## Thursday, November 28, 2013

### Bitcoin will probably keep on skyrocketing

Bitcoins' deflationary problem

Disclaimer: I own no bitcoins while I am writing this blog post.

Bitcoins, a virtual currency, were proposed in late 2008 and started in early 2009. Those who have been in it for a year or more have seen remarkable returns – something like 8,000-percent returns in a year. Most of us have missed this opportunity but it's completely plausible that the rise will continue. I want to discuss these issues.

The bitcoin is a unit of wealth stored in the solution to a difficult mathematical problem that needs the brute force of computers to be found. So some computers are working to "mine" this new kind of gold. This gold, when found, may be transferred to others. You may sell actual products for bitcoins.
My experience is zero but perhaps, I would recommend you an online wallet service like coinbase.com
People who use the bitcoin payments often store the full blockchain – a multi-gigabyte public ledger that remembers all the transactions in the world (this will surely get highly impractical if the currency keeps on expanding). How many coins are there? How is their number evolving with time? What will the bitcoin-dollar exchange rate be doing?

The defining rules say that when you find another solution of the mathematical problem, you will not necessarily earn the same number of bitcoins as when you solved a previous one. Computers may be getting faster, more computers may be ordered to do the calculations, and so on, so this could lead to inflation.

Instead, you only get the proportionate amount of bitcoins that were "mined recently" so that the total number of bitcoins increases with time according to a well-prescribed rule. In fact, the total number of bitcoins in the circulation is designed to increase and converge to 21 million in such a way that each 4 years, the deviation from the ultimate number 21 million is being halved.

Currently, the number of bitcoins that exist is 12 million. To calculate their number $$N(Y)$$ at moment $$Y$$, the year treated as a continuous number, roughly speaking$Y = {\rm year} + \frac{\text{#(day) in a year}}{365+\epsilon},$ where $$\epsilon=0$$ or $$1$$ for a non-leap year or a leap year, you may use the simple formula$N = 21,000,000 – 9,000,000 \times 2^{-(Y-2013.907)/4}$ This formula has the property that it goes to $$N\to 21,000,000$$ for $$Y\to \infty$$ and is equal to $$N=12,000,000$$ for the current "date", 2013.907 or so. Don't be picky about the precise numbers, they don't really affect anything because you don't need to know the overall number of bitcoins precisely. The relative changes are more important.

The current value of one bitcoin is $$\1,000$$ or so which means that about 12 billion dollars are currently stored in all the bitcoins of the world. This represents the increase by 1 order of magnitude in half a year (a bitcoin was worth below $$\100$$ sometime in Spring 2013) and 2 orders of magnitude in less than 2 years.

Most of us don't even know how to install the client to own bitcoins etc. so this is a "different world" and the profits we haven't realized are hypothetical. But the people who installed the software a year ago and spent their bitcoins instead of saving them must feel upset: the wealth would have risen 10 or 100 times in half a year or less than two years.

Is there some potential for a further growth of the bitcoin's value? I think that the answer is almost certainly Yes. Why? Well, if we ignore all the traps to be disussed later, the bitcoin may be used as a currency and the limited number of bitcoins that may exist – at least after much more than 4 years when the deviation from the ultimate number 21,000,000 becomes negligible – guarantees that there can't really be any positive inflation if the bitcoins are used.

Unless the bitcoins are banned or otherwise destroyed, an increasing number of people in the world will think that it's a good idea to store an increasing percentage of their wealth in the bitcoins. And it's just a fact that this percentage is so far negligible, despite the 8,000% growth of the bitcoin's rate in the last year or so.

The average adult in the world approximately owns $50,000 of wealth but he or she has around$2 in the bitcoins so far, using the current exchange rate. I find it sensible to imagine that the percentage will increase so that the average person will store not $2 but e.g.$200 in the bitcoins. Another growth by two orders of magnitude sounds totally plausible. Or more than that. Or fewer. There's no hard science to predict the future rates because it will depend on people's psychology and mass hysteria.

The exchange rate could jump to this value – \$100,000 per bitcoin – immediately. But the people who already have bitcoins just don't want to spend too much, partly because someone may "kill" the whole bitcoin industry, partly because they have some psychological prejudices telling them that a "growth rate faster than XY is too much". But if the industry isn't killed, the fast and persistent growth of the value may continue.

With this growth of the bitcoin's value, prices expressed in bitcoins will keep on plummeting. In other words, one will have deflation. If you expect deflation to continue, it is stupid to spend the money (bitcoins). It's better to use something else to pay for your food – for example, to sell your house and use it to cover your expenses.

That's why everyone would like to sell his services for bitcoins but almost no one will actually offer his bitcoins for actual goods and services. No one will pay predetermined wages in bitcoins (because the value of such a salary might go insanely up); no one will borrow bitcoins, either (no loans!), or make other commitments to pay XY (fixed number) bitcoins in the future. This hoarding of the deflating currency (bitcoins) means that the currency isn't really circulating much. It's only used as an alternative part of people's portfolios for savings and its alternative status is here forever (unless the project is abandoned or killed).

However, someone may be hypothetically offering bitcoins for your work. At some point, the total amount of wealth stored in bitcoins may stabilize – which will pretty much mean that the deflation/inflation rate will return back to sensible values around 0%. At that moment, the bitcoins might look as a usable currency for your business.

That will only be the case for a while because it's ultimately clear that the zero inflation isn't good enough. If you know that the prices will deflate in the long run, it's better to hoard the bitcoins. At some moment in the future, the number of newly produced bitcoins will be more or less negligible and the people or companies that already own the bitcoins will keep them. It's likely that they will be very rich but there will be almost no trading.

Ultimately, people may realize that a deflating currency has no long-term future and it won't replace dollars etc. due to the hoarding. So the currency may start to lose the attractiveness and value, too. Assuming that the industry isn't killed, I do expect another impressive growth but I ultimately do expect the fad to go away, too.

One thing will be here to stay: volatility. Whether you like it or not, the prices (and wages) expressed in ordinary currencies are stable because the numerical values are written in various contracts and on stickers – and it's not straightforward to rewrite the stickers all the time – or because someone actively "regulates" the currency (e.g. by inflation targeting or GDP targeting).

If all the calculations of the prices becomes electronic and dynamical, there's no real reason to expect stability of prices (or wages). There's no reason to expect any stability of prices (or wages) expressed in the bitcoins. No one is targeting the inflation rate. No one is doing anything else that would keep the value of the bitcoins nearly constant (relatively to products and services or at least relatively to well-known currencies which are quasi-constant). So if there's nothing that would remove the volatility, the volatility will exist.

To summarize, I do think that it's a fair idea to be using electronic currencies and "automatic" policies that determine their values. But I think that the "macroeconomics" behind the bitcoins is flawed. Because it has absolutely no mechanisms that would encourage price stability, the overall prices – and the value of the bitcoin – will never be stable and this is a huge vice for a currency that would like to be used in most everyday and other transactions. A good electronic currency should be based on some mechanisms that pretty much target the inflation rate.

But people will probably have to learn about the bitcoins' lethal vices in the hard way and the price of the bitcoin will probably continue for quite some time – by some order of magnitude or more – before the bubble pops and people abandon it. There will be many winners, especially the winners who already exist, but there will also be many losers.

Now, you would surely love to know what the maximum value of a bitcoin will be and when it will be reached. I don't know. But I think it will be very high and it will appear when a vast majority of the bitcoins are already "out there" – simply because people are irrational or stupid and the very fact that with the current rules, bitcoins can't be used as the world's main currency, will remain controversial for years (it will be as controversial as it is today). So lots of people will play with the idea that it could be a main currency which means that a sizable fraction of the mankind's wealth may be moved to the bitcoins before the maximum is reached.

But most responsible companies and market players won't translate their prices and debt and incomes to bitcoins. Because only a minority of the wealth will be stored in the bitcoins at every moment and because the price of the bitcoin will always be dominated by speculators who don't have any "objective guide" what the price should be, the volatility will be here forever.

I want to say one more thing. The very concept that "new bitcoins may be mined" was designed in order to allow an increasing number of people to be "sucked" into this Ponzi scheme. But for the actual value of the bitcoin, it's pretty much irrelevant already today because over 50% of the ultimate bitcoin number already exists. So the potential for drop due to new bitcoins is less than a factor of two while the speculative bubble may still add orders of magnitude. For this reason, it's already a good approximation to neglect the fact that new bitcoins are being mined. This fact is only important for the marketing – to make new people interested. The real question is what percentage of mankind's wealth people are ready to store in an alternative currency that may never become the "main currency" due to its intrinsic volatility which prevents it from being accepted by people and companies who have to financially plan (i.e. by almost everyone).

1. "Because it has absolutely no mechanisms that would prefer price stability, the overall prices – and the value of the bitcoin – will never be stable and this is a huge vice for a currency that would like to be used in most everyday and other transactions."

This is actually a major advantage of virtual currencies compared to the currencies we currently use - the separation of different functions of money into separate currencies. While the US dollar has to take on the burden of providing all functions of money in one currency (and all regular currencies work the same way), virtual currencies can be specialized to provide a better quality function each. BitCoin is great for storing value, since it is inherently deflationary and therefore makes for a good, safe investment. Some kind of spending power stable Freigeld (like E-gold but with much more pronounced demurrage) is a great medium of exchange since economic agents are naturally compelled to make it flow while the prices would remain stable - it would serve as the national fiat currency. And so on. Free market will ensure the exchange rates between such currencies will coincide with the value people attach to their individual functions, since the functions are not lumped together which obscures their individual value. The growing legal acceptance of these currencies (BitCoin in Germany etc.) is a great victory for free markets as it brings a much needed competition to a part of economies dominated by central banks/commercial banks/governments whose domination stems from neo-feudal practice of granting monopolies on the creation of money.

2. Dear Milan, you're completely wrong. It is a great disadvantage for a currency if it isn't "versatile". The U.S. dollar is important exactly because it may play all the roles.

It's nonsensical to "use" a currency that only exists "for the savings". One shouldn't really call it a "currency" at all.

3. Aside from money laundering
and speculation, I don’t really understand the goal of Bitcoin. If one
wanted to undermine a corrupt, authoritarian government, one could live
naturally in a commune sharing goods and services as necessary. I fix
kids, etc.

Bitcoin is worse than the gold standard in that the creation of currency
corresponds even less to the amount of economic activity.

4. Off topics -- Happy Thanksgiving Lubos!

5. As with Maple Leafs, Krugerrands, and whatever the US is shilling (debased alloy?), threatened governments will issue a smorgasbord of competing electronic currencies rendered "safe" by high level encryption. Nothing escapes from NSA security. Obamacare databases are not subject to hacking. US utilities have the best most secure customer billing software in the world.

Putting all your eggs into electronic baskets affords ease and economies of Central confiscation. Hoarding, being in physical possession of primary value, will be declared illegal (again).

6. It is really funny reading the musings of physicists and mathematicians in the area of economics. There is a right and wrong answer to almost any question in math and science, whereas there are no correct answers in the soft "science" of economics. Money only has value because we believe it does, and propounding theories of why or how is missing the point. All we actually know about economic matters is, if I may quote a great (and dead) American, prediction is hard, especially of the future.

The thousand dollar bitcoin today may be one dollar tomorrow. But the dirty little secret is that the dollar could go the same way, at any time. What really matters is what these ephemera called money can buy. How many loaves of bread, how many gallons of gasoline? Loaves of bread have cost millions of dollars before - in fact hyperinflation is pretty common in history. The only real enduring store of wealth is real property, and even that can be taken away by government fiat. Less enduring forms are useful items that can be stored and people will always need. Some people believe that gold metal has a magical property that makes it have enduring value, but it seems to me that there are no really safe bets. But what do I know? All I know is math, and there are some missing variables that prevent me from finding a solution.

7. There are tons of stories about the great Fermi but my favorite is this:

When working late in the lab after everyone else had gone home Fermi sometimes needed a special part. When this happened he would turn on the lights in the machine shop and make the part.

It is our great good fortune that Fermi married a Jewish woman and had to flee Italy for the US when Mussolini enacted his anti-Semitic laws.

Personally, I had the good luck to take a course in nuclear physics taught at Berkeley by his colleague, Emelio Segre, also one of the century’s top experimentalists.

8. Inspiring, Gene!

9. Happy Thanksgiving, Justin!

10. Sorry but assigning price and value to things and financial products isn't a mysterious religion without any rules where every answer is as good as any other. Much of it is determined by rational thought, especially in some respects. Whenever the trading is fast and efficient, certain imbalances are removed almost immediately. Economics isn't as hard a science as particle physics but that doesn't mean that rational thinking about economics doesn't produce more accurate answers than irrational thinking.

11. Bitcoin is not the only one: http://coinmarketcap.com/ - and some have built in inflation.

From the alternatives, or altcoins, I'd like to point out Primecoin which doesn't compute useless SHA256/scrypt/whatever hashes as the other do, but chains of primes (which could be more useful?) http://primecoin.org/static/primecoin-paper.pdf

Another interesting one is Namecoin, which can serve as an alternative to DNS.

Zerocoin is not yet released solution which should make transactions anonymous by using zero knowledge proof.

AFAIK the other altcoins are just clones with somehow different parameters.

12. To see that pompous fools and dimwits with too many powers sitting at too influential places, such as editor positions of well known journals, always existed and that they could not obstruct progress in theoretical physics at the end, makes me looking a bit more optimistic into the future too ... ;-)

That Femi's analysis of waves on a string already earnded him his PhD is very cute among other things, I like this :-)

13. Economics is a religion with math, Michael. Economists generally don’t even have
the most basic concept of the scientific method, and that’s why
it’s easy to understand how they can be unphased by little things like
the fact that their models are little more than ideological trash that
bears only a cursory resemblance to reality.

14. Thanks, Franta. The existence of this competition may have a huge impact on the maximum price that any of these coins reaches.

15. LOL, do you think that the purpose of Bitcoin is to undermine corrupt governments of the world? ;-) An interesting thought.

16. This is just facile sloganeering, not an argument.

17. Lubosi, feel free to design one :) It's historical opportunity and whole world is your laboratory. From technical point of view, bitcoin protocol is based on cryptographic primitives known for decades (cryptographic hash function and asymmetric cryptography). You can easily predetermine inflation as a function of time and number of nodes in the network, but you'd have hard time to tie it to some other real world value, like basket of products or local currency. One of the main principles of bitcoin is absence of central authority - majority of nodes must agree on everything.

I don't think bitcoin was designed as a currency in the first place, clearly it behaves like a Ponzi scheme. First adaptors get paid by the late ones. Sure, it's hard too bootstrap new currency in a fair way, but bitcoin originator owns 10% of all coins and since the only intrinsic value of bitcoin is ability to pay for drugs semi anonymously, criminals have probably also high stakes in it. And FBI owns 144000 BTC sized from Silk Road administrator, which is 1.4 % of all coins at the moment :)

18. The digital realm has enabled free sharing of information across the
globe, Lubos. Now with Bitcoin and other crypto currencies, it will enable
FREE trade across the globe. These are examples of people empowerment,
not empowerment of the oligarchies or the status quo. If we want to
evolve as a species on this planet, we have to cooperate as people, not
as governments or as nations. Digitizing trade and information is
leading us there, slowly but surely. At least that's my hope, if Bitcoin is for real. But I'm still holding my breath on that.

19. Beautifully put and so stipulated. Why do you then insist that massless boson photon vacuum symmetries (arXiv: 0912.5057) are exactly fermionic matter (quarks) vacuum symmetries without ever having performed a definitive empirical test? Massed physical theory is patchworks of parity violations, chiral anomalies, symmetry breakings; Chern-Simons repair of Einstein-Hilbert action. String/M-theory has some 10^500 excess vacua. Fermionic matter has a vacuum symmetry leak at the starting line.

"Nature abhors bureaucracy and leftwingers." A garden must be plowed, prepared, fertilized, watered, planted, tended... weeded and pruned. Physical theory is growing Heracleum mantegazzianum - big, pretty, toxic to touch.

20. Characteristics is incredibly effective in generating greatly different results from a greatly specific statistical viewpoint. Even if two things or phenomena or ideas or paradigms look as different as you may get, these two things or phenomena or ideas or paradigms may be relevant by a small change, a simple and apparently simple less indication, and they may eventually be alternatives of the very same actual equations.

21. I think it is several religions, and more arithmetic than math. They can predict nothing. They can't even predict the past.

22. I can't disagree Lumo. But one thing stands out - there has never been a currency with fixed supply. All currencies have always inflated supply throughout history, so this might be an upcoming event without precedent.

23. The most apparent "legitimate" reason to use these is illegal transactions. That works best only until governments act. The clever "mining" idea seems to be a presently legal - or at least hard to prosecute - combination of money laundering and Ponzi scheme.
I assume those who will make the most off this are those who use the coins to sell contraband and avoid bank records. Those who will lose the most are those who watch the rising value and think that they are making money.
This is not much different than the crazy trading in debt securities that led to US economic travails some years ago. There big insurance and banks bought and sold bad debt in form of mortgages. People fueled this by buying houses at inflated prices. Eventually it collapsed. The folks who bought the houses at inflated prices that they never could afford lost. The rich executives who created it kept their vast earning including bonuses and many retired with their golden parachutes. The US government bailed the insurers and banks out so they could figure out new ways to make profits. They have of course. Utilizing the reduced expectations of the newly impoverished they have been able to cut labor costs.
Here the drug dealers and biggest schemers will retire. The ordinary people will go broke and the Government will hire a lot of new spooks to figure out how to prevent this, thus giving the companies who need to facilitate transactions new business opportunities.

24. A new euphemism for logorrhea, thanks!

25. Bitcoin's sustainable value being 0 is an argument that metallists would use against any fiat currency. While chartalists would also claim that bitcoins should be valued at 0 since they are not backed by tax obligations imposed by governments, modern currencies earn their value mainly through being a legal tender, which bitcoin is in e.g. Germany, as I mentioned. Therefore, just like the USD, bitcoin has value because people assign it value - that's how the value of everything in the world is created. Since a currency like bitcoin draws much of its value from network effect, it is quite natural that it's value has been skyrocketing up to now - the more people believe in its viability, the more viable as a currency it becomes. It is a self-reinforcing loop. Once the belief in its viability becomes widespread, its further changes in value will be driven by its inherently deflationary nature, resulting in gradual increase in purchasing power over time, acting as a good way for storing value. Of course, if the confidence in population is lost along the way, its value may very well plummet to zero just like the value of say Reichmark (though both would have collector's value with their physical representation, that wouldn't make them a currency).

Using a stable currency for denominating a debt and using a nonstable currency for repaying it ensures the value of the debt will remain the same no matter the price development of the nonstable currency because debts are denominated in nominal value. Even simple usage of a currency similar in nature to SDR (a currency basket) as a unit of account that has exchange rate with every currency but cannot be actually exchanged would ensure more stable values of obligations. The currency risk would be much lower than in case of denominating the obligation in any other foreign currency, since the value of this SDR2 would be very inflexible in comparison, nullifying exhcange rate volatility. Furthermore, unlike in case of denominating a debt in local currency, the real price of the obligation would not fluctuate according to inflation or deflation.

26. "If you expect deflation to continue, it is stupid to spend the money"

Serious question: which of the following is true about people in the US in much of the 19th century (in particular from after the civil war to 1879, when every year but one saw deflation relative to the previous year, averaging rate of about -4% per year)

A. They were stupid and spent money anyway
B. They switched to something other than dollars for their transactions
C. None of the above

27. Fine but none of these sentences will cure Bitcoin's vices that make it unusable as a main currency.

28. Dear Werdna, the deflation rate was below 2% between 1870 and 1890 and even smaller in the 19th century average

https://en.wikipedia.org/wiki/The_Great_Deflation

It was OK for the U.S. because it was beginning to industrialize but the deflation was harmful to the U.K. businesses because U.K. was already industrialized. See the page above.

Guess which of these two countries in the late 19th century had more similar conditions to the present in the West.

29. legal tender, which bitcoin is in e.g. Germany
False, Bitcoin is not legal tender in Germany. The German authorities have merely recognized Bitcoin as a form of "private money" and stated their position on legal and tax aspects of Bitcoin, but that does not make the currency legal tender. For example, one may not demand to discharge one's tax obligation via Bitcoin.

I don't know of any country where Bitcoin is legal tender.

30. I'm waiting for the Time magazine cover story with a picture of some guy
kissing a Bitcoin, Lubos. That's when I'll know it's over. ;~)

31. Where can you buy bitcoins is what a lot of group asking. at who-accepts-bitcoins.com you can see an overview of all bitcoin websites who furnish bitcoin. Also you can download your notecase and do such much. Simplified for bitcoin lovers!

32. Again, sorry for my delay in responding to this but I don't believe that you are correct claiming that the uncertainty in inflation is completely independent of its rate. It is, in fact, more true to say that the higher the inflation rate the higher the uncertainty about the inflation rate. The reason for this is pretty simple, the higher the inflation rate, the more likely it will be to turn a good investment into a bad one. This riskiness then feeds into people's pricing decisions by driving inflation upwards(people have to raise prices to be sure of earning an adequate return when investments are riskier). Wikipedia's entry on hyperinflation discusses this in detail.

Cheers ,:)

33. Dear Shawnet, I think that your reasoning is circular. You want to establish that they're not independent but you're really assuming it.

What you call "high inflation" in your examples is some combination of "high inflation and high uncertainty" which of course has high uncertainty, too.

But that doesn't mean that there can't be exactly equivalent situations with a high inflation and small uncertainty - a targeting policy for a high inflation will do exactly that.

Similarly, the inflation is tiny, around 1%, in most of the West today but the uncertainty of the expectation is substantial, I think. It seems totally plausible to me that there will be a 5% inflation in the U.S. or eurozone in 2015, for example.

34. Sorry my friend but I fear that you still have a serious problem because you have not defined inflation in a way that can be measured. If you think of inflation as an increase in prices, which I suspect that you do, there is no way to actually measure that increase. If the increased supply of money, which is really what inflation is, drives up the price of Manhattan or London real estate, health care, bonds, stocks how do you really measure the effect, particularly when many of those asset classes may have been overvalued to begin with? How do you measure something as simple as the price of vegetables or processed cereals? After all, there are various grades and portion sized that the BLS people are not very good at dealing with. And what if the increase in the supply of money offsets the expected price declines that come from productivity improvements?

Sorry my friend but you assume much more knowledge than you have and are making the same type of error that is made by the AGW proponents as they create inadequate models with incomplete information.

35. Dear Vangel, deflation is nothing else than a negative value of the inflation rate.

It can not only be measured - as the change of the price of a basket of products - but it is actually importantly measured in all countries in the world. For example, the inflation was 1.3% in OECD countries in October:

http://www.livemint.com/Politics/n6cCQ8AUaqktpVzEbxPjFP/Inflation-slows-to-13-in-October-in-OECD-countries.html

Why don't you look at basic things in the real world e.g.