Topic: Mises regression theorem is inconsistent (Read 9420 times)

Hello, what I exactly claim  is that regression theorem is not necessary to explain why the market grants value to any medium of exchange.  It would be like saying that any economic good needs to have a previous utility to actually have any utility!.  

Regression theorem is not only unnecessary to explain bitcoin´s value.  It´s also unnecessary to explane commodity money´s value or fiat currency´s value.     A commodity which is seem as useless today, could be subject to a newly discovered utility as a medium of exchange and then become valuable just for this new and unique utility.

What Carlos Bondone claims (and I agree) is that Monetary utility is valuable enough by itself.  The real problem for Mises is that since his monetary theory does not separate the concepts medium of exchange (currrency) and money he couldn´t accept that the market may grant value to a medium of exchange which is not a present good (i.e. money), but the fact is that the market uses credit (which of course is not a present good) as medium of exchange, and the market is perfectly capable by itself to calculate the value of credit.

Maneulgar,  this is a really interesting point.  I'm not sure I can disagree with you at this point.  I want to think and read about this for a while and then will post my thoughts back on here.  It would have been impossible for Mises to predict the emergence of a scare money source out of non-scarce items (bitcoins which are based on electrons, or digital ones and zeros) so does his Regression Theorem fail?  Food for thought. 

Careful, Menger said that usually, all commodities that became money throughout history had a non-monetary use before becoming money.  It is was just an historical observation, but he did not establish that as a necessary precondition for money.   Only Mises, with his regression theorem, established it as a necessary precondition.


As I said from the beginning, I don´t think it´s mandatory for anything to become money to have a prior non-monetary use.  Besides, having a non-monetary use does not guarantee that the good is not demonetized in the future and probably lose most of its value.  Nevertheless, having more than one utility is a nice property, which will make that new money more valuable and stable.   Namecoins would fullfill the regression theorem for sure.  I agree with you.

I was thinking today that even if the goldbugs are right about Bitcoin, then surely Namecoin (or perhaps something more mature) can emerge as a successful currency and fill the same role. Because if the criticism of Bitcoin is right (i.e. that unlike gold, you can't do anything with a bitcoin), and this inevitably means that its value will fizzle out over time and that there will always be a bagholder left at the end (which I don't necessarily accept), then Namecoin doesn't suffer from this problem. The argument is that because gold and silver can always be used in electronic equipment and other industries, it will always have some demand in the economy, so there's never a bagholder. A red herring I suspect, but oh hum... they might be right. At the very least Bitcoin will be a nice empirical test of that, because it effectively controls for the centralized nature of other fiduciary currencies (I won't call Bitcoin a fiat currency, because that term relates to the force of government).

The argument from the Mises' Regression Theorem and Menger goes that a money good must have demand for some non-monetary utility, at least initially in history, in order to even become money. Then it will be used in barter trades and the very first prices will be set. From then on, it is only the most divisible, fungible, etc, goods that are being traded that will emerge to become de-facto standard currencies. People will accept gold as payment because they know it stores value well, and they can always sell it on (i.e. purchase other goods with it, and divide it up for change if necessary) because there is always some demand for it in the economy, because somewhere down the line it has non-monetary utility for someone.

But if money naturally emerges from a state of barter between goods that provide utility, then Namecoin can work as a currency by this logic. Namecoins are identical to bitcoins in the way that the are mined, and in the way transactions occur, but you can do something with a namecoin: register and hold a .bit domain name. This is like land in the sense that no two people can own the same parcel of land, and no two people can own google.com, but it has utility. You enjoy a monopoly over the land you own, and over the domains you own. They are especially useful to poker sites and bittorrent sites, who could potentially have their domain names seized by the government (because they have control over the root DNS servers), but they will be safe with a .bit name. The only person who can control it is the one who owns the private key for the namecoin, which will be stored and backed up somewhere in Sweden or something, where the US government has no jurisdiction to launch a raid even if came down to that.

So if they're right, Bitcoin will eventually die (once all the hype settles down), but Namecoin could potentially live on because it does provide non-monetary utility. I think it satisfies what these guys want to see in a currency. Because of this, it could eventually emerge as a competing currency in the same way that gold is purported to have arisen as money: naturally and spontaneously. It could fill the role of Bitcoin, even if Bitcoin can't. So who cares? Any way you shake it, distributed peer-to-peer cryptocurrencies could very well be revolutionary

Just a thought.

(watching -- sorry for the noise)

This is just enough for any agent to give something in exchange for Bitcoins. How much? whatever that agent is willing to give on exchange is a valid market price.  Once that first voluntary exchange is performed, you have your first market price for Bitcoins.

Need for any prior non monetary use? Absolutely not.

The point is, how many fiat currencies have been tried to impose, and therefore setting their value by decree, that have failed completely from the beginning?

The point is that whenever issuers have tried to impose a fiat currency, they looked extremely careful to the market, because they know that they cannot defeat the market (they would like to, but they know they can´t, it´s too expensive).  First they sound out the market to retrieve a reasonable market price, and then they publish that price.  Are they setting the price or is the market? Of course they will pretend that they did set the price and all their propaganda will say so, but that´s just propaganda, not real facts.

And yes, monetary theory since Mises is not correct (Hayek did honestly acknowlege this).  In my opinion only the monetary  principles of Carl Menger (the founder of the austrian school) and more recently Carlos Bondone  are correct.

If you need a prior price to excahange anything, then nothing could have ever became exchangable for the first time.  Because this has happened all time through history whenever an utilitiy was first discovered for a good.  

The utility of medium of exchange is valuable enough by itself, no need of any previous utility.    Wouldn´t you like to buy something that you suspect that its single utility is to become a generally accepted medium of exchange before everyone else realise it? I would, and I would pay just for that.

Wrote a very long answer to this that got lost. Anyway a summary. I don't think this is a complete explanation. Monetary utility comes from exchangability. People exchange there goods for money because money have a higher marginal utility being highly exchangable. BitCoin in its very early stage did not meet this criteria, you couldn't buy much of anything with it, also the regression theorem is there to explain where this monetary utility of money comes from in the first place. It states that you should be able to trace the value of money back something value when it was not a medium of exchange.

I really wouldn't call BitCoin in its very early stage a medium of exchange. Maybe a expected future medium of exchange. I think the initial value came from hopes of increased privacy and just plain nerdy fun. The value was minuscule at the time but that doesn't matter and this would satisfy the basics of the regression theorem.
Mises explination of it does however need some work for it to explain BitCoin properly but that is besides the point the theorem holds...

Yes, everything does comply with the regression theorem or it would not exist. That is what regression is...


No, I doubt that is why he developed it. The regression theorem explains the emergence of fiat and credit currencies also, or the theorem would be invalidated and useless to make any point.

Since it does apply to every currency that does and ever existed and thus is not invalid it is however useful to explain that the origin of money has to happen in a free environment and that you can not create money by decree without pegging it to something the free-market already established a comparable price on.

Since no one was forced to use BitCoin at a pegged value to something else at gunpoint it must have regression or it would not have value or prove the regression theorem wrong. But it can't prove it wrong because the regression theorem is based on a set of very simple a priori truths.

If it is can actually be disproven by a post priori fact Austrians have far bigger problems then the regression theorem being invalidated. The entire methodology and basic principles of Austrian economics would have be thrown out...

I don't think I can improve much on the explanations on how BitCoin regresses given here already by me and others. They are good enough for me not to throw out Austrian economics and I think I will leave it at that...

It is very much economics. There are material resources involved getting distributed between humans by human action. Having fun is a perfectly valid economic reason for a choice of how to use resources. To not be economics it would have to be 100% immaterial or completely internal to one individual. BitCoin uses plenty of material resources from the general market. Someone buys electricity and computers and network infrastructure from someone to have fun with BitCoin, it is economic activity.

Actually I would say that the ultimate goal of all economic activity is to have fun. Having fun or increasing the capacity for future fun thru collaboration and material factors is what economists mean when they speak of utility or a good...

What the regression theorem says is that for a commodity to become money, it needs a prior non monetary use.  It´s not about historical prices because since the very first moment that a good has an utiltiy and therefore is exchanged, then it has a price.

The issue here is that bitcoins are first exchanged and consequently they have a price because they are seen as useful.  And they are useful because of its monetary utility.  There is no need to find any other prior utility, and besides they do not have any other.

The first time a price is set for something is a normal market task.  Whenever anything is subject to have any utility (no matter if it is monetary or not) then the market will begin to work to set a price for it.  This has happened to all economic goods in history.

That's because the $20 bill isn't worth $20, it represents $20.  Don't confuse the symbol for the thing with the thing.  People value the symbol at $20, but the physical representation of the thing, when destroyed, is no longer distinguishable as the symbol.

Actually, it would not be in economics, but rather praxeology, the general science of human action, of which economics is simply a subset of action.

Pre-existing demand is necessary for a good to become a medium of exchange, and this demand had to have existed because otherwise no one would be trading electricity for mining coins. I think that theoretically, demand by one person is enough for a medium of exchange to be created. If 0 people demanded coins, then 0 people would have mined them initially. All it takes is one miner to change that. Unfortunately, the Austrian School disagrees. Well, this one guy does, anyway: http://www.lewrockwell.com/blog/lewrw/archives/89471.html

There is still need for a regression theorem to explain how bitcoin became (or potentially will become if you don't think it is) money. Without the regression theorem, you have no explanation of the current $30 prices on Mt gox. The regression theorem explains to us that "exchange value" of any commodity or money, need to have a historical reference of past prices. No one would buy a $30 bitcoin if they hadn't seen what others were prepared to pay for them.

In austrian theory it is. If it's a part of human action, it's economics.

Very well thought out.  Thanks for the link!

http://forum.bitcoin.org/?topic=583.0

Previous topic on this idea. User "xc" has an excellent anlaysis on the subject.

Processing power is an input for BitCoins, it is a necessary element for creating them, not a previous use.

Information (bits) and processing power of course have many other economic uses.  But if you use them to create BitCoins which you are owning just for fun or hobby with no intention of economic exchange, there is nothing economic about it.

subscribed

Regression Theorem requires a previous economic use.  Owning or using something just for fun or hobby (no intention at all to pay for it or to sell it) is not within economics.


Again, unit of account or trading are monetary purposes.  If you use physical gold coins you just need a mask to achieve anonimity, Gold coins are not trazable, the only disadvantage oposed to bitcoins is that anonimity is not built-in.   But any bearer-money is more anonymous than nominative money.  That´s why drug-dealers use suitcases full of dollar bills. This is not new, anonimity has been something valuable for currencies very long ago.


The issuer may call for an auction, so the market will set the value of the currency.  It´ll always be the market who sets the value of currencies  regardless of the issuer, and that happens everyday when agents exchange currency for goods or services at new prices.

Status markers have always been commodities. This was the primary use for gold until somebody invented speaker jacks :-)

I agree, that´s why when I talk about credit or fiat I always use the word currency, because I don´t consider them as money (appliying Carlos Bondone Monetary Theory), but we have to acknowledge they are currency (medium of exchange).

- Currency: Generally accepted medium of exchange.  

- Money: Present good which is accepted as medium of exchange.

Fiat and credit currencies definitely are not present goods, because they are credit (future goods).  But like it or not, they are used as a medium of exchange, so they qualify as credit currencies.

Regarding mental value of bitcoins, I don´t agree that much.  Software is not tangible, the same as bitcoins, but that does not mean that software is just a mental concept.  Software is something real that renders a service.  BitCoins are real and also render a monetary service, and that´s enough to qualify as money (no need of regression theorem).

In order for something to become money in a free market it also needs the properties of good money. Fiat could never become money because it lacks the neccessary properties. It easily counterfitable (look at the government), and it's not durable for two things. In theory though, if it did had the right properties it probably would evolve into money without government support. Fiat even has some physical commodity value (the use of the paper), unlike bitcoin which only have some mental commodity value.

I think you're not regressing far enough.  Bitcoin is a utilization of existing computing power.  Digital bits are the tradable commodity that has been bootstrapped to become a medium of exchange.

@rahl & @mother of another

It is very weak because with arguments like that everything would comply with the regression theorem.  Mises developed the regression theorem b1ecause of his rejection to fiat and credit currencies oposed to what he believed to be "sound money".    But in the real world we use credit as currency like it or not, and we use it because of its monetary utility, not becuase of the law, taxes or any value set by decree.

In fact most credit or fiat currencies are very weak as store of value, and appart from that, the market will set the value of the currency, regardless of what value the issuer tries to set at the begining or during the life of the currency.

You are undervaluing fun by a long, long way.

The regression theorem in its essence just says that if you can bootstrap a tradeable commodity it will become money.

Geek fun is as good way as any to bootstrap, or pizzas, or alpaca socks ..... value is pretty agnostic on origins once the ball is set in motion.

I don't think owning it for fun is a weak argument in this case. There seem to be plenty of people here which are interested enough in the crypto and p2p technologies to participate in the project as a hobby to satisfy there technological interest.

I think the first actual trades of goods in bitcoin did not happen because bitcoin was easy to exchange, or because it was a safe preserver or value. It was very poor in booth of these regards in the beginning. Actually it wasn't even durable, cause a lot of coin was being created in relation to the money base and for those not tech savy enough to understand how the protocol work it is a new and not well tested technology so they probably where not very concerned about durability either.

Where it would have had value is however as an accounting tool. People could trade goods in one place and USD in another and it would be impossible to identify who traded what goods yet the system still keep track of the rewards for sales and transfer them to the USD marketplace. That is not what a currency is for and not a requirement anyone have made on a sound money...

All of this put together means we have regression. It is just not as straightforward as jewelry -> gold -> credit
 

I am not sure what you mean by not necessary to justify its condition. Try issuing completely unspecified credit notes and an see how many will accept them. It is very much necessary to tie the initial issuing of credit notes to some established price and create regression...

Bitcoin had original value as nerd points. The bigger the number you see in your balance, the better you feel about yourself.

Hello Rahl,

a) You can view the first uses as you wish, but to me all uses you are describing are monetary.  Except owning it for fun, which is a very weak argument.

b) All those characteristics are an improvement in relation with other currencies. I agree. But being better currency does not mean is not currency.

c) Credit currency does not invalidate regression theorem.  It just demonstrates that is not necessary to justify its condition of currency.  If you are interested you can read more on this here: http://www.carlosbondone.com/pdf/Theory_of_Economic_Relativity.pdf pages 112 to 114.

The main difference between credit currencies and money (understood as present good) is that the first one is created while the second one is produced.  Credit can be created on unlimited quantities with minimal effort, as long as there is somebody willing to borrow.   

Still the first transactions can be viewed as bartering in a new commodity. That someone wanted to have because of it's unique value that you can enforce your property right in it very easily and without anyone knowing who you are or just because someone was thought it was nerdy fun. Either way it would fullfill the regression theorem.


But the way BitCoins are held and exchanged adds alot of value on top of what would just be expect to get from it as a medium of exchange. This is separate utility built into the system. BitCoins exchangability and the way they are owned and exchanged are different things. Even if you put everyone with bitcoin on an island so they can all still meet eachother easily they probably would not use it if they had to carry it on them and trade face to face, so there is some other value there that makes people want to hold and use it.


Yes, and when the liability is issued it's value is specified in something else that has a price. It can be fraudulent or debased after but that doesn't matter for the regression theorem since such action will only affect the future price of the credit currency thru normal laws of supply and demand. I still don't see how credit currency has anything to do with invalidating the regression theorem.

Still the first transactions can be viewed as bartering in a new commodity. That someone wanted to have because of it's unique value that you can enforce your property right in it very easily and without anyone knowing who you are or just because someone was thought it was nerdy fun. Either way it would fullfill the regression theorem.


But the way BitCoins are held and exchanged adds alot of value on top of what would just be expect to get from it as a medium of exchange. This is separate utility built into the system. BitCoins exchangability and the way they are owned and exchanged are different things. Even if you put everyone with bitcoin on an island so they can all still meet eachother easily they probably would not use it if they had to carry it on them and trade face to face, so there is some other value there that makes people want to hold and use it.


Yes, and when the liability is issued it's value is specified in something else that has a price. It can be fraudulent or debased after but that doesn't matter for the regression theorem since such action will only affect the future price of the credit currency. I still don't see how credit currency has anything to do with invalidating the regression theorem.

Yes, the network effect and the P2P is very interesting.  The underlying technology is obviously new, but I don´t think it is something new from a monetary point of view.  Except for anonymity, physical gold and silver worked in a very similar way and they are also decentraliced money.

Of course, only the market can decide if finally the initial project known as BitCoin becomes money, but the initial objective of the project was to provide an anonymous and decentralized medium of exchange and store of value (i.e. money), or at least I have no knowledge about any other non-monetary objectives.   Anonymous and decentralized are very nice characteristics, but they are specific characteristics that were specifically designed for its purpose as a medium of exchange


Payment implies exchange, money is medium of exchange. BitCoin is not VISA (payment service), they are exchanged for goods.  The payment is performed through the delivery of BitCoins in exchange of goods and services.
It doesn´t matter where you place your ask order if it is not filled.  Bid and Ask orders are subjective valuations, price arises from real exchange (executed transactions).



Credit currencies are currencies because of its monetary utilities other than store of value (in fact that´s their weakest utility as currency). The great difference between money and credit currency is that the first is a present good used as currency, and the second is someone else's liability used as currency.

That something is intended to become money does not necessarily mean that it is money.


Not it is not. It is payment service utility.
Yes, but you would not place the first sell offer in the market at 0 but probably at your cost or higher unless you where really stupid.


It does explain why credit currencies have value. They are initially issued in promises of X that already have a market price. So there is plenty of regression there with a very strong link to previously established prices. Bitcoin has a very weak link to previous established prices but provides additional payment service utility which has value in itself.

There is network effect of money to consider here also. The fact that bitcoin protocol is P2P may have imbibed it with an inherent potential to be money from the beginning, regardless of the regression theorem. It is a new and interesting line of reasoning.

BitCoin was developed as a currency from the begining.  Just look at its name bitcoin, and the paper that originated them "bitcoins: a peer to peer cash system".

Transaction privacy is a monetary utility.   Costs are irrelevant for market value, I could spend lots of energy and work to produce something that nobody wants, so no matter the cost its market value will be 0.  Market values drive costs, not the opposite.

When I talk about Industrial use I mean any non monetary use, which includes ornamental uses. Jewerly is a pretty strong industry.   Gold had been used for ornamental purpose well before it was used as currency.  Mises claims that for a good to became money it needs to have a non monetary prior use, this is what his Regression Theorem is all about.

But I claim that Regression Theorem is not necessary, and it does not work not only with BitCoins, it doesn´t explain well why credit currencies have value.  Carlos Bondone´s monetary theory, which is based on Austrian School founder Carl Menger,proposes a stronger monetary theory, which prefectly explains the BitCoin phenomeon and in regard of this theory BitCoins do qualify as money.

great to get these links to old interesting forum posts. I'm a relative noob so haven't seen many of them.
(off topic - but this shows the limitations of forum infrastructure. The collective memory is being diluted. Probably as many new members in the last 2 months as there has been in the previous 2 years.)

Sorry, I first tried to use the old thread, but the system suggested me to start a new one because the old one had been quiet for a while.

Please that is not what I am saying at all. "frame of reference", "can", "initial" ... maybe you missed reading all of those words. It helps a lot to know what shit that has no market price did cost you when you are trying to figure out what to sell it for it is pretty obvious...

And yes some people will have a much higher expectation of future value, some will be happy getting there cost back and some just want get rid of it for anything they can get. Bit cost does have an effect on where the initial bidding can start.

Interesting ideas being floated here, just going to follow this thread.

That is just wrong. The ashes of a $20 bill don't have $20 as a lower limit to value.

There is a place where this has been discused already: http://forum.bitcoin.org/?topic=583.0

Dont start new threads, use the old one.

Yes, but anchoring BC in scarce physical goods like energy and processors did create scarcity and a frame of reference (ie everyone knows there own price in other stuff for the cost they paid for them and it can set a lower limit anyone will be willing to do the initial barter at).

Cost does not create value. Cost is paid because value is expected.

No don't think it violate the Mises regression theorem, it is just that it evolved incredibly fast so the first stages where barely noticeable.

BitCoin was not actually created as a money, but a scarce virtual good that can be owned anonymously. The first transactions with BitCoin would have been barter and not indirect exchanges. People mined some coin and used it to barter to something else, and only gradually there are more and more indirect exchanges where BitCoin is a money taking place.

The fact that a owner of a unit of BitCoin good can not be directly identified and that it is virtual add tremendous utility value to BitCoin. Also a bit of the initial value can be derived from the nuseanse and cost of creating it. Mises theorem is simply and understandably just not adapted to a virtual economy so it is difficult to see, but there is utility value in BitCoin that is not purely monetary. It don't think Mises ever claimed something first have to be used in heavy industry, gold didn't have any industrial value before it was used as money. It was just pretty to make ornaments from and scarce. With BitCoin instead of wanting shiny things to give the money initial value people wanted something which is virtual and who's owner is very difficult to identify...

Hello,

I´ve been thinking a lot about this.  I think I know very well the regression theorem of money from Mises and certainly BitCoins violate this theorem.

The issue here is that I don´t think that the regression theorem of money is necessary to explain why currencies come to existence.   It´s not true that currencies need previous industrial utility to become currencies.     Credit has been used as currency for a long time and it has no industrial utility.

I´ve written a brief post about this on my blog, please visit if you are interested.

http://eleconomistaprudente.wordpress.com/2011/06/06/bitcoins-and-mises%C2%B4s-regression-theorem/

P.D.  Nevertheless, it wouldn´t hurt at all for BitCoins to have a non-monetary use.  Maybe some cryptographic application as unique tokens?