 0:13:30.439 bank account, you've download a random app onto your phone 0:13:30.480 --> 0:13:32.760 from an app store, and there's like, yeah, yeah, yeah, 0:13:32.760 --> 0:13:34.960 I've totally read the terms of service. Tick this box, 0:13:35.000 --> 0:13:38.240 move on, happily. Most people do not read the terms 0:13:38.240 --> 0:13:40.880 of service. Do you think people are going to start 0:13:40.880 --> 0:13:42.960 paying more attention to these kinds of things because of 0:13:43.000 --> 0:13:45.440 these collapses, because of how many folks have lost money 0:13:45.480 --> 0:13:48.800 as a result of them. You know, after reading the 0:13:48.920 --> 0:13:53.160 Examiner street Port, I might be more inclined to read 0:13:53.720 --> 0:13:57.560 terms of service, but you know, realistically, we are so 0:13:57.720 --> 0:14:00.679 busy with our lives, right with everything we do. I mean, 0:14:00.720 --> 0:14:03.920 who has the time to read you know, the tens 0:14:04.040 --> 0:14:07.280 or hundreds of pages in there. And plus they're not lawyers. 0:14:07.360 --> 0:14:11.040 We you know, not everybody can understand the legalist that's 0:14:11.120 --> 0:14:14.120 in the terms of service. And I think it's a 0:14:14.200 --> 0:14:20.680 real problem because you know, I imagine that if a 0:14:20.800 --> 0:14:24.920 CEO of a company during a YouTube video tells you 0:14:25.040 --> 0:14:28.320 that this is how things are, you know, most people 0:14:28.400 --> 0:14:31.120 will will trust what this person will say and sign 0:14:31.200 --> 0:14:34.080 up for the service, imagining that this is what they're 0:14:34.080 --> 0:14:37.800 going to get in the end. But that's certainly not 0:14:37.840 --> 0:14:42.320 the case, as Celsius showed us. And this is exactly 0:14:42.320 --> 0:14:46.720 why various regulators around the world are starting to emphasize 0:14:46.720 --> 0:14:51.680 this idea of should there be better consumer protections for 0:14:51.720 --> 0:14:57.960 people who are investing in crypto because the assumption that 0:14:58.440 --> 0:15:02.720 companies are doing appropriate risk management, or even that customers 0:15:02.760 --> 0:15:06.720 are doing appropriate personal risk management is turning out not 0:15:06.800 --> 0:15:12.600 to be founded on reality in some of these cases. Absolutely. 0:15:12.800 --> 0:15:15.560 I mean, over the last year, so many companies went 0:15:15.680 --> 0:15:21.760 bankrupt and so many hundreds of thousands of users lost 0:15:21.880 --> 0:15:25.600 money here in the US because of this. I think 0:15:25.720 --> 0:15:30.640 this left basically a lot of the regulators here with 0:15:31.640 --> 0:15:35.320 you know, essentially a black eye. You know, where were 0:15:35.360 --> 0:15:39.640 they when all of this stuff was happening now. I 0:15:39.680 --> 0:15:41.600 started this episode by saying that we're not going to 0:15:41.640 --> 0:15:43.480 talk about FTX, and people will be like, oh, why 0:15:43.560 --> 0:15:47.880 fts is so interesting? What should people understand about these 0:15:47.960 --> 0:15:52.760 other companies in the crypto ecosystem, companies like Celsius and 0:15:53.120 --> 0:15:56.280 why they are and continue to be important in order 0:15:56.320 --> 0:15:59.920 to really get a sense of how this overall mark 0:16:00.160 --> 0:16:04.400 please operates. You know, like any other markets, crypto I 0:16:04.440 --> 0:16:10.480 think thrives on on essentially leverage. You know, a lot 0:16:10.520 --> 0:16:15.400 of people enter this market via leverage, where they would 0:16:15.880 --> 0:16:18.760 you know, put some money down on a crypto exchange 0:16:19.080 --> 0:16:23.120 and make bets with you know, a much much larger 0:16:23.160 --> 0:16:26.520 amount of money. If they lose their bed, they're sort 0:16:26.560 --> 0:16:30.240 of collateral, if you will, will evaporate. So a lot 0:16:30.280 --> 0:16:33.240 of people have been doing this encrypto for a while, 0:16:34.400 --> 0:16:37.720 for a number of years, and what the scrypto lenders 0:16:37.720 --> 0:16:44.000 allowed people to do is to even do more complex 0:16:44.360 --> 0:16:48.640 sort of borrowing and blending strategies and allowed more people 0:16:48.760 --> 0:16:55.720 to access to leverage. And you know, while I would 0:16:55.760 --> 0:16:59.880 say most most of the scrypto lenders win bankrupt last ye, 0:17:00.440 --> 0:17:03.720 not all of them. The fact of the matter remains 0:17:03.800 --> 0:17:07.080 that I think people will find out a way to 0:17:08.600 --> 0:17:12.320 gain leverage in this market as they have another as well, 0:17:12.960 --> 0:17:16.320 and you know, I think it's just behooves regulators to 0:17:16.640 --> 0:17:19.520 stay on top of this. Well. On that cheery note, 0:17:19.720 --> 0:17:21.680 thank you very much, Olger. It's always a pleasure to 0:17:21.720 --> 0:17:24.080 have you on the show. It was my pleasure. I 0:17:24.119 --> 0:17:28.760 so appreciate you hapving me on. That was Bloomberg reporto Ogokarf. 0:17:29.240 --> 0:17:31.120 You can find more of her reporting on the Bloomberg 0:17:31.200 --> 0:17:34.520 terminal and on Bloomberg dot com. And if you're interested 0:17:34.560 --> 0:17:37.080 in all things bankruptcy, because come on, we talk about 0:17:37.119 --> 0:17:39.080 it a lot on the show, you should check out 0:17:39.080 --> 0:17:42.280 a new newsletter from our colleagues. It's called The Brink. 0:17:42.800 --> 0:17:45.199 It's as exciting as it sounds. You can find it 0:17:45.280 --> 0:17:53.399 on Bloomberg dot com. This is Bloomberg Crypto, a daily 0:17:53.400 --> 0:17:57.680 podcast from Bloomberg and iHeartRadio. For more shows from iHeartRadio, 0:17:57.920 --> 0:18:01.159 visit the iHeartRadio app, Apple pod Casts, or wherever you 0:18:01.240 --> 0:18:05.240 get your podcasts. Send us your comments, questions or suggestions 0:18:05.240 --> 0:18:10.879 for the show to Crypto at Bloomberg dot net. The 0:18:10.960 --> 0:18:14.919 supervising producer of Bloomberg Crypto is Vicky Vergelina. Our senior 0:18:14.960 --> 0:18:18.560 producer is Janet Babin. Our producers are Mohammed Farouk and 0:18:18.560 --> 0:18:22.280 Sharon Barrero. Our associate producers are Ty Butler and Moses 0:18:22.280 --> 0:18:26.600 on Them. Desta wonder At is our engineer. Original music 0:18:26.720 --> 0:18:31.480 by Leo Sidron. I'm Stacy Mariaschmaal. We'll be back tomorrow.