The 21 million coin myth

[joe]

We need to put together a think tank to figure out where we go once the coins reach close to 21 million. Transaction fees will not work because large sites will form that have a huge bitcoin bank account, and money will be sent between users of the site instead of actual transactions on the bitcoin network. The only solution is to lift the 21 million cap. Remember, the 21 million cap is not set in stone. At the end of the day, the rules are set by miners. We have a handful of mining pools that collectively control over 50% of mining activity easily. Miners have the option to agree on a percentage rate of increase beyond 21 million. If the percent rate is too high, the confidence in the currency goes down and so does its value. If the rate is too low then miners are not paid enough to support the network.

These are just starting points of discussion but weshould get this out in the open sooner instead of later. The mining pools are going to be a powerful force in the monetary policy of the bitcoin network in the future.

[bittersweet]

No. Start a new transaction chain if you want new rules. I'm fine with 50% of miners leaving Bitcoin and starting their own currency.

[eturnerx]

Yawn. The market will sort it out. Basically the miners left will just stop accepting cheap transaction fees - but will be limited by the competition amongst miners. Once bitcoin hits the max-coin level, if their are big banks doing their own internal transfers then surely bitcoin will be worth so much relative to fiat that even 0.005btc could buy a small car.

[Timo Y]

...until eventually, NO transactions happen on the bitcoin network, and the block chain is simply a historical archive proving how much money each of those banks owns.

[eturnerx]

They'll still need miners to stop rogue claiming ownership (inwards transfers) of coins claimed by one of the banks. And there'll still be the people wanting anonymous exchange.

[bittersweet]

You would still need it for transactions between banks.
There could be thousands of banks, you can't really tell what will happen.

With fast googling I found out only in USA there is 9459 banks or savings institution:
http://wiki.answers.com/Q/Total_number_of_banks_in_US

[joan]

We need to put together a think tank

Maybe we can do that in 2035 or something ? We'll still have 5 years to think about it.

[realnowhereman]

...until eventually, NO transactions happen on the bitcoin network, and the block chain is simply a historical archive proving how much money each of those banks owns.

Making it unprofitable for miners, reducing the number of them, reducing the difficulty hence making it cheaper to run a miner, profitable again and hence making the transaction fee lower, making more transactions likely.

Does anyone seriously think that if there were no transactions and no profit, there would still be a terra hash mining network?

If Bitcoin does collapse, it won't be because of mining.  That is self regulating.  It will be because no one finds it useful, and so doesn't do anything with them.  At which point they will decide to move their bitcoins back to more useful dollars, and the price will drop as the currency de-monetises.

On the other hand, if it is useful to be able to perform online transactions, remove the need for banks, debit cards, paypal, and physical wallets; there might yet be a future in Bitcoins.  Let's see shall we?

[AllYourBase]

We need to put together a think tank to figure out where we go once the coins reach close to 21 million. Transaction fees will not work because large sites will form that have a huge bitcoin bank account, and money will be sent between users of the site instead of actual transactions on the bitcoin network. The only solution is to lift the 21 million cap. Remember, the 21 million cap is not set in stone. At the end of the day, the rules are set by miners. We have a handful of mining pools that collectively control over 50% of mining activity easily. Miners have the option to agree on a percentage rate of increase beyond 21 million. If the percent rate is too high, the confidence in the currency goes down and so does its value. If the rate is too low then miners are not paid enough to support the network.

These are just starting points of discussion but weshould get this out in the open sooner instead of later. The mining pools are going to be a powerful force in the monetary policy of the bitcoin network in the future.

Bernanke, is that you?  How'd you get in here?

[dust]

If transaction fees prove to be an insufficient incentive for miners, banks may chose to mine at a loss.  Currently, banks pay for hardware (vaults) and guards to secure their assets. Mining at a loss would be similar to this.

[BombaUcigasa]

Transaction fees will not work because large sites will form that have a huge bitcoin bank account, and money will be sent between users of the site instead of actual transactions on the bitcoin network.

Goood gooood, you are pretty smart man. Except for you know... those large sites and banks will need to have smaller fees than the network. Will they be able to do it? Will there be serveral competitors?

Remember, transactions will be made in the blockchain, because you require trust, this is actually the better money solution, what makes you think that account numbers backed by nothing else but pure database cell rows will be trusted more than peer transactions? And transactions in the blockchain will require taxes, that everyone agrees to, especially those that can supply blocks. They decide what the tax is, and what fees they include on what transactions they bind.

How low can we go? Let's find out! In 2038 something... but keep trying, you almost cracked it man...

[joe]

Bernanke, is that you?  How'd you get in here?

You can call me Ben.

First, let's understand transaction fees.
Transaction fees encourage miners to donate CPU to the network, which is a good thing because the premise behind bitcoin is that if we have a large network then attackers can't control more than 50% of the network. To understand this, assume that there were no transaction fees (and no 50 BTC reward). Then the miners would disappear.

Now that we understand what fees are for, let's explore alternatives. One alternative is inflation beyond 21 million coins. Right now, miners pay themselves 50 new coins per block. Why? Because the original bitcoin client has this allowance built into it, and all bitcoin users agree to accept these new coins. In several years this scheme will have the block reward dry up.

Once the reward gets small and transaction fee income goes down as people get savvy, miners will disappear and the network will be in serious trouble.

Miners will, however, have the option (and they have the option right now, too) to change the 21 million rule, and reward themselves whatever they want for each new block. No change to the official bitcoin client is required, because most mining is done by mining pools which can make the change to their own modified clients.

But wait. Won't miners just reward themselves insane numbers of bitcoins? No, because the general bitcoin public will modify their clients to detect these unfair coins and reject all coins generated on that branch. This places downward pressure on the reward choice. In essence, a market process will take place as individual miners weigh how many BTC they want to reward themselves per block with how much they can get away with without driving bitcoin users (and, therefore, bitcoin miners) to the branch point before the unfair reward block.

[eturnerx]

I'm boycotting any pool that changes the reward structure from that orginally planned. I suspect I'll not be the only one. Deflation is a key part of the bitcoin currency.

[Findeton]

Math is easy. Price/Difficulty rate should double every four years for mining to remain as profitable. That would do the trick, without the need of any transaction fee.

[malditonuke]

Bernanke, is that you?  How'd you get in here?

You can call me Ben.

First, let's understand transaction fees.
Transaction fees encourage miners to donate CPU to the network, which is a good thing because the premise behind bitcoin is that if we have a large network then attackers can't control more than 50% of the network. To understand this, assume that there were no transaction fees (and no 50 BTC reward). Then the miners would disappear.

Now that we understand what fees are for, let's explore alternatives. One alternative is inflation beyond 21 million coins. Right now, miners pay themselves 50 new coins per block. Why? Because the original bitcoin client has this allowance built into it, and all bitcoin users agree to accept these new coins. In several years this scheme will have the block reward dry up.

Once the reward gets small and transaction fee income goes down as people get savvy, miners will disappear and the network will be in serious trouble.

Miners will, however, have the option (and they have the option right now, too) to change the 21 million rule, and reward themselves whatever they want for each new block. No change to the official bitcoin client is required, because most mining is done by mining pools which can make the change to their own modified clients.

But wait. Won't miners just reward themselves insane numbers of bitcoins? No, because the general bitcoin public will modify their clients to detect these unfair coins and reject all coins generated on that branch. This places downward pressure on the reward choice. In essence, a market process will take place as individual miners weigh how many BTC they want to reward themselves per block with how much they can get away with without driving bitcoin users (and, therefore, bitcoin miners) to the branch point before the unfair reward block.

You're saying that miners can more easily increase the block reward than increase transaction fees.  It's the other way around.  People are on board because of the stability of the bitcoin supply.  Take away that, and it's a worse proposition than an already existing inflationary fiat currency.

[AllYourBase]

Miners will, however, have the option (and they have the option right now, too) to change the 21 million rule, and reward themselves whatever they want for each new block. No change to the official bitcoin client is required, because most mining is done by mining pools which can make the change to their own modified clients.

Bitcoin is the only deflationary currency I'm aware of, and is a central part of its success.  If you want inflation, use any of the several hundred government currencies available.  Or fork the block chain and make one that competes.

[Phenomenon]

Once the reward gets small and transaction fee income goes down as people get savvy, miners will disappear and the network will be in serious trouble.

As the BTC economy grows, the number of transactions and thus the amount collected in transaction fees will grow.  Assuming BTC is still around and going strong, by several years from now I would expect the amount of transaction fees per block to be at or over the 50 BTC level.  Just think if any significant portion of online transactions was done in BTC, you are talking about millions of transactions per day, and at even 1 million transactions per day and a .01 fee per transaction, you're talking around 70 btc in transaction fees per block on average.

[caveden]

No change to the official bitcoin client is required, because most mining is done by mining pools which can make the change to their own modified clients.

You need to change every full-client implementation, not only miners.