{ "episodeId": "SLP122", "speakers": { "stephan": { "name": "Stephan Livera", "role": "host", "tag": "STEPHAN" }, "planb": { "name": "PlanB", "role": "guest", "tag": "PLANB" } }, "segments": [ { "speaker": "stephan", "time": "00:09", "start": 9.05, "text": "Hi and welcome to the Stephan Livera podcast focused on Bitcoin and Austrian economics. Today, for episode one hundred and twenty-two, we are speaking again with Plan B. But first, let me introduce the sponsors. So firstly, Kraken, one of the world's longest standing, highly regarded Bitcoin exchanges. They bring a high quality platform with With high trading volume, low fees, and no minimum or hidden fees, they have twenty four seven support, they offer the best in class accounting reconciliation and reporting services, and recently they just announced the Kraken Pro mobile app. So Kraken Pro delivers all the security and features you love about Kraken Exchange, but in a beautiful mobile first design for advanced Bitcoin trading on the go. So you generate an API key from the Kraken website, and then you use that to interact with the app. So also there's Kraken OTC desk for the more private and personalized service for large block trades a hundred k USD or more. Kraken also support margin trading up to five x long and short, and there's also futures, up to fifty times leverage to benefit from price swings or hedge your price risk. Go to kraken dot com. This episode is also brought to you by Unchained Capital. Unchained Capital is a Bitcoin financial services company empowering customers with unprecedented financial freedom and control. All of their products and services are built on the foundation of multi-sig and their approach to The collaborative custody gives users control over their private keys, as well as the benefit of a financial partner and financial services. So Unchained have two or three vaults, they're a great option for those thinking about how to best secure their Bitcoin for the long term, and if you ever need to access liquidity but without selling your Bitcoin, Unchained's collateralized loans offer a unique option. All the Bitcoin is stored on-chain in dedicated multisig addresses, it's never rehypothecated, and you can share in the security of your Bitcoin by holding one of three keys. I'm really impressed with Excellent services and releasing valuable content and open source tools, so learn more and sign up at unchained-dashcapital dot com. Next, check out GiveBitcoin dot io. It's the easiest and safest way to get your friends and family into Bitcoin. Take it from me, I've given Bitcoin to people before and they lost it because they just simply didn't know what they were receiving. That's why I saw huge value in GiveBitcoin, which allows you to time lock your Bitcoin gift for one to five years. Every month for the first year, GiveBitcoin delivers a lesson of their world class curriculum put together Together with input from many well-known Bitcoiners including Safdean, Matt O'Dell, Jan Pritsker, and others, I'm also an advisor with a small equity stake assisting with the curriculum. You can also get Bitcoin as a present for birthdays, Christmas, bar mitzvahs, graduation, and weddings, so put Bitcoin on your wish list at givebitcoin.io. I really think givebitcoin.io can have a positive impact on Bitcoin adoption and understanding, so I'm excited to have them as a sponsor. Last but not least, Bitcoin Outlet. Bitcoin Outlet delivers rare and extraordinary merchandise to warriors Bitcoin. Outstanding design isn't blindly slapping your logo on any object available at Bitcoin Outlet. Every product they carry is a work of art with a thoughtful design, in keeping with the ethos of Bitcoin. All products created at Bitcoin Outlet are limited edition. Once a product sells out, that's all folks. When you purchase something from Bitcoin Outlet, you'll be one of the only people in the world who have it. Bitcoin Outlet is a sister company to Canada's Bull Bitcoin, so both companies are Bitcoin maximalists through and through. Bitcoin Outlet only supports Bitcoin. This core belief has led them to To align with other, other unapologetically maximalist companies. So if you want a rockstar merch from a designer with some actual moral compass and unwavering maximalist views, head to BitcoinOutlet dot com and grab yourself some of the best damn apparel on the planet. Use coupon code Livera for twenty-one percent off anything in store. So today the interview is with Plan B. His first two interviews were on my show, episodes sixty-seven and eighty-six. He is a pseudonymous Bitcoin quant whose work has really been blowing up lately. And so there were some responses to his His modeling work, and I thought it would be good to have him back to talk through some of the responses and his thoughts on them. So in this interview, we talk about co-integration, what is it, why is it important, and importantly, we also talk about what's the steel man case against his model, and then we also talk about what he's working on these days and what's next for him. I'm sure you'll enjoy this interview. Here it is. Plan B, welcome back for the third episode. Thanks for having me back. there's been a lot happening since the There's been a big blow up in your following and you've had many more translations in your articles. How's that been for you?" }, { "speaker": "planb", "time": "04:38", "start": 277.53, "text": "It has been great. again, I'm still waiting for the moment that, that all this, stops or, or levels off, but, I guess it's not, not now yet. so in, indeed, lots more people, read the articles. I think the, translations help enormously with that, for example, the Chinese, translation, was, was fun. One, 'cause I never knew that people in China, cannot read Medium articles, 'cause it's, it's fenced off by the firewall, so everything has to be on WeChat. But since that article is on WeChat, I, I guess it opened, the, the article and the content to, to lots more people. Same with, India. I, I, I never knew there were six, languages in India. And, Well, there's only one translation, translated in, I don't even remember what, what, what the exact language was, but sixty million people living there, that can now read the article. So, yeah, it, it has been phenomenal. And, well, I hope it goes on for a little while." }, { "speaker": "stephan", "time": "05:45", "start": 345.43, "text": "Yeah, that's awesome. So let's just, set some quick context. So listeners, check out the first two episodes. They are sixty-seven and eighty-six. But suppose we can- And just give some high-level context around the idea. So what you're essentially doing is some quantitative modeling on this relationship that has existed, between the stock-to-flow ratio of Bitcoin and the price of Bitcoin. And I suppose just like, well, let's call out some of the high-level numbers that you are, in some sense, predicting, or at least the model is predicting, let's say. So there are a couple different numbers, and I believe you've got the twenty- Twenty twenty halving, so approximately May twenty twenty. Now, doesn't mean it'll hit exactly that number then, but roughly, you've got fifty five thousand for then. I think it's roughly four hundred thousand for the twenty twenty four halving and the, and then three million for twenty twenty eight. Is that a fair summary you would say?" }, { "speaker": "planb", "time": "06:45", "start": 405.47, "text": "Yeah, that, that's correct. And, that, that's the numbers that were in the original article that I wrote in March, this year. and maybe to go one step back, so I came to Bitcoin from a, investor perspective. I'm a traditional investor working at a large, listed, investment company in the Netherlands. So, yeah, I approached this market not from a technical point of view, but from a purely investment point of view, and, I noticed the lack of, quantitative models, econometric models, if you will. there was a lot of technical analysis, but yeah, the- The model that I made is actually quite simple, trying to capture the price with a measure of scarcity, and I found the stock to flow measure, quite useful, the stock to flow measure that is, mentioned by Tragedy Mayor, but also by Seifried and Amos, of course, in his book, and, the model I made is a, A, a, a linear regression, on the, logarithmic values of both price and, and stock to flow, and then you come up with a very nice, fit, a high R squared but, but what's even more important is the, the co-integration, we might talk later on that. But, but you're right. So the model, predicts, or models the, the last ten years historical values of, Bitcoin stock to flow and, and Bitcoin, price. And because of the tight fit and the co-integration, well, it's, it's at least reasonable to believe that, it will hold for, for one or two, more halvings, a-and the numbers you mentioned are, are absolutely correct." }, { "speaker": "stephan", "time": "08:38", "start": 517.93, "text": "Right. One cl- one clarification on that point. So I have seen some different numbers, and it might be just good to clarify for the listeners. For example, for the next halving for May twenty twenty, approximately, I've seen a fifty-five thousand number And I've also seen a one hundred thousand number. Could you just clarify for the listeners where those two different numbers are coming from? As I understand, one of them is sort of only feeding the model the first few years of data. Is that correct?" }, { "speaker": "planb", "time": "09:04", "start": 544.17, "text": "Yeah, that's correct. And that's, there's some confusion about it, so I'm glad to explain it, So the fifty-five thousand number for next halving in May twenty twenty, that was the number from the original model in the original March, two thousand nineteen article. and, and that number was, was, quite high, i-in, in my, opinion. I thought, well, we're now at, the time of the writing, it was, three or four thousand dollars, so fifty, fifty-five thousand seemed like, well, A very high number, and I didn't want people to get o-over o-optimistic about it, so, I rounded the parameters a little bit. So, if, if you read the article, you see the, the formula is, i-is with nice round numbers. zero point four times stock to flow, to the power three. And so zero point four and three are the round numbers, but if you, If, if you would, would take more digits or you would, you would get a little higher number, but then of course it becomes, less, less easy to, To explain and to communicate. So the model, evolved from, from March, this year, until wh-where it's now. and of course, more data was added, the, the data of the first fifty-five, thousand model was, until December two thousand eighteen. and now of course, we have, nine more months, to take into account. Also, I have, have more earlier numbers or data, data archaeology. where the, yeah, I, I, I do have numbers from September two thousand nine right now. So if we take that into account, just the new numbers, you get to values somewhere between sixty and ninety thousand, and those are also the numbers that all the other teams that replicated my model, come up with. so there's, there's one more model, and it's, it, that's where the hundred thousand comes from. That's the third model, and I personally prefer that one. I like it not because it's a higher number, but because that model was made on only, pre-November two thousand twelve data. So before there was any halving. on only the first four years of, Bitcoin data, and yeah, that model predicted basically the ten x jump in price Of two thousand twelve halving and the ten x jump in price from the two thousand sixteen, halving. So I'm, I'm really," }, { "speaker": "planb", "time": "12:00", "start": 719.91, "text": "yeah, I, I'm, I'm really fond of, out-of-sample performance, so I, I'd rather take a model that's, that's done on less data but performs very well out-of-samples so on new data, than, than, than taking all the data and fitting it as good as possible. So, so those are the three models, fifty-five, sixty to ninety thousand, and a hundred thousand." }, { "speaker": "stephan", "time": "12:23", "start": 743.17, "text": "Fantastic, thank you for that clarification. Also, it might be interesting and valuable to talk about the timing. So typically, what we have seen looking historically is the first halving was roughly mid two thousand and twelve, and then two thousand and thirteen was where we saw the big bull runs happen. Then the next one was in mid twenty sixteen, and then again, we saw the bull runs happen in twenty seventeen, or one, rather, one big bull run, let's call it. So would you say then that if you were to copy paste that same pattern, the halving next time around will be mid twenty twenty, but actually the run up will come in twenty twenty one? Would you say that's a fair statement, or what's your view there?" }, { "speaker": "planb", "time": "13:05", "start": 784.92, "text": "Yeah, I think that's a fair statement, but, but formally the model, doesn't Say anything about it. So there, there's no short term prediction, there's no, prediction of the all time high or the next peak. There's also no prediction of the low. the only thing that you can take from the model is that, you, you, you have the co-integration and, and it will stick very close to the stock to flow, level and, so somewhere before, say, Christmas two 2021, the price should be above a hundred thousand if we take that third last model, or above fifty, fifty-five thousand if you take the original model. So that's in fact the only thing that you can, you can take from the model. And where the next half thing is, or where the next, all-time high is, sorry, yeah, it, it, it's guessing, but, but the way you said it is correct. normally, or at least last two halvings, the market didn't react immediately on a halving, but it, it lagged a little bit. You can, you could even model that lag, but, but I, I wanted to keep the model simple, so, yeah, that's," }, { "speaker": "stephan", "time": "14:24", "start": 863.52, "text": "Fantastic. So let's jump into a little bit around some of the responses to your model, and then one of those is around co-integration as well. So as I understand And, there's a gentleman named Fraudstar, and I think his name is Nick, he wrote an article, attempting to, basically saying something like falsifying the stock to flow model. could you help us break that down a little bit? just give us an overview on that?" }, { "speaker": "planb", "time": "14:49", "start": 888.95, "text": "Yeah. Yeah, that, yeah, he's, by the way, he's an Australian, guy. Oh, really?" }, { "speaker": "stephan", "time": "14:54", "start": 894.42, "text": "I should, I should talk to" }, { "speaker": "planb", "time": "14:55", "start": 895.36, "text": "him. Yeah, yeah, yeah, you'd like that. so Nick Frauder indeed." }, { "speaker": "planb", "time": "15:02", "start": 902.32, "text": "He, he did a, a, a very important study, on, on the, on the model, and he, and, and he's, he's an interesting guy, 'cause he, he's saying, he's very strict in, in learning and, and, and the a-academic approach. So he's, he's saying like, the only way you can learn is by falsifying things. You can't verify things. You can, you can only- Falsify things, so it's true until you falsify it, and that's why he called his, his, research \"Falsifying the Stock to Flow Model,\" because that was his intention, and his aim. So he, went after it and basically did the same exercise I, I did, came with the same conclusions, But then also raised the question, the same question I did, like, okay, you can have a, a very high R squared, especially in a, logarithmic, domain But it could be spurious, it could be a false regression, it could be a, a correlation not causation, a, a meaningless, correlation that we, we measure. so he took, he took one step more, and that's crucially important. he checked for co-integration, and co-integration is, the two, variables, stock to flow and price in this case, stick together. So the difference, but-- so he's, he's actually checking and testing and modeling the difference between the two, and if that difference stays stationary, that's a statistical term, but if, if it stays- Around zero and doesn't wander off too far, all is good and the two, series are co-integrated, which is very rare. So, basically if, if you find that two series are co-integrated like he did, that means that the, relationship that you find, the, the high correlation that you find, aren't spurious. So, that's why he concluded That he could not falsify the sock to flow model. So in, in that way, the title is a little bit misleading, but it was a, a very important, moment also for me, in, in, in the development of this model." }, { "speaker": "stephan", "time": "17:27", "start": 1047.1, "text": "Great, let's break that down a little bit. So from, fraudster Nick's article, he mentioned a few different, if you will, assumptions that must be true before you can, do this technique. He mentioned the use of log scale, and so he's commenting that, and, and as you mentioned, it's log of stock to flow ratio and the log of BTC price. Can you articulate why that is?" }, { "speaker": "planb", "time": "17:52", "start": 1072.0, "text": "Yeah, it-- you, you need a, a linear relationship if you do a, a linear regression fit. So obviously, if you look at the stock to flow, but also price is even more, more important. If you look at the price of Bitcoin On a, a, a normal linear scale, it, it's, it's, it goes exponentially up and, and, you don't see any details, b-before 2017 if you plot it that way. so to get more detail and to, yeah, to get more meaning, you, you have to transform the data to, massage it into a, a, a linear line. So that's, that's, and, and massage, Might sound as, as manipulate, but it's just a transformation. So you transform the data, you try to see if, if a lo-logarithmic regression is, is useful or, or an exponential function is more useful or if it can be transformed in a linear shape. And, and if you do that, with stock to flow and, the price, then you get this beautiful, linear, A shape that you can fit with the linear regression, so b-b-basically what that means is that the, that, that the regression that you, fit on those data can be transformed in a power law, so that's the function I use, so the, the zero point four times stock to flow to the power three, that's a power law, and, that basically, and, and that ha- Has also to do with the logarithmic scale, that models change, so not, not so much the, the level, but the change of stock to flow and the change, of price, they are proportional, so for example, if the stock to flow goes two X, the price goes ten X, and that, that's, that's what the, the function says, the power law function, and that is true on a very small scale, so if it goes, three X or Or, or point three x, it, it doesn't matter, this power law stays the same. And that's the, the power of, of, of this logarithmic, transformation." }, { "speaker": "stephan", "time": "20:13", "start": 1212.94, "text": "Excellent, thank you for that. And there were some other terms that you, that were used in order to, you know, as, as, as he says, falsify the stock to flow model. He mentions the terms, normality in error. Can you articulate what that is?" }, { "speaker": "planb", "time": "20:27", "start": 1227.29, "text": "Yeah, the, the errors that, those are the, the difference between the predicted values and the actual values and, and the study of the analysis of how those errors look is very important because there's all sorts of assumptions that are made about it, so, they can't be, autocorrelated, for example. And, yeah, they should be normal. So if you have the ideal model, what's left? So what's not explained by the model should be white noise, as we call it, so that's, that's normally distributed random noise, all the other factors out there impacting the price, but, on average, leveling Out to zero, and, and, and if you find, normal errors, normally distributed errors around zero, then you know you have a good model. That's," }, { "speaker": "stephan", "time": "21:26", "start": 1285.92, "text": "great. And now, one way I've heard this explained, around, I guess, speaking to the broader idea of co-integration, is that it's, it's like a rubber band effect, that as one moves further away from the other, so in this case, Bitcoin price moving further away from the model, then it sort of rubber bands back to the model. Is that one way to think of this?" }, { "speaker": "planb", "time": "21:51", "start": 1311.14, "text": "Yeah, and, and Nick has this nice, classic- The story about the, the drunk and his dog. And, a-and, you know, the drunk going out with his dog on a leash. the drunk, wanders in a, a sort of random fashion, and the dog has to go with him, 'cause he's on a leash. But sometimes he will be on the right, sometimes he will be on the left, but he can't go further than the leash, and come closer. So you don't know where the dog and the drunk are going, but you do That, that is, yeah, it's, it's a classical example, and but it exactly states what, what core integration is, and that is very useful from, well, multiple perspectives, but, but investment being one. And, and in this case, it's, it's, it's even, more interesting because we know of course that the stock to flow isn't wandering like a drunk, it, it's, it's, it's, it's predetermined. So for predict-prediction purposes, this is- Almost too good to be true." }, { "speaker": "stephan", "time": "22:57", "start": 1376.72, "text": "Yeah. Right, yeah, and he, he mentions at the end of the article, he's saying, \"In short, Bitcoin is the drunk, and stock to flow is the road home.\"" }, { "speaker": "planb", "time": "23:04", "start": 1383.57, "text": "Yeah, that's it, that's it, and, oh, you, you should have him on your, podcast, 'cause he, he, he can explain this a lot better than I do." }, { "speaker": "stephan", "time": "23:11", "start": 1390.73, "text": "Oh, fantastic, yeah, I might, I might have to do that. let's talk about some of the other responses to your work. So" }, { "speaker": "planb", "time": "23:25", "start": 1404.65, "text": "Yeah, so, Berger, Nick and, and, and I sort of worked together, so we were-- we came into this separately, but our work was so related and, and, so now we're working with, three of us together. yeah, Berger actually, took the same path as, as Nick. He was very skeptical of the good fit and the high R squared, so he wrote an article, he verified the model, came with the same results, but really focused on the spurious, the, the possibility that it could be a spurious regression, which, which of course is a good point. And, and later in his second piece, just after Nick wrote about co-integration, he verified that co-integration as well and even did some extra studies, so, you can check co-integration with a several, several tests, several, different tests. He did like, like three or four, and came with the same conclusion. So we all three now have a shared, conclusion that we stand by, where- We, we all began with being very, very skeptical, and so I see a lot of people, in that first stage right now, where all three of us began a couple of months ago." }, { "speaker": "stephan", "time": "24:48", "start": 1488.0, "text": "Great. And, one other one I think might be interesting for you to touch on, and this is more like a very outside of the, the ordinary squares approach that you have taken, there was one by HC Burger one, and it's a, basically a time-based model. Now, this one was a little interesting because he was basically predicting, based on this model, that the price would reach one hundred thousand per Bitcoin no earlier than twenty twenty-one and no later than twenty twenty-eight. So, in some way- It was, it was a more conservative estimate. Have you had a chance to look at this modeling work and do you have any comments on that?" }, { "speaker": "planb", "time": "25:26", "start": 1525.89, "text": "Yes, it's, it's an interesting study as well, and, we talk too. The, the time models are, are classic. So instead of stock to flow, you take, a look at time as an, an, an explanatory, variable for, for the price of Bitcoin and, and the, the narrative then is, it's like adop- adoption that takes, a- Time to, to play out, and that's what you, what you see in the price, which is very reasonable. And in fact, it's, it's where I began modeling as well before March, this year. So the very first models I made were The exact same models as, HC Burger, did with, with the time, and we have the, the same, parameters as well. So, so it's, yeah, it's definitely interest-- it's an interesting model. the only thing is, first, there, there is a big difference in, in predictions between stock-to-flow model and this time-based model. Like you said, the time-based models, are, are lower in prediction. So, and it's very interesting to see, maybe not this halving, but especially after twenty twenty-four halving, the models really, deviate from each other Each other, so we'll see which one is right. Very exciting. but more importantly now is that, what, what I don't like about the, the time model And, and why I personally abandoned the model and, and jumped to stock to flow is two things. First, the parameters aren't stable. So the first time model is actually made in two thousand fourteen. It's a, it's a classical chart. I have it somewhere in my, I, I will tweet it out later. it's, it's a, Green line, green price line with a red curve, the logarithmic curve, in it, and I think it was updated later by, Tude Meister. So," }, { "speaker": "stephan", "time": "27:30", "start": 1650.05, "text": "yes, this is a well-known one. Tude shared it, I think." }, { "speaker": "planb", "time": "27:32", "start": 1652.27, "text": "Yeah, yeah, you, you shared it as well, I think. And, and, and what you see there is the prediction that was made in two thousand and fourteen, and it was much higher than the actual price that we see right now. So the, the, the red prediction line is Real price, i-in the updated chart by, by two, so, and then if you look at the model that was fitted on all the data, so including the, the data from two thousand fourteen till today, that H-H-E Burger is using, then you see a much lower model. So, so the model came down significantly, and of course, now it really fits well, but, my prediction would be that it will have to be, updated again. after two thousand twenty-four halving. And, so the stability of the parameters is a key thing for me, 'cause I don't want to update the model, every year, 'cause I'd like to, to use it for prediction and investing. and, and like I said, the, the model that I prefer, the, the third incarnation of the stock to flow model, if you will, is made only on two thousand twelve data before there wasn't any halving, and it's still working today, so without any change on parameters. So, For me, is a crucial thing." }, { "speaker": "stephan", "time": "28:50", "start": 1730.07, "text": "Yeah, that is a really fascinating thing, and it speaks to whether there is some kind of, again, it's not praxeology, it's not economic law, but potentially in a, in an Austrian term, you might say something like timology, which relates more to like human psychology in response to, markets, let's say, that or psychological aspects that lead to human action. And as you, as you were saying, this is before even there was a halving. Now, I think it would be good It's good to talk about one critique that I have seen where some people are leveraging, are laying a critique against the stock to flow model idea by saying, \"Well, look, the stock to flow model, it explains the price only by reference to supply and not by demand.\" What's your thought on that?" }, { "speaker": "planb", "time": "29:35", "start": 1774.74, "text": "Yeah, I, I, I get that critique a lot, and I understand it, 'cause if, if you have an economic background, you are taught by, with, with the theory that, that prices are, are made on markets, but- Buyers and sellers, and there has to be, supply and demand, and where they meet, that's, that's where the price is. So a model only on supply is weird in that sense, and, and that's true. I mean, I understand that critique. But the thing is, and Nick Zabo, talked about it in the, in the tweet, this week as well. The Veblen," }, { "speaker": "stephan", "time": "30:13", "start": 1813.09, "text": "good idea, right?" }, { "speaker": "planb", "time": "30:14", "start": 1814.07, "text": "Exactly. So it, it could very well be that Bitcoin is a Veblen good, and a Veblen good is, diamonds or Rolex, watches would be an example. It's a good that where demand increases, where price increases, and that's totally- That's hundred and eighty degrees different than, than, demand and supply no-- laws, normally work at that you are taught at, at school. Because normally if prices rises, rise, demand, will, will drop and not, and not rise. But there is some indication that Bitcoin is behaving like, like a, a Veblen good, and that would be-- that would explain, so, so in, in, in, in that way the demand is directly caused by- The, supply or the stock to flow, if you will. But, but even, even more important than that, so that might be the case, but even more important than that, and it's a general point, demand isn't in the model, but lots of other factors are also not in the model that are important for the price. So for example, if we look at the past ten years, when China did this banning, oh, China bans this, China bans that, yeah, it has an effect on price, price goes down. China bans aren't in my model. So when the government, the SEC or the, the CFTC or whatever, cracks down on, on Bitcoin or, or puts a foot article, in the media, it has an effect on price. Is it in my model? No, it's not in my model. So there's like Tens, tens of variables that are really important that are, but that aren't in my model, and still that means, that doesn't mean the model is wrong or the model is, is, is not useful, because it's just a model. It's, it's a model, and it's, and it's really a simple model. And I don't know who, who the quote is from, but, the saying is all models are wrong, but some are useful. And, yeah, and I think that's, that's true. I mean, if we At, at the power law, and that's, that's why I think it's, it's interesting to have a non-spherical, a power law here. there's, Bitcoin obviously is a complex system, very nonlinear, very dynamic, with, at least seven network effects, miners, investors, merchants, developers, all, and, and, and, and regulators and countries, everything, has an impact on price, so to model that would be- impossible. And, in complex sys-systems, what you see, is sometimes there is an underlying structure that is very simple. So there is a complex reality, but a simple underlying structure, and that is, what a power law can, can grasp. So in a way, I think the stock to flow I don't know how exactly, of course, but in a way, that's what I measure. The stock to flow captures all this underlying complexity and, and including demand for, for whatever reason." }, { "speaker": "stephan", "time": "33:24", "start": 2004.31, "text": "Right. And, and maybe another way to frame that might be something like, there are many factors, but it's just that stock to flow happens to be the dominant one right now, and that isn't to say that it will always be the dominant one, as you mentioned. Maybe the model breaks down after twenty" }, { "speaker": "planb", "time": "33:40", "start": 2019.94, "text": "twenty-eight. Exactly. Yeah. And, and, and there's one more thing, the demand, b-b-because if something has a high stock-to-flow ratio, there's, there's not much things that have a high stock-to-flow ratio. Gold has it, Bitcoin has it, diamonds have it. basically it could be a definition of money, right? People use high stock-to-flow things for the function of money and, and the demand for money of course is unlimited There's always demand for money, so, so, you could assume that to be there. It, it's a very, very rough assumption, but you could assume in a money model, which my model is, that, that, demand is there, or at least that it's captured through the stock to flow." }, { "speaker": "stephan", "time": "34:30", "start": 2070.33, "text": "A, another point that I think I've seen you make is this idea that, and it's, I guess, harkening back to Nick Zabe with the unforgeable costliness, but as you were mentioning You were saying it's all about energy, and ultimately it's how much energy does it cost you to make Bitcoin versus any substitutes, right? So even in the case of gold and platinum and palladium and so on, there are some elements of substitutability amongst those other metals, but fundamentally it's about how much energy does it cost you to make them." }, { "speaker": "planb", "time": "35:00", "start": 2099.69, "text": "Yes, there, there are some other critics, that also go in this direction. I think it's Stephan Livera or, or, I forgot his name, but, so he's also very much into- To, pricing Bitcoin in watts, and, and that's a very interesting concept, which I think is correct, as, as you said, like, like gold is scarce because it's very expensive to mine it, and, and the same with Bitcoin, same with diamonds, and, and we can also see the opposite of that, right? If, if, if money, is very easy to make, l-l-like our current fiat money. Then, then it goes wrong, and, and we saw that in, in Zimbabwe very clearly, and, and more recently in Venezuela, you can just print the, the money as a government and spend it, but in the end, it will not, that won't last long, and, and it won't end well. and, and of course, this is- Well, actually, the reason why I came to Bitcoin, the whole quantitative easing, experiment that is done by central banks today, nobody knows how that will end, and some Fear it won't end well, me being, one of those people, of course." }, { "speaker": "stephan", "time": "36:16", "start": 2176.5, "text": "Of course, I, I think so. one other big critique that might be good to address at this point is the \"why isn't it priced in?\" idea. Now, one suggestion could be that most or everyone who knows about it is already invested as much as they reasonably can. But what's your view there?" }, { "speaker": "planb", "time": "36:35", "start": 2194.85, "text": "Yeah, if I were to give a steelman argument against the model And I'm constantly looking for good arguments against the model, 'cause that's how I learn and that's how I know I'm investing the correct way. If I were to give a steelman argument, it would be this argument. The, the relationship is true, and if the information is out there as it is since, since March, the, the, and, and or even earlier about the halvings, the halvings should be priced in, in a reasonably efficient market. So this would be my ultimate argument against the model, and it's, it's the one thing I don't understand why it isn't priced into, it's not priced in the halvings. And it was also Nick Szabo's ultimate argument against The stock to flow model in his tweet last week. So, yeah, wh-why that is, I, I don't know, but I can guess, and, and I learned something this week a-about this. So f- my, my first reaction would be, okay, markets aren't efficient. Bitcoin is small with this, with its hundred and fifty billion market cap, it's small, so maybe it's not efficient. But on the other hand, if you look at currencies and Bitcoin, it's very efficient. So you can't buy bitcoins with dollars dollars and then convert it, sell them in euros and then sell the euros for dollars and then make a profit or something. That, that could be done in the early days, but, but now you have exactly the, the foreign exchange rates on those Bitcoin prices. So markets are pretty efficient. yeah, I don't think that's, that's the thing. The other thing would be, okay, the information isn't out there, for everybody. For example, new investors won't know immediately about the stock-to-flow model. So They don't know about it and, and they don't price it in, and maybe some of the current investors don't know yet about the model, although the model spreads like wildfire, so I think most of the investors right now know. But then the, the other argument, and, and that's the simplest argument, and I think that's the, the, the true one, is that there's lots of people that don't believe in the, in the model, in the stock to flow relationship. So I put a tweet out yesterday Yesterday with, three well-known Bitcoin people that are against the stock to flow model with good arguments as well. I, I don't think they hold, but th-th-there are good arguments that you can, that you can follow and, and believe, and, and then you would, you would be against the stock to flow model and not invest, In, in, into a, stock to flow price, r-relation-relationship, and in that view, the halvings would be less important than what I think they will be. So I guess it's, it's just A normal market situation where there's not much people that really believe in the model, which is very interesting and a nice investment opportunity." }, { "speaker": "stephan", "time": "39:38", "start": 2378.26, "text": "Yeah, that's, well, for everyone who is a believer, I guess that's, that's their opportunity. Let's turn now, one interesting topic that you were discussing was also this concept of downward difficulty adjustments, or stated in other words, Bitcoin bull markets seem to start at difficulty bottoms. So what's your take there?" }, { "speaker": "planb", "time": "39:58", "start": 2398.34, "text": "Yeah. so what I learned from, from my, career in, in investing, and, and so, I'm a traditional investor, we do mortgages and bonds, and, and those are billions of dollars, so, so big deals. What I learned there is that it's, it's very interesting to look at what the, what the big buyers and sellers are doing, so the people that move billions of dollars have very, good research very smart teams, great access to low prices, they have, they have it all, and they ha- they are at the top of the spear, if you will. So looking at what they are doing Might give you an edge just by following, them. And, and the same is true with Bitcoin, in my opinion. miners are the ones that, that are really invested in, in, in, in Bitcoin 'cause they invested in hardware and, and, and, mining these days is, is, really only profitable if you, if you are a professional industrial miner, so you need access to very cheap electricity below Five, four, three cents, almost free electricity. So you need to, to be next to a, so a water dam or, or, or something else that where, where there's excess energy, otherwise you, you can't make, bitcoins profitably. And you need the latest and the latest hardware, so the seven nanometer, chips, specific chips. Yeah. So, so the, the miners are, are big players, and of course, they're big sellers. So they, they make bitcoins, and, and once they make them, they have to To, to, to make good for the, for the electricity costs, and, and the capital expenditure on, on the miners. So it is interesting to see what, what the miners are doing. And, in bull markets, you see, and, and I'm not talking causation here, but just, just an observation, you see, the hash rate and the difficulty go up. So, so, and that means that miners are being added, mining hardware is being added to the network. or, or the latest chips are being bought by miners and are increasing the, the hash rate and, and thereby the security by the way of the network. but so miners are investing and the opposite is true in down, markets. So in the latest bear, bear market, and you saw that in, in last two bear markets as well, the two bear markets before this one. Y-y-you see at a certain point that, that miners are switching off their, their old equipment because the price of Bitcoin is too low and their miners aren't profitable anymore. So it costs them more electricity than it, than it, it, it, gives them in, in revenue. And that's why they switch them off. And you can measure that, so you can see a drop in difficulty. And it's very rare, it only happened like- Well, a couple of times, in the last ten years, and each-- and if you, if you look, if you, if you look at the, the lowest point of that, of that difficulty decline, I call that the difficulty bottom And mark that point from, from where it starts, going up again, so miners are, are, are, are getting positive and investing again. That Point in time, has been the start of bull markets in the last through three, three times. So it's, it's very interesting to, to look at the big guys. It's by no means a statistical, hundred percent verified and secure, thing like the stock to flow model, but it's just a v-- I find it a very interesting, observation." }, { "speaker": "stephan", "time": "43:54", "start": 2634.47, "text": "Excellent. And just for the listeners, the prior difficulty downward adjustment As you point out, they were late 2011, early to mid, well, early 2015, and most recently December 2018. So just for context for the listeners there. Turning now to Your mission of trying to merge the worlds of Bitcoin and professional money management. Now, some of your work online, you've shown things like the Nasim Taleb-influenced barbell portfolio of holding, say, one to five percent of Bitcoin and then the rest in cash, and showing that this has a much better return versus risk profile and obviously a much higher Sharpe ratio than the traditional sixty forty stocks and bonds or other traditional investments. Have you had any luck convincing- Professional money managers on that point." }, { "speaker": "planb", "time": "44:45", "start": 2685.12, "text": "Yes, I think I have, and, you might have noticed indeed that I'm shifting, my focus a little bit towards in-investors and investor podcasts as well, like, The investor podcast with the Preston Fish or, the interview I did with Real Vision and Rowel Paul lately, the gold, versus Bitcoin thing. So I'm focusing on, on investors because that's also, yeah, my, my background. And they are very, very skeptical, at least they were last year, the year before that, maybe a couple of months ago. But something is starting to change, and, you see the very forward-looking guys, the hedge funds and, let's say the, the group that, that Raoul with Real Vision is, is, is, targeting, that group is, is getting it. if, if you listen to Raoul, he's also saying that most of the people are almost All the people that he knows, the CEOs of these, the CIOs of these hedge funds, are already invested in Bitcoin personally. So that's where it all starts, I think. so, so you see it there. You see also, Anthony Pompilio, of course, with his company, getting customers in the more traditional domains, the pension funds and the, the endowments. So that, that, that's a step. And You also see, if you look at stock to flow model, which was, discussed, but on, on CNBC, the other day, you see a real turn in thinking there, which is, remarkable. and the other day you had this, this, German, Landesbank, the Bayern Landesbank, who, verified the stock-to-flow model and actually published the results to, to their clients. So you can see traditional, parties that, that, that used to just, just, tell all the, the mainstream media, fud, Bitcoin is for criminals, Bit- Bitcoin is a fraud, Bitcoin, boils the oceans. You're seeing that turn slowly but surely, and I can give one example of my own, company, where I work Nobody, except the quants of course, that do the work, but nobody of the investors really wanted to know anything about Bitcoin. but the example that, that, that's really well, received is the following, The Bitcoin futures markets are very interesting at the moment. If you look at the price of futures at the CME in Chicago or the ICE Buct, system in, New York, physically settled, you see that the future market is in contango. That means that the future prices, the prices next month and over two months and even over three to six months, are higher Than the spot prices, so the current prices, and that opens the door to a classical," }, { "speaker": "planb", "time": "48:08", "start": 2888.17, "text": "carry trade construction. it's, it's, it's, it's not complex, I'll, I'll explain it. It's, it's like you have a Bitcoin position, you do buy Bitcoin and have a Bitcoin position, and then you sell that same position For delivery over one month or two months or, or six months. So you, you buy it and you immediately sell it against this higher future price, so you lock in A certain profit, a, a, a guarantee profit. So in a way, that's, that's a risk-free profit, if you," }, { "speaker": "planb", "time": "48:45", "start": 2924.54, "text": "that there are some risks, but they're, but they're small. So you can do the same thing with gold, buy gold and futures, sell the gold and then rinse and repeat every month, you won't make one percent every year in return, which is low, but which is consistent with gold being very stable and, and, and risk-free. But if you do But with Bitcoin, like the way I described on CME or BOK, you can make like twelve percent per year, and that, and that's almost risk-free. So if, if you pitch that example To a traditional, money manager, his first reaction will be, \"That's not possible.\" And then you, when you show him the quotes, the actual quotes He'll be like, \"Okay, I have to know more about this.\" And that's the example I use nowadays for people, with a traditional investing mindset that really gets them thinking and into Bitcoin. Actually, I know three guys who, personally bought Bitcoin after this, this example who are professional money managers." }, { "speaker": "stephan", "time": "49:54", "start": 2993.81, "text": "So the proliferation of the carry trade idea, and I think, we touched on this in some of our earlier episodes as well. One, question I've got around that is the contango, as you mentioned, the future price being higher than the spot price. Part of that may just be driven by people having this future expectation of a rise in Bitcoin's price, so they're buying it, and is that part of what's driving that overall contango scenario?" }, { "speaker": "planb", "time": "50:20", "start": 3019.53, "text": "Yes. yeah. There's two components. There's the actual- cost component. So with gold, for example, you see the same thing. If you have physical gold, then you have to store it and you have to insure it, and that costs money. So, that's why the future price of gold, should be a little bit higher, you know, buyers of, of gold in the future, W-will like the idea of not having to pay for storage and, and insurance, next month or two. So, so that real cost of storage and insurance explains part of it. The other part is speculation of, what, like you say. So if there's lots of people that expect the price of Bitcoin to go to, well, let's say fifty-five thousand, then, they can, they can, they can, take on leverage, multiply By their, their, investing results, by, by buying futures, and, and they will play the futures game. So, so, yeah, in, in a way, and, and that of course, if they buy, futures, that, that drives up the price, of the, of the futures and also creates contango. so in a way, it's, it's the carry trade that I just described, describes, is, is funded by the people that are very, very optimistic by, and, and are willing to carry, to carry that risk of Bitcoin. In, in a way, it's, it's, it's, it's like the prelude to a option market, which I think Bock is also, introducing, which is very logical. So, so the, the people that think about fifty-five thousand Bitcoin in, in one or two years, like me if, if they're pro-professional money managers, they think of it, of it like an option. They think they have a certain, probability that that fifty-five thousand scenario is happening. That probability might be low, might be like ten percent or twenty percent, and the other scenario being it goes to zero, eighty percent. so that's sort of an option structure, but still they would, they would buy that, that future, and, and someone else isn't, isn't willing to take that even- That, that ten or twenty percent, chance of a fifty-five, percent, fifty-five K scenario, so they rather have the cash and carry where they make a certain, let's say ten percent per year And it's, yeah, it's, it's a great way of futures markets bringing together speculators and more risk-averse, investors." }, { "speaker": "stephan", "time": "53:07", "start": 3187.42, "text": "Excellent. Plan B, do you have any other related projects or other things that you're working on at the moment?" }, { "speaker": "planb", "time": "53:14", "start": 3194.1, "text": "Yeah, there, there is, there is lots of projects and, and, there's some I can mention, maybe. So, so in the summer I was, I was busy with all the translations of the article, which is, which is great Great, and there's twenty-four now. But right now, lots of people are replicating the models and, and building, real-time versions of models in, On websites, there, there's, Digitalic, a Swiss guy who made a very nice website, and I'm, we're sort of working together, and I'm thinking of not making, my own charts anymore, I'm just using his charts, 'cause they are, evolved in even better charts than I ha- than I can make, and they're real time, so, so that's very nice. You might also have, have seen the art thing that I, I'm involved in. We, the charts that I make Inspire me, but also some of my followers that they want high resolution prints and stuff. So I, yeah, and, and that's, that's, that's great. I, I have that same desire and, so, so I decided to take it one step further and, and, commissioned an artist that, that is, by the way, one of my followers and approached me, on Twitter. So we met face to face and, and she's now making an artwork of, of, of one of the charts. And, and I think that's That's important 'cause Bitcoin is more than just numbers and, and investments, it's, it's, it's a movement, and, and it's very important that it's not only programmers and investors, are involved, but also, the art world. So, so science has to meet art, and, and yeah, it's, it's something I'm really, really excited about, the other thing, you might have seen, that is pure science is chain analysis. So we talk- Talked about miner, capitulation and difficulty drops and that, that kind of stuff. You can extract a lot more data from the, from the blockchain than only difficulty or stock to flow ratio. You can, yeah, you, you have three hundred and gigabytes of data that you can analyze and extract and, and, yeah, I'm, I'm diving into that with new computers, new ideas and, yeah, with a more, more high frequency The trading, point of view and the things that you find are actually very, very interesting. I haven't, tweeted or written about it, I'm sure that, that will come and, and, yeah, that absorbs a lot of my attention at the moment, but it's, it's very interesting what you can, what information you can get from," }, { "speaker": "stephan", "time": "55:57", "start": 3357.24, "text": "From, from the chain. That's awesome. It sounds great. I'm really, looking forward to hearing about, what you've got, coming up next. I suppose just as a final comment for the listeners, can you just let them know what should they be watching and thinking about, just to understand if the model has broken down or, is there anything else? are there any other people who you would like to hear from?" }, { "speaker": "planb", "time": "56:19", "start": 3378.78, "text": "yeah, the, the model is, is, it's very simple. so, so, If it doesn't, show in, in, in the next, one or two years, if it's gone, then the model breaks down. And, so my point, it's a bit conservative, I'd like to see it earlier than that, but if, Bitcoin isn't above a hundred thousand or the fifty-five thousand, d-depending on the model you'd like to use, if it's not above that, that numbers, before Christmas, twenty twenty-one, then, yeah, the model is in real trouble. And and pro-probably so am I, because I, I made it, but, but, yeah, it, it, it is a possibility of course. Let me, let me finish with that. It's just a model, it's, it's not a guarantee for quick profits, and, it, it can be wrong. So, so, yeah, please, please watch that. And in the meantime, there's lots of things going on around the world, so, so, especially also in the investor world that slowly starts to understand Bitcoin and starts to invest in Bitcoin, and I expect a lot of, a lot more investors piling in, and the halving around May 2020 Might of course help with that if prices, do indeed rise like they did last two times and like, is predicted by my model, that would create the setting, where at least the, investors that believe in Bitcoin, I want to pitch it to their investment committees, at least those people will have, better charts, to, to show there and, and have a, a somewhat better story Than they have now, 'cause now you could still argue that Bitcoin is dead and, and it will never go, to the all-time high of twenty thousand again, but once that point is reached, I think we'll see Yeah, an exponential, rise in, interest of, investors." }, { "speaker": "stephan", "time": "58:32", "start": 3512.05, "text": "Fantastic. So, Plan B, just, let the listeners know where they can find you and follow you as well." }, { "speaker": "planb", "time": "58:38", "start": 3518.42, "text": "Yeah, so, so I'm on Twitter at hundred Trillion US dollars, hundred with a one o o, and then trillion USD. And my article is on Medium, it's called, Modeling Bitcoin's Value with Scarcity. but most of the time you can find me on Twitter, and all the DMs are open, so, please, please reach out. I, I like it, and I still can manage to answer at least all the DMs. So," }, { "speaker": "stephan", "time": "59:07", "start": 3547.1, "text": "yeah, well, thank you very much. It's been a fascinating conversation. Thank you. If you want the show notes or the transcript, go to stephanlivera dot com. There's also a link to subscribe to the show there. If you want to support the show, share it with your family and friends or in your chat groups, and email me if you've got any feedback or if you want to sponsor the show, stephanlivera at pm dot me. Thanks for listening, and I will see you in the citadels." } ] }