Unable to dynamically match supply with demand

This post will look at an amalgam of ideas touched on by Eli Dourado in a post several days ago regarding Bitcoin.  This includes volatility, cross-border payments, nemo dat, settlement finality and machine-to-machine transactions.

I also answered a few frequently-asked-questions that have been emailed to me that intersect with some of the same ideas.


On Sunday Eli Dourado posted a response to Noah Smith and JP Koning both of whom previously discussed why bitcoin has not become a medium-of-exchange.

I don’t want to turn this into a post solely on volatility so if you’re interested in other ideas, skip to the next section titled cross-border payments.

The problem with Dourado’s analysis on volatility is that it does not look at what the actual causes of volatility are, the core of which is a perfectly inelastic money supply.

What does it mean to have a “perfectly inelastic money supply”?  In short, irrespective of the quantity demanded, the money supply itself does not change or shift.  For a Bitcoin-like network, its supply is programmed to remain static irrespective of external conditions.  While some advocates and enthusiasts consider this a feature, it is a bug if bitcoin wants to be used as a modern medium-of-exchange.  Why?  Because the only way to reflect changes in demand is through a change in price, which as described below, is done so via volatility, often violently.

And consequently, determining what the elasticity of demand could be is effectively impossible due to the opaqueness in both the exchange and OTC markets, which partly explains the unpredictability around cryptocurrency prices in general.1

In contrast to Dourado’s view, Robert Sams recently provided a more cohesive look at the fundamental reasons for why, despite the creation of new “liquidity” venues, uncertainty cannot be removed in a similar manner:

volatility 1

volatility 2

volatility 3

The three slides above appear in an April 2015 presentation by Sams.

Yet it is Sams’ short white paper on stable coins that probably, succinctly, describes the issue of future uncertainty with present day prices:

It is the nature of markets to push expectations about the future into current prices. Deterministic money supply combined with uncertain future money demand conspire to make the market price of a coin a sort of prediction market on its own future adoption.  Since rates of future adoption are highly uncertain, high volatility is inevitable, as expectations wax and wane with coin-related news, and the coin market rationalises high expected returns with high volatility (no free lunch).

In other words, at present bitcoin’s price inelasticity of demand means bitcoin’s price isn’t a function of the availability of bitcoin or, for that matter, demand for it.  This makes bitcoin vulnerable largely to the machinations of prognosticators (e.g., pumpers), not tangible market forces.2

Below are a few other questions that have hit my inbox related to volatility which tie into the ideas addressed by Dourado and others above.

Some short Q&A on volatility and prices

Visible volatility appears to have declined in the past 5 months, why?

One possible explanation relates to the inelasticity argument: if traders “feel” that this is a good price and there is no motivation or incentive to trade, thereby moving it up or down, it will tend to stay there (i.e., trading based on sentiment).

Another potential explanation for why there has been less volatility in the last couple months could be that as participants have left the market, there has been less demand from speculators due to a lack of interest and thereby a corresponding lack of volume.3  We may not know for sure what the actual trading volume is at exchanges in aggregate for years to come.

For instance, contrary to the Goldman report, the Chinese RMB does not compromise 80% of the trading volume; this “volume” as discussed by Changpeng Zhao (former CTO of OKCoin) were a combination of internal market making bots, wash trades and tape painting.4  If there was a legitimate increase in demand from speculator then there would have been corresponding increases.  Maybe “whales” will return again after Fed tightening or concerns over Greece.  Or maybe not.

In addition, VC funded companies like BitPay are stating on record that they absorbing some (all?) bitcoins onto their balance sheet, this likely in the short run reduces some of the volatility but is not sustainable.

Why not?

Because with roughly $400,000 – $800,000 in trade volume per day that BitPay processes, it simply does not have the cash on hand to absorb all of the incoming bitcoins for more than a few weeks at most.  Thus, despite the claims (video) from Jason Dreyzehner — that BitPay tries to keep all of the bitcoins that they process — after talking with several contacts at large exchanges, it turns out BitPay does in fact sell bitcoins in bulk to exchange and OTC partners.  See also, A pre-post-mortem on BitPay.

Another common question I have received: with a string of “positive” developments lately such as GBTC, new exchange infrastructure, and more VC funding, why hasn’t bitcoin’s price risen?

It hasn’t risen in part because of elasticity.  Bitcoin’s value can be susceptible to external factors, but it does not need to be if there is inelasticity of demand.  In that case, steady prices amounts to Newton’s First Law.5

In addition, thus far there is no compelling reason for:

1) Consumer-based transactional demand.  To most consumers in developed countries, trying to use bitcoin is an added friction, so they are not interested in doing that.  What are the demographics of a bitcoin owner?  Based on several sources we know what the owner demographics are: a North American / European male in his early 30s, they have access to other payment platforms and own bitcoins primarily as an investment, not virtual cash.6

2) Speculative demand has not increased (yet) because it is now an old story for some active traders — they know what a “bitcoin” as an asset is and how to get it.  As Nathaniel Popper (from NYT) discussed a couple weeks ago at Plug and Play, editors and writers at large media companies are tired of the same stories, these Bitcoin companies need to now go execute which few have actually done.

What about the new exchange companies and liquidity providers being added to the market?

As noted above, as of this writing the price of bitcoin is largely a function of speculative demand still.  Companies like Coinalytics have looked at the on-chain data to show that there has not been much of an increase in on-chain usage or demand from above-board commercial entities.7 Perhaps that will change.

Therefore if consumers are not participating, bitcoin is left with movements dictated by changes in the unpredictable demand curve (and appetite) of speculators.  There are startups that provide different types of instruments: SolidX, LedgerX, Mirror, Tera Exchange and Hedgy but none has likely gotten much volume and only have limited capital to absorb the continual bitcoin production rate of miners and other sell-side participants.  Again, maybe this will change over time.

What if bitcoin adoption were to proceed more aggressively in non-currency applications (real-time securities settlement, for e.g.), what is the impact from that on bitcoin’s price?

First off, the Bitcoin network is not a real-time securities settlement, at most it clears one batch in roughly 10 minutes — not real-time.  But if we are truly defining post-trade finality in terms of title transfer, Bitcoin itself cannot do that with off-chain assets.  Why not?  Because Bitcoin’s validators — in this case mining pools — have no control over off-chain assets.  Title still resides and is controlled off-chain, out of the purview of miners.8

Ignoring that for a moment the main reason why watermarked methods have seen a surge in interest is so that a company (or financial institution) does not need to buy gobs of bitcoins in order to represent socially-recognized value on the edges (houses, cars, airplanes, boats) — thus since watermarking takes a small fraction of a bitcoin, even in aggregate it probably does not add much demand to bitcoin itself.  Whether that is a secure method for transferring value is another topic altogether.9

On this point I also spoke with George Samman, co-founder of BTC.sx and weekly contributor to CoinTelegraph.  In his view:

When talking about settlement and clearing the sheer size – in dollar terms – of the FX and equity markets, it makes a 51% attack on watermarked assets much more of an eventuality than a probability simply because it’s now worth the effort to do so.  Why?  Because the increase in aggregate asset value transferred on a blockchain incentivizes attacks.  In fact a new paper suggests that an attacker does not even need 51% to achieve their goals.

How might Bitcoin help FX traders and arbitrageurs more easily and quickly align their books and execute a global strategy?

As of June 2015, probably none. The market simply is not deep or liquid enough compared to the multi-trillion dollar FX space.  Even if we took the volume of Bitcoin exchanges at face value — that operators are not exaggerating their numbers which we know they are10 — you would need volume to increase by several orders of magnitude before FX traders probably are interested in using it either as a vehicle or as part of their “global strategy.”

According to Bitcoinity — which uses self-reported volumes — total global bitcoin trading volume over the past 24 hours amounted to 312,532 BTC (~$78 million), though 70-80% of that is likely market making bots and wash trading.  For comparison, according to the BIS, in April 2013 the daily FX turnover globally was $5.3 trillion.  This number has stayed roughly the same over the past several years.11

What impact can the BitLicense have now that it has been finalized?

Again, I’m one of the few people that thinks the BitLicense is not a bad thing — it may seem expensive but if a Bitcoin company provides the same good and service as a traditional company then it would make sense to have them liable to the same type of compliance — why do they get an exception just because of the word Bitcoin?  With that said I do think that it could bring in more players who believe this now provides regulatory certainty.

For example, I am looking forward to seeing how Gemini impacts the network now that there is a legitimate exchange you can “short” bitcoin on — it may provide a new incentive to destabilize the network in order to gain.

For perspective I reached out to Raffael Danielli, Quantitative Analyst at ING Investment Management.  In his view:

The points made in Robert Sams recent post are worth looking at. It is a reason to be wary of a professional exchange such as Gemini. Also it adds to the volatility problem. It is probably just a question of time until we see some hedge fund disrupt the network somehow while profiting from it with a massive short. The incentives will be in place sooner or later.

Honestly, I believe that the misconception about volatility (“it will go down over time”) might blow up in the face of many people. The argument that Robert Sams makes is strong. As long as supply cannot be dynamically adjusted to match changes in demand expectations (essentially what the Fed is trying to do) volatility is unlikely to decrease.

It is worth pointing out that a trader can currently “short” bitcoin on Tera Exchange and Crypto Facilities via their forwards contracts (and swaps in the case of the former).  So far the only participants interested are miners for obvious reasons (though it is unclear if anyone involved is generating much revenue yet).  It is also unclear what the incentive for doing a swap is too, with the inability to predict or model exchange rate changes months into the future.

I also reached out to George Samman once more.  According to him:

It is more about the implied volatility which for bitcoin, is always higher than other asset classes and the reason I believe this is because bitcoin is still a giant unknown. Bitcoin continues to trade mainly on sentiment and technicals as well, and this in turn makes it by nature a more volatile asset.

I would also say to the disappearing volume on exchanges it has to do with a lack of trust, hoarding by deep pockets, and its been going off-exchange.  For example LocalBitcoins volume hit record highs in May, while volume at the biggest exchange and the one used by the most active traders use, Bitfinex, has declined steady all year long.12

Kraken, the San Francisco-based crypto currency exchange, is launching a new “DarkPool” option for volume traders who want to buy and sell coins in larger orders. Typically, large orders in the exchange swing the price of bitcoin dramatically, but with the new dark pool trading option, it lets people or institutions order in a way that the rest of the market does not see. Think of it as a level of privacy for people buying or selling bitcoin in bulk. The service will cost users an addition point-one-percent on orders.

Kraken is not the first exchange to bring a “dark pool” to market.  In 2013 Tradehill launched a service called “Prime” that purportedly acted as a “dark pool.”  In addition, one of the attractions to LocalBitcoins may be that it does not require traders to provide identification (via KYC); its volume could decline if it tried to comply with similar KYC/AML/BSA requirements that many other exchanges do.

Cross-border payments

Dourado’s explanation for how credit card processing work is not fully fleshed out.  For a more detailed explanation I recommend readers peruse two posts from Richard Brown found below in the notes.13 In short, Dourado’s explanation for the alleged value proposition between Bitcoin versus a credit card ignores the biggest difference: there is no native credit facility or lending ability on the Bitcoin network.

At best the comparison should be with debit cards.  In addition, in his example, not only is there unnecessary foreign exchange fees in moving into and out of bitcoin, but transactions do not occur instantaneously (even zero-confirmations take longer than a card swipe).  Furthermore, the current Bitcoin network is unable to handle everyone wanting to use bitcoin today (there is a continuous backlog of unconfirmed transactions, sometime measuring into the thousands).  One thing he could have mentioned is that that foreign exchange trades may offset merchant fees, but he did not (yet).

For instance, Dourado states:

You may use a payment processor such as BitPay to instantly convert the bitcoins you receive into dollars. I may use a wallet that instantly converts dollars to Bitcoin at the time I want to make a payment. We both have trust relationships with intermediaries, but because the transaction and settlement occurs on the blockchain, we no longer have to trust the same intermediary.

There is no reason to use Bitcoin itself to do this.  Since users on both ends of the transaction are not only identified but they also need to “trust” a trusted third party, they could just as easily use a different payment method.  And empirically they do, hence one of the reasons why JP Koning wrote the first post in the first place.  In practice, Bitcoin as a payment system is just an added friction: why go from USD->BTC->USD when a user can simply bypass this artificial friction and pay in USD?

Dourado does not provide a cost-benefit analysis nor does he explain why credit card companies work the way they do (see again Brown’s posts in the end notes).   Instead, he discusses the example of unbanked and underbanked, stating:

This is relevant when thinking about bringing the next few billion people online and into the global economy. These people will not have credit histories that are accessible to the same intermediaries that I am set up to use. They may have local intermediaries that they can use, or they may be willing to use Bitcoin directly. If that is the case, they will be able to enter into the stream of global commerce.

In my lengthy book review on The Age of Cryptocurrency I explained 3-4 reasons for why Bitcoin probably is not the savior of the unbanked and underbanked.

One of the reasons is volatility, another is compliance and customer acquisition costs.

One more is the fact that nearly all venture capital (VC) funded hosted “wallets” and exchanges now require not only Know-Your-Customer (KYC) but in order for any type of fiat conversion, bank accounts.  Thus there is a paradox: how can unbanked individuals connect a bank account they do not have to a platform that requires it?  This question is never answered in the book yet it represents the single most difficult aspect to the on-boarding experience today.

Thus contra, Dourado and others, Bitcoinland has recreated all of the same types of intermediaries as the traditional financial world, only with less oversight and immature financial controls.

In terms of “rebittance,” in practice, what ends up happening in these emerging markets is that local residents attempt to cash out into their local currency, irrespective of whatever cryptocurrency funds were originally sent with.14 It is highly recommended that readers peruse analysis below in the notes from Yakov Kofner who studies this at SaveOnSend — looking at actual data such as margins and fees15 And again, maybe this will slightly change through the efforts of Align Commerce, Coins.ph and BitX but it has not yet.

Continuing Dourado writes:

We will finally have a unified global financial system to which everyone will have access. Capital controls will become impossible, or nearly so.

Unlikely via Bitcoin, perhaps through other distributed ledger systems being developed (with mintettes).  The above statement may be the hopes and dreams of many Bitcoin investors, but recall the drama surrounding Coinbase this past February when the leaked pitch deck (pdf)  — which highlighted Bitcoin’s ability to bypass sanctions on Russia — ended up in the hands of regulators.  The head of compliance at Coinbase ended up leaving and the startup was on thin ice (maybe still is?).16

Settlement finality

Another quibble with Dourado’s piece is based on his statement:

So in order to do apples-to-apples comparisons, we might want to examine other systems of final settlement. One such system is cash. Cash of course has some limitations, chief among them that it is not possible to send cash online without an intermediary.

The problem with this is that cash in the real world is given exception to nemo dat and bitcoin is not.  I tried pointing this out to him on Twitter, to which he responded with one word: “Absurd.”  Nemo dat is the legal rule that states that Bob cannot purchase ownership of a possession from Alice if she herself does not have title to the possession.

And it is not absurd.

In fact, as described two months ago, when talking to attorneys such as Amor Sexton, Ryan Straus and George Fogg we learned that one of the problems facing bearer instruments like bitcoin is that many of these virtual assets do not have clean title — that they are encumbered.  What this means is that while the Bitcoin network itself may provide settlement with respect to the transfer of private key credentials, on the edges of the network in the social ‘wet code’ world, the title to these credentials could be non-final.

This means that because of how trusted third parties such as Xapo or Coinbase originally pooled and commingled (e.g., did not segregate) customer deposits, some customers may unknowingly end up with encumbered bitcoins.  Whether anyone litigates on this issue may be a matter of time as Mt. Gox may have practiced the same behavior with pooled deposits.

Ignoring this could impact the bitcoins you may have.  Did you mine the coins yourself or did you buy them through an OTC provider like Charlie Shrem?  There is currently no method of “cleansing” these virtual commodities from previous claims.  Thus, as described earlier in this post, while settlement finality is a potential benefit of distributed ledgers, it probably needs to be integrated within the current custodial framework in order to be effective. 17

Machine to machine

My last quibble regarding Dourado’s piece is where he states:

Direct settlement also means that machine-to-machine transactions will be possible without giving your toaster a line of credit or access to your full bank account. What new inventions will people create when stuff can earn and spend money?

The core innovation around Bitcoin are censorship-resistant cash and its decentralized ledger — thus trying to merge costly pseudonomity with the KYC of a traditional financial system and then innovate on top of that seems like a one step forward and then one step back.

Therefore it makes little sense for why Dourado, Antonis Polemitis, 21inc and others continue to bring up machine-to-machine as if it is the “killer app” for Bitcoin.  What is the need for proof-of-work in these cases?  I briefly looked at this in Appendix B: why can’t prepaid cards be used to pay for the same service?  If parties — or washing machines and toasters — are known, what benefit does this asset provide that cannot be done with other systems?  Why do you need to insert censorship-resistant virtual cash in a transaction that ultimately will need national currency on both sides of the transaction?

Furthermore, even if machine-to-machine transactions somehow did take off and the Bitcoin blockchain was used, it would quickly become bogged down due to block size issues.  For more on this point, it’s worth reviewing the two most recent posts from TradeBlock below in the notes.18


It is unlikely that many early adopters or those who believe static money supplies are a feature, will find any of the discussion above of merit.19  Yet, as Noah Smith pointed out again yesterday, bitcoin’s volatility may need to become “boring” (non-existent) if it ever were to become a viable medium-of-exchange.  However as described above, there are multiple external factors for why this may not occur including the fact that there is no current method to automatically, trustlessly rebase the purchasing power in Bitcoin.

Last fall Robert Sams published a short paper (pdf) proposing one solution, via a “stable coin” — an idea that has subsequently been explored by Ferdinando Ametrano20 and may eventually be emulated in projects like Augur and Spritzle.

Whether or not this feature is adopted by the Bitcoin community remains and open question.  What is probably not an open question is whether volatility will ever disappear for a perfectly inelastic money supply, particularly one without a type of rebasement mechanism.

[Acknowledgements: thanks to Raffael Danielli, Justin Dombrowski, Yakov Kofner and George Samman for their feedback.]

End notes:

  1. See What is the “real” price of bitcoin? and Too Many Bitcoins: Making Sense of Exaggerated Inventory Claims []
  2. I would like to thank Justin Dombrowski for bringing this point to my attention. []
  3. Readers may be interested in Low Volatility and The Shanghai Composite Are Killing Bitcoin by Arthur Hayes.  Note that you can have liquidity from underlying demand as a transactional cryptocurrency, but that does not seem possible to coordinate with a limited, decentralized money supply in the Bitcoin model. []
  4. The Goldman Sachs report used self-reported numbers from the exchanges themselves.  See 80% of bitcoin is exchanged for Chinese yuan from Quartz. []
  5. I would like to thank Justin Dombrowski for this insight. []
  6. See New CoinDesk Report Reveals Who Really Uses Bitcoin as well as the the leaked Coinbase pitch deck (pdf). []
  7. See The flow of funds on the Bitcoin network in 2015 and A gift card economy: breaking down BitPay’s numbers []
  8. See: Consensus-as-a-service as well as No, Bitcoin is not the future of securities settlement by Robert Sams and On the robustness of cryptobonds and crypto settlement by Izabella Kaminska []
  9. See also: Will colored coin extensibility throw a wrench into the automated information security costs of Bitcoin? and Can Bitcoin’s internal economy securely grow relative to its outputs? []
  10. See Too Many Bitcoins: Making Sense of Exaggerated Inventory Claims []
  11. See Daily FX volumes hold above $5 trillion in Feb-CLS from Reuters []
  12. George Samman suggested interested readers look at a presentation he made for Coinsetter last week, starting at slide 93. []
  13. A simple explanation of fees in the payment card industry and Why the payment card system works the way it does – and why Bitcoin isn’t going to replace it any time soon both from Richard Brown []
  14. See The Rise and Rise of Lipservice: Viral Western Union Ad Debunked []
  15. Western Union: permanent leader of international money transfer? and Does Bitcoin make sense for international money transfer? both from Yakov Kofner []
  16. In addition, while an organization like a government may not be able to totally eliminate Bitcoin itself, they could likely severely reduce its use by imposing such absurd punishments that most would fear to use it.  But that is a topic for another post. []
  17. See also: No, Bitcoin is not the future of securities settlement by Robert Sams []
  18. Bitcoin Network Capacity Analysis – Part 3: Miner Incentives and Bitcoin Network Capacity Analysis – Part 4: Simulating Practical Capacity from TradeBlock []
  19. This concept, of static money supplies, is not an unknown idea for central banks.  David Andolfatto, VP at the St. Louis Federal Reserve, pointed this out in his presentation last month. []
  20. Slides and video from Ametrano’s March 2015 presentation []

Leave a Reply

Your email address will not be published. Required fields are marked *