Gold collapsing. Bitcoin UP.

[Carlton Banks]

The 6-7 properties of money were written thousands of years ago, people. Technology, including "anti-money" technology, has come on a little since then. Privacy is a property of money since at least some point in the 20th century.

Furthermore, cryptocurrency brings a whole new set of (so far) necessary properties. There's no need to compartmentalise "classical money" properties from those which have been added during it's subsequent evolution.

I think it is important to determine what you "have" to have in a money instrument vs. what is "nice" to have vs. what "derives" from other properties. When arguments on which altcoin might be better start, the differences in opinion seem to often come down to this. I don't think the properties of ideal money have changed just because we are using a digital medium now. I also think with the right "have to have" properties, that you can layer on top of that the "nice to have" aspects. Anyway the market will decide.

So, new innovations in anti-money tools don't change the fundamental characteristics of money itself?

My personal view is I do not believe so.

The reason is I have not seen any anti-money tools that are not defeated by the six main properties working together.

Gold was attacked by anti-money tools (bank notes) because it was not very portable, and as a result people stopped using it directly and governments were able to circumvent gold over time. Bitcoin is very portable and perfectly fungible, this makes anti-money tools harder to deploy. For anything I can imagine they will throw at Bitcoin, it seems trivial to layer additional usage mechanisms on top of the protocol as needed. If bitcoin wasn't portable, fungible, divisible, etc this wouldn't be the case.

@ bolded: I would argue that intangibility and being cryptographically driven are far more important properties for cryptocurrencies to counter anti-money technology. Intangibility is definitely a new development in money (people remembering doesn't necessarily count, figuratively and literally ), and it is not derived from the original 7 at all.

It will be interesting to see how the opinions of monetary theorists of this century will eventually settle on this issue (I have no doubt it will become a relevant debate, even if the conclusion is against an expanded group of properties)

[1714216323]

1714216323

[1714216323]

1714216323

[1714216323]

1714216323

[HeliKopterBen]

Bitcoins are not interchangeable as long as they have an identifiable history.

Bitcoins are not interchangeable if a recipient decides not to accept them, whether because of an identifiable history or some other reason.  Monero is not interchangeable if a recipient decides not to accept them, whether because of a lack of identifiable history or some other reason.  It all boils down to which currency will be the most widely accepted, but I will concede that monero is technically more fungible. 

As far as gold being an example, American Eagle gold coins are considered legal tender in the United States and guaranteed by the US government.  As a result, they sell for a slight premium to other gold coins, effectively making gold not completely fungible. 

[smooth]

Bitcoins are not interchangeable as long as they have an identifiable history.

Bitcoins are not interchangeable if a recipient decides not to accept them, whether because of an identifiable history or some other reason.  Monero is not interchangeable if a recipient decides not to accept them, whether because of a lack of identifiable history or some other reason.  It all boils down to which currency will be the most widely accepted, but I will concede that monero is technically more fungible. 

Of course. All I was talking about is fungibility, not other merit.

As far as gold being an example, American Eagle gold coins are considered legal tender in the United States and guaranteed by the US government.  As a result, they sell for a slight premium to other gold coins, effectively making gold not completely fungible. 

Gold items that trade at a premium for whoever reason are certainly not fungible with other gold. Same for fine handmade jewlery, antiques, etc. Just being gold doesn't make it fungible.

[smooth]

Correct about diamonds. Incorrect about cryptocurrncies. Fungibility means that units are interchangeable. Bitcoins are not interchangeable as long as they have an identifiable history. And in practice participants such as Coinbase are taking that history into account in evaluating coins and users.

If Coinbase was the processor of BTC, then I would agree with you. But Coinbase is not.

The P2P and mining network is the processor of BTC. The P2P and mining network will always consider BTC to be perfectly interchangeable, 1 BTC is always equal to 1 BTC regardless of history, law or anything else. The P2P and mining network will always consider 1 whitelist BTC to be of equal value to 1 blacklist BTC, and process transactions accordingly. This maintains perfect fungibility.

If you argument is that entities (goverments) will try to make individuals value certain coins (whitelist) more over other coins (blacklist). However if they can dictate usage of bitcoin, they can dictate usage of anything, including Monero. Either way it does not matter though, to the P2P network BTC are all still identical, this is fungibility.

If Bitcoin was not perfectly fungible, then you'd have a situation were the P2P network and the blockchain processes 1 blacklist BTC as somehow equaling less than 1 whitelist BTC. That is impossible with the current network and would require a hard fork.

Well I disagree with your definition, as I explained earlier. Fungibility exists within a social context. If nobody cares about the history of Bitcoins, then they're fungible, but if people do care, they aren't. I have some Bitcoins that were mined by me. I'm not interested in trading them for Bitcoins that for all I know might have been stolen from bitstamp and blacklisted by btc-e or others, or might otherwise cause me problems in the future. Not even for a small premium. That's not fungibility by the accurate definition.

[smooth]

For example, even with non-reused addresses, compliant participants may report their addresses to a risk-scoring service. If you go outside that "system", boom your coins risk score goes way up.

This is not sustainable in the long term.

All that's required is that users put one non-reporting entity between every transaction between reporting entities. A non-reporting entity could be anyone or any business.

How are you going to get people to do that? Joe Sixpack either buys coins from Coinbase or gets paid in those coins by his employer using a compliant payroll service, those coins go into his coinbase or circle or other compliant wallet. He spends those coins at Overstock or some other retailer using a compliant payment processors.

There is zero incentive for any of these parties to put a non-reporting entity in the middle of their transactions. At best it adds inconvenience and extra cost for no gain. At worst it impairs the acceptability and therefore the value of their coins, which is actually an incentive to not do it.

Coinbase et al. may attempt to keep their users inside a walled garden that is only allowed to transact with other people in the same walled garden, but they won't get away with it in the long term.

I'm not so sure. That's certainly now how things are going now.

[Peter R]

to me, Bitcoin is fungible when you think in terms of UTXO's.  

which half of 1BTC is tainted from joining 0.5BTC taint with 0.5BTC pure from 2 UTXO's?  when you then split that 1BTC into two 0.5BTC payments, are they both uniformly tainted or is there a discrete 0.5BTC that is tainted?  how do you distinguish?

yes, an address can be tainted if identified with a drug dealer.  but that's not the same as UTXO's, which is where the real action is.

I think the mixing of UTXOs that occur during each transaction makes white- or black-listing unworkable in practice.  Each transaction, the taint from the input coins mixes and spreads evenly across the outputs. Taint slowly diffuses across the UTXO set such that nearly every coin--even a freshly minted coin--becomes a certain shade of gray.  

But then what are the rules for accepting "gray" coins?  I don't think it's possible to implement a 100% "white coin only" policy without simultaneously killing the currency.  The reason is that coins are always blacklisted in retrospect.  By the time the coins are blacklisted, taint will have already diffused across previously-white coins.  If these coins are suddenly un-spendable, then the currency system is pretty useless.  

So that means that there will be some threshold for how gray your coins can be before they are unacceptable for certain purposes.  And if that happens, this just creates a market for "coin whitening."  Say, some of my coins are whiter than the minimum…well my wallet can add in a bit of black coins for a slight profit!  Or someone wants to pay for something and only has dark gray coins?  Well, they could use a service like xmr.to except instead of converting moneroj to bitcoins, it converts dark gray coins to light gray coins for a small fee.  

I think it would be so easy to work around any tainting initiative that tainting would be ignored, abandoned, or never attempted in the first place.

[smooth]

to me, Bitcoin is fungible when you think in terms of UTXO's. 

which half of 1BTC is tainted from joining 0.5BTC taint with 0.5BTC pure from 2 UTXO's?  when you then split that 1BTC into two 0.5BTC payments, are they both uniformly tainted or is there a discrete 0.5BTC that is tainted?  how do you distinguish?

yes, an address can be tainted if identified with a drug dealer.  but that's not the same as UTXO's, which is where the real action is.

I think the mixing of UTXOs that occur during each transaction makes white- or black-listing unworkable in practice.  Each transaction, the taint from the input coins mixes and spreads evenly across the outputs. Taint slowly diffuses across the UTXO set such that nearly every coin--even a freshly minted coin--becomes a certain shade of gray. 

That depends on the rules. It may be that the taint is contagious. Coins mixed with black coins become black, or white coins mixed with non-white coins become non-white. It will be in your interest to avoid getting black/non-white coins mixed into your wallet, and certainly regulated participants will do that (by banning/reporting/freezing when it occurs).

At least to the point where having this mixing occur means you will be subject to greater scrutiny, hassles, and expenses (such as your Coinable account being frozen), which means you will want to try to avoid it even if the consequences aren't complete loss of funds or worse.

There is no "natural mixing" that occurs, really. Inputs to a transaction are all controlled by the same party, unless you are using a mixing protocol like coinjoin. If that becomes widespread and very effectively implemented (to avoid any distinctive signatures) it might provide some plausible deniability that the inputs are controlled by the same party, but that isn't the case now. Even so that doesn't address the issue people wanting very much to avoid undesired taint, which gets right to the matter of fungibility.

BTW, it doesn't really make sense for a freshly minted coin to have a shade of gray. Not sure what you are talking about there. Coins are either born clearly white/legal (mined via a legal, possibly-regulated process) or born black (mined otherwise). If you are in the former category it will be in your interest to avoid mixing your coins with less-white ones. Hence my example that I have zero interest in trading my clearly-legal self- mined coins for unknown coins. The logic of hoarding (reverse Gresham's Law) prevents the sort of graying/mixing you describe.

So that means that there will be some threshold for how gray your coins can be before they are unacceptable for certain purposes.  And if that happens, this just creates a market for "coin whitening."

That will likely be treated as intentional money laundering and raise something of an inconvenience and potential loss of some value to the level of criminality. Most people won't do that, and certainly businesses won't. Easier to just stay within the white system to begin with. It also doesn't fix fungibility because I won't want to accept darker-than-par gray coins at par. That's extra cost to me for whitening.

[Peter R]

to me, Bitcoin is fungible when you think in terms of UTXO's.  

which half of 1BTC is tainted from joining 0.5BTC taint with 0.5BTC pure from 2 UTXO's?  when you then split that 1BTC into two 0.5BTC payments, are they both uniformly tainted or is there a discrete 0.5BTC that is tainted?  how do you distinguish?

yes, an address can be tainted if identified with a drug dealer.  but that's not the same as UTXO's, which is where the real action is.

I think the mixing of UTXOs that occur during each transaction makes white- or black-listing unworkable in practice.  Each transaction, the taint from the input coins mixes and spreads evenly across the outputs. Taint slowly diffuses across the UTXO set such that nearly every coin--even a freshly minted coin--becomes a certain shade of gray.  

That depends on the rules. It may be that the taint is contagious. Coins mixed with black coins become black, or white coins mixed with non-white coins become non-white.

My second paragraph spoke to this point.  If coins mixed with black coins become black, then every coin is at risk of turning black after it's already been accepted for payment.  So I don't think the currency system works at all at that point.  

There is no "natural mixing" that occurs, really. Inputs to a transaction are all controlled by the same party, unless you are using a mixing protocol like coinjoin.

Let's imagine a bitcoin thief sends three "black" outputs to three recipients whom were not involved with the original crime.  The first recipient then uses this black output along with two other white outputs to purchase something from Overstock.  The output that Overstock receives can now be thought of as gray and is an example of "natural mixing."  

BTW, it doesn't really make sense for a freshly minted coin to have a shade of gray. Not sure what you are talking about there. Coins are either born clearly white/legal (mined via a legal, possibly-regulated process) or born black (mined otherwise).

It depends how you define taint.  Imagine that a miner mines a block that awarded him 1 BTC in transaction fees.  Later it's learned that those transaction fees were from coins involved in a crime (and retroactively blacklisted).  Is the 26 BTC coinbase reward now "white," "black," or "gray?

If the answer is "white," then this is a great loophole for whitening coins.  If the answer is "black" then this currency system is probably useless.

[cypherdoc]

The coins remain traceable on the blockchain.

The coins are traceable, but the link between the coins and to whom they belong is less so.

That's irrelevant to the extent that blacklisting or whitelisting (or the prospect of such) impair the value of those coins.

Not true. There have been numerous attempts to blacklist and track coins from every failed exchange in Bitcoin's history all the way back to MyBitcoin. Yet we all probably have those coins in our wallets to various degrees.

Do you live under a rock? BTC-e already showed Bitcoin has no fungibility and coins can be blacklisted.

yeah, and what has happened since Evo?  nothing as far as i can tell.  coins continue to move freely.

let's do it this way.  show me evidence of a blacklist of coins that is currently operational and effective.

[cypherdoc]

 The reason is that coins are always blacklisted in retrospect.  By the time the coins are blacklisted, taint will have already diffused across previously-white coins.  If these coins are suddenly un-spendable, then the currency system is pretty useless.  

exactly right. 

also, think about the edge case where an attacker decides to send blacklisted coins to every known address in the system.  are all the UTXO's in those addresses now tainted?  given my emphasis on UTXO's being the more valid way to conceptualize this topic, blacklisting addresses doesn't even make sense.

[NewLiberty]

Who would have thought BofA would recommend gold?

The summer months offer a lose-lose proposition for risk assets: either the macro improves and the Fed gets to hike, which will at least temporarily cause volatility; or more ominously for consensus positioning, the macro does not recover, in which case EPS downgrades drag risk-assets lower.

http://finance.yahoo.com/news/bank-america-markets-twilight-zone-123352515.html?soc_src=mail&soc_trk=ma

In a note sent out this morning, Bank of America Merrill Lynch has a warning for investors:

Investors remain trapped in “The Twilight Zone”, the transition period between the end of QE and the first rate hike by the Fed, the start of policy normalization... until (a) the US economy is unambiguously robust enough to allow the Fed to hike and (b) the Fed’s exit from zero rates is seen not to cause either a market or macro shock (as it infamously did in 1936-7), the investment backdrop will likely continue to be cursed by mediocre returns, volatile trading rotation, correlation breakdowns and flash crashes. For this reason we continue to advocate higher than normal levels of cash, adding gold and owning volatility in mid 2015. Given extremities of liquidity, profits, technological disruption, regulation, income inequality… potential for a cleansing drop in asset prices cannot be dismissed . Most likely catalysts: Consumer, Rates, A-shares, Speculation, High Yield.

[NewLiberty]

Do individuals have the freedom to refuse coins with taint that they do not like?

Politicians have certain restraints, and perhaps some religious people might not want to associate themselves with money from some sources?

The public block chain ledger offers new ways to organize this use of money that was previously not so possible to authoritatively accomplish.  Is it a better money for this reason?

 The reason is that coins are always blacklisted in retrospect.  By the time the coins are blacklisted, taint will have already diffused across previously-white coins.  If these coins are suddenly un-spendable, then the currency system is pretty useless.  

exactly right. 

also, think about the edge case where an attacker decides to send blacklisted coins to every known address in the system.  are all the UTXO's in those addresses now tainted?  given my emphasis on UTXO's being the more valid way to conceptualize this topic, blacklisting addresses doesn't even make sense.

[cypherdoc]

Correct about diamonds. Incorrect about cryptocurrncies. Fungibility means that units are interchangeable. Bitcoins are not interchangeable as long as they have an identifiable history. And in practice participants such as Coinbase are taking that history into account in evaluating coins and users.

If Coinbase was the processor of BTC, then I would agree with you. But Coinbase is not.

The P2P and mining network is the processor of BTC. The P2P and mining network will always consider BTC to be perfectly interchangeable, 1 BTC is always equal to 1 BTC regardless of history, law or anything else. The P2P and mining network will always consider 1 whitelist BTC to be of equal value to 1 blacklist BTC, and process transactions accordingly. This maintains perfect fungibility.

If you argument is that entities (goverments) will try to make individuals value certain coins (whitelist) more over other coins (blacklist). However if they can dictate usage of bitcoin, they can dictate usage of anything, including Monero. Either way it does not matter though, to the P2P network BTC are all still identical, this is fungibility.

If Bitcoin was not perfectly fungible, then you'd have a situation were the P2P network and the blockchain processes 1 blacklist BTC as somehow equaling less than 1 whitelist BTC. That is impossible with the current network and would require a hard fork.

Well I disagree with your definition, as I explained earlier. Fungibility exists within a social context. If nobody cares about the history of Bitcoins, then they're fungible, but if people do care, they aren't. I have some Bitcoins that were mined by me. I'm not interested in trading them for Bitcoins that for all I know might have been stolen from bitstamp and blacklisted by btc-e or others, or might otherwise cause me problems in the future. Not even for a small premium. That's not fungibility by the accurate definition.

that's your choice.  but i expect that you continue to accept USD bills routinely despite them being covered in drug taint.

likewise if most of the actors in the Bitcoin system don't care or more importantly understand that tracking tainted addresses makes no sense in light of the fact that UTXO's are what gets joined and split in essentially non-trackable ways then what you decide personally to do doesn't matter.

and this is what i see as happening; at least for the last 6 yrs.

[smooth]

Correct about diamonds. Incorrect about cryptocurrncies. Fungibility means that units are interchangeable. Bitcoins are not interchangeable as long as they have an identifiable history. And in practice participants such as Coinbase are taking that history into account in evaluating coins and users.

If Coinbase was the processor of BTC, then I would agree with you. But Coinbase is not.

The P2P and mining network is the processor of BTC. The P2P and mining network will always consider BTC to be perfectly interchangeable, 1 BTC is always equal to 1 BTC regardless of history, law or anything else. The P2P and mining network will always consider 1 whitelist BTC to be of equal value to 1 blacklist BTC, and process transactions accordingly. This maintains perfect fungibility.

If you argument is that entities (goverments) will try to make individuals value certain coins (whitelist) more over other coins (blacklist). However if they can dictate usage of bitcoin, they can dictate usage of anything, including Monero. Either way it does not matter though, to the P2P network BTC are all still identical, this is fungibility.

If Bitcoin was not perfectly fungible, then you'd have a situation were the P2P network and the blockchain processes 1 blacklist BTC as somehow equaling less than 1 whitelist BTC. That is impossible with the current network and would require a hard fork.

Well I disagree with your definition, as I explained earlier. Fungibility exists within a social context. If nobody cares about the history of Bitcoins, then they're fungible, but if people do care, they aren't. I have some Bitcoins that were mined by me. I'm not interested in trading them for Bitcoins that for all I know might have been stolen from bitstamp and blacklisted by btc-e or others, or might otherwise cause me problems in the future. Not even for a small premium. That's not fungibility by the accurate definition.

that's your choice.  but i expect that you continue to accept USD bills routinely despite them being covered in drug taint.

Sure because literally no one cares or has ever cared about the drug taint on bills. That's not quite true with respect to serial numbers, so there is a bit more of a concern there, but not so much in practice.

likewise if most of the actors in the Bitcoin system don't care or more importantly understand that tracking tainted addresses makes no sense in light of the fact that UTXO's are what gets joined and split in essentially non-trackable ways then what you decide personally to do doesn't matter.

and this is what i see as happening; at least for the last 6 yrs.

You can't say that no one has ever cared about bitcoin blacklisting, whitelisting, tracking etc. It is unquestionably the case that actors such as Coinbase (which may well be the largest actor in the entire Bitcoin system by reasonable measures) do care about it. It is also clear that discussion and proposals for blackisting, whitelist, etc. have repeatedly occurred, etc. Finally it is certainly the case that the original BitLicense proposal had language about banning tumblers/mixers, etc. (I don't know about the current status).

You can speculate about how that might play out in the future but that is exactly what you are doing (I'm doing so too, I'm just acknowledging more uncertainty about how things will work out).

[thebigtalk]

I want to keep up with the discussion but the pages of this thread is just enormous so I'll share my opinion after reading a few pages (skipping some). Gold will be a permanent, physical item that is being largely traded and the vast majority of its demand. Although, the power of BTC is skyrocketing (or so it seems) cryptocurrencies are still too hard to predict. Gold might be collapsing today but who knows in the future right?

[cypherdoc]

Correct about diamonds. Incorrect about cryptocurrncies. Fungibility means that units are interchangeable. Bitcoins are not interchangeable as long as they have an identifiable history. And in practice participants such as Coinbase are taking that history into account in evaluating coins and users.

If Coinbase was the processor of BTC, then I would agree with you. But Coinbase is not.

The P2P and mining network is the processor of BTC. The P2P and mining network will always consider BTC to be perfectly interchangeable, 1 BTC is always equal to 1 BTC regardless of history, law or anything else. The P2P and mining network will always consider 1 whitelist BTC to be of equal value to 1 blacklist BTC, and process transactions accordingly. This maintains perfect fungibility.

If you argument is that entities (goverments) will try to make individuals value certain coins (whitelist) more over other coins (blacklist). However if they can dictate usage of bitcoin, they can dictate usage of anything, including Monero. Either way it does not matter though, to the P2P network BTC are all still identical, this is fungibility.

If Bitcoin was not perfectly fungible, then you'd have a situation were the P2P network and the blockchain processes 1 blacklist BTC as somehow equaling less than 1 whitelist BTC. That is impossible with the current network and would require a hard fork.

Well I disagree with your definition, as I explained earlier. Fungibility exists within a social context. If nobody cares about the history of Bitcoins, then they're fungible, but if people do care, they aren't. I have some Bitcoins that were mined by me. I'm not interested in trading them for Bitcoins that for all I know might have been stolen from bitstamp and blacklisted by btc-e or others, or might otherwise cause me problems in the future. Not even for a small premium. That's not fungibility by the accurate definition.

that's your choice.  but i expect that you continue to accept USD bills routinely despite them being covered in drug taint.

Sure because literally no one cares or has ever cared about the drug taint on bills. That's not quite true with respect to serial numbers, so there is a bit more of a concern there, but not so much in practice.

likewise if most of the actors in the Bitcoin system don't care or more importantly understand that tracking tainted addresses makes no sense in light of the fact that UTXO's are what gets joined and split in essentially non-trackable ways then what you decide personally to do doesn't matter.

and this is what i see as happening; at least for the last 6 yrs.

You can't say that no one has ever cared about bitcoin blacklisting, whitelisting, tracking etc. It is unquestionably the case that actors such as Coinbase (which may well be the largest actor in the entire Bitcoin system by reasonable measures) do care about it. It is also clear that discussion and proposals for blackisting, whitelist, etc. have repeatedly occurred, etc. Finally it is certainly the case that the original BitLicense proposal had language about banning tumblers/mixers, etc. (I don't know about the current status).

You can speculate about how that might play out in the future but that is exactly what you are doing (I'm doing so too, I'm just acknowledging more uncertainty about how things will work out).

the way i see it, the market has already beaten down most of your concerns.

yes, there has been lotsa discussion but i'm just looking at the results.  and that is that there is no formal blacklisting services that exist yet that i know about.  and if there is one, i'm sure it's not significant.  do i care when about origin when i buy coins nowadays?  no.

also, i view what CB is doing differently.  i do believe they are dedicated Bitcoin advocates and know and care deeply about fungibility.  it's just that they don't want to go to jail so they track the first hop and i'm sure they'd cooperate with any legal investigation.  but that doesn't mean they are in favor of blacklisting.  in fact, i'm quite sure they aren't.

[smooth]

My second paragraph spoke to this point.  If coins mixed with black coins become black, then every coin is at risk of turning black after it's already been accepted for payment.  So I don't think the currency system works at all at that point.  

We'll see. Maybe it doesn't work at all, or at least ends up being somewhat similar to traditional banking where there are extensive KYC/AML requirements. You realize if Bitcoin has any chance of becoming globally dominant with trillions in market cap than the market is pricing that probability at roughly 0.1% (or possibly much less if there are non-dominant, non-zero outcomes), right?

There is no "natural mixing" that occurs, really. Inputs to a transaction are all controlled by the same party, unless you are using a mixing protocol like coinjoin.

Let's imagine a bitcoin thief sends three "black" outputs to three recipients whom were not involved with the original crime.  The first recipient then uses this black output along with two other white outputs to purchase something from Overstock.  The output that Overstock receives can now be thought of as gray and is an example of "natural mixing."  

It is the Overstock customer who screwed up here, and and Overstock may well reject the purchase, or require additional documentation and reporting. Let's say Overstock rejects the purchase and sends the coins back. The customer now has mixed gray coins that may be difficult to use for anything, instead of white coins that were clearly usable.

By contrast if the customer had kept the black coins segregated and rejected them (sent them back and demanded "real" coins instead), this problem wouldn't occur. This in turn will cause this customer to be a lot more careful who he receives coins from and under what conditions and they get mixed together. Which is precisely not the definition of fungibility.

You are assuming, I think, that coin control doesn't exist or isn't used. But people will use it if they have an incentive to do so. If I receive my salary via a clearly-legal regulated (whitelisted) process I probably won't want to taint those coins and create unnecessary problems for myself by mixing them with coins I received somewhere more questionable. If I accept questionable coins at all, I'll keep them separate. It then becomes unclear whether anyone will accept these lemons.

BTW, it doesn't really make sense for a freshly minted coin to have a shade of gray. Not sure what you are talking about there. Coins are either born clearly white/legal (mined via a legal, possibly-regulated process) or born black (mined otherwise).

It depends how you define taint.  Imagine that a miner mines a block that awarded him 1 BTC in transaction fees.  Later it's learned that those transaction fees were from coins involved in a crime (and retroactively blacklisted).  Is the 26 BTC coinbase reward now "white," "black," or "gray?

If the answer is "white," then this is a great loophole for whitening coins.  If the answer is "black" then this currency system is probably useless.

There is not really a "loophole" because we were discussing in explicitly-legal, possibly regulated mining process which would certainly in-effect (or actual practice) whitelist the coins. That would not include deliberate money laundering.

[smooth]

the way i see it, the market has already beaten down most of your concerns.

yes, there has been lotsa discussion but i'm just looking at the results.  and that is that there is no formal blacklisting services that exist yet that i know about.  and if there is one, i'm sure it's not significant.  do i care when about origin when i buy coins nowadays?  no.

also, i view what CB is doing differently.  i do believe they are dedicated Bitcoin advocates and know and care deeply about fungibility.  it's just that they don't want to go to jail so they track the first hop and i'm sure they'd cooperate with any legal investigation.  but that doesn't mean they are in favor of blacklisting.  in fact, i'm quite sure they aren't.

1. Clearly the market hasn't decided anything. Either Bitcoin is a long, long way from its potential, with many pitfalls and unexpected twists ahead, or it won't ever be very significant.

2. The "Chainanalysis" uproar was two months ago. As far as I know that company is still in business and still working on tracking and risk-scoring, and presumably still has customers or expects to. I very much doubt they are the only company working on these things. I also very much doubt we've heard the last about tracking, whitelisting, blacklisting, bans on mixing/tumbling, etc.

3. Coinbase apparently tracks more than one hop. I'd guess they use services like Chainanalysis, too, but I don't really know. It doesn't really matter what they are in favor of as a matter of opinion, it matters what they do.

[Peter R]

There is no "natural mixing" that occurs, really. Inputs to a transaction are all controlled by the same party, unless you are using a mixing protocol like coinjoin.

Let's imagine a bitcoin thief sends three "black" outputs to three recipients whom were not involved with the original crime.  The first recipient then uses this black output along with two other white outputs to purchase something from Overstock.  The output that Overstock receives can now be thought of as gray and is an example of "natural mixing."  

It is the Overstock customer who screwed up here, and and Overstock may well reject the purchase, or require additional documentation and reporting. Let's say Overstock rejects the purchase and sends the coins back. The customer now has mixed gray coins that may be difficult to use for anything, instead of white coins that were clearly usable.

...

I think I see what you're saying Smooth, but I think you're missing the point I'm trying to make: the black listing can occur after the transactions took place.  So suddenly Overstock is stuck with coins that it accepted according to the rules (they appeared to be "white"), but that are now suddenly gray or black.  And if this can happen, then either the currency system is essentially unusable or the blacklisting efforts are largely ignored.  

[smooth]

There is no "natural mixing" that occurs, really. Inputs to a transaction are all controlled by the same party, unless you are using a mixing protocol like coinjoin.

Let's imagine a bitcoin thief sends three "black" outputs to three recipients whom were not involved with the original crime.  The first recipient then uses this black output along with two other white outputs to purchase something from Overstock.  The output that Overstock receives can now be thought of as gray and is an example of "natural mixing." 

It is the Overstock customer who screwed up here, and and Overstock may well reject the purchase, or require additional documentation and reporting. Let's say Overstock rejects the purchase and sends the coins back. The customer now has mixed gray coins that may be difficult to use for anything, instead of white coins that were clearly usable.

...

I think I see what you're saying Smooth, but I think you're missing the point I'm trying to make: the black listing can occur after the transactions took place.  So suddenly Overstock is stuck with coins that it accepted according to the rules (they appeared to be "white"), but that are now suddenly gray or black.  And if this can happen, then either the currency system is essentially unusable or the blacklisting efforts are largely ignored.   

Again, or they just act conservatively and try to avoid accepting coins that place them at this risk, which means rejecting coins of unclear or uncertain origin. If your coins come from some part of the explicitly-legal the walled garden (Coinbase, Bitpay, etc.), they will welcome you with open arms. Otherwise they may take a pass, even if you coins are (or might be) perfectly fine.

I don't disagree the currency system may become unusable under these conditions, or have limited potential. Fungibility is very important. It just isn't assured by the technology in any way I see.

EDIT: Also I would imagine a method for Overstock to legally separate the gray coins in the instance you describe. They could take 1.5 BTC which contains 0.5 BTC of blacklisted coins, split off 0.5 and send it to some government-blessed escrow or claims fund, with the remaining 1.0 BTC being white again. Of course Overstock still takes a 0.5 BTC loss here, which again is going to make them want to be careful what coins they accept. So, again, not fungible.