Bitcoin’s belated bust

by John Quiggin on November 23, 2018

It’s been quite a big week in cryptocurrency markets. The price of Bitcoin has fallen close to $4000, down from a peak of nearly $20 000.

As a longstanding sceptic of cryptocurrencies, it might be thought that I would be taking a victory lap. After all, I have previously written that “Bitcoins will attain their true value of zero sooner or later, but it is impossible to say when.” With the Bitcoin price having fallen by 75 per cent, it might seem that my prediction is well on the way to being justified.

Unfortunately, the second part of my statement, about the impossibility of predicting timing has been proved definitively correct.. I wrote this in 2013 when Bitcoins were valued at around $100, and the total market capitalization was a mere billion dollars. A single wealthy individual could have driven the price to zero by short-selling.

Five years later, and despite the price collapse of the past few months, Bitcoins are selling at nearly 50 times the price I criticized as excessive. Moreover, as cryptocurrencies have proliferated, Bitcoin now constitutes only a fraction of the total market. The capitalization of the cryptocurrency market as a whole is fluctuating still close to $100 billion.

Yet this massive valuation is built on nothing. The idea that Bitcoin, or any of its competitors will provide a new and superior means for buying and selling goods and services has been tested to destruction. Nearly a decade after the currency was launched, the use of Bitcoin in purchases is modest, and rapidly declining.

The upsurge in ‘cryptocurrency’ markets in late 2017 was premised on the idea that, rather than being currencies, blockchain tokens like Bitcoin constituted a ‘store of value’. Given the plunge in prices since the December peak, this looks pretty unappealing.

More importantly, if a security which does not constitute a claim on anything except a pointless calculation can be turned into a store of value, financial market valuations of all kinds become totally arbitrary.

The idea of cryptocurrencies as stores of value seems now to be dying away. The last refuge of the defenders the claim that, whatever the weaknesses of individual cryptocurrencies like Bitcoin, the underlying idea of the blockchain is an innovation comparable to the creation of the Internet. By analogy, it is argued, the current cryptocurrency bubble should be seen as a rerun of the dotcom bubble of the 1990s.

Even this claim is looking shaky. Many cryptocurrencies advocates point to Bitcoin alternatives such as Ethereum and Ripple as exemplars of useful implementations of blockchain. But these currencies have followed the bubble-and-bust pattern of Bitcoin. Ethereum has fallen 90 per cent December peak of nearly $1400. Ripple has fallen more than 80 percent.

Meanwhile, although a variety of institutions, from stock exchanges to central banks, have announced blockchain projects, few if any have seen the light of day. The obvious problem is that, for most purposes, centralised sharing of data through a trusted intermediary is more reliable than any algorithm based on decentralised anonymous voting.

It’s premature to write off blockchain completely. But a comparison with the World Wide Web is instructive. The first web browser was publicly released in 1991. Ten years later, it was estimated that there were more than 500 billion documents on the Web. The first conceptualization of blockchain was that of the pseudonymous Satoshi Nakamoto in 2008, which was followed by the introduction of Bitcoin in 2009. Ten years later, useful application of blockchain remains a vague promise.

While the Internet was of real social value, the associated stockmarket bubble was not. In inflation-adjusted terms, the NASDAQ composite index is still below its peak level of May 2000, and most of the dotcom darlings have long since disappeared. But at least, unlike the case of cryptocurrency, there was some realism to the story.

What do bubbles such as that in Bitcoin, and the early dotcom bubble tell us about financial markets? Assuming, as seems likely, that the true value of zero is eventually reached, we can say that these markets aren’t totally untethered from reality. On the other hand, as Keynes is apocryphally quoted as saying, markets can stay irrational longer than you can stay solvent. My 2013 claim that Bitcoin has a true value of zero is looking more credible every day, but anyone who bet on it back then would have lost their money.

Financial markets are one of our most important institutions, but they don’t work very well. It follows that massively rewarding the participants in those market is a waste of resources that could be better employed in addressing real social needs. We understood this lesson in the decades of widely shared prosperity that followed World War II. We need to relearn it today.



oldster 11.23.18 at 11:24 am

John, you might want to warn readers that the link at “fits comfortably” will take them to a sketchy page. It immediately started asking my machine for information and wanted to push notifications at me and I got the hell out as soon as I could.

Sketchier than lots of other sites on the web? No. But not what I was expecting.


MisterMr 11.23.18 at 12:31 pm

“More importantly, if a security which does not constitute a claim on anything except a pointless calculation can be turned into a store of value, financial market valuations of all kinds become totally arbitrary.”

I don’t really agree with this: most stores of values are quite arbitrary.
If you see this from the opposite point of view: if you assume that all capital assets have an intrinsic value, and this value is somehow related to the income they generate or that is needed to create them, you have an implicit hard-ish proportion between income and wealth: it means that there is a total, technological limit to the wealth that people can accumulate in aggregate.

Is this hard proportion really realistic in a world where wealth is mostly a social phenomenon?


Murali 11.23.18 at 4:11 pm

Not to jump too much on my hobby horse, but it seems to me that the quickest way to rein in the financial sector is to eliminate limited liability. Limited liability adds noise to price signals and increases the amount of slack between the actual value of a good and its price. Suppose I invest $100 in 10 different randomly chosen startups at t1. At t1, I have $1000 worth of shares. Suppose the companies are just as likely to make a profit as a loss and at the same magnitude. Thus, if the value of the shares in one company increases by $12, it is just as likely that the value of the shares in another company decreases by $12. We can expect that the total value of all my shares will remain at $1000 at some later point in time t2. In the long run, I can at most expect to break even. Limited liability means that the value of a share cannot drop below 0. However, if there is a floor on losses but no ceiling on profits, any sufficiently large random set of investments can be expected to be profitable (or more profitable than the actual value of the investments). That makes the financial sector more lucrative than warranted by the value of the things invested in. Of course you are going to get an oversized financial sector this way. And we can clearly see why it is inefficient. Limited liability creates negative externalities for those entities the LLC has liabilities towards. If my company owes you $12000 but liquidating it would only yield $10000, you are still $2000 short which you cannot recover from me because of limited liability. That’s a $2000 externality right there created by limited liability.

I don’t know how to implement negatively valued stock. (who would buy it? you would have to pay someone to have it!) But if it could be implemented (Perhaps creditors could garnish the wealth from anyone who holds the relevant negatively valued shares.) that would cut down on the general attractiveness of the financial sector. And, it would do so only by the extent to which the financial sector is inflated because of the negative externalities it inflicts.


CJColucci 11.23.18 at 4:30 pm

As Keynes famously said, the market can stay irrational longer than you can stay solvent.


Anarcissie 11.23.18 at 4:56 pm

It seems to me that fiat currencies also lack intrinsic value, yet many of them seem to function. The problem with Bitcoin and its kind, as currencies, appears to be the lack of state force backing them up.


RJB 11.23.18 at 6:00 pm

Two questions for JQ:

1. If you add in the value of cryptocurrencies for criminal enterprise (and ignore the social costs of those enterprises), is the 10-year performance more comparable to the first 10 years of the web? More generally, how would you incorporate such markets into your analysis?

2. Are you sure about NASDAQ’s performance since 2000? this chart suggests it is up, inflation adjusted, by 15%. I don’t know how they account for the fact that some successful NASDAQ firms move to other exchanges/indexes, so they might be understating or overstating performance of a NASDAQ portfolio in 2000.


BruceJ 11.23.18 at 6:05 pm

“Cryptocurrency: a medium of exchange that combines all the disadvantages of a hard currency with all the disadvantages of a fiat currency”

Blockchain itself may, in the end become a useful tool (it does provide an un-forgeable, verifiable audit trail), although most of the claims of it’s boosters are are fantasy, bordering on the delusional.

Uhh…cities don’t work that way…


CC 11.23.18 at 7:44 pm

Let’s not lose sight of the iteration that must happen with money in order to make incentives better and more humane!

@ Murali:

Limited liability making a price floor is brilliantly obvious. Thanks for sharing.

I’m not sure how to solve for negatively valued business stock either. For individuals though this sounds like a mutual credit currency.

There is cryptocurrency project creating one. The conscientious readers on here will definitely appreciate the ideas tied to this project.


L2P 11.23.18 at 7:51 pm

“It seems to me that fiat currencies also lack intrinsic value, yet many of them seem to function. The problem with Bitcoin and its kind, as currencies, appears to be the lack of state force backing them up.”

Dollars can be (must be, actually) used to pay taxes, so they have a floor value: the sum for all tax obligations. That’s less than the market value of all dollars in circulation, but it’s above zero. If the market of dollars ever goes below the market value of tax debts, you could make a killing in arbitrage. So there’s a value there no matter what.

Bitcoin has nothing. It’s value is literally whatever someone will pay for it, which could be zero.


eg 11.23.18 at 8:19 pm


The intrinsic value of a fiat currency is that the citizenry is obligated to pay its taxes only in said currency. Since citizens need it, they will exchange labour and commodities for it.

The cryptocurrencies as you point out lack this state sanctioned feature (unless some sovereign were foolish enough to accept them in lieu of the currency over which it exercises a monopoly) and operate rather as commodities. Like “the barbarous relic” they have whatever value those who want them are willing to exchange for them. Presumably there may be underworld networks for whom their idiosyncratic features may retain some usefulness?

As for blockchain, perhaps there will prove uses in the form of ledgers of some sort, but the notion of using it to replace fiats is risible.


Brian Hanley 11.23.18 at 9:06 pm

This old paper on Bitcoin from 2013 agrees in principle.

However, as I have thought about it and watched how cryptocurrency is being used over the years I have modified my evaluation. This aspect is discussed in the latter paragraphs here.

My evaluation of the cryptocurrency technology has not improved however. Blockchain and the Satoshi proof-of-work algorithm that requires ridiculous resources to be consumed in order to scale remains the case study in abominable software design that is oblivious to capacity and performance issues. The cost of electricity to create and support bitcoin alone is greater than the market cap of bitcoin. Cryptocurrency as it is implemented is an environmental disaster that should be killed for that reason alone.

Cryptocurrency has serious issues in addition to the technological problems. It isn’t as anonymous as it is cracked up to be, simply because it’s a public record, and some location information can be inferred, as well as having the meta-data of linkages visible.

Kieran Healey has an excellent discussion of the use of metadata here to take down the American Revolution.

Nor is cryptocurrency secure. By my calculations, when I first published the above bitcoin paper, at least 1/4 of bitcoins had been lost or stolen. Since then, we have had the huge heist at Mt Gox. And when theft or loss happens, there is absolutely nothing you can do about it. On top of this, the system as a whole can be taken over at any time by anyone with the computing resources to throw at it. The only reason this hasn’t happened yet to take it all is that the primary players have decided not to because it isn’t in their interest. However, both of these factors makes the idea of “smart contracts” using blockchain, or almost anything else obviously ridiculous if you can so easily lose it or have it stolen so easily. If someone could take title to real property en masse by taking over the blockchain system, that would happen.

I strongly suspect Mark Karpeles in the Mt. Gox theft, because it is impossible for “offline media” to be hacked from outside. It would be easy for anyone with physical access to the storage media. The FBI estimated that 60% of “hacks” are really inside jobs by people with physical access. I think it is probably much higher than that. I also wonder if the Winkelvoss Twins’ bitcoins kept in a vault on flash drive and other media haven’t been copied and stolen as this MT. Gox heist shows a unique vulnerability of cryptocurrency. You can have it, and keep it on your media, but if someone copies it and spends it first, then it’s theirs. It is as if paper currency could be perfectly copied in seconds by someone who just looks at it, and then if they spend it, your copy is no good anymore.

Based on watching bitcoin over the years, I think that Coinbase in particular is exerting itself to maintain a market price and an orderly market. As long as the market cap doesn’t get too high, and the currency doesn’t get too popular, a few wealthy private parties like the Winklevoss twins can do this. (I also think that severe market intervention is visible in charts I made of EU bank activity on the futures markets for fiat currencies to keep the Euro down. And a veteran trader told me there is no end to what a central bank can do in this regard because they can just roll over their futures contracts for very little expenditure, while the little guys have theirs expire and have to pay off the loan.) This market intervention by Coinbase is something that tells me the price of bitcoin isn’t going to be zero if they can avoid it. A store of value can last as long as a market maker is able to keep it going.

Aside from the observation about Coinbase as market maker, I don’t think that the valuation of bitcoin is correctly zero. As RBP alludes to above, criminal use needs to be accounted for. Bitcoin is still the best method for conducting generic illegal monetary transactions over the internet. No, it’s not as anonymous as claimed, but to crack it requires sophistication and resources such as few agencies have. Plus, what you get is indicative, but not perfect.

It took the FBI a bit over a year from start to finish to catch Ross Ulbricht with a passel of agents and significant tech resources, just 2 years after he launched Silkroad.

That kind of expenditure of resources is not practical for the small fry. So, people buying and selling drugs and things are mostly safe, which gives it value until something better comes along.

I differentiate generic from non-generic criminal transactions on the internet. I believe that many (if not all) of the truly bizarre payoffs on Kickstarter and similar sites are really criminal transactions. For instance, when someone makes $70,000 for videoing themselves making potato salad, when people with much more interesting and useful things can’t get $10 I don’t believe people are really that nutty. Simiarly, when someone trades a pencil for the title to a house, that’s also probably a means of paying on a secret deal. I used to room with a man who became a pretty significant kingpin in the drug industry, and later shared a house with the grandson of a Sicilian Don who had run Western US distribution of cocaine for them. I can tell you that both of these guys were smart, creative, and operated like any other business. (The grandson quit the business when he got religion. Said he saw what it did to people and told the family he couldn’t do it anymore.) However, you can’t run a distribution business to the average guy with that kind of creative thought. The average drug user is not terribly bright. They just want their stuff – right now!

What the utility for secret transactions means is that a primary utility of bitcoin in the developed world is as a method of exchange for illegal transactions. I haven’t put time into figuring out how to calculate the pricing of such a thing. Would it be related to the fiat currency valuation of all the product? Is it valuation by cutting the risk? Bitcoin, in this sense represents a first in economics I think. We have had smuggling and illegal transactions forever. (Colonial smugglers protesting the East India Company getting their tea tax repealed was the real reason for the Boston Tea Party.) But in bitcoin we have a currency that should derive much of its valuation from this one sector. I think that should be interesting.

Ethereum/cryptocurrency has another valuation source that is similar to what Ben Franklin was doing in colonial America printing debt notes that circulated as currency. When I look at Ethereum and the crypto offerings, what I think I see is young people unable to get funding for their ventures who are creating a kind of debt note that is backed by the enterprises. The USA’s SEC has agreed and ruled that these are stock offerings. (Albeit they are different from other stock offerings. See below.) I suspect this use of cryptocurrency is driven in significant part by the rise in inequality between generations. So this is a legitimate use that should have some level of quasi-stable value. Ethereum was created for this purpose, and Vitaly did cure one of the serious problems of cryptocurrency by creating to a proof-of-stake over a proof-of-work system. In concept, proof-of-stake can be conceived as similar to how fiat currencies work – we all have a stake in the currency working, and the government enforces that by using it to collect taxes. I think Vitaly’s innovation is a clever address to cryptocurrency’s most fundamental problem. Because of this change, Ethereum (and it’s clones) doesn’t have to use the electrical budget of the UK in order to succeed. This was a big step.

I will also point out that Ethereum, which was created to be essentially stock certificates, is different from stock certificates in that it’s really a basket of stock certificates. This too should be interesting to economists to study. This has strong similarity to Ben Franklin’s debt notes that circulated as currency. In that sense, you could model the younger generation as being colonized by the 1% who withhold access to monetary instruments.

Then there is the use of cryptocurrency in the developing world, Africa specifically. I have gotten a view of this through African Facebook friends. Mostly they are engineering graduates, bright, capable, and extremely frustrated by the deeply corrupt tribal politics of their government. For instance, a set of Nigerians who don’t want to play the Nigerian scammer game to make money are using cryptocurrencies in a way pretty similar to how Ben Franklin’s printed debt notes were used in the colonies when the British Government could not (or would not) provide enough money through the banking system for the level of commerce the colonials required. Some of them are making their own crypto currency, recognizing that it’s smarter to create out of nothing than to plug into the bloated price of bitcoin. One has developed a quasi-automated trading system for playing a large set of cryptocurrencies. So this too is a legitimate use. It is similar to what Vitaly did with Ethereum. Just like him, these Nigerians recognized that making their own certificates from scratch was more sensible. However, I have not studied the African ecosystem in detail. What I have are just indicators. I have heard that in worse of countries than Nigeria, that people use bitcoin for transactions and to store value because volatile though it is, it’s still better than their fiat currency.

There is still the caveat for bitcoin specifically that it is going to run up against its hard limit soon. That will necessitate a new algorithm for keeping it going, because the miners won’t be able to mine anymore and that infrastructure is huge. Can it weather that? I don’t know.

(And for all you Aussies, Happy Thanksgiving, and I give thanks for the Thanksgiving eve rain that put out the huge fire near here. We have some 20,000 fire refugees. Mostly gone to family and friends, but not all.)


J-D 11.23.18 at 9:49 pm


I don’t really agree with this: most stores of values are quite arbitrary.


It seems to me that fiat currencies also lack intrinsic value, yet many of them seem to function.

Having read John Quiggin on this subject before, I think I can give you at least part of his response: fiat currencies have a foundation for their value in their perennial usefulness for discharging taxation liabilities. Others may (at least in theory) have a choice about whether they will accept your dollarpounds in payment, but the government which issued them will always do so, and people do have to pay tax to the government. Therefore …

The problem with Bitcoin and its kind, as currencies, appears to be the lack of state force backing them up.

Just so.


Howard Frant 11.23.18 at 10:31 pm


Gold has some intrinsic value (jewelry, electrical connections, etc.) but surely most of price reflects demand for it as a store of value. Or is that wrong?

I think what you say in the last paragraph is basically true, but I’m not sure that it “follows.”


Priest 11.24.18 at 1:11 am

Owning stock in a company that does not pay dividends doesn’t feel all that different than owning bitcoins or whatever. I have shares in a very profitable company that have lost 50% of their value in the last year because of investor fears about growth faltering several years from now. Say you think the current sentiment is wrong – should you buy “undervalued” shares now (P/E ratio is under 8)? The problem is, even if you are correct, you will only realize a gain if prevailing sentiment changes. So in 2024 company X continues to be profitable, but by then there are some other fears holding down the price. Investors that didn’t buy now won’t feel the opportunity cost and you won’t be rewarded for your judgement. The asset is based on something of real value (a profitable business), but what actually matters is what humans think about the future, not the objective facts of the moment.

Makes me want to go drop a stack of chips on red at the nearest roulette table, at least I know I’ll get paid off if the wheel spins my way.


Matt 11.24.18 at 1:18 am

Something to add to the pile, from Kevin Drum:
(Probably the last part is the most directly relevant.)


Rapier 11.24.18 at 2:34 am

As long as crypto currencies are valued in regular money then they will be subject to the same forces which inflate and deflate other assets. When assets have to be liquidated, sold that is, to get actual money, ie. a bank deposit, in order to preserve capital or to pay off debt, then crypto’s like everything else’s price will drop. The stupendous inflation of crypto has been directly caused by the stupendous inflation of most financial assets and actually the great increase in the amount of money itself.

Cyrpto is a particularly speculative asset. Owned in very large part by those with very short time horizons. A recipe of volatility on a grand scale, and that’s before general liquidity in the global system was shrinking, as it is now.


John Quiggin 11.24.18 at 5:26 am

oldster @1 Thanks for the alert. I’ve never noticed any problems with the site, but I have lots of blockers. I’ve deleted the link.

Various: As stated, the value of fiat money is the fact that the issuing government accepts it as payment of taxes. Conversely, a key strategy of colonial rulers trying to force subjects into the monetary economy they controlled was to impose taxes such as the Hut tax imposed by Britain in Africa.

When a valuable commodity such as gold is used as a currency or store of value, demand is increased (sometimes greatly) and the value rises accordingly. But without inherent value (including guaranteed acceptance by a state or ruler) the process can’t start.

On illegal transactions, I’m unclear whether this value remains substantial after the closure of Silk Road. Monero is supposedly the most secure cryptocurrency and its value remains trivial, while Bitcoin, where the authorities seem to be able to trace transactions fairly easily if they choose, is much larger.

Thanks for other comments which I will follow up when I get a bit more time.


Chetan Murthy 11.24.18 at 5:33 am

Just a small comment about tech, b/c this thread is really about the economics and politics of cryptocurrency, not so much about tech (and I think that that’s the right focus).

I’m a distributed systems & data guy, and the claims made about blockchain (BTC, ETH, Plasma, Casper) are all bullshit. It’s not “decentralized” (look up “DAO hack”) and it’s not possible to scale it up to the tran-rates of the real-world, nor to the data-set-sizes of the real-world. There are numerous technical problems, and it’s fair to say that the people who designed these systems were either wilfully or just plain ignorant of those problems, and the history of solutions. The *only* thing one can say in favor of these systems, is that they came up with a novel “solution” to the “consensus problem” (as long as you redefine what the problem is *HA*!)


Chetan Murthy 11.24.18 at 5:35 am

Brian Hanley commented! That’s GREAT! Everybody who has even a passing interest in cryptocurrency should go read his paper: The False Premises and Promises of Bitcoin which (if I’m not mistaken) is what he references in “this old paper from 2013”.


Jeff R. 11.24.18 at 6:33 am

Bitcoin was created by people who read Cryptonomicon and concluded that the bits about the gold were completely unimportant.

Also, it’s tough to get just how dumb Bitcoin is without understanding Tether.


Peter T 11.24.18 at 11:20 am

Money is a system of transferable debts (eg, treasury notes are a debt of the state, corporate notes are a debt of the company issuing them, much credit is a debt on the bank/card company).

So the question for Bitcoin is – against whom is it a debt? And what is their credit? The answer to the first would seem to be a miscellaneous collection of individuals, and therefore the answer to the second is that their credit is doubtful – certainly nothing like the credit of a state, a bank or a large company.

This does not mean their value is zero (after all, even IOUs have some value), but it makes a value of $4,000 seem far too high. South Sea shares and Peruvian molasses mines spring to mind.


hix 11.24.18 at 1:28 pm

There was no way to short sell bitcoins in 2013 right? The future that trades right now is also rather illiquid.


MisterMr 11.24.18 at 2:24 pm

@Rapier 16

” The stupendous inflation of crypto has been directly caused by the stupendous inflation of most financial assets and actually the great increase in the amount of money itself.”

While I agreethat the increase in the price of cryptocurrencies was driven by thye factors that drive other bubbles, I don’t think that the idea of an increase in the quantity of money is a rasonable explanation, nor that the increase in the price of capital assets can be called inflation.

Inflation is then generalized increase in the price of stuff and, according to some theories, it might be caused by an increase in an ill specified “quantity of money”, however the first effect of this would be an increase in nominal GDP (whichn is total income).

On the other hand, financial assets like bitcoins represent “wealth”. If the same factor, an increase in the quantity of money, influences both asset prices [wealth] and nominal GDP [income], then the wealth to income ratio would be stable and we would never have bubbles, or an increase of the public debt to GDP, or similar phenomena.

In reality this kind of phenomena like bubbles or increasing debt to GDP ratios are a situation were the price of assets [wealth] increases faster than nominal GDP [income], so it can’t be caused by an increase in the quantity of money however defined but rather as an effect of changes on the saving rater or unbalanced savings or whatever.

So the idea that the increase of the price of bitcoins is due to an increase in the quantity of money is misleading, it would be better to say that there is a saving glut and various government entities are forced to ramp up deficits/keep down the interest rate (both actions that are usually considered an increase in the quantity of money) to keep the economy going; but the increase in the price of bitcoins is due to the saving glut, not to the money printing (however defined).


William Meyer 11.24.18 at 4:01 pm

The notion of a “store of value” is so vague as to suggest it is deliberately trying to obscure something. How about defining “value”? I would suggest that it is essentially “the ability to coerce other people’s labor in the future.” This coercion is not limited to the state making people pay taxes in currency. It also includes the whole private property regime (also backed by state force) and the consequential fact that the vast majority of humans come into the world without owning financial assets, fertile land, game-rich forests, or useful commodities. To survive the significant majority of people must do the bidding of the smaller group of people who control the world’s productive assets, or starve. So a monetized private property regime is best thought of as a forced work do-as-I-say scheme. As such, it is something of an advance over previous forced work social schemes (feudalism, classical slavery, debt bondage) but clearly identifiable as one more highly coercive experiment in solving collective action problems. I wonder if human beings will ever develop less coercive means to solve such problems?


Gregory J. McKenzie 11.24.18 at 9:54 pm

John is right to compare bitcoins to gold when it is used as money. Gold coins usually have a face value that is well below its market value. Like ordinary shares they can be traded up and down at the will of speculators. Short selling will provide overall volatility but panic selling can exaggerate this at times of extreme uncertainty. Money has other tests for its usefulness. Generally accepted, standard of deferred payment just to name two other common acid tests for ‘good’ money. All crypto currencies need to be tested in many ways not just how good they are as a store of value.


Peter T 11.24.18 at 10:22 pm

Further to my comment above, it occurs to me that bitcoin and similar are not trying to replace fiat currencies, which are backed ultimately by states. They are trying to replace gold – the most universal credit token. Hence “mining”, the stress on scarcity, anonymity, lack of state backing. I doubt they will succeed but, even if they do, this is very much a niche market.


Rapier 11.24.18 at 11:55 pm

RE Mister Mr

Inflation is always a monetary phenomena. There is a cognitive problem when it comes to asset prices, especially stock prices because it has been taught that stock prices are never said to inflate, they ‘increase in value’. This is propaganda in its purest form.

Bonds and credit instruments, assets as well, have been inflating for 33 years but that is over. When interest rates rise then the prices of existing debt falls. Period. It is deflation. The 33 year decline in interest rates ended in 2016 when rates were the lowest in the history of the known universe. $6Tn or so is ‘invested’ at a negative rate. An ‘investment’ which destroys capital.

Eventually central banks will print again, and again, and yet again. To prevent asset deflation.


Alan White 11.25.18 at 4:29 am

The novel Galapagos is one big reason to sing praises of Vonnegut. Therein he does the ultimate metacriticism of the valuation of all monies as the result of human big brains leading to the destruction of civilization by war over (in some sense) artificially created values. Can’t say he’s off the mark–Deutsche or otherwise. Ok, it glosses over empirical realities of post-barter societies, but it still strikes a chord with anyone like me who looks at a dollar bill and even a gold brick and ponders its so-called “true” worth.


Brian Hanley 11.25.18 at 6:24 am

@Chetan – No clue why none of my links worked. The html looked fine.

Bitcoin’s demise was inevitable because it requires more and more electrical power to “mine” a bitcoin. It was obvious that this party had to end when the last bitcoin was mined, absent a redefinition of how the system worked. But it was equally obvious that in the real world, the party ends before then. The electricity and installed computing cost of “mining” has always been marginal. In fact, there is evidence that China’s government has been conned into supporting the mining of bitcoin by fraud. China subsidizes electricity for farming. And pays the bill for government entities. Farm electricity is the equivalent of 2 cents a kwh in China. That’s how a number of bitcoin farms have been making money.

And now, as bitcoin “farms” go bankrupt, the system rapidly becomes moribund because everything was paid for by those “miners”. The whole system has been a parasite on the miners.

That is a huge problem with “blockchain”. All uses aside from cryptocurrency are parasitic, tolerable by the miners only so long as it was more work to erase them than they lost in computing costs.


Ebenezer Scrooge 11.25.18 at 12:08 pm

This is a topic I’ve actually thought about:
1. People need liquidity. Currencies provide liquidity. It is the satisfaction of this need that gives fiat currencies their value. Money launderers need untraceable liquidity. Bitcoin provides this untraceable liquidity. They may continue to keep some value for money launderers, as long as their are a few chumps at the exchanges.
2. Taxation is merely a source of liquidity. Pretend that Norway’s sovereign wealth fund is big enough so the Norwegians stop raising taxes. They can still run their own fiat currency. The normal spending of the state should be sufficient to support the transactional hegemon that in turn supports fiat currency.
3. There is a big distinction between bitcoin and blockchain. Blockchain (especially permissioned blockchain) is a mere technology that supports a registry. It’s no more or less exciting than stressed silicon–another technology used to support running a registry. Only the technologists should care about a better black box.


Lee A. Arnold 11.25.18 at 1:13 pm

Brian Hanley @11, great, fascinating comment, thank you!

It helps to be very simple about it. Anthropologically, money is just a common thing that you will agree to receive in a transaction with a stranger, because you KNOW that you can use it in the next transaction with another stranger, who will also agree to receive it. –This is the primary value, the psychologically immediate value, the elemental anthropology of money.

The secondary value in anthropology was status valuation in public ritual, and these are somewhat still with us, although far less important now.

But money was then given (by economists and philosophers) tertiary expressions of “value” which are now held to be primary, i.e. some additional expression of why you will receive it. Well-known examples, from the discussion above:

Money must constitute a claim on something (JQ, top post). Or, #2 money has value as capital that can generate more money. Or, #3 money should be valued at the “actual value” of the good it is swapped for (i.e. money should measure “scarcity”). Or, #5 money has value because a central authority can enforce its uses. Or, #10 money has value because it can also be used to pay taxes.

So: scarcity, authority, store of “value”.

It helps the understanding to boil down the cryptocurrency phenomenon in this way. Cryptocurrency attempts to provide the function of “scarcity” (so that it can match the “scarcity” of real goods in exchange) by the difficulty of its origination (requiring difficult computation and large energy resources). This is not unlike the ancient difficulties in finding gold or in minting metal coin.

Cryptocurrency attempts to provide the function of “authority” by a blockchain (a decentralized authority, because everybody holds the entire ledger).

The next thing, store of “value” (or equivalents such as “claim on labor” and “transferable debt”), would come only after somebody agrees to exchange a cryptocurrency for something real, because he or she knows it can be used in the next transaction. Cryptocurrency won’t be widely accepted until there are enough other people who will do this. It’s really a simple tautology.

Problems in acceptance may be insurmountable. 1. The system is too complicated for non-experts to trust. No non-expert will trust computers with their private information, nor should they ever. No non-expert will ever assume that blockchain authority cannot be hacked. The “origination of scarcity (leading to value) by the difficulty of a big calculation” is a fairly otherworldly concept, so far.

2. Cryptocurrency is being introduced for usage not in the real sector of consumer transactions (where it might be introduced as e.g. a consumer coupon). It is largely in the financial sector, where it occupies the function of another paper asset, but one that is unrelated to an underlying flow of payments, such as debt, equity, derivative. It is superfluous. Speculation in it, is untethered from underlying accounts.

Comment #11 describes cases when cryptocurrency finds a useful niche as an alternate payment arrangement where the reigning possibilities for payments are unsuitable: because unavailable, or because the governmental authority is crooked, or because criminals themselves are in search of hiddenness.


SusanC 11.25.18 at 1:15 pm

If we assume that bitcoin will crash eventually, but we dont’t know ecactlywhen, then this acts a bit like inflation. The expected value of my bitcoin tomorrow is integrated over a probabilty distribution that includes a crash in the next 24 hours. Quite possibly, expected value at time t tends to zero as t tends to infinity.

But … this is also true of fiat money. If a put a dollar bill in a box under my bed its expected value decreases over time. The asymptotic decline to zero doesn’t stop people using fiat money in the interim.


Anarcissie 11.25.18 at 4:13 pm

William Meyer 11.24.18 at 4:01 pm @ 24 —
It is interesting to speculate on how an actually voluntary monetary system might be set up for general use, but I suppose it would be off the subject here (the far more scandalous doings of Bitcoin).


Scott P. 11.25.18 at 4:19 pm

If a put a dollar bill in a box under my bed its expected value decreases over time. The asymptotic decline to zero doesn’t stop people using fiat money in the interim.

First, most people don’t keep their dollars in a box, but in some kind of interest-generating vehicle, so it’s not quite true to say its expected value decreases over time.

Second, inflation is generally slow and predictable, unlike bitcoin fluctuations. When inflation is high and unpredictable, people _don’t_ hang onto fiat money. Examples of this are legion.


Birdie 11.25.18 at 5:05 pm

Easy to see how the requirement to pay taxes in state-sanctified specie encouraged the transition to a cash economy; likewise I have heard suggested that paying soldiers deployed far from home in coin encouraged the use of cash in certain segments. What keeps the dollar stable is the Fed intervening actively to maintain value. The dollar is exceptional, but these days the IMF attempts the same globally. All of which of course is exactly the kind of structure cryptocurrency was designed to do without. But it turns out there was a reason ….


Collin Street 11.25.18 at 7:06 pm

Blockchain is actually useless, though, because it relies on a consensus that can’t be produced if more than a third of nodes are bad-faith actors. Better to run a central and a local registry and complain loudly if there’s a discrepancy: a central point of control is a single point of monitoring and rectification.


John Quiggin 11.26.18 at 5:53 am

@4 As far as I can tell, Keynes didn’t actually say this. But it sums up his view.

@6 I wrote this a while ago, anticipating the drop, so the claim about NASDAQ may be out of date by now

@36 I’m increasingly convinced that blockchain is at most, a marginal option, perhaps useful in some limited contexts, but not generally


peterv 11.26.18 at 9:31 am

@Collin Street #36:

Technically, a central database may well be better. It is also more vulnerable to external malicious attack than a database replicated in multiple locations.

But the major motivation for the current great interest in blockchain by large companies is not technical – it is ultimately due to anti-trust law. If a collection of western companies in the same industry wish to share their common data in a centralized database, in general no single one of the companies may hold that database, This is because it will contain commercially sensitive data on their competitors, something prohibited by competition laws in most western jurisdictions.

So, the companies may jointly decide to ask a third party to hold it, such as an accounting firm or a law firm. But such a third party will charge for this service and may also seek to take commercial advantage of the industry-wide information in the database. A distributed ledger, or something with the same features, will likely be cheaper to establish and run, easier to control, and will allow the participants, not the third party, to reap any commercial advantages of the information it contains.

This is why so many major banks, insurers, energy companies, etc, are participating in consortia to create industry-wide platforms. The end result of these collaborations may not technically be blockchain platforms, but blockchain got the participants in the room together in the first place. Even that act – meeting together – normally requires legal advice to ensure anti-trust laws are not breached.


politicalfootball 11.26.18 at 2:13 pm

Does it matter that Ohio is willing to collect taxes in bitcoin? Presumably, government endorsement in this fashion changes the equation for bitcoin, doesn’t it?


Trader Joe 11.26.18 at 4:04 pm

@36 and @37

I don’t agree with writing off block chain as a ‘marginal’ option. Its probably a marginal option in the context of a mass-market, everyone participates sort of thing but in business to business transactions there are already a lot of really interesting and really valuable uses. There are a number of applications involving cross border transactions of goods where the block-chain effectively documents chain of control of inventory (i..e from manufacturer, to truck, to boat, to warehouse dock, to truck, to receiver dock) and the chain allows for release of payments (often in different currencies) at each step of the chain.

There are also a number of examples within certain types of insurance transaction – for example when a large asset like an aircraft is destroyed there can be 30-50 different insurers on the risk, a block chain format can help identify when each needs to pay and has paid his appropriate share based on the ascribed losses.

Real-estate transactions, in the long run, will almost certainly be blockchained thereby eliminating the dubious need for title insurance.

These won’t be “paypal” mass market applications, but I think they go well beyond marginal and they are real right now, not some vague promise for the future.

I agree with the points made on Bitcoin though would note that Wall Street has invested many millions in being able to process and trade in things which, if for no other reason, could tend to insure their survival for longer than expected regardless of all the good points made.


MisterMr 11.26.18 at 4:55 pm

RE Rapier 27

I’ll try to explain my objection again:

Suppose that the USA has a yearly nominal GDP of 100$. This represents the total income, aka the total amount of stuff produced and spent, in a year, in nominal terms.

Furthermore people in the USA own a certain quantity of assets: hoses, bitcoins, picasso paintings etc. This represents the total amount of wealth.
Let’s say that in our example the total wealth in the USA is 200$.

This means that there is a wealth to income ratio, that in my example is 2/1, meaning that the total wealth is equal to 2 full years of production. This ratio changes during time and has changed recently, this was one of the main aspects of Piketty’s book and was discussed also on this blog (not everyone likes Piketty’s explanation).

Now, suppose that starting from my example of income [ngdp] 100$, wealth 200$, there is some inflation, meaning a fall in the purchasing power of the currency, while the total stuff produced is still the same.
You might end up in a situation where total income is 200$ nominally, and total wealth is 400$ nominally, but both the income and the wealth represent the same stuff that they represented before: it’s just that 2$ now represents what 1$ represented before, because of inflation.
In this situation the wealth to income ratio is still 400$/200$ = 2/1 because both wealth and income inflated nominally.

But this is not what is happening today, in facts inflation in income is quite low while the price of various assets goes up, so we are in a situation where, from income:100$ – wealth:200$, we arrive in one where income is still 100$, but wealth is 400$.

So there is a change in the wealth to income ratio, that is now 4/1.

This is a situation where some factor is pushing up the value of assets and wealth relative to income, while income satays unchanged; I’ve my opinion on what the factor is, however certainly it’s not inflation which would pump up nominal income too.
Speaking of inflation or money printing only muddles the water.

Also, importantly, this is not a weird phenomenon but something that is happening continuously in the whole world since at least 1970, so it’s something that should be studied.
Take picture 3.1.1 in this study (that I think still comes from Piketty’s group) for example:


Collin Street 11.26.18 at 8:39 pm

Eh: what I wrote is why blockchain is impossible (not resistant to plausible threat models)

The reason blockchain is useless is ’cause even if it worked all it is is a mechanism for gathering snd distributing third-party-witness electronic signatures for your documents, and that’s not actually a problem needing a distributed solution very often: the parties care (but two-party signatures prove everything they need in the event of a dispute) and local regulators might care (but they can be cced the documents and date-stamp them themselves) and remote regulators _might_ care but their requirements can be met with signatures endorsements from the local regulators.

This is even more decentralised than blockchain in that there isn’t even a central registry, and it’s the _current_ technology. Blockchain adds nothing, here or anywhere else.


Chetan Murthy 11.26.18 at 9:25 pm

The focus of most comments has been on the economics and pragmatics of cryptocurrency — the most salient use of blockchain. A couple of people have mentioned other (enterprise) uses of blockchain, and made arguments for why it’s effective there. These arguments are incorrect. Points worth noting:

(1) blockchain enthusiasts make much of the fact that blockchain is a “replicated database” as opposed to a “centralized database”. The comparison being between a blockchain app running on N computers all around the world, and a centralized app running on a single server, using a single Mysql server instance. This is an incorrect comparison. There are a number of important replicated databases in the world, running on N computers (viz Google Spanner). The only thing they lack (from the enthusiast’s POV) is that they are centrally administered. We’ll deal with that below.

(2) blockchain apps ARE centrally administered. Enterprise blockchain apps use “permissioned” protocols (e.g. PBFT) and those protocols involve managing a list of members allowed to participate. This is a *centrally-managed* list. To run any replicated database, all the copies must be running IDENTICAL versions of the system. If a bug is fixed in one server, it must be rolled-out to ALL OTHER servers in a timely manner, or chaos will ensue. Both of these require central administration. When you consider that (e.g. in Trader Joe’s example) there can be many, many businesses involved in a blockchain app, it’s clear that there’s no way they could all agree on the code, business rules, access permissions, for the app — it’d be an interminable group-grope meeting — and so the only way these apps come into being, is when some one or few leaders build the app and attempt to sell it to the others.

(3) So what are these blockchain entrepreneurs so het up about? Example: the DTCC (Depository Trust Clearing Corp) is the registry for all (most?) securities in the US. Every stock trade is eventually “settled” by moving data between rows in databases at the DTCC. They charge a pretty penny for their services: $1.4b last year, IIRC. And because of various human issues, it takes several days after a trade, for that trade to be reflected in the DTCC’s systems. Blockchain entrepreneurs want to replace the DTCC’s role, with a blockchain app. Here’s the thing: they’re going to charge for their services, only less than the DTCC. So all the DTCC has to do, to snuff those guys out, is lower their prices. Many of these clearinghouses are owned by their customers (banks). But they charge what the market will bear, and if a eompetitor emerges, they’ll just lower their rates. And this pattern repeats all over the financial sector. The story is similar when it comes to “time to settle trades”. If you look into the details, you’ll find that the reasons are manifold, but NEVER concern

To sum up: There are other replicated databases; the difference is they are centrally administered. But the idea that a blockchain app is not centrally administered (in the enterprise context) is foolish: they all are, because bugs must be fixed, corrupted data must be repaired, new members must be enabled and old members banished. All of this doesn’t change, just because we wave a magic blockchain wand.

Expertise notice: I designed the Hyperledger Fabric (HLF) transaction lifecycle, and consulted for Cisco on their blockchain platform.


Chetan Murthy 11.26.18 at 9:30 pm

Argh, I forgot (4) even if blockchain works, it’ll never work for applications that don’t fit in a single computer. There are strong limitations to scaling up blockchain to really high transaction-rates, and to really large data-set-sizes. And these have nothing to do with the scaling limitations of bitcoin and ethereum.

I’ll put it baldly: It is near-impossible to build sharded “permissioned” blockchain systems that actually work, if you’re really committed to the thesis that any computer in the system could be run by a malicious actor (which is, after all, the BASIS of blockchain’s claim to being valuable). And the only way to scale up transaction-rates and data-set-sizes, is via sharding.


hix 11.26.18 at 10:52 pm

Money launderers need untraceable liquidity. Bitcoin provides this untraceable liquidity.

Thats at least disputed on both counts.


Torsten Suel 11.27.18 at 1:45 am

@39: no, it doesn’t. The taxes you owe are still denominated in US dollars. They just hook up a bitcoin processor that translates bitcoins into dollars for them at whatever the current exchange rate is, plus probably a decent fee.

So, yes, the whole thing is baloney. Even if there was a value in block chain technology, you cannot invest in that technology by buying bitcoins, any more than you can invest in amazon by buying lots of products from them.


Chetan Murthy 11.27.18 at 3:16 am

I’m sure at some point, somebody is going to point out that a blockchain is an immutable ledger, where a database is a mutable thing. Or some such. So to just forestall that, I thought I should point out that EVERY database has an “immutable ledger” — we call this thing the “database log”. It is append-only, and is only ever discarded for space reasons (in some heavily-regulated shops [which I will not name], they never discard the logs, storing them for regulatory compliance&audit). The real question one should ask is: “if the blockchain is a database log, then what is the corresponding database?”

Answer: Bitcoin’s database is a single relational table, with four columns (transaction-id, output-number, script-validating-ownership, amount-of-bitcoin). With a primary key of the first two columns.

Ethereum’s database is in two parts: a per-“account” part with a few counters and an amount of ether credited to the account, and then a key/value hashmap, with a two-part primary key: (account-number, 256bit-integer-address) and 32-byte value.

Every blockchain can be analyzed this way, and they all fit into this model. Because they really all are just degenerate databases.


John Quiggin 11.27.18 at 3:30 am

Chetan, these comments are very helpful. Please, keep them coming.


Chetan Murthy 11.27.18 at 3:35 am

John Quiggin @ 48: I do wish more people would read Brian Hanley’s paper and discuss -that-. I’m not an economist, but geez he seemed to really nail the case for “Bitcoin is never gonna matter” pretty well.

The tech aspects are amusing when they’re not depressing (that anybody believes the rubbish). The economic and juridical aspects are …. *fascinating* and really (to me) lay bare the assumptions people make about law, the way it works, and the way they wish it worked.

I had a conversation [due diligence] with the CTO of a major Ethereum startup, when we were describing our system to him. He remarked that our system intrinsically involved a single company that would hold all the legal exposure for a screwup in the software. And that he felt that this was a bad thing: that one of the “design goals” of smart contract companies, was to construct things so that the company pushing the software was NOT legally liable for any screwups.

I mean …. this is ridiculous on its face, and yet it’s more-or-less the way all these “public blockchain” companies work.


Collin Street 11.27.18 at 4:50 am

Hang on, blockchain is basically everybody signs every transaction, innit? rofl.

(so processing scales roughly as the square of the usage: in the real world, n log n is the worst scaling performance anyone will accept)


Faustusnotes 11.27.18 at 12:11 pm

I would like to applaud chetan’S comments, i really would, but shard is a really nice word, and chetan turned it into a verb. “Sharded,” seriously!? I can’t even! I was suspicious when I first saw it and thought it was a misspelling of shared, but then cynical me thought “wait no this is computer science and blockchain, maybe they actually did verb that noun” and cynical me was right.

You should be ashamed of yourself chetan!


mpowell 11.27.18 at 2:48 pm

JQ, I’ve always agreed with your take on bitcoin, but I’ve never really appreciated the broader argument you are trying to make about market efficiency or government policy.

First, let’s talk about market efficiency. You are trying to make the argument that since there are two big bubbles we can retro-actively identify in financial markets, this proves they don’t work very well. But just because you predicted these markets would collapse in the future and then they did does not prove that the probability this would happen was 100%. Those markets may very well have been perfectly rational, given the universe of available information at that time. Certainly the current market valuation of tech companies supports the notion that the internet was a major disruptive event and a speculative nasdaq valuation may have been justified given that the winners and exact dynamics were far from clear.

Second, I’m not sure exactly what argument you are making about the policy implications of market inefficiency. It’s not clear why this is necessary or sufficient to indicate we should return to policy similar to post-WWII. I didn’t understand that there was much central planning in US economic policy at the time. And even if that is the goal, I don’t see the proof that these examples prove that ‘experts’ can do a better job of assigning market prices. There was quite a consensus among ‘experts’ that housing was over priced in the 2007-2008 period. But to me this looks more like an example of experts forcing their conclusions on the market through the actions of the fed with fairly disastrous consequences. In reality, limits on house building in productive areas of the country have been driving both prices and rents up to tremendous values and that process resumed once the economy returned to healthy growth. Disrupting the market pricing mechanism only discourage the investment required to expand inventory and allow prices to return to more manageable levels. So overall, I am quite suspicious of the alleged superiority of experts to determine market prices even if the 2000 stock bubble and recent bitcoin bubble really can be described as definite market inefficiencies. But it depends quite a lot on the particulars! There are plenty of cases where I am in favor of market intervention by the government to address well understood problems with the dynamics of resource allocation by price (let’s say, health insurance markets for example). I just don’t see these examples as closely tied to arguments about efficiencies in the market prices for financial assets.


Anarcissie 11.27.18 at 4:11 pm

MisterMr 11.26.18 at 4:55 pm @ 41 —
A theory I like: Besides fiat money, you could have a different kind of money, call it ‘credit money’, which is not issued by the government — for example, I write a check or a promissory note. The interesting thing about this other kind of money is that it can be kept away from the ‘real economy’, that is, the proles, who still have to do labor to get the products (goods and services) produced by labor, which they need or want. In their world, as long as the price of labor is stable, the prices of things derived from labor are stable. Thus, there is little or no inflation in their economy. The ‘credit money’ can be reserved mostly for the rich — no one will lend a prole much money, and if they do, it will be at a high interest rate. Thus the things in which the rich are interested — real estate, equities, collectibles, high-status education, novel financial entities, politicians, and so on, can and do experience rapid inflation apart from state issuance (‘printing’). In effect the prole realm and the rich realm have two kinds of money with limited crossover.

Bitcoin is a kind of parody of this situation. But it is not so far from, say, the fundamental value of Facebook shares, or the ‘value’ of real estate which was worth $100,000 in 1990 and is now worth $2.5 million because it is in Long Island City, the site of an ongoing gentrification blitzkrieg.


Orange Watch 11.27.18 at 5:24 pm

I do wish more people would read Brian Hanley’s paper and discuss -that-.

I read Hanley’s paper and am grateful for the opportunity to have done so – it really was enlightening. Speaking strictly for myself, the reason I have no inclination to discuss it looks a lot like the reason he couldn’t get it published in economics journals – once you read it, it seems incredibly obvious; there doesn’t seem to be much more to say beyond “blockchain evangelists are necessarily lying to either themselves or everyone else”.


Bill Benzon 11.27.18 at 7:09 pm

I wonder, is there a correlation between having faith in the blockchain and believing in the coming Singularity/Rapture when computers will become super-intelligent?


Brian Hanley 11.27.18 at 8:06 pm

As Chetan points out, the blockchain is a pretty simple database construct in terms of what it delivers. If you want to delve into detail of the issues with setting up a cryptocurrency system based on current primary algorithms, I recommend Tim Swanson’s books. They are info-dumps, but they do cover the ground well.

This one is quite good and goes down into the weeds. My paper mostly sticks to the conceptual errors. The Anatomy of a Money-like Informational Commodity: A Study of Bitcoin

I also want to stress that over time, my thinking has evolved on cryptocurrency. I don’t see it primarily as a scam anymore, although it arguably is that. I see it as something created to address a need, and I think that need is an effect of inequality. I know too many people now who are in the 20-40 age range that have used bitcoin or other cryptocurrency to start an enterprise.

This article capsulizes that thinking. – The real meaning of cryptocurrency.

Yes, cryptocurrency is a kind of monstrosity. It is self-contradictory, ridiculously inefficient, conceptually not even up to being sophomoric, it’s all that. But when there is a demand, things appear to serve that demand. I think we need to look closer at what that driver is and see if there are ways we can either meet it, or evolve some current system so it can be met. I’m working on a new type of banking – have been since 2012. But maybe cryptocurrency can also evolve. To do so, at minimum, it must centralize. It must dump the voting algorithm. It must end proof-of-work. It must in some form end scarcity. Maybe that can be done.


eg 11.27.18 at 8:20 pm

@ mpowell 51

“Those markets may very well have been perfectly rational, given the universe of available information at that time.”

Perhaps you can help me with this premise. How is a market run by and for human beings, all of whom are themselves irrational, supposed to be even imperfectly rational, let alone “perfectly rational”?


Trevor 11.27.18 at 9:04 pm

Setting aside the dollar value of Bitcoin, and the transparently post-hoc “store of value” use case, the one thing I’ve seen out of crypto apologists that’s actually persuasive is claims you can use it to make credible pre-commitments that are reasonably hard to break. You can write an Ethereum contract that’s reasonably guaranteed to run according to a certain set of open-source instructions and hard to renege on, and the possibility of all of a currency’s users enacting a hard fork of a protocol adds another, weaker check on someone taking over a network. Those interact and neither is ironclad, so you see project like the DAO not actually proceeding fully autonomously or according to code, but even defeasible constraints can be valuable. Even Bitcoin is kind of an instance of this – it’s strongly committed in the code to running this deflationary monetary policy, which is what attracted gold bugs and Fed conspiracy theorists early on, before it really found its home as a black market currency.


Chetan Murthy 11.28.18 at 1:35 am

Trevor@37: Trevor, thank you for bringing this up. Each of the “arguments” for cryptocurrency need to be separately defeated, and this is one of them.

TL;DR this capability (like everything else in cryptocurrency) turns out to be real only in the simplest of cases, which are useless for anything real.

First, let’s note that all the public cryptocurrencies are based on “permissionless” blockchains, and hence there is no legally responsible party for errors or corruption. If something goes wrong, the code is the law. And this is one of the selling points of these systems — they call it “censorship-resistance”, where “censorship” refers to governments (also judges) attempting to control which transactions are allowed to be applied, hence to control who has what coin. This is a *selling* point.

It’s true that Ethereum can “lock up” coin until some condition is met. There are two problems with this approach:

(1) The simplest problem: if there’s a bug in that smart contract, it sucks to be the person who loses out. And there’s nothing they can do, b/c “censorship-resistance”. No judicial remedy. The assumption is that every party will either read the smart contract code like a CHAMP, or they’ll hire such champs to read for them. I’m sure you can see how this is insane.

(2) the “condition” can only be something -algorithmic-. Basically, something like “If you solve this sudoku puzzle, you can have the ether locked up in this smart contract”. [The condition has to be something that’s efficiently checkable, of course, but that’s no big deal.] If you want a condition dependent on external conditions (“this money will be released on-demand to Joe, if the temperature in Houston is greater than 120F on any day in 2019”) you need to interact with external processes (called “oracles”) and … this gets complicated. Of course, those oracles are run by third parties, and the integrity of Ethereum itself says NOTHING about the integrity of the third parties running the oracles. Of course there’s no legal framework for forcing compliance from them, because the entire Ethereum ecosystem was aimed at avoiding lawyers.

To sum up: this capability to “pay when some condition becomes true” turns out to be just like all the other capabilities and promises of cryptocurrency and “permissionless blockchain”:

Demonstrate some simple case of the capability, and use that as proof that the general case, or the scaled-up version, is equally feasible, when in fact, no such thing has been shown, and it’s easy to show that the latter is impossible.

It happens again and again: throughput scalability, data-set scalability, programmability (of Ethereum[1]), and the capability you cite.

[1] The Ethereum VM is programmed in bytecode; there are a number of languages that compile to it, but the most “reasonable” is called Solidity. It’s fair to say that whoever came up with the VM is an idiot, and should be sentenced to take senior-level programming-languages and first-year graduate-school programming-language semantics. Probably also compilers at both levels. B/c they sure don’t show any competence at these things.


Chetan Murthy 11.28.18 at 2:13 am

Another in a seies of “field notes” about blockchain and bitcoin:

(1) BTC & ETH are traded on the futures exchange (in, I presume Chicago). Needless to say, the systems they use are centralized databases. Probably IBM mainframes, but maybe Oracle.

(2) The various “exchanges” (like Coinbase[1]) use centralized databases, too. How do we know this? Well, I had an “interview” with a company that does market-making for their “token”, and they explained that they basically provide liquidity, responding to bids to buy/sell their token on various exchanges (Coinbase, etc) and right now they respond in a few seconds at most. But they’re aiming for milliseconds soonish. B/c that’s the speed with which exchanges complete trades for their users. Of course, another reason we know this, is that BTC is capable of a whopping 7 transactions/second.

Two observations come to mind:

(A) These transactions are obviously recorded on BTC/ETH blockchains — b/c the latency there is at least 10min, and more like 60min.

(B) obviously for the vast majority of users, their only interaction with the blockchain will be via these crypto exchanges. For those users, the fact that the blockchain is “decentralized” is utterly moot: they’re dealing with a single entity, a corporation, and hence what matters is their contract with the exchange, and what rights it gives them.

It should give us pause, that NOT ONE crypto exchauge uses blockchain technology to record their transactions in real-time. NONE of this stuff is even suitable for the use of the main (really, “only”) way in which users interact with cryptocurrency.

If there weren’t billions of dollars being spent on it, if it weren’t consuming more electricity than Ireland (as of Jan 2018, nearly a year ago), it would be farcical.

[1] I’m leaving aside that these exchanges aren’t really regulated, and play pretty fast-and-loose. They’re on a path to regulation, so let’s just assume that all that will work out fine. Even though there have been cases where exchanges collapse and customers are left with …. *nothing*. Because “their crypto” was actually owned by the exchange.


Chetan Murthy 11.28.18 at 2:21 am

Brian Hanley@55: two responses:

(1) I’ve emailed Brian separately to discuss his position that cryptocurrency speaks to a need for currency-creation. It’s probably not a good use of the readers’ attention, to discuss our differing thoughts on ICOs (“Initial Coin Offerings”) here.

(2) Brian brings up thoughts on how cryptocurrency systems might evolve to become more … not-insane:

But maybe cryptocurrency can also evolve. To do so, at minimum, it must centralize. It must dump the voting algorithm. It must end proof-of-work. It must in some form end scarcity.

Here’s the problem: If you remove proof-of-work, you end up with …. centrally-administered consensus protocols (like PBFT or Paxos). These are well-understood (and Paxos is widely-deployed for decades), but also are … (wait for it) … NOT censorship-resistant. They’re basically just digital currency. And we already have a digital currency. It’s called Federal Reserve dollars. Almost all dollars are held in Federal Reserve accounts — only a small portion of them are actually represented as *cash*. If we were really serious about some sort of digital currency, the path to get there is for the Fed to allow individuals to open checking accounts at the Fed, and provide online banking. Sure, there are details, but at some level, that’s how to do it.


nastywoman 11.28.18 at 4:42 am

”before it really found its home as a black market currency”.

And we thought that ”Bitcoin” was invented to be ”THE black market currency”?

And that’s what it is – and always will be – and sometimes there is more demand for THE black market currencies and sometimes there is less – and that’s how it’s value changes…


Chetan Murthy 11.28.18 at 7:58 am

Argh! In my comment #59 above, I meant to say:

(A) These transactions are obviously NOT recorded on BTC/ETH blockchains — b/c the latency there is at least 10min, and more like 60min.

By which I meant, that if an exchange can process trans for its users in a few seconds, and it takes many minutes to commit those trans to BTC’s blockchain, then we can safely infer that the exchange is NOT using a BTC blockchain to record each of those trans.


Chetan Murthy 11.28.18 at 8:10 am


You are *precisely* correct. The only real demand for BTC is from crime: illicit goods (drugs, arms), ransomware, capital flight, etc. Studies have shown that when capital flight increases (the example I remember was about Argentina), then the BTC price increases. And similarly with decreases.

nastywoman is right about something else: this is more-or-less the “mission statement” of Bitcoin. B/c they’re very proud, very demonstratively boasting, about “censorship-resistance”. They’re proud of the fact that errant/fraudulent BTC trans are immune to judicial remedies. In what world, would people who aren’t *intrinsically* computer-literate and superhumanly error-free, use such an insane system? It’s just *asking* to be robbed by more-clever actors. And again, with no ability to go to court to get compensation.

The latest story for BTC is that it’s not aimed at retail transactions (from real users), but rather for “settlement”. If this worked-out, it would mean that BTC trans came only from banks (a necessary characteristic in order to keep the # of trans/sec down to something the BTC network can sustain). But those are -precisely- the users who have *no need whatsoever* for BTC, and who, by dint of being regulated by the Fed, couldn’t take advantage of censorship-resistance at all. All the immense costs of BTC, all the insane limitations of the BTC software and protocols, and for *nothing* — no gain. Some make the argument that BTC is needed for settlement b/c interbank transfers can take -days-. But they forget that the reason those transfers take days, is roadblocks imposed by intermediaries who make money on those roadblocks. Clearly one can do better — and the fact that I can walk into an ATM in Boston and withdraw money from my account in San Francisco, tells us that it’s been a solved problem for my entire adult life.

BTC has no valid purpose as a settlement system, either.


Raven Onthill 11.28.18 at 11:20 pm

The hope of the original cryptoanarchists was that the advantages of cryptocurrencies would be so great that people would abandon regulated bank currencies in their favor, therefore making taxation impossible. So far the cryptocurrencies have demonstrated, yet again, why governance is valuable to a workable economic system. This is not news to anyone who knows monetary history, but there was that arrogance, that belief that their thinking was superior to all previous thinking on the matter.

Anarcissie, the price of labor is not stable; it responds to supply and demand and is as subject to inflation as anything else with a price. This was worked out by Keynes in his famous General Theory, which is probably as near to certain as anything in the social sciences.


Ebenezer Scrooge 11.29.18 at 1:02 am

I’m pretty much with Chetan, except on two points. First, the DTCC is a coop, managed by its members. Its profits are distributed to the same folk who pay for it. In other words, they may be a monopoly but they can’t really overcharge anybody.
AFAIK (I’m a banker, but not an omniscient one), bankers are hoping that blockchain can disrupt some inefficient industry path dependence: especially correspondent payments. The secret sauce (if any) isn’t blockchain, but disruption.
Second, central bank accounts are not digital currency. The trick to paper currency (or any negotiable instrument) is that it can be used for payments without the involvement of the issuer. A digital currency would presumably have the same property. Central bank accounts are ordinary bank money, not digital currency. Central banks are just ordinary banks, albeit not profit-maximizing.

Comments on this entry are closed.