Cryptoasset Valuations: Past, Present and Future - Chris Burniske @ TOKEN2049

[Music] thank you for coming in after lunch so I'm Chris pernitskiy I am a partner at place holder and I am co-author of crypto assets which you just heard about and today I'm gonna be talking about crypto asset valuations something near and dear to my heart and something that I got started on at the end of 2014 actually at arc when greyscale approached arc about the Bitcoin investment trust and potentially putting Bitcoin in to publicly managed funds of course arc being a registered investment advisor required that we really understand the operating metrics the security model the potential for future appreciation of Bitcoin which really got me started on the valuation path reflecting on now versus then I think the most important takeaway is that we're even here having this conversation for a long time in the magic internet money community we were laughed at crypto assets were worse than tulips at least with tulips you got a flower with this you get bits on a database and so the fact that we're having the conversation is heading in the right direction it's still extremely early there's lots of opportunity for people to get involved and so I'll kick it off with talking a little bit about place holder which is where I am now I remain involved with arc but place holder is a venture capital firm that invests in decentralized information networks incentivized with the token we chose a venture capital model as it is a 10-year committed fund which allows us to really focus on the developers even though a lot of these crypto networks have valuations that are in the tens of millions or hundreds of millions or even billions many of them our seed stage teams that are still facing seed stage problems and so as a high high touch VC we come in and really go to work for the network and now on to what I'm going to be talking about if you agree that crypto is a new asset class which I certainly - then unique value drivers are going to require new valuation methodologies and I won't bore those of you in the audience today that have heard me talk about coupe des a new asset class but I did write a paper on it in 2016 with coinbase and G Dax where I go through the many ways in which we can differentiate asset classes and one of those is value or our value drivers and the value drivers are different here at least for how I think of crypto assets they don't have cash flows so you can't value them using a discounted cash flow model as you could equities and bonds the scarcity is much more predictable and programmed programmable which arguably gives us a better shot at valuing these than most commodities and so on and so forth that said not everyone agrees with me so professor damodaran who's really one of the godfathers of modern valuation wrote in his blog I don't believe that cryptocurrencies are now or ever will be an asset class you cannot value it or invest in it you can only price it and trade it so clearly professor damodaran and I disagree I've exchanged some emails with him I have a lot of respect for him and I hope to someday be able to convince him otherwise to get into that today we're going to be talking about two two kinds evaluation relative valuations verse intrinsic valuations so intrinsic valuations or fundamental valuations are really where you investigate a asset in a vacuum and you're trying to price that asset based on some fundamentals of it irrespective of what mr. market is doing whereas with relative valuations and this is what is most commonly used in crypto right now you are either pricing an asset based on its historical pricing or relative to other assets so I'm going to start with relative evaluations and I'll really focus on relative valuations for much of today because given the nature of the market it's probably the more useful of the two right now although over time we will mature and better develop intrinsic methods the most common relative valuation technique used today is now work value.i is the term network value because these are networks with native crypto sets they are not companies with stocks and thus I don't use the term market cap and network value is helpful but it can also be dangerous and it can be dangerous if we're not clear about which network value we're measuring or if the asset issuers in particular are disingenuous about the amount of their token that's currently circulating versus total supply so to give an example of that I'm going to show a tweet from Nick Carter or a couple tweets so I'll read it so whenever V is the latest version of something I call the market cap game he should have said it the network value game but that's ok part of the appeal of these Forks is that they get to claim a supply of X million coins but of course only a fraction circulate so they get a glamour spot on coin market cap with minimal liquidity why would anyone bid up the price because a fat market cap gets you a nice spot on all the ranking sites and if you do it for long enough investors FOMO into your coin it's a marketing strategy and you have extremely high leverage due to the tiny float so for example I could create another Fork of Manero and I could adjust the supply so that there will be a trillion or narrow tokens and I may only issue 10 million of those and if I have enough well friends then I can manipulate the price so that it's about $1 a coin in that case while the circulating supply may only be 10 million the total supply would be a trillion at which point I could claim to be the progenitor of the most valuable crypto asset on the market and it sounds like a joke but it happens and it happens more regularly than than most people think I was in a analysts meeting last year with a very reputable team that spoke about how their network value was actually within reach of bitcoins and they used their total supply as opposed to their circulating supply and for many of the analysts that were in the audience which were traditional financial analysts they didn't necessarily know the differ and a lot of people continue to not know the difference and it's systemic to the point where I'm going to show an example using coin market cap so with coin market cap you can rank different crypto assets they have coins versus tokens I think they define tokens as a crypto asset that operates on top of another network and so I'm going to use tokens as the example and you can rank them by circulating supply or you can rank them by total supply which I'm going to do in a moment and the difference is drastic so if you look at the top tokens by network value this is the number-one-ranked to rank three rank for rank and five ranked the number one rank by circulating supply is cos at 3.8 billion and the number one ranked by total supply is something called Internet of Things coin XO T which I'd never heard of but purportedly has a 30 billion network value and so I wanted to dig in and say okay of these top five tokens by total supply what percent of that total supply is actually circulating and if you go into coin market cap of the top five at least when I ran these numbers on March 16th I could only find two that had disclosures or facts around the circulating supply and that was exchange Union and veritasium in the orange here we have the total supply and in the blue you have the circulating supply so exchange Union has a total supply of 18 billion versus a circulating supply of 12 million and veritasium has a total supply of 12 and a half billion versus a circulating supply of 250 million so what does that mean if you're an investor that means that veritasium investors can expect a 50 X dilution at the whims of the development team and exchange Union investors of 1500 X dilution I pulled these numbers from coin market cap I'm not weighing in on their validity but this is a very publicly available and highly used resource that I don't think people really have the full context for how to use yet it's not just the tokens it's not just the total supply assets it's also the mind assets where you have to take into consideration the rate of inflation or total supply versus current circulating supplies so to do that we can look for example at Bitcoin versus - so expected supply inflation over the next year for Bitcoin is about 4% it's about 9% for BA - both of them are pretty moderate - launched in 2014 so it's a younger Network than Bitcoin which launched in 2009 but for example we can compare that to Z cash which launched at the end of 2016 and the expected supply inflation over the next year for Z cash which is about 75% I'm gonna dig into Z cash a little bit more before I do I'm gonna say I'm using Z cash because I have a tremendous amount of team respect for the team and the network and I the team really welcomes frank discussions i ran this past suco and zuko said thank you very much for bringing a lot of attention to this very important issue and what I'm gonna show is what you can expect the Z cash network value to look like in the future based on inflation plus your expectations around that inflation again using network value as a relative valuation tool so here's what Z Cash's network value has looked like since inception it briefly crossed the 2.5 billion dollar mark it's right around in the billion dollar range and so if we look over the next 10 years Z cash follows the same supply inflation schedule as Bitcoin with having's every four years currently pumping out 50 z cash every 10 minutes that we'll have after the first four years or 210,000 blocks to 25 in another four years it'll have to twelve and a half so on and so forth and so if you expect Z cash to stay at least when I pulled these numbers at three hundred dollars a coin this is what you can expect Z Cash's future network value to look like so over there you're at about five billion network value by the end of 2028 totally plausible but if you're a crypto asset investor getting a 0% expected rate of return for the next 10 years you're probably not doing a very good job and so you want some kind of expected rate of return and that rate of return should be based on the amount of risk you're taking so within the equity markets your expected rate of return is typically about 8% and so if you were to expect that rate of return from Z cash by the end of 2028 you would expect a network value that's over 12 billion dollars which is about 10x from where we are today again totally plausible for Z cash but just important to think about these expectations because when people will look at the ranking sites they'll say oh there's so much potential for future future appreciation in Z cash because it's so much smaller than Bitcoin without thinking about the fact that inflation has already baked into your expectations now again if you're getting an 8% rate of return on Z cash you're also probably not doing a very good job because crypto assets are significant significantly riskier than equity so you want to be getting a higher rate of return I typically use somewhere in the 30 to 50 percent range annually for for my discount rates within crypto if I use the the middle of that 40 percent then this is what I can expect the Z cash network value to look like at the end of 2028 so that is right about 200 billion dollars this is again just running a model based on expected inflation which is baked into the software and my expectations as an investor and so what that means is that at a 40 percent annual rate of return over the next ten years you are expecting Z cash to be bigger than than Bitcoin is today totally plausible but are you aware of your expectations implicit and explicit as a crypto asset investor some people will say well 40 percent annual rate of return that's a lot that's unrealistic I I don't expect that if you look at what 2017 was a 40 percent annual rate of return over ten years is about a 30 x4 ether holders in 2017 they multiplied their their wealth if they were able to hold by about a hundred ex start of the year to the end of the year so there's a lot of people if especially if you got involved in 2017 that have come to expect continued asset appreciation like what we had in 2017 now 2018 has certainly been a wake-up call but again what I'm trying to drive home is that supply inflation and expected rate of return are baked into every future Network value and so if you're using network value as a relative valuation tool that's fine you just have to be aware of its strengths and weaknesses network value is one tool that's just one number but a lot of relative valuation tools are actually ratios and when you have a ratio you have at least two numbers and the most common ratio is where you have some kind of price metric and you are calibrating that price metric for some kind of utility metric and the most commonly used is network value to transactions ratio the NVT ratio also known as the crypto p/e ratio the reason some people refer to it as the crypto p/e ratio is if you think about companies as being valued as a function of their profits then the core utility of a company at least in the eyes of the stock market is to pull profits out from the service it creates crypto assets at least how I think of them don't have profits they are actually exchanging a native asset a unit of account in exchange for the goods or services of that crypto Network and that's represented in the transaction volume and so in that case the denominator should be transaction value and in the numerator we have network value so we went through at the start of the presentation all of the assumptions and things we have to keep in mind around network value and we have to do the same for transaction value because suddenly we've doubled the complexity to get this ratio I'm going to break down how NBT is created with some graphs so if we start with Bitcoin and bitcoins Network value here's a beautiful chart log chart and the correction hasn't been as big as the the media makes it out to be at least on a log basis this is what bitcoin has looked like since it first listed on Mount Cox and we had a more or less reputable exchange price now this is the numerator in the network value two transactions ratio the denominator the transaction value looks like this and already we see okay there's some kind of resemblance so maybe transaction value going back to what I said about transaction value being the core utility of a crypto network maybe that is supporting the Bitcoin price and so we can overlay those two metrics on the left hand side we still have network value log scale so we've got Bitcoin over a hundred billion and then on the right hand side we have transaction value log scale so Bitcoin got close a couple days to processing ten billion it's more in the in the billion dollar range and so if I divide the blue line by the orange line then I get the NVT ratio and that's what this is now this specifically is an MDT ratio that was optimized by Dimitri college King you can search for his medium post but what he did is he took the daily network value and divided it by the 90 day trailing average of transaction value and he writes in his piece why he did that what I'll point out here is the the average is right around 80 and so bitcoin is right around in that zone if you look at different MDT ratios for example you can go to coin metrics and plot this MVT ratio you can also plot sort of what was the standard NBT ratio originally which was just taking the rolling average of daily network value divided by daily transaction value you'll see with that ladder ratio we're actually breaking out which means these assets are getting pricier because while network value has come down transaction value has come down more likely because still a fair amount of transaction value is tied to speculation edge traders shuttle assets between exchanges but I bring it up because even within a single MVT ratio depending on how you calculate it you can get different results this shows Bitcoin and just Bitcoin on a longitudinal basis but you can also use these ratios to compare across crypto assets but if you're going to do so then you have to be really careful about your price and utility metrics and that the assets that you're comparing merit being compared for example when I was a by side equity analyst I wouldn't compare General Electric with Facebook they're totally different companies with very different business models with very different growth profiles and the same goes for crypto assets and so we can look for example at comparing NBT ratios of UTX o based crypto currencies so u TX o being unspent transaction output which is the way in which all of these assets keep track of their balances so actually litecoins e cash D cred - in Bitcoin cash are all originally Forks of Bitcoin Sun that have been more heavily modified than others and these for the for those of you that are familiar with some my other work around taxonomy x' these are all specifically crypto currencies so a crypto currency being something that targets being a universal means of exchange store value unit of account which I would argue most crypto assets are not actually targeted at being crypto currencies the majority right now are more along the lines of crypto commodities we've got a fair amount of crypto tokens and this new vertical of crypto collectibles but this I specifically chose UT XO based crypto currencies and what we see is okay there there appears to be this comfortable range in the 20s you've got light coin which is kind of the so typical is gold and so maybe that's why it is valued less on a relative basis Zi cash I find interesting and Zi cash this is actually just for the non shielded transactions so Zi cash is true NBT ratio is probably actually lower than this because it doesn't include the shielded transactions D credit is one of your sexier cryptocurrencies going after the autonomous cryptocurrency idea and then you have - and Bitcoin cash which are elevated and when I say elevated all that means is that if you go back to what this mu tape MVT ratio means takes network value a price metric and divides it by transaction value a utility metric and so it's saying okay on a relative basis of transaction value how much is the market paying for these assets so the higher the ratio is the more the market is valuing that asset and that can either be because the market is irrational or overpaying for something or it can be because there's something defensible or something fundamentally different about these assets from the other assets something that I've been experimenting with that I didn't have time to put into a chart is you can actually have more than two variables within a lot of relative valuation ratios so for example I say here relative valuation ratios don't have to stop at two inputs and can value supply-side strength too so just going back we could calibrate for example these nvt these we could calibrate these nvt ratios for growth metrics so for example maybe - is growing much faster than any of the other Crypt of assets and so if - is growing 50 percent in transaction value or unique addresses whereas litecoin is only growing 10 percent totally hypothetical numbers but then if I divide NBT here by 0.

5 that would yield 134 for this ratio and if I divide litecoin at 23 by 0.1 that yield 234 this ratio and so then actually on a growth basis - would be undervalued and litecoin would be overvalued and so that's what I'm what I'm trying to get out with we don't have to stop at two inputs and I would call that ratio an N vtg ratio or someone cannot can come up with a better name but basically getting into this idea that we can take a existing relative valuation ratio and calibrate it for growth there are other ideas for example there's Network value - trading and transacting volume historically I've excluded trading volume from these ratios because if you look if you think about a crypto assets transaction value as a representative of its GDP than FX foreign exchange numbers are excluded from GDP and thus I have historically excluded it from the NBT denominator but for things like Bitcoin and arguably arguably etherium which are the reserved crypto assets for the ecosystem then the exchange into other assets could be viewed as a a purchasing of a good or service with a reserve crypto asset and so for example if you plot bitcoins env TT it's getting cheaper and cheaper and cheaper over time as it's used more and more and more as a reserve crypto asset there are others like pressed price to Metcalfe's law ratio there's a blog by clear blocks out there that digs into different variations of this metric that I find interesting so this again uses network value as the numerator but then different variants of Metcalfe's law where Metcalfe's law is squaring the number of nodes on a network represents the value of that network he uses Metcalfe's law for transaction value but also daily unique active addresses which I found to be very interesting and I think people should explore more and then this idea of you can investigate the supply side so all these three are demand-side metrics to show okay how much are people using this network this capital expenditure to network value is a supply-side metric to say okay what is the capex that's gone into Bitcoin relative to its network value so what is the dollar amount spent on Hardware relative to the dollar value stored and you could do that across proof-of-work assets you could also do it for proof of stake assets and look at the percent that is staked as a function of the network value onto intrinsic valuations so intrinsic valuations is actually something that I'm more interested in than relative valuations I find relative valuations to be interesting but it's not the holy grail of the space really I think that either we're totally wrong and we can't intrinsically value these cryptic assets or someone will win a Nobel Prize on being able to really accurately price these things over time and I think we have a pretty good shot at being able to accurately value crypto assets over time just based on the fact that all of this data is free open and transparent and the very nature of a blockchain is to monitor value flows with a precision that we really haven't had available within macroeconomic macroeconomic models to date so the most common intrinsic valuation method and one that I've put forth for a while is to use MV equals P Q so the equation of exchange within monetary theory and traditionally the equation of exchange is used to map out the flows of a currency within an economy and how the size of the monetary base changes the pricing of goods within an economy so on the left hand side we have M which is the size of the monetary base for example for the US dollar monetary base is about 4 trillion V is the velocity so the number of times per year that occurrence e turns over within its monetary base for the US dollar it's about five right now so for the US if you have four trillion in in a monetary base turning over five times a year that's twenty million in in volume that turns over and that equals this P Q and P Q within the way this equation is traditionally used represents the GDP of an economy where P is the average price of a basket of goods in that economy and Q is the quantity of goods in that economy historically how this ratio is most commonly been used is you assume velocity to be flat and you assume Q to be flat or at least not not mapping to the other variables and so as you increase the monetary base so - you must increase the price of the goods which is represented by price inflation I use it in a admittedly unorthodox way where I solve for the necessary size of the monetary base in order to support a crypto economy of size PQ at velocity V there have been some have who have claimed to debunk the use of this equation in this manner I haven't been convinced I'm always open to strong and rational arguments but as of yet have not been convinced and haven't seen a better model put forth with this PQ I'm gonna dive into that right now and how you can project that so this is a graph taken from a model I open-source if you search crypto asset valuations it should be the top hit on Google and within that medium post I have an in-depth blog and also this model about a fictitious coin called AI net which represents again a fictitious coin within a virtual private network network and within that model what I do is I project out the price and the quantity of gigabytes provisioned through this VPN network and the price represented in blue here cost per gigabyte is really just a run-of-the-mill technology cost decline curve and I think I used a 20% annual cost decline curve and the queue is a little bit more involved where you have to make assumptions around the total addressable market and then the penetration of a crypto network in provisioning that good or resource within that market and the economy peak you will grow salt so long as your unit growth outstrips your cost decline curve now just because an economy is growing does not mean that the necessary monetary base to sustain that economy has to grow because we have this big elephant in the room velocity which is one of the most important and as of yet still not well understood variables not only within crypto economics but within traditional monetary theory and to give an example if I hold all other variables constant but I just play with a velocity of 20 versus velocity of 50 versus a velocity of 500 the necessary monetary base to sustain the exact same economy looks fun it looks very different at a velocity of 20 we have a monetary base that needs to be roughly 350 million dollars at a velocity of 50 a monetary base just under 150 million and at 500 a velocity of 500 you have a monetary base that just needs to be a tad over ten million I bring this up because these are digital native assets and that means that the friction to transact them is less than the friction to transact a physical asset and we're nowhere near an equilibrium stage for these assets yet but as we get closer and closer to equilibrium stage I think we will see pretty drastic increases in the velocity of a lot of assets and so if you're not careful as an investor you could get caught in these velocity traps where you could have a crypto network that provisions a very large and necessary crypto economy of size PQ but if the velocity is super high then the actual monetary base necessary to sustain that economy can be low and again monetary base here is synonymous with network value and to back out the price of the token you just divide the necessary monetary base by the number of tokens in the float and if you've projected that out ten years from now then you discount back that utility value then to the present to get your discounted expected utility value which is how markets are priced or assets are priced there has been a really noteworthy extension to this theory that I originally put out by Alex Evans I highly recommend reading this blog it's called unvalued velocity and monetary theory where alex builds upon what i put forth within my original crypt asset valuations post and it is involved it requires reading a couple of times he also open sourced the model so you can see that but his primary extensions were to place a crypto asset in the context of a broader pool which allows you to decouple the demand for a crypto asset as money from the demand to use it to gain access to the commodity or whatever good or service is provisioned within the network he also incorporated payment patterns and what all of this allowed him to do is to end dodging ously calculate velocity which is a fancy way of saying the model calculates velocity based on the maturity of the crypto network as opposed to velocity being a arbitrary input that is indexing off of the the little bit of history that we have based on crypto economics so I highly recommend reading Alex's post I'm going to wrap up shortly before I do I want to say thank you to all of these data providers who I used in this presentation they don't make a lot of money from do this but it's a really important service so coin market cap of course we all know blockchain dot info has the best stats out there on the Bitcoin network crypto compare amazing service and then also lesser-known but to newer kids on the block on chain effects who has this great metric that shows network values for 2050 so sort of getting into my earlier discussion and then also coin metrics so coin metrics I believe you can download on chain transaction volumes for about 50 crypto assets now which is really amazing and they also have ways for you to calculate NVT ratio and compare them across script of assets to finish all models are wrong and some are useful and so everything that we went through today is an oversimplification of reality but one that we use to inform ourselves and they're more useful than they are dangerous so long as you're careful of your assumptions both implicit and explicit and so I encourage all of us including myself to better think about ok based on our experiences thus far in the crypto asset markets what am i overly optimistic about what am i under estimating and how is that going to influence future network values with that thank you very much I'm Kris Pearn a ski at placeholder' BC and I will be signing books if you would like me to sign your book in the coffee break afterwards [Applause] [Music]