Quote from: AnonyMint on November 07, 2013, 05:19:38 PM

If the altcoin has sufficient demand (not even any where near Bitcoin) to generate a liquid exchange, then the altcoin can be converted on the fly to pay those who accept only Bitcoin.

So an altcoin doesn't need ANYONE to switch from Bitcoin, it only needs to grab a small percent of the market of new users. Bitcoin only has 350,000 users. There is a long way to go to 7 billion.

The key though is the altcoin needs to present something truly useful that Bitcoin can't copy. So it has staying power.

Agreed. The success of an altcoin is very unlikely to depend on it beating bitcoin (something most late-to-the-party altcoin peddlars don't seem to appreciate). If liquid exchanges exist for a coin then it is usable in every context bitcoin is.

Quote from: justusranvier on November 07, 2013, 11:38:39 PM

Friction always imposes costs. No how matter how good you make your exchange, it will always be more expensive than not needing to make an exchange. A line is the shortest distance between two points.

This is only true when there is a dual-price bid/offer spread. Traders will be willing to pay the par value of the currency they're buying, plus a little bit extra for the added value they perceive in holding one currency rather than another. In a fee-less exchange, if there is more demand for one currency than another then costs will only be imposed one way (from the less desirable to the more desirable), with savings made in the other direction. Now in most cases it is natural to assume that demand for bitcoin will be higher than demand for any particular altcoin and therefore the cost of friction will always fall upon the shoulders of the person trying to spend altcoins on bitcoin services - but this is not necessarily the case. In some cases the friction will be negligible, making truly liquid exchanges a (pseudo-)reality for some alt-coins.

The issue then is whether there will be liquid exchanges for a coin, which depends on whether or not there is any incentive to hold the coin. Anonymint says that for there to be any incentive to hold an alt-coin then:

Quote from: AnonyMint on November 07, 2013, 10:46:19 PM

then the altcoin has to be appreciating faster than Bitcoin, else there will be a pernicious downward spiral.

This is not true. Faster appreciation of value of a coin is what will encourage purely speculative investors to hold/ditch the coin. But there are other reasons for holding certain coins, depending on the features they offer.

Simply offering something that bitcoin can't do however is not sufficient to imply that there is a good reason to hold the coin, because people might prefer to trade out of bitcoin just to use that feature then trade back into bitcoin afterwards, making adequate liquid exchanges unviable.

Let's consider then some current and possible future features of alt-coins:

***************************

Quicker transaction times - if coins with quicker transaction times were only purchased when needed, i.e. bitcoins were exchanged for them, then the wait for the exchange transaction would negate the gain in efficiency - thus people would hold onto coins if quicker transaction times (with the associated orphan-risk) was thought a desirable enough feature.

Smaller blockchain - the point of wanting a smaller blockchain is so that coins can be stored locally rather than having to trust a centralised service. If someone only exhanged bitcoins for this alt-coin when they wanted to use it, they would either have to store those coins centrally or store them locally with a large or else a centralised blockchain, defeating the point of having a smaller blockchain - thus people would hold onto coins if a smaller blockchain (with the associated risks of a likely smaller network) was thought a desirable enough feature.

In these two cases, if the feature was a desirable one there would be a reason to hold on to alt-coins and thus liquid exchanges. However, I personally do not feel either is enough of an improvement to produce a large enough demand.

Coloured coins - there is little difference between switching between coins of different colours within an alt-coin and just buying coins of a certain colour directly with bitcoin. Therefore there is no reason to hold them IN GENERAL: however if there is a particular colour of coin that is worth holding for reasons of usability and/or appreciating value, particularly one that doesn't have its own blockchain/distribution system, then there will be a reason to hold the alt-coin and thus liquid exchanges.

Anonymity - this is likely to (and WILL) be a very attractive feature of some alt-coin (the only reason bitcoin would even dream of implementing it would be if not implementing it would cost them SIGNIFICANT market share and threaten its future dominace, which is unlikely due to the network effect and the fact that most bitcoin users nowadays don't feel the need for ABSOLUTE anonymity). Many services which rely on or prefer anonymity would accept this currency directly, and users only using crypto for those services would hold the currency directly rather than buying bitcoins. Liquid exchanges would be very likely - decentralised, anonymous exchanges would make users of the anonymous currency even more confident.

Proof-of-stake - this provides its own incentive for holding coins, if and only if the growth rate of the held balance * the appreciation rate of the coin's value is greater than the appreciation rate of bitcoin (or some other coin or commodity). Whether the coins will find any significant direct use beyond being an occasional speculative opportunity for bitcoin investors is doubtful, but this alone is sufficient for liquid exchanges and therefore makes them spendable on bitcoin-enabled services.

Stable value - a coin with some sort of built-in automatic monetary policy that adjusts the amount of coins in circulation to keep the price stable would I think be able to attract a modest market share of risk-averse individuals who want a reliable, decentralised store of value. Also, having a stable price will give merchants/service providers an alternative currency to hedge their prices in. They could still use bitcoin as their primary mode of payment, but rather than calculating the bitcoin price based on a fixed dollar/euro price, they could set a fixed "stablecoin (e.g.)" price. Such a coin would function more like a currency than a commodity (which arguably is how bitcoin is behaving right now), though this is unlikely to make it topple bitcoin as people will still generally think in "dollars/euros" and see even a stable cryptocurrency as a medium of exchange. There is still an incentive to hold it as a store of value however.

Different inflationary principles - (demurrage, interest, different/no volume ceiling). There may be some limited situations where a coin with different inflationary principles from bitcoin is desirable, however this is extremely unlikely for the simple reason that any digital currency that doesn't automatically stabilise it's value or else sufficiently peg its value to some commodities/services will be seen as a secondary medium of exchange for other currencies rather than as a true currency in its own right. The article pointed out the flaw with currencies that provide a disincentive to hold on to them - people won't buy them in the first place. What about a currency that provides an incentive to hold on to coins, such as paying exponentially increasing interest (a flat rate of interest would increase the money supply uniformly accross all balances, leaving everyone in the same position value-wise)?. Provided it could gain some initial value, and interest was earned at a decent rate, there would be more demand to buy than to sell so liquid exchanges would be likely (and frictionless in the altcoin-to-bitcoin direction). How much of a userbase something like this could get in order to establish a value however (competing not just against bitcoin but also proof-of-stake coins) I'm unsure. When the supply of bitcoins (eventually) reaches its limit, then perhaps some less scarce altcoin (LTC for example) will become useful for sub-satoshi price differentiation (though even if bitcoin reached 1000x its current value 1 satoshi would be worth only $0.03 so the need for this will likely be very limited).

Name/value pairs (i.e. Namecoin) - obvious: if you want to keep the name/value pair (which is why you bought Namecoin in the first place), you have to hold the coin.

New features (e.g. messaging layer, completely re-imagined contracts) - a new coin built from the ground up, using the principles of bitcoins but offering completely new services, would function like an extension to the bitcoin protocol thus it is likely people would only trade out of bitcoin and hold these coins if they had an immediate need for one of these extended services. Thus liquid exchanges are less likely.

***************************

Of course, all of these features will likely appeal only to informed users who keep abreast of developments, know what they want and consider their options - there is not going to be any massive pull for "ignorant" users (meant in a non-condescending way).

Quote from: justusranvier on November 07, 2013, 11:38:39 PM

Any feature that is so heavily demanded will be added to Bitcoin, because that is the most economically valuable solution.

The other factor that determines whether a feature is sufficient to make an alt-coin viable is whether it is easily integrated into bitcoin. Factors that can be implemented on a client level (and don't require changes to the protocol), such as coloured coins and smaller blockchains, are never going to give an alt-coin a unique advantage. In general, most new features which work well enough on alt-coins will likely be integrated into bitcoin, but there are four exceptions. Ideas that are controversial (like ZeroCoin/anonymity), impossible (like changing the reward system to proof-of-stake), involve a trade-off (like quicker transaction times at the expense of more orphan blocks), or aim to achieve something at odds with one of bitcoin's core principles (like different inflationary or distribution principles).

TL;DR:
In order for an altcoin to have long term success, it needs to offer unique features which a) bitcoin will not replicate and b) provide in-and-of-themselves an incentive to hold onto coins.

[Apologies for the length of the post.]