Topic: A new way to calculate "fundamental" BTC value (Read 1478 times)

True, Bitcoin isn't a company. Crypto is an entirely new asset class and it's not clear how to define its "intrinsic value." The point of my proposal is to come up with a new way of thinking about its valuation by making an analogy between crypto and a stock share that provides a dividend. The problem then becomes to determine the magnitude of the "dividend" provided by crypto (not including the profit/loss that result from bitcoin price fluctuations).

For the sake of this thread, I am looking for a way to measure the intrinsic worth of bitcoin as a payments system, independent of its worth as a currency or as a store of value. It is a useful component of a payment system and this is what creates demand for bitcoin. This demand creates buying pressure and it is the price that would hypothetically result from this degree of pressure that I'm interested in measuring. This is independent of the buying / selling pressure generated by bitcoin investors or speculators.

So, summary of my method to measure intrinsic value of bitcoin-as-a-payment-system: bitcoin is useful as a payments system; this use creates demand for bitcoin; this demand results in upward pressure on the price; the price that would result from this upward pressure is modeled as the price of a share that provides a "dividend" to "shareholders" (ie bitcoinholders). The "dividend" takes the form of what is effectively an "interest payment" made to the bitcoin holder by those who wish to use it as a payment system, but do not use it as a store of value. The average "interest payment" is determined by the spread on the various bitcoin exchanges. The calculation is performed by inputing the volume of transactions over the payment system per unit of time (eg per year), multiply that by the spread on exchanges to get the effective interest payment, and divide by the number of bitcoins in the system, to calculate the dividend per bitcoin per unit time. Then multiply by whatever you think would be a reasonable P/E ratio to get the final estimated intrinsic (taking into account use as a payment system only) worth of one bitcoin.
 

What profit?
You have the opportunity to make 30%/year on some bitcoin investments which is bullish for the price of Bitcoin.

It is obviously that now it is only a question of speculation , when all this will finish the price will "drop" and we can start to spend our bitcoin.

20:1 is massive bubble territory, i.e. where the stock markets are now.
Bitcoin also isn't a company, show me a currency with a 20 P/E ratio, what is the P? a currency doesn't make profit!

OK but what about the fundamental use of store of value?

I have been pondering whether we might predict that the future velocity of bitcoin should be higher than present-day fiat, lower, or the same.

The central question is: WHAT DETERMINES VELOCITY? I am thinking along two lines:
1. Velocity can be limited by technology
2. Velocity can be limited by people hoarding money

Is there one approach that covers both of the above factors? Or do they need to be considered separately? idk. But here are some thoughts:

1. The velocity of bitcoin in a world where it is used as a payment system will be different than the velocity of bitcoin in a world where it is used as a currency. In the first world, every bitcoin transaction involves conversions: person A converts fiat to bitcoin -> bitcoin sent from A to B -> person B converts bitcoin to fiat. These currency conversions cause resistance to movement. This can be thought of as a technological limitation, ie the transaction is expensive to do because the ecosystem has not fully evolved to take full advantage of superior crypto technology. In the second world, fiat has been cut out of the loop. The resistance to movement will be reduced, and so the velocity will (perhaps) be higher. Higher velocity means lower price of bitcoin, if we keep other variables (transaction volume per unit time) the same. Paradoxically, this means that the value of bitcoin COULD DROP when we make the transition from bitcoin-as-payment-system to bitcoin-as-currency. I say "could drop" because in such a scenario, the increase in velocity will probably also be accompanied by dramatically increased volume of bitcoin transactions per unit time, which would have the effect of INCREASING the price of bitcoin. So the question becomes: which factor changes more: increase in velocity? (which will depress BTC price) or increase in transaction volume per unit time? (which will elevate BTC price). Probably the latter, but ... don't know.

2. If bitcoin is used as a store of value (unlike fiat), and if we assume that investors will want to keep their long term investment in cold storage, then that will effectively reduce its velocity since a big chunk will be locked up in cold storage. (Alternatively, we could model this as a smaller money supply, with cold storage bitcoin being not involved in the calculation of bitcoin velocity.) Under long-term steady state conditions (like 50-100 years in the future), we may assume that bitcoin's value will increase (in real terms) at a rate that approximately equals the growth in the economy (assuming that size of economy ~ volume of transactions per unit time, with constant velocity). That means that keeping your wealth in bitcoin would be expected to give approximately the same yield as keeping your wealth in an index fund that approximates the entire economy (like a Vanguard international fund). So, where would you store your wealth? Here's what I think:

- you would keep as much bitcoin-as-bitcoin as you think you will need for the sake of liquidity
- if you have excess bitcoin savings, your number one preference would be to lend it to other individuals / businesses who need liquidity and are willing to pay interest. This would increase your stock of bitcoin and therefore be BETTER than hoarding bitcoin, and therefore BETTER than hoarding stock market shares.
- you would NOT buy shares in the stock market unless:
--- you want to diversify risk (hedge against the possibility that, eg, some altcoin knocks bitcoin from its pedestal)
--- you are a VC, meaning that you believe you have a better than average eye for businesses that will outperform the rest of the market. If you're just some guy who picks blue chip stocks and holds for 30 years (a la Motley Fool), you can expect to do basically the same as if you just kept it all in bitcoin ... and you'd have the disadvantage of being less liquid. So: keeping wealth in bitcoin is (marginally) better than investing in the stock market.

Getting back to the question: WHAT DETERMINES VELOCITY? In my thought process above, people keep most of their wealth in bitcoin. They don't buy shares of stock (not like we do today), but they do LEND BITCOIN AT INTEREST to companies that need liquidity.

I am thinking that in crypto-utopia, the resistance to movement due to technical problems is no longer the limiting factor. Instead, the velocity of money is determined by how much of their wealth people keep in the form of liquid bitcoin. And this would be proportional to their uncertainty in encountering unexpected expenses. The more uncertain you are about tomorrow's expenses, the more liquidity you will need on hand to cover unexpected expenses. If you have no uncertainty whatsoever about your future expenses, then you don't need to be very liquid at all: you can time your bitcoin loans to mature at precisely the right time to cover your planned expenses.

Risk taking (in the overall economy) ==> increased need for liquidity ==> increased demand for bitcoin ==> increased value for bitcoin

Ahh, I'm getting tired. My apologies for a long and sprawling post. May have to edit it later. One point I am trying to build up to:

Wouldn't things be a lot more stable if we just tethered shares in the Vanguard International Stock Market (or Vanguard Global Fund or whatever is the broadest measure of the worldwide economy) to bitcoin, and used that as the worldwide currency, instead of free-floating bitcoin? We expect the value of bitcoin (under long-term steady state conditions where bitcoin is the world's reserve currency) to change roughly in proportion to the growth in the economy, but that is based on a number of assumptions that may or may not be true.

ty


Thanks for pointing out the link oda; I had not seen it before. Your application of the quantity theory of money makes sense to me. I find it intriguing that you and I came up with very similar estimates (you calculated a value of $2000/BTC, I calculated a value of $6800/BTC -- see edit of post #5) even though our methodologies were completely different.

Similarities in our approaches:
First, we use similar values for the annual transaction volume: you used $200B per year, and I used $360B per year. (If I had used the same value you used -- $200B instead of $360B -- then my final calculated answer would have been ~ 3800 / BTC, not too far away from your $2000 / BTC estimate.)

We also use similar values for the current money supply of bitcoin: you assume 10M bitcoins, and I assumed 10.5M. (Actually I assumed 21M but then I effectively changed it to half that when I threw in a factor of 2 later in the analysis.)

Differences in our approaches:
You basically applied the quantity theory of money with the assumption that the Bitcoin velocity could be approximated by setting it roughly equal to the velocity of USD M2, as measured by the Fed. More on that in my next post.


I wouldn't say that I set the velocity to infinite. My approach makes no assumptions at all about the upper or lower limits to the velocity of bitcoin; it simply does not show up in my calculation. ( I could use my final derived valuation of bitcoin to calculate the velocity using the same equation you used, using the quantity theory of money, and I'd get something pretty similar to the value you got. )


I like your use of the word friction. The spread in the exchange rates is sort of a proxy for friction of movement of bitcoin through the market. Could it be that velocity is to a large extent, determined by this friction? I'd say yes, it makes sense to say so, in a model where bitcoin is used as a payment system rather than a currency (in which case friction would be determined by other factors -- maybe miner fees -- rather than exchange spreads).

So instead of looking at velocity, my method takes as input the typical spread on a BTC/USD exchange, which I set equal to 1%. This is effectively a measure of how much people are willing to spend (more appropriately: how much they do in fact spend) for the privilege of borrowing bitcoin for however long it takes to do their transaction (whether that be one millisecond, 15 minutes, whatever). The fact that people pay real money to borrow bitcoin is what gives bitcoin intrinsic value in my model.

So your method takes as input the velocity of bitcoin, whereas my method takes as input the spread on a bitcoin exchange. What on earth do these two values have to do with one another? and how the heck did we arrive at almost the same answer for USD value of BTC -- better than a factor of 2 apart? Two possibilities:
1. Blind luck on my part.
2. There is some sort of deep connection between the velocity of bitcoin and the liquidity of the exchange markets. Intuitively, I can see that such a relationship might exist, along these lines:
bigger spread on the exchanges <==> decreased liquidity of exchange markets <==> higher friction of movement of bitcoin through the financial system <==> reduced velocity
I've never seen anything written about such a relationship, although I'm not an economist and not familiar with the literature. Maybe a relationship exists and is well known; I don't know.


One of the things that makes the concept of "velocity of money" difficult is defining what we mean. If a single bitcoin transaction is instantaneous, but a typical bitcoin only gets transacted 10 times per year, then is its velocity infinite? Or is it 10 per year? I think we mean the latter. So no, I'm not setting the velocity to infinite. But there are other aspects of the definition that are even more confusing (see next post).


I'd agree that my friction analysis places a lower bound. And I'm thinking that an estimate of velocity of money is another lower bound. I'm contemplating whether these two estimates in theory should yield the same result.

NEXT POST: more thoughts on the velocity of bitcoin

Interesting approach, BTCtrader71. Thoughtful analysis, much appreciated.

That said, you're throwing out the baby with the bath water a bit here. I did my own estimation of the 'transaction-based minimum valuation of the network' in a thread a while ago (link), and I understand that probably the most crucial (and difficult to estimate) parameter is the velocity.

Your solution, mathematically at least, basically boils down to setting the parameter to 'infinite', then grating some 'friction' that supports the network minimum valuation you derive in the end.

Problem is, like any currency, Bitcoin does have some velocity of money, even if it is harder to estimate, and it almost certainly isn't unbounded, i.e. transactions that involve actual goods or services are almost certainly /not/ completely instantaneous like you treat them.

So in my opinion, the conclusion to draw from your 'friction' analysis is providing an absolute lower bound to valuation, on top which the next level of valuation (derived from an estimate of velocity of money on the network) has to be placed. Agreed with that view?

The 0.01 refers to the 1% due to slippage. The 10% was already baked in to the calculation: 10% of 500 billion gives 50 billion

Speculators can be wrong!
Consider that, going by market reaction and commentary by traders, many speculators felt Silk Road closing was bullish. They work under the assumption that the more Bitcoin is regulated and controlled, the more it will resemble a respectable investment vehicle and therefore skyrocket in valuation.

Speculators believing so will trade accordingly, and suffer losses if it turns out that black market trade and other illicit uses are one of Bitcoins key competitive advantages. If this is the case, cleaning the Bitcoin ecosystem up will cause fundamental demand for BTC to shrivel and leave users with a currency that's a kind of nifty but cumbersome way of avoiding credit cards.  OTOH, if the black market continues to thrive and grow, speculators betting on legitimacy will miscalculate fundamental support levels and sell too low.

Thanks for making this topic!

Isn't 10% 0.1 * and 500 billions 500,000,000,000 or am I confused?
0.1 * 500,000,000,000 / 21,000,000 = $2,380

Makes total sense to me but I'm just wondering how the speculative selling can drive the price down so much and so quickly when we have a fundamental value established. When the speculators put a sell order eventually they will need to buy back. During the course of driving the price downwards I presume there are bound to be someone who panic sell forcing price to dip lower and thus the speculators can afford to buy back at lower price.

Summary: the intrinsic value of bitcoin-as-a-payments system can be modeled as if bitcoin were a share in a company and paid out dividends, distributed to all of its shareholders. The amount of dividend per unit time is calculated by taking the total volume that gets transacted through the network and multiplying by the percent that would be lost through slippage (say, 1%).

Let's use that in some other calculations.
Remittance
Assume that bitcoin has captured 10% of the $500 billion per year remittance market. This results in yearly dividend of:
0.01 * 50,000,000,000 / 21,000,000 = $23.80
IOW, from remittance market alone, your one bitcoin should be valued (intrinsic value) as equal to a share that produces a whopping $23.80 in yearly dividends.

Let's do the same for some other markets.

eCommerce
assume bitcoin captures 10%
eCommerce is estimated to be $1.3T per year
0.01 * 130,000,000,000 / 21,000,000
==> one bitcoin like a share that yields annual dividend of $61.90

black market transactions
assume bitcoin captures 10%
black market transactions is estimated to be $1.8T per year
0.01 * 180,000,000,000 / 21,000,000
==> one bitcoin like a share that yields annual dividend of $85.71

Total: one bitcoin has intrinsic value equal to a share in a company that yields an annual dividend of $23.80 + $61.90 + $85.71 = $171.41

This is an underestimate given that only about 13.8 million coins have so far been produced, and who knows how many are lost and effectively don't exist. To be fair, let's double the final number: $171.41 * 2 ~ $340.

So, how much would you pay for something that produces $340 per year? Well, I suppose that depends on how long you think it will keep producing. Let's say that we have faith that it will exist for 10 years: so, roughly, one bitcoin should have intrinsic value (bitcoin-as-payments-system) of about $3400.

Think of the $3400 per BTC as being like the floor (after bitcoin has penetrated 10% of the above markets). The ceiling would be determined by other factors; I'm still pondering how to model them. To begin with, I don't know how to theorize a relationship between intrinsic value and speculative value. Or whether there even is one.



EDIT: using a traditional P/E ratio of 20 (instead of 10, which is what I effectively used above when I assumed it will produce for 10 years), and an estimated $340 "earnings per year" per bitcoin, gives us an estimated bitcoin valuation of $6800/BTC.

The more I think about it, the more this analysis makes sense. When bitcoin is used as a payments system, every dollar that gets transacted through the bitcoin network translates into roughly one cent (estimating the cost of slippage to be 1%) that gets injected into the market. That one cent exists because collectively speaking, the people involved in the transaction need to "buy high, sell low" for the privilege of using the bitcoin system. That one cent goes to bitcoin holders, so can be modeled as a dividend that is distributed evenly to all bitcoin holders [1].

Note that I am considering the intrinsic value of bitcoin as a payments system as opposed to bitcoin as a currency, the difference being that users of bitcoin as a payment system go through the cycle of fiat -> bitcoin -> fiat once per transaction, but users of bitcoin as a currency just don't use fiat at all. The intrinsic value of bitcoin-as-a-currency will need to be calculated using a different method than the one I am proposing here, just like the value of bitcoin-as-a-store-of-value or bitcoin-as-speculative-instrument will also need to be calculated differently.

[1] The fact that the one cent "dividend" goes to bitcoin holders who buy and sell on exchanges and is not actually distributed evenly to all bitcoin holders, like a real dividend would be, does not matter for our purposes; we can pretend that it gets distributed evenly. What really happens is that the one cent drives up the price of bitcoin, thus benefitting all bitcoin holders whether they are active market players or not; so it is "as if" the one cent does get distributed to all bitcoin holders, in proportion to the amount of bitcoin they own, like a dividend.

There are lots of variables that go into the calculation. One of the corollaries seems at first to be counterintuitive: as market liquidity improves, and slippage decreases (below 1%), then all other variables being equal, the fundamental (i.e. intrinsic) value actually decreases. The key here is: all other variables being equal. Because of course, they're not, actually. If loss due to slippage decreases, then the cost of using the network decreases, making the volume of transactions per unit time increase, which in turn increases the intrinsic value.

Seems legit. I'd assume 80% speculative and 20% intrinsic value. But, you may be more right than me.

I have seen lots of people attempt to calculate how much one BTC should be worth based on various assumptions, e.g. that within X years a certain percentage of transactions will be performed using bitcoin instead of fiat, or using MV = PQ, the equation of exchange from the quantity theory of money. (http://en.wikipedia.org/wiki/Quantity_theory_of_money#Equation_of_exchange). There are a number of difficulties, one being that it is hard to get a meaningful measurement of bitcoin's velocity. So, I thought of another method that is a little more tractable.

First, separate the demand for bitcoin into two categories: 1) using it as a currency and/or payments system ("fundamental" value), and 2) using it as an investment / store of value ("speculative" value). I will be proposing a way to calculate its fundamental value alone, without taking into consideration what its speculative value is.

I am going to make a simplifying assumption: that whenever anyone uses bitcoin in a transaction, three things happen:
1) person A (e.g., the customer) converts fiat to bitcoin in the precise amount needed for the transaction;
2) person A transmits bitcoin to person B (e.g., the merchant); and
3) person B converts bitcoin immediately to fiat.
(These assumptions break down if you assume person A is an early adopter who is cashing out bitcoin bought long ago; but that assumption cannot hold forever, bc eventually the early adopters will deplete their stash. )

Note that two things are happening here simultaneously: person A is buying bitcoin at market value, and person B is selling bitcoin at market value (not necessarily on the same market, although we are assuming they do it simultaneously). If you were to execute simultaneous buy and sell orders (market orders, not limit orders) on an exchange, you would lose money through slippage. I'd estimate this comes to about 1% (assuming the size of the transaction is not large enough to move the market or generate lots of slippage). Essentially, what that means is that person A and person B are collectively "paying" a fee of 1% of the transaction that goes directly into the market. In terms of moving the market, it would be equivalent to buying bitcoin. Buying how much bitcoin? The amount that you lose through slippage. Note that I am not taking into account miner fees or exchange fees, because those amounts do not get injected into the market as demand for bitcoin, so to speak.

Therefore, every bitcoin transaction can be conceptualized as market demand for bitcoin, currently about 1% of the value of the transaction itself. IIRC, BitPay recently stated they process $1million per day in transactions. Let's assume that BitPay accounts for one fifth of all transactions-as-payment-system, so there are $5million per day total. That translates into $50,000 of bitcoin demand per day, or $18M per year; at current prices, this is an effective demand for 217 bitcoins per day attributable to its fundamental value.

Clearly a demand of 217 bitcoins per day is small potatoes compared to 3600 bitcoins generated per day. But that number was certainly a lot smaller a year ago, and will almost certainly be a lot larger a year from now. If we assume bitcoin usage triples every year, then in about 2 years, the fundamental effective demand will be roughly equal to, and will roughly cancel out, the supply from miners -- and that's assuming that miners dump 100% of their coins immediately. (This calculation also assumes price is constant.)

To sum up, the current "fundamental" value of bitcoin is equal to that which would be (hypothetically) generated by a demand of $18 million worth of bitcoin per year. I'm not sure if this comparison makes sense, but if you think of this as analogous to earnings of a company (clearly it's not the same thing, since bitcoin is not a company; the question is whether it makes sense for our purposes to make the analogy), and plug in the typical P/E ratio of 20, then we get a total valuation of about $360 million. The current valuation (market cap) of bitcoin according to http://coinmarketcap.com is $3.2 billion, roughly 10 times that. So we can conclude, based on these calculations, that at current prices ($230/BTC), about 10% of bitcoin's value is justified as its fundamental value based on current usage as a payments system, and the remaining 90% is speculative.

What do you guys think? Does this approach make sense to you?