Bitcoin Forum

Hey guys,

Why does the number of bitcoins in circulation follow a log-curve, and not linear (or perhaps of another shape)? What are the advantages of this approach?

My understanding is that is to create a diminishing return like that found in a resource like gold. That is to say, there is always more gold to be mined, but it gets harder to produce an ounce every year. That keeps upward pressure on the price.

It's a log-curve because the block reward that miners get for solving a block halves every 4 years.

If you want to understand why this is the case I would recommend reading the original whitepaper.

It's a natural result of log math (when you find something, there are less left to find, so it gets harder)

but I always suspect it's to get the technology to evolve to a point where it can the most efficient it can be.

If it was a straight line we'd all still be doing it with a PC like Litecoin.

Litecoin has block reward halving every 4 years as well, but not even FPGA yet. I believe to get the technology to evolve it must be profitable, which Litecoin is not, so no FPGAs yet

Litecoin uses the Script algorithm, which makes heavy use of memory, to keep it on PCs and stop people gaining a mining advantage by using FPGAs/ASICS. They compensated for this by making the blocks easier to solve which strangely makes a 51% attack less likely.

It is due to the manner in which new BTC are minted. Originally, block rewards were meant to 1) introduce and distribute the coin, and 2) reward the miners for processing transactions in creating the blockchain - important to incentivize mining when value was next to nonexistent in the beginning.

As RodeoX says, by mimicking a natural and finite resource like gold, it provides a larger workable base of coins much sooner for a developing market than a linear distribution, keeps upward pressure on value via scarcity, and rewards early adopters.

The value of each crypto coin will ultimately be determined by the number of transaction we see in the system so it's important to get stores using them to get them in circulation.

Eventually the only reward will be the transaction fee, which currently many mining pools do not share with their miners which sux!
https://en.bitcoin.it/wiki/Comparison_of_mining_pools

Yes, it is possible to make ASICs with big relative fast memory included, to outperform GPUs in power consuptions, but who is going to spend up to 1 mil USD fr development when the Litecoin mining is not much profitable? Ony preorder based investing could do it, but there is little profit to be made

That was the point, to make it so expensive, coin production couldn't be concentrated into the hands of the few.

Its already this way and will continue. Bitcoin is not better than other alt-coins, it is just most used, best merchant support, biggest development team and highest hashing thus most secure. That is what gives Bitcoin such value

Maybe because it was predefined like this due to wanting to be similar to supplies like gold I think?

I believe that in time, we'll see considerably larger rewards exclusively from transaction fees than we have ever seen from new coin in block rewards. Within the last few weeks, I had seen a block reward at Slush's pool that included between 6-8 BTC in transaction fees alone for a 30+ BTC reward. I've forgotten the exact figure now... In any case, BTC mining will be spawning an entirely new professional industry - only instead of being called "miners", we'll call them payment processors - and we are currently in the early stages of such evolution.

The value of BTC, or any currency for that matter, is not so much based upon the number of transactions being processed, but rather supply and demand.
Group A doesn't want to spend more than X-1 < X (BTC value) < Group B won't sell for less than X+1

This is for the best.

I believe bitcoin supply curve is halving every 4 years as an experiment. It could be 1 year, 10 years, only -25% or whatever else. No one know what would be best, so lets hope the initial Satoshi decision was ok  :)

It should be. The less, the better.

But so far it doesnt look like transaction fees could be enought. We will see...

Thanks everyone,
so I see that the general consensus is that Bitcoin mimics a precious mineral. From your explanations it does seem natural, but why is it better than a linear distribution?

With a linear distribution the demand/supply ratio would still grow as Bitcoin would gain popularity (and thus difficulty would increase), so the price would still go up and bitcoins would remain scarce and difficult to obtain. Early adopters would still be decently rewarded. And since bitcoins are divisible, I don't see a problem with whether there a a lot / a few coins already in the circulation.

But a log-distribution benefits the early adopters even more and makes it less and less probable with time that a newcomer will mine, thus compromising the decentralization, doesn't it? I mean, this probability would already be dropping with the difficulty rising, and they half it every 3 years on top of that...

As I was typing this, I thought that perhaps this is designed to soften the transition from free transactions to necessary transaction fees? Perhaps they were afraid that an abrupt change would shatter the network in some way? (though not a good enough for me reason still...)

What do you think?

If you have unlimited supply, Bitcoin would be priced at 0.01$ and not 1000$....

Sorry, I implied that we cut off the supply at the same time (~2140) with the same eventual amount of bitcoins.

The block reward needs to be a sufficient incentive to provide enough proof of work to keep the Bitcoin system secure until transaction fees have enough value to take over. If the block reward was too high, the price would keep dropping and everyone would be playing "hot potato" with their Bitcoins. If the block reward was too low, the network would be insecure and unreliable.

My guess is that Satoshi underestimated how successful Bitcoin would be and as a result there is much more mining than there needs to be.

If due to this by-design artificial scarcity the dollar price of BTC increases 10,000x, then by definition the value of us dollars has decreased by 1/10,000. If BTC is to be the only decentralized crypto-currency and be used by everyone, then the net worth of the broad population must become devalued by 1/10,000.

After only a small percentage of the broad population has purchased Bitcoin, the price will be several doublings higher than it is now and the net worth of rest of the broad population will be devalued by the reciprocal of several doublings. Five doublings is 32x. Bitcoin has already risen from pennies to $350, which is roughly 15 doublings.

So logically one can only conclude mathematically that Bitcoin's supply curve was either chosen to create a war between early adopters and the broad population of the world, or it is to motivate the creation of alternative crypto-currencies.

Bitcoin is the antithesis of the moolaching coin of the people.

While the value of dollars relative to Bitcoin will drop, the value of the dollars themselves won't change very much. For example, dollars will still buy about the same amount of gold, the same number of apples, the same number of Euros, and so on.

Bitcoin can only capture value it creates. Alternative currencies will have to create value in order to capture it.

In computing, having an undetermined machine state is bad. If there wasn't a limited amount of Bitcoin that will be produced, the size of the variable holding it would eventually overflow, memories would overflow, etc. A compiler can see that if mining were to continue forever there would be more bitcoins than molecules in the universe, and optimize out the indeterminate state by removing mining entirely. It is easier to program without bugs when there is a maximum number of Bitcoin base units, we know how many digits need to be shown in the user interface and such.

As for why this particular curve? It's better than 50 BTC for 8 years and then no rewards after. In hindsight, the halving times should have been longer so that half the currency wasn't already mined before it really became popular, but you could not have anticipated the adoption rate of the currency - it could still be a few cryptographer's experiment, or it could have gone gangbusters the first year.