Recent Posts

Thursday, January 11, 2018

Bitcoin Valuation Part II(a): What Might Work

We need to distinguish between a fair value and what price an instrument trades at. Economic and financial theory is almost invariably about the first part, and the second is not viewed as respectable. However, Bitcoin as an instrument is an excellent example how observed prices are actually determined by the actions of human beings, and not mathematical models. If we wanted to have a more formal approach, we need to look at the valuation of Bitcoin on a macro basis, and not arguments about the valuation of one Bitcoin on the margin.

Anyone looking for a price target for Bitcoin might as well stop reading right now. For reasons described below, I will not give anything that resembles an investment recommendation in Bitcoin. An addition issue is that the information that would be useful for Bitcoin valuation is not really in the public domain. This article is a continuation of Part I, where I described what will not work. Unfortunately, I had not written the whole set of articles, and what I described as "Part II" has ended up as multiple parts. Hence, this is Part II(a). (Link to Part II(b).)

An Instrument Trades at What Someone is Willing to Pay for it (or Sell it For)

If I had not signed my life away with non-disclosure agreements, I could point out some unusual movements in Canadian dollar fixed income relative value relationships (outside of the Financial Crisis, where everything was unusual). Also, I could explain exactly why those movements occurred.

That said, the only way to validate my story would be to subpoena a lot of market makers, and there is absolutely no reason for anyone to do that. Therefore, almost all academics would completely ignore my story, and would rather explain it based on some model or another. An economist would grumble about supply and demand curves. A financial academic would make up a story about risk premia. A physicists that dabbles in finance would come up with a story how prices are set by the random movements of investors, just like atoms (or whatever).

The fact that academics will insist on the wrong answer, in order to sound "scientific" -- probably tells us all we need to know about science in the context of finance. However, they have a point: there is no way of advancing our knowledge of finance by just saying that prices are whatever investors transact at. Therefore, we want some kind of valuation model.

Additionally, the observation does not help you much unless you have access to information on flows and positioning. Market makers in financial markets are in centre of a web of information gathering, and so they attempt to use their knowledge of positioning in a way to push weak hands into unfavourable trades. However, once that information flow is cut off, the trading advantage disappears. (Historically, there were a number of star proprietary traders that went off to form hedge funds and fell flat on their face. My understanding is that hedge fund investors have smartened up on this front, and so this happens less often now.)

Nevertheless, there is an important principle to keep in mind. Financial markets are not one person, one vote, rather one dollar, one vote. Models that pretend all investors are equal is an egalitarian fantasy. The movements of the big investors are what matters. When applied to Bitcoin, that means we need to think about the behaviour of the big holders of the instrument.

We Need to Think Macro

If we do want to attempt to come up with a valuation metric, we cannot use incremental analysis (what can we do with one Bitcoin?). That was the subject of Part I. Instead, we need to look at the entirety of the Bitcoin economy, and use its interactions with the rest of the world, to get a handle on valuations.

This is similar to the valuation of equities that do not pay a dividend. The control of the corporation presumably has some value (unless creditors take over), and so we need to come up with an enterprise value (dividing by the number of shares to get a share price target). This is not really that radical, but we need to be careful. One idea would be to compare the total capitalisation of Bitcoin to the equity market capitalisation some other payment network. Unfortunately for that idea, ownership of one Bitcoin does not give you any ownership of transaction fee profits; transaction fee profits (or losses) are earned by the miners.

One added complication is the relationship of Bitcoin to other crypto-currencies. One common valuation metric that is quoted is the relative size of market capitalisation of the various crypto-currencies, and the role of Bitcoin allegedly is to act as a senior currency among the group. This implies that some relative value investors will trade in a way to keep relative prices stable (in some sense) between the crypto-currencies. We then have to value the entire crypto complex as a whole, and then attribute a portion to Bitcoin.

The difficulty with that approach is that unless there is a guaranteed conversion privilege (which I believe does not exist), any investor that attempts to preserve an off-market parity is acting exactly like a central bank that is attempting to defend an unsustainable exchange rate peg. It will be successful -- until it runs out of ammunition. For this reason, each crypto-currency (that is not convertible to something else) has to be valued on a stand-alone basis if one wants to be safe.

Enterprise Value Does Not Work

The first thing to keep in mind is that the Bitcoin architecture requires the continuous use of energy. (One could presumably print out a hexadecimal dump of the blockchain, but that would not be a particular useful store of value, other than as a curio.) Until there is a significant Bitcoin "real economy" (as discussed in the previous post), those energy costs would be denominated in some fiat currency. The miners effectively need to raise fiat currency to pay their utility bills. This could be done by drawing down existing fiat currency holdings, which implies a fiat currency drain to finance positions in Bitcoin. Unless the miners have extremely deep pockets in fiat currency (a hard-to-estimate question), they will need to sell Bitcoin that ends up in the hands of some investor that wishes to buy Bitcoin using fiat currency. (Yes, there could be some intermediaries along the way.) If we want to believe that Bitcoin will survive in some form of steady state (a hidden assumption in valuation exercises), it seems that Bitcoin needs to attract net fiat currency inflows to cover energy costs. Therefore, if we wanted to think of the entire Bitcoin complex as a consolidated entity, it has an expected negative cash flow in fiat terms, under the plausible assumption that there are no utilities that wish to hold ever-increasing amounts of Bitcoin.

(It is entirely possible that creative utilities will offer electricity contracts that are fixed in Bitcoin terms. All we need are creative investment bankers that can structure a product that allows them to find investors to take the other side of the utilities' hedges. However, such financial engineering has just laid off the existing risk to some other party, and so the underlying flows still exist. Such hedges will have a finite life span, and when they roll off, the new contract terms would be exposed to the future market price of energy. In any event, it is entirely possible that the miners believe their own press releases, and would not consider hedging their energy cost risk.)

As an aggregated entity, "Bitcoin Inc." is going to have negative cash flow in fiat terms. However, that just tells us that using an enterprise valuation metric on the aggregate Bitcoin economy does not work. All it needs for a positive valuation is a reason for continued net inflows, which plausibly exist (as I will discuss in the next part). However, this aggregated analysis suggests that the physical intuition that an energy price spike is dangerous for Bitcoin can be easily understood in financial terms.

To summarise the argument in my follow up article (I hope the last), Bitcoin's valuation depends upon what price existing holders can demand, yet keep net flows balanced with the energy cost drain.

No Recommendations From Me

There is no way I could offer a recommendation to buy or sell Bitcoin, which is one reason I cannot offer a fair value target. The reason is that I am unaware of any legal precedents involving Bitcoin in an investment environment, nor is there is any reason that such precedents would apply both in my home jurisdiction and the legal jurisdiction of my readers.

Take a very simple example why this area is a legal minefield for advisors. How does Bitcoin fit into one's plans for leaving an inheritance?
  • If the method of access to your wallet is secret, your wallet dies with you.
  • If you provide a means to describe how to access your wallet, you are at risk of it being cleared out while you are alive. Bitcoin was designed to evade control by governments; the court system is a branch of government. The instrument is designed to make it extremely difficult (if not impossible) to seek legal redress.
I am sure some blockchain genius allegedly has this issue all sorted. However, given the reports of lost wallets by relatively young people, I have severe doubts that the proposed solution is going to survive the myriad of known risks that we face in the real world  (home fires, mental incapacity, etc.).

Link to the final installment, where I cover valuation in more detail.

(c) Brian Romanchuk 2018


  1. Check this out:

    1. Similar to one idea in my next post. Using Bitcoin as an intermediary does create a floor value, but only a floor.

  2. I agree with much of the discussion above concerning price dynamics at the last sale for bitcoin or a stock without dividends. However in a growth stock it is somewhat possible to guess whether the market is going to reverse trend from averaging up to averaging down, or vice-versa, based on techno-fundamental analysis, unless it is a very small cap stock being manipulated by boiler room or other sales and deliberate information tactics. In bitcoin I have no idea what type of fundamental information is available to a buyer or owner. I once owned a small cap growth stock that went up 169% in a few months and I sold when the CEO went on CNBC to "pump" the stock and I figured the large investors would be "dumping" on this "liquidity event". One never knows for sure but I played this stock just right at the time many years ago.

    I have an off-topic question that I hope you might address in a future blog post. A question is asked in this video around the 38 minute mark by a woman who wants to know if there is a shortage of collateral in the US system (I think this is known generally to economists as the Triffen dilemma but I'm not sure)?

    The bond trader answers this question with a statement that I do not quite understand, and in the background, another expert seems to be nodding approval. If you have the time and inclination to address the subject I would find it helpful and perhaps it might be of value to you or your readers to consider this topic. If not I appreciate the blog efforts you make in general.

    1. Well, there’s a lot of technical analysis around bubbles. Some people make money off of that. Not my cup of tea.

      Ugh - 38 minute mark? Hopefully I can jump to it easily. Not a fan of watching videos; I can read a transcript a lot faster.

    2. I think you can jump right to 38 minutes on youtube controls. The question is about a shortage of Treasuries prior to the global financial crisis being a causal factor in the crisis and the answer has something to do with market pricing patterns of Treasuries and forward Fed funds projections that supposedly demonstrates whether such shortage or surplus exists.

      Growth stocks trade on bubble dynamics every day and the market as a whole does not need to be in a "bubble" as you probably know. By studying growth stock price dynamics one can acquire many case studies in overshoot and collapse patterns of a price series.

    3. I was in the audience for that talk.

      He pointed out that you can buy a Treasury, and hedge out the interest rate risk, and earn a positive spread over the derivative (I assume OIS swap) rate. However, if there was a true shortage, they would be expensive versus swaps. That’s a pretty good argument, although I would have to think whether one could argue that the entire cash curve is expensive. (That seems implausible, but I would need a lot of data that I no longer have access to.)

    4. I appreciate the reply and not sure if you want to take further questions in this tread. I get that if there is a shortage of Treasuries they should be more expensive relative to a surplus but I don't understand how one can diagnose the difference and whether profits can be made consistently or whether one must get on the "right side of the market" relative to some shift that manifests in the future and puts one in the money versus in a cut-loss exit in this context?

    5. The way the trade works is that you buy a Treasury, and hedge the interest rate risk with a swap. You earn the short rate plus the positive spread you locked in. In other words, you have a t-bill with enhanced yield. If there were a shortage of Treasurys, you should not be able to get such a sweet deal.

      (Ahead of maturity, you get mark-to-market profits based on the movement of the spread between the swap and Treasury. In practice, you want to pocket those gains, and not the slow carry profit.)

    6. The quote below is from a reference discussing LTCM strategies.

      "In trading terms, a bet that the spread between swap interest rates and government interest rates will widen is called being long the spread. A bet that the spread between swap interest rates and government interest rates will shrink is called being short the swap spread. A long swap trade profits when swap spreads widen. A short swap spread trade profits when swap spreads narrow."

      So if this is the correct logic then which type of bet will pay off under a surplus of Treasuries? Which would pay off under a shortage? Or is the logic more complex?

    7. That quote does not help. Unless you are trading swaps, it is a serious mistake trying to keep up with swap trader lingo. I worked as an analyst in products traded against swaps for a long time, and learned the best way is to explain exactly what you are doing.

      The first problem is that different makets use different conventions as to what is subtracted from what. By itself, saying “the spread is 20 basis points” is meaningless, unless you get the jargon exactly right, so that readers know what convention you are using. However, if I write “The 10-year Treasury is trading 20 basis points cheap to OIS swaps.” any reader who has a chance of understanding anything I write knows that the 10-year Treasury yield is 20 basis points higher than the 10-year OIS swap rate.

      All the speaker said is that Treasurys are cheap (yield more) than swaps. This would not happen if there was a shortage of them.

      To make money with a spread trade, you are betting on the spread going up or down. Since “up” or “down” depends on your sign convention, you need to know how the instruments work to know what your trade structure is.

      However, if you think that Treasurys are going to get more expensive versus swaps, you buy Treasurys, and pay fixed in swaps. Any other phrasing is asking for trouble when you inevitably get the sign convention wrong. I do not recall anyone other than traders using “short” or “long” (or buy or sell) with respect to swap spread trades - including bond portfolio managers that worked earlier as market makers. (The fact that the working language was franglais also put a premium on clarity.)

    8. Well that's clear as mud although I don't fault you for my difficulty following the conventions or your efforts to describe the customs in those markets. It seems to me a spread would widen regardless of the sign convention and it would narrow regardless of the sign convention applied. So a bet placed on the change in a spread would either pay off or get liquidated over some time horizon. I know Hyman Minsky focused on money market conditions that trigger a financial crisis and Perry Mehrling writes papers showing the spreads before, during, and after the global financial crisis. I guess more education is required on my part to solve this riddle to my satisfaction.

      I know that the Treasury wrote a paper about Sallie Mae in which it said student loan asset backed securities (SLABS) had "rich valuations" at one time when organizations skimming profits off the government loan guarantees may have also been lobbying Congress frequently to argue that subsidies were not sufficient to profit from those loans. I recall thinking a rich valuation meant more money could be made by packaging SLABS because the rates of interest were comparatively high and the risk of default was comparatively low due to government default insurance payments. How that applies to a rich or cheap price of Treasuries relative to other available strategies I still don't know.

    9. The spread can be either positive or negative, and the payoff depends upon the absolute change. There’s a big difference from going from 20 -> 30 versus 20 -> -30 (although yes, narrowing is always in the same direction). Writing about spread widening/tightening therefore is dangerous in the swaps/govvie space. You’re only on safe ground in using tightening/widening in CDS, or in corporate bonds where the spreads are normally quite positive versus whatever benchmark you measure spreads off of.

    10. Thanks. I'll refer back here after doing some research for a thesaurus (Greek - treasure; treasury; or storehouse) of such terms (perhaps not easily obtained if swap experts don't publish such resources).

  3. I think I'll just stick to the simpler greater fool theory.

    1. The greater fool theory shows up (sorta) in the next part...


Note: Posts may be moderated, and there may be a considerable delay before they appear.

Although I welcome people who disagree with me, please be civil.

Please note that my spam comment filter appears to dislike long "anonymous" posts. I get no warning about this, and only go through my "spambox" infrequently. The best bet it to keep comments short, and if you think the spam filter struck, let me know with a short comment.