the end of hypervolatility?

[ThomasV]

Volatility has been decreasing over the last 6 months.


http://bitcoincharts.com/charts/mtgoxUSD#rg180zvzlztgSzm1g10zm2g25zxzi1gCVolatility

This indicator is normalized, so it does not depend on the usd/btc price, but only reflects variations.
(the indicator is dependent on the time scale, do not compare results obtained on daily charts with hourly charts)

My interpretation is that the supply of bitcoins is now distributed among much more people than a
few months ago. We saw a lot of massive sales, where a single trader used to sell 10k or 20k bitcoins
at once. When this happened, the coins that were sold were distributed among many buyers; the result
is that the number of people holding bitcoins is now much higher. Of course there were a few massive
buys too, but much less; buyers typically bought smaller amounts, especially during the downward trend.

This distribution of bitcoins among traders is only one factor explaining volatility.
There are other factors, such as media attention and adoption, that are not predictable.
However, the observed decrease of volatility should be linked to the redistribution of bitcoins among traders.

Of course bitcoin is still very volatile, compared to other currencies or precious metals.
However, the decrease of volatility should facilitate trade.

[Revalin]

Note that most of the volatility was during the bull market.  I worry that everyone's just holding their breath waiting for the slide to end, and once they believe it has, they're going to send it off into another speculative bubble.  I hope that it will go low enough to finish breaking the spirits of even the biggest fools, but I'm not counting on it.

[idontknow]

My interpretation is that the supply of bitcoins is now distributed among much more people than a
few months ago. We saw a lot of massive sales, where a single trader used to sell 10k or 20k bitcoins
at once. When this happened, the coins that were sold were distributed among many buyers; the result
is that the number of people holding bitcoins is now much higher. Of course there were a few massive
buys too, but much less; buyers typically bought smaller amounts, especially during the downward trend.

So if that's true, how long until people stop complaining about the early adopters?

[hugolp]

So if that's true, how long until people stop complaining about the early adopters?

They wont. The few people complaining about early adopters dont really care and use it as a way to attack Bitcoin.

[Revalin]

Possibly never.  Even if all coins were redistributed evenly among all current users, if the userbase expands 10x, then we will be the Early Adopters With An Unfair Advantage.  Then the next 10x growth causes those users to be the Early Adopters.  If it ever becomes a major world currency it might put an end to it, but I'm afraid this problem will prevent BC from ever achieving that kind of widespread success.

The solution I'm considering is dynamic inflation: adjust mining rewards to pin the price to a currency basket, at least until it's an established and stable currency.  That doesn't prevent devaluation so you'd have to be conservative with the adjustments, but it would prevent the "early adopters" problem.

This probably won't happen with Bitcoin since everyone wants to get rich quick, but an alternate cryptocurrency might do it and succeed because of it.

[idontknow]

The solution I'm considering is dynamic inflation: adjust mining rewards to pin the price to a currency basket, at least until it's an established and stable currency.  That doesn't prevent devaluation so you'd have to be conservative with the adjustments, but it would prevent the "early adopters" problem.

But then nobody would have any incentive to get in early. Having an incentive to get in early is needed for establishing and growing the user base, no?

[Revalin]

Why?  I think right now far more people see it as a pyramid scheme (it's not, but there's a similarity: everyone's trying to push for wider adoption because it'll grow their own value; and if people quit buying in they lose value) and that's incentive to stay away.

If NewFairCoin has a fairly stable value, is readily exchangeable to fiat, and inherits BC's anonymity and decentralization, I think people would adopt it.  Perhaps slowly (you won't see a mad rush like earlier this year), but who cares in the long run?

[hugolp]

Possibly never.  Even if all coins were redistributed evenly among all current users, if the userbase expands 10x, then we will be the Early Adopters With An Unfair Advantage.  Then the next 10x growth causes those users to be the Early Adopters.  If it ever becomes a major world currency it might put an end to it, but I'm afraid this problem will prevent BC from ever achieving that kind of widespread success.

The solution I'm considering is dynamic inflation: adjust mining rewards to pin the price to a currency basket, at least until it's an established and stable currency.  That doesn't prevent devaluation so you'd have to be conservative with the adjustments, but it would prevent the "early adopters" problem.

This probably won't happen with Bitcoin since everyone wants to get rich quick, but an alternate cryptocurrency might do it and succeed because of it.

It wont happen because its impossible to do with a decentralized currency. This has been discussed to eternity in the forums.

[Revalin]

My idea: miners add the current exchange rate to every block.  Anyone cheating will simply create orphans, so you don't have to enforce a specific quote source; each new miner just has to agree that your quote was fair +/- a little bit (just like they agree now that the transactions you signed were accurate).  From there you just apply a smoothing algorithm to it (perhaps a mere SMA, but I have more sophisticated methods in mind) and adjust inflation / demurrage to compensate.  I can think of similar methods to handle other aspects of monetary policy in a decentralized way, but I don't want to get too far off the subject here.

If this is debunked, can you link me to where I can read more?

[hugolp]

My idea: miners add the current exchange rate to every block.  Anyone cheating will simply create orphans, so you don't have to enforce a specific quote source; each new miner just has to agree that your quote was fair +/- a little bit (just like they agree now that the transactions you signed were accurate).  From there you just apply a smoothing algorithm to it (perhaps a mere SMA, but I have more sophisticated methods in mind) and adjust inflation / demurrage to compensate.  I can think of similar methods to handle other aspects of monetary policy in a decentralized way, but I don't want to get too far off the subject here.

If this is debunked, can you link me to where I can read more?

What exchange rate do you use?  Bitcoin to dollars? Bitcoin to euros? Bitcoin to pounds? Bitcoin to rembini? A basket of all the previous ones? Who decides if the index of the baskets need to change because the fundamental economic situation has changed (the euro has gone down, the dollar has lost reserve estatus, etc...). And then where do you get the data from? MtGox? TradeHill? Others? Combination? What happens if some of those companies goes down? Who is going to trust in a currency that needs constant changes and agreements to work? What happens if the miners dont agree? How do you deal with the uncertainty?

[idontknow]

My idea: miners add the current exchange rate to every block.

Just wondering, how would you choose WHICH exchange rate or which basket? For some people the price in USD is irrelevant, for example.

Ah, once again someone beat me to the post. I really need to upgrade this computer!

[netrin]

The way that Brazil broke hyperinflation was to post conversion rates to a virtual currency (unidade real de valor). While the old currency was still in force, people began thinking in URV until a new "Real" currency replaced the old with stable URV parity. Allowing miners to 'agree' on a URV value/difficulty would just transfer speculation from speculators to miners. It would certainly be an interesting case study for econ academics, but I would expect it to produce more volatility and manipulation, not less.

[Revalin]

What exchange rate do you use?  Bitcoin to dollars? Bitcoin to euros? Bitcoin to pounds? Bitcoin to rembini? A basket of all the previous ones? Who decides if the index of the baskets need to change because the fundamental economic situation has changed (the euro has gone down, the dollar has lost reserve estatus, etc...).

The idea is to pin it only temporarily until it is established well enough to stand on its own merits.  To that end a USD+EUR basket is adequate, but honestly, the specific basket doesn't matter: ANY reserve currency is a couple orders of magnitude more stable than Bitcoin has been.

The strength of pinning will be adjusted as the number of coins in circulation increases.  For example, it will aggressively pin until the market cap is $1B; become gradually more conservative with adjustments from there; tapering down to zero pinning once the coin supply reaches $1T.  After that it gets to float.

And then where do you get the data from? MtGox? TradeHill? Others? Combination? What happens if some of those companies goes down?

Miners may use any source they wish for their quotes.  Blocks that do not have a reasonable quote will not be accepted by the other miners, and are therefore orphaned.

Who is going to trust in a currency that needs constant changes and agreements to work?

Take a look at any fiat currency.  People trust them pretty well even when central banks are fooling around with them.

Who's going to trust Bitcoin when it can't hold value within an order of magnitude for 6 months at a time?

What happens if the miners dont agree? How do you deal with the uncertainty?

The same way as when miners don't agree on a transaction: each faction mines the hell out of whichever version of the truth they believe in; whoever has more hashrate wins.  There can also be auditing by the network: invalid quotes are not relayed, the same as invalid transactions are not relayed.


Edit: It is also possible to adapt the monetary policy (for instance, changing the basket weighting) by extending this technique to create a "decentral bank", but it's more complicated.  I'm planning to write that whole idea up in another post after it cooks in my mind a bit longer.

[hugolp]

The idea is to pin it only temporarily until it is established well enough to stand on its own merits.  To that end a USD+EUR basket is adequate, but honestly, the specific basket doesn't matter: ANY reserve currency is a couple orders of magnitude more stable than Bitcoin has been.

The strength of pinning will be adjusted as the number of coins in circulation increases.  For example, it will aggressively pin until the market cap is $1B; become gradually more conservative with adjustments from there; tapering down to zero pinning once the coin supply reaches $1T.  After that it gets to float.

Again, your system is too open to uncertainties and not practical to implement.

And even supposing it was possible (its not), Bitcoin has changed price a lot mainly because its a very small market. Trying to adjust inlation because of some exchage rate wont change it too much. Any big influx of money will make it go up and any big sell off will make it go down quick. As Bitcoin grows this will disappear.

Take a look at any fiat currency.  People trust them pretty well even when central banks are fooling around with them.

Who's going to trust Bitcoin when it can't hold value within an order of magnitude for 6 months at a time?

Fiat currencies are not "trusted", fiat currencies are accepted because the government impose them. If the governments were ot forcing their fiat currencies nobody would accept them.

And there is lot of people trusting bitcoins (including me).

[Revalin]

Again, your system is too open to uncertainties and not practical to implement.  And even supposing it was possible (its not),

How is it impractical or impossible?  I don't know if it's economically sound, but the mechanism I described is very simple to implement.

Bitcoin has changed price a lot mainly because its a very small market. Trying to adjust inlation because of some exchage rate wont change it too much. Any big influx of money will make it go up and any big sell off will make it go down quick. As Bitcoin grows this will disappear.

I don't consider fluctuations inherently a problem.  Like you say, that's the nature of a small market.  The problem is the wild-eyed speculation that coins will be worth 1000x in a few years if we just push this thing hard enough - I think that's undermining Bitcoin terribly.

My goal isn't to prevent all price fluctuation.  I'd definitely want a very heavily-smoothed inflation algorithm, to let the market do its thing (arbitrage, market making, forex, etc) over the short to mid term.  The inflation is to limit to long-term price speculation to prevent bubbles during deployment - training wheels until it's big enough to ride on its own.

Fiat currencies are not "trusted", fiat currencies are accepted because the government impose them. If the governments were ot forcing their fiat currencies nobody would accept them.

That's a fair argument.  Let me take a different angle:  Why is this less trustworthy than Bitcoin?  It's preprogrammed inflation, not at the whim of a central authority - which is a lot like Bitcoin - but it adds a stabilizing outside influence.  It also adds uncertainty, but is that uncertainty bigger than that currently surrounding Bitcoin's future?

And there is lot of people trusting bitcoins (including me).

You do, but take a look at the market cap: that's a pretty direct measure of the global trust of Bitcoin. 

I trust Bitcoin to function as designed, but the design may be economically flawed.  How are we going to achieve widespread trust if we can't get past the current small-scale problems?

[idontknow]

My interpretation is that the supply of bitcoins is now distributed among much more people than a
few months ago. We saw a lot of massive sales, where a single trader used to sell 10k or 20k bitcoins
at once. When this happened, the coins that were sold were distributed among many buyers; the result
is that the number of people holding bitcoins is now much higher. Of course there were a few massive
buys too, but much less; buyers typically bought smaller amounts, especially during the downward trend.

So if that's true, how long until people stop complaining about the early adopters?

2 years later: People are still complaining about the early adopters.

Oh, and the end of hypervolatility you say?

[cr1776]

Love the post - in 2015 there will be the same complaints except 2013 will be included in the early adopter range. 

The same thing happened in 1995-1999 with domain names. And the "early adopters" who bought land in <wherever> that is now worth much more.  Intel and Apple in 1986.  etc.

2 years later: People are still complaining about the early adopters.
Oh, and the end of hypervolatility you say?

[coinft]

Possibly never.  Even if all coins were redistributed evenly among all current users, if the userbase expands 10x, then we will be the Early Adopters With An Unfair Advantage.  Then the next 10x growth causes those users to be the Early Adopters.  If it ever becomes a major world currency it might put an end to it, but I'm afraid this problem will prevent BC from ever achieving that kind of widespread success.

The solution I'm considering is dynamic inflation: adjust mining rewards to pin the price to a currency basket, at least until it's an established and stable currency.  That doesn't prevent devaluation so you'd have to be conservative with the adjustments, but it would prevent the "early adopters" problem.

This probably won't happen with Bitcoin since everyone wants to get rich quick, but an alternate cryptocurrency might do it and succeed because of it.

Being an early adopter is a major driving force, and such an alternative currency will fail if it doesn't promise to reward early adopters well enough.

By the way there is no need to redistribute. Just buy some bitcoins now, and to the people in the next 10x increase you'll be an early adopter too.

[BBazaar]

ThomasV has it mostly right and volatility is decreasing.  Following the April pop I had pegged BTC around 75 and it will settle somewhere in that range.  You don't need to adjust anything yet.  As Bitcoin gets adopted by more individuals, processed in more transactions, and finds it's way into everyday life you will see stabilization and continued growth.

You can't stop inflation.  Yes it's built it, but even if it wasn't, economies don't grow without any inflation.  No inflation means no growth.  The same concepts that apply to basic public finance and economics will apply to bitcoin.  Bitcoin might be an online currency, but it is a currency just like all FIAT and will obey the same principles. Currency allows for economy and economy will also always behave according to basic fundamentals.  In fact, it's behavior will closely resemble other modern day economies purely by association in this day and age.

All I'm saying is, don't go messing with code and mining just yet.  The process has gotten it this far, you would be tweaking something that isn't broken yet.

It's worth pointing out that in every economy, early adopters are taking the most risk thus reap the most reward.  Another basic fundamental.  There would no incentive without it.

[Shaltuum]

My idea: miners add the current exchange rate to every block.  Anyone cheating will simply create orphans, so you don't have to enforce a specific quote source; each new miner just has to agree that your quote was fair +/- a little bit (just like they agree now that the transactions you signed were accurate).  From there you just apply a smoothing algorithm to it (perhaps a mere SMA, but I have more sophisticated methods in mind) and adjust inflation / demurrage to compensate.  I can think of similar methods to handle other aspects of monetary policy in a decentralized way, but I don't want to get too far off the subject here.

If this is debunked, can you link me to where I can read more?

That would actually require a centralization of Bitcoin.