Topic: Mises' Regression Theorem. (Read 12985 times)

May. It's like declaring that water is good for fire because people have tried different liquids for the task, and water is proven to be the best.
However, it does not follow, that the characteristics of water have nothing to do with this problem.
I think it is objective.

I don't know for sure, you'd have to ask him. I just figured it was either that he took real financial transactions very seriously or he didn't want people to get too invested in it so that he could make breaking changes if needed or as the developer, he perhaps wanted to stay independent of the financial side of it. To this day I've never bought or sold any bitcoins from/to him and I have absolutely no idea how many bitcoins he has. But the way I looked at it, the idea was too golden to have it sit in obscurity. Earning money wasn't my goal and I'm not a libertarian; I just wanted this revolutionary technology to get the exposure it deserved. Factoid of the day (that last one counts as yesterday): I chose this user name and marketed my now retired website toward the End the Fed movement, libertarians, the Tea Party movement and related movements because I figured that was the best way to get the word out. But I'm not part of any of the above and I don't have any problem with fiat currencies, the Federal Reserve, the current banking system or the current majority party in the United States or the general direction the current government is moving. There are some things I don't like about government and a few decisions from the current administration that I strongly disagree with, but I'm a very traditional USA democrat, so I'm quite content with the general direction in which things are moving.

Yes, I understand that completely. However, bitcoins create a complete public transaction graph showing who has transacted with who. This gives away lots of correlation information. Someone else drew some nice graphs in another thread.

The obvious initial identity leak is in purchasing coins. Since making fiat payment often comes with identity information you have a potential leak there. Also since on purchase, coins are likely transmitted in a single transaction to a single bitcoin address, it becomes an anchor account tied to your paypal account. (depending on the precautions of the market of course)

A second obvious identity leak is in spending the coins for hard goods. As someone previously pointed out, this ties your real address to the bitcoin address the payment was made from.

Also vendors and users publish well known bitcoin addresses tied with either a person, organization, behavior, idea, or product. Say for example, "Donate to "xxxxx" if you want to support anonymous cheeseburgers delivery." Donating "tags" your account as belonging to a carnivore just as efficiently as a carnivore tag on YouTube.

If you trade from a single bitcoin address to multiple well known addresses, it simply corrolates more information about you. Say a naive noob bought 50 BTC and had them all sent to his shiny new bitcoin address. He then sent 10 BTC to anonymously overthrow the government. Send 1 BTC for a cheeseburger, then later decided he wanted a cool "send me bitcoins if you like me!", in his forum signature, and on his facebook page. Poof, he has now tied his anti-government sympathies, cheeseburger support, forum account to his real world facebook page. Just by being a noob. If you created a second anonymous donate account, but then naively transfered those coins to your main account, you are just as screwed. The many to one, then one-to-one pattern would be obviously apparent.

Now say, both parties took lots of precautions and created lots of intermediate accounts for plausible deniability. But unfortunately decided that access to the kiddiepron site costs 123.45 BTC.  Now it is a simple matter to search for all transactions of 123.45 BTC. Then you simply follow the graph backwards and forwards until you find previous or future transactions that collate to the real world.

Generating coins is anonymous, so un-traded coins are safer. Single use bitcoin addresses are safer. Trading entire blocks of coins at once so there is no forking or merging is safer. Always trading in standard size amounts is safer.

If someone created a "trusted laundry service" (danger word) where you could transfer 100 BTC to the service and have it copy 1 BTC to each of 100 new addresses from it's coin cache, (not your previous transaction), that would be even safer.

But in all cases, the number of private keys that people have to track is going to get huge. This means the likelihood of key-loss goes up.

Red: That's the whole point of the multiple addresses.  You can send from any of them, thus they can't be linked together.  Having everyone know about every transaction is required of the algorithm to keep everything secure.

If they can build an Onion router, I can build a cheeseburger router!



Actually, this is the one feature of bitcoin I'm not enamored with. By having every transaction available to everyone whether they need to know or not, It does make traffic analysis and correlation attacks trivial. I don't think people understand how easy that is for the determined. Digital goods don't help in a lot of cases, because vendors still feel entitled to your name and address.


If one was to build a large cache of bitcoins, they could create their own bitcoin laundry service. That will probably be worth a considerable amount of N-value in the future.

That's pretty cool. What was Satoshi wanting to wait for?

Factoid of the day: I willed Bitcoin's monetary value into existence. Satoshi was all like, "This is beta, don't use it for real financial transactions." And I was like, "Dude, this is friggin' awesome! I think I'll give it monetary value." And then I did. Look it now, with the all the monetary valuing. Brings a tear to the eye, I tell ya!

How do you plan to remain anonymous when it comes to actually shipping the product to you? You have to give them a real-world address in order to receive anything. Worse, that location can then be tied to your Bitcoin sending address. Even if you try to obfuscate the Bitcoin address, traffic analysis can probably tie it to your other accounts unless you're absolutely paranoid about maintaining separate online identities. The fact that all transactions are public knowledge would make that kind of analysis rather simple.

Services and digital goods should be safe enough, provided you work through a anonymizing system like I2P or Tor.


I think it's safe to say that any currency capable of establishing itself from the ground up, without coercion, is just as capable of re-establishing itself after a so-called "loss-of-confidence" attack. Assuming that Bitcoins were to become an established currency in the first place, that's not something I'd lose any sleep over. As for the non-monetary value component, it's enough that people want the commodity and are willing to pay the opportunity cost of producing it (direct exchange). Even if that demand is based on nothing but novelty, it's still possible for it to evolve into widespread indirect exchange, transforming the commodity into a currency.

Finally, the resilience displayed by even fiat currencies highlights the difficulty of carrying out any serious attack against an established currency. Marketability has a sort of momentum attached to it. Moreover, most of the attacks which work against plain fiat currency, such as massive overprinting and fractional-reserve lending, are not applicable to Bitcoin. You can't mint arbitrary amounts of bitcoins without paying an approximately equal cost for the CPU capacity, and while a bank could set itself up as a lender of bitcoins, issuing notes to increase the apparent supply, anyone could trivially see their reserve balance just by examining the block chain. There isn't very much scope for loss-of-confidence to begin with.


I'd say that the commodity part of gold's value is pretty much all there is at the moment. Who actually uses gold for indirect exchange, outside of the currency market itself? What goods or services are denominated in gold? However, most of that non-monetary value has nothing to do with gold's material properties; it's all based on exactly the sort of non-material interest which Pippin is discounting regarding Bitcoin.

And how big is the commodity part of gold's value? I guess it is pretty small, making it irrelevant for stabilizing the price, but I don't know how to find out and put  a figure on it.

Is there some minimum value that N has to have? Are you sure that N is exactly zero for Bitcoin?

I'm eagerly awaiting "loss of confidence attacks".

fruidian? Or intentional?

I'll contest you on this point. But I'll concede if I'm just being ignorant in your terms. :-)

I have dollars in coins/bills, dollars in my checking account, and dollars on plastic cards. In monetary value they are all worth the same. In my case one Wendy's double cheeseburger.

However, if I pay for the cheeseburger using a bill or a coin, I have some semblance of anonymity when the fat police come around. (non-monetary value) If I use a check, or plastic dollars, I lose that advantage. However, if I want to buy my cheeseburgers over the internet, I have only non-anonymous choices, and the fat police are going to track me down. So, if I could buy a cheeseburger using bitcoins over the internet, I would have the same monetary value as a plastic dollar PLUS the anonymity value of a coin.

Therefore, by my logic (which could be out of touch with your terminology) it has to be some value other than M-value.

According to Nenelod he is NOT doing that.  he has as many coins as he needs and is on to the next step in his project.

M-value, monetary! Anonymity in this case is an advantage of the medium of exchange! It has no meaning beyond the scope of the exchange process. Look, Red. Suppose you find out that you can stuff Bitcoins instead of buckshot in the shells for you shotgun. This would be non-monetary use.

I think you are right about anonymity as a main engine of Bitcoins. If they take off, they'll do this exactly as a system of instant and anonymous money transfers, not as a saving tool.

Actually, I want them because they claim to be anonymous. That was in the title of the white paper, but it seems to be the least examined feature.

Is it still a priority? It's a huge source of N value.

Once again, it wouldn't. A direct quotation from "Study Guide to Human Action"
What industrial use can there be for human feelings like "Hey, this is a cool concept"? How are you going to eat ideas?

The only truth here is that they are decentralized. In all others points it's exactly opposite. It's pure fiat. The notion of the non-fractional-reserveness is a sheer nonsense: Bitcoins should rather be called "zero-reserve", because they have not even a fraction of tangible goods to be backed with. This is the problem. Though it does not mean Bitcoins can not serve as a means of exchange. They can, but with a huge breach in their soundness. 


Decentralization, non-fiatness, non-fractional-reserveness are all properties of money. So they present exactly the monetary value. Though above, I tried to show that they are not.

And if we find a far cheaper substitute for gold, which accomplishes all those aspects of N?  It's N value would disappear, leaving only a small amount of value for "purists" who only deal with gold, etc.

Bitcoins are a novelty.  People want them because "Hey, this is a cool concept".  They want them because of it's decentralized, non-fiat, non-fractional reserve nature.  That would be non-monetary value.

A possible solution to this problem might be to back bitcoins with nodes. That is, anybody who wishes to solve problems can purchase CPU time can use the bitcoin network to do so. Of course, how do we verify that the problem is something that people want to solve? Another problem with this solution is probably something that I didn't think of. So mine is probably a bad idea.

If anybody got any idea on how to imbues bitcoin with non-monetary property or at least fix the weakness, please say so. We might come up ways to increase bitcoin's value as money.

No, it's not. You guys switched from the crucial point - an existing pre-monetary value of the thing that becomes money, to the secondary one - the process (barter) by which this thing emerges as a universally accepted medium of exchange. It is like declaring that water is good for fire extinguishing because people tried different liquids for the task, and water proofed itself the best one. It doesn't follow, however, that water properties have nothing to do with the matter.

This is the weakest spot of Bitcoin. In "Human Action" Mises said that valuation of money (true money) always includes two components -  valuations of its monetary and non-monetary properties (F = M + N). What does this mean for gold, for example? It means that if, suppose, some nerd, having been hoarding a huge pile of gold, suddenly releases it on the market, he, theoretically, can destroy the M component. Suppose even people would think that gold is not a good means of exchange anymore. But he never can destroy the N component, because it resides in the properties of gold that have nothing to do with the exchange process. Gold is needed as a consumable material for jewelry, medicine, electronics etc. So immediately after the intervention stops, the N component, so to say, kick-starts the process of imputing the monetary value to gold. The question is: where is non-monetary component of Bitcoin? There is none. That's why Bitcoin is vulnerable to all kinds of "loss-of-confidence" attacks. I think that this is precisely what Nenolod is doing right now.

Of course they arose out of barter.  Bitcoins were traded for their novelty, and value as a decentralized system.  People thought "Oh hey, this is a cool idea, lets try it out."  "Oh hey, send me 5 bitcoins so I can check it out." etc.  Then, as it became more widespread, merchants began providing goods and services for it.

Granted, the evolution of Bitcoins is quite unique compared to other currencies, but I still say misers theorem holds here.

The OP brings up a point that I had originally when I saw this (though not quote Mises' regression theorem).  It is a real issue b/c money is not an intellectual process.  Your property is a physical extension of your efforts and as such any mucking with it promotes a visceral response.  I'm not sure that the act of bartering clock-cycles for digits constitutes a large enough opportunity cost to satisfy most people's definition of property.  It may.  We're too early in this cycle to determine that yet.  But, if Bitcoin's acceptance rate never really takes off, this problem will be a contributing factor; likely a big one.

That's really not barter. If you strapped some logs together with some vines would you say you bartered them for a raft? Someone else has to get something in order for it to be barter.

We could say that we build coins because they are valuable to us.

IMO, the value could be attributed to the way the Bitcoins are both created and distributed - it is completely decentralised and anonymous. People value fairness and their anonymity, so it should hold that Bitcoins gain value from this.

That said, it is reason to doubt adoption. Time will tell.

Speculative buyers can give an unbacked currency an initial market value.

I think an unbacked currency such as bitcoin might arise in a free market through speculative buyers trying to make a profit by supporting some exchange rate, and then promoting the adoption of the currency in order to drive up the value of their initial investment.  As long as the reasons for adopting the unbacked currency are compelling enough, and the exchange rate that the speculators initially support is reliable enough, I don't think this scenario is at all far-fetched.

Furthermore, purely ideological support also plays a role in initial adoption.

It's an interesting question, but I can already get many things for BitCoins, everything really, since I can get USD for them. I love some Mises, but if things happen that a theory says don't happen then the theory is wrong, not reality.

I don't think it's to much of a stretch to say that some people started offering valuable goods for coins because of a future expectation that the coins would be a money. I mean they are limited in quantity, easily transferable, and private. They have desirable attributes. Okay, my final (for now) answer is that they actually were a valuable thing as soon as they were created. They have properties people want therefore they have value.

So, bitcoins didn't gets it value from barter. Gold, and other traditional money have industrial or decorative functions.

But bitcoins seem to have no previous application whatsoever and thus doesn't satisfy Mises' regression theorem. Is this a problem that bitcoins doesn't arise out of barter?

If it is a problem, what people can do to fix this?