What happens when Bitcoin’s block subsidy runs out?

Miners are an essential component of the Bitcoin blockchain — they package transactions into blocks and publish them, making the included transactions valid. Without miners there are no new blocks, no transactions are processed, no bitcoin can be transferred, and hence the system becomes pointless.

To have a shot at publishing a valid block of transactions, you have to throw a lot of specialized computing hardware at what amounts to a lottery — the more hardware you have, and the more efficiently it runs, the higher the odds are that your mining outfit will meet the conditions for publishing the next valid block.

No one is going to do this on a significant scale for free, hence Bitcoin includes an incentive called the “block subsidy”, which gives miners who find a valid block the right to mint a bit of bitcoin for themselves. But this subsidy halves every 210,000 blocks (approximately every four years).

It started at 50 bitcoin, and is now at 3.125 bitcoin. Fortunately for Bitcoin, as the subsidy has gone down, the price of bitcoin in dollars (which are still required to buy hardware and pay electricity bills) has gone up.

Eventually the Bitcoin blockchain subsidy is going to be worth so little that miners will need some other incentive for continuing to throw hardware and energy at Bitcoin to keep it running.

At the moment, the only argument I have seen is that, in the future, transaction fees will form that incentive. But let’s back up, and look at how Bitcoin has managed to incentivize miners so far. It is always good to start with some definitions: what is a block subsidy, what is a miner, and how has Bitcoin mining developed over time?

A brief history of Bitcoin mining

Satoshi Nakamoto had a lot of problems to solve for Bitcoin, the main one being the prevention of double-spending in a decentralized manner without a central authority keeping track of every spend. Two other problems were:

• How to distribute the bitcoin, and
• How to encourage people to secure, maintain, and grow the network

These two distinct problems were solved using the same mechanism — reward people in bitcoin for running the miners that make the whole system tick (literally). This is the block subsidy. By reducing the reward by approximately half every four years or so, this additionally ensures that there is an upper cap to the total amount of bitcoin that can be mined.

Geek power

In the very early days of Bitcoin, people ran miners because they thought it would be interesting to do so. There was no immediate profit, because bitcoin didn’t have much if any of a dollar value, and certainly not enough to cover the time, effort, and electricity bill to do so. There is even evidence that in the very early days, Satoshi Nakamoto had by far the most mining power, but throttled it to give other geeks a chance to mine blocks.

Mining for profit

Then things changed. With enough interest and the arrival of Bitcoin exchanges, the fledgling cryptocurrency achieved a market price, and the amount of hardware and electricity thrown at mining it became approximately proportional to the costs versus the dollar gains of doing so. People even put time and effort into inventing more and more efficient Bitcoin miners, first with GPU software, then with FPGAs, and then with custom ASIC designs, which are still being worked on and improved to this day.

Since about 2013, Bitcoin mining has been a commercial industry, and has continued to expand in that capacity.

Mining profits explained

Mining bitcoin makes money. It is a gamble in the short term, but to this day, in the long term it has generally been profitable.

At the moment the Bitcoin block subsidy is 3.125 bitcoin, has been since April 19th of this year, and will continue to be so for the next three years or so. That means that when a miner finds a valid block (which happens about every 10 minutes on average), that miner gets about $300,000 worth of bitcoin at today’s prices.

This makes buying mining hardware and paying the electricity bill to run it worthwhile over the medium term. Over the long term the expectation that the price of bitcoin will keep rising means most of the miners keep running even if they’re making a loss in the short term.

As a side-effect, all these Bitcoin miners ensure that transactions are processed and the network stays secure. But every four years, the block subsidy halves (or rather, is divided by two and rounded down to the nearest eighth decimal place), which means eventually the block subsidy will be zero.

Miners also make a profit from the transaction fees that are paid by people submitting transactions. And unlike standard banking fees for sending money somewhere, the transaction fee is not dependent on the amount of bitcoin that you transfer. Whether it’s 1 satoshi, or 100,000 bitcoin, the transfer fee is determined by market forces — if lots of people are submitting transactions the fee goes up, and if hardly anyone is moving bitcoin around, the fee is low.

As I write this, the Bitcoin transaction fee converts to about $3.50 in satoshis (a satoshi is 0.00000001 bitcoin). Back in April this year it briefly peaked at $128, but usually the cost per transaction is a few dollars.

The thing is, you can only fit about 3000 to 4000 transactions in a Bitcoin block, and so, being generous, at the moment miners only make an extra $14,000 from the transaction fees per block. We can make an assumption that that’s the market price, and will be for the foreseeable future.

What is clear, is that the subsidy is the main reason for mining activity today, and fees account for 5% of the gains. Cancel the subsidy, and most miners will stop mining.

Where to next?

As the chart in the first section shows, the increase in Bitcoin’s dollar price has more than made up for the halving, encouraging ever more and more hash power to be deployed onto the system. When Bitcoin first started, the miners were computing 10 million hashes per second. Now that number is 634 billion billion hashes per second.

There is bound to be a turning point at some time in the future though. Just to make it economically viable to maintain the power of the network requires the price of Bitcoin to double every four years. The various governments and central banks of the most prevalent currencies have been extremely helpful with their “print now, pay later” monetary policies over the last decade, but at some point economic realities will bite Bitcoin, and the returns for running miners, processing transactions, and securing the network will flatten, then drop precipitously from some fateful halving onward.

Exponential price growth forever?

Looking forward to the end of the century, in the figure below I provide some initial hypothetical values for the Bitcoin price after each halving, just as a thought experiment (note: this is NOT a prediction!).

The hypothesis is “what if the price of Bitcoin doubled every halving from now on, as well as the average transaction price?”:

A Bitcoin price of $34 billion in 2100 would mean that the market capitalization of Bitcoin would be $700,000 trillion dollars, which is 50% more than the total estimated wealth of the entire planet. This seems infeasible.

Our first conclusion: ever increasing value of Bitcoin in dollars is not going to sustain Bitcoin very far — perhaps to some point in the 2040s at best. We moved from “geek enthusiasm” to “block subsidy profits” as a motivator for mining, but at some point a new incentivization model will have to apply if Bitcoin is to survive.

Transaction fees to the rescue?

There are some convincing arguments that present a problem that could occur when the block subsidy in Bitcoin is regularly less than the transaction fees for that block, based on recursive reasoning logic (“we know that they know that we know…”) which is currently used by miners to determine which chain to mine if there are multiple choices: each miner mines the chain that they think all the other miners are going to mine too. That currently means they all mine the longest, strongest chain, because it provides the most rewards. Why this might break down when block subsidies are too low is too complicated to go into right now, so I will save it for a future article, because it is rather interesting.

Instead, let’s look at the problem with transaction fee incentivization purely from an immediate financial reward perspective.

Firstly, increasing the number of transactions that can be put in a block won’t help. If people are willing to collectively put forward $15,000 in transaction fees per block, increasing the number of transactions that can go into a block will simply proportionally decrease the cost-per-transaction. The total block transaction fee reward should remain approximately the same.

Secondly, if transaction fees were to rise to hundreds or thousands of dollars, then Bitcoin would only be useful for moving millions or billions of dollars of value around. So much for a peer-to-peer electronic cash system.

Our second conclusion: even if transaction fees were to rise proportionally with the price of Bitcoin, they are never able to keep up with block subsidies as an incentivizer.

Thoughts for the future

Given that block subsidies are eventually unsustainable, and transaction fees fall far short as a replacement for the block subsidy incentive, we have to look for other means to maintain blockchain integrity.

Is it therefore at all possible for the Bitcoin network to be secured by miners that are not in the game for the direct profit that mining used to provide?

Noting that making predictions is hard, especially predictions about the future, here are some possible scenarios I thought up in the sauna last night, that could save the day:

1.Significant amounts of money have been poured into Bitcoin and supporting infrastructure by companies and exchanges, resulting in profitable businesses. It thus makes sense for them to run their own mining rigs at a loss to ensure the independence of the network, picking up the slack as direct mining profits decrease. They have a derived incentive to ensure the underlying asset of their marketplace continues to exist.

2.Bitcoin mining moves from being an arms race between independent mining pools to becoming a multi-lateral détente between nations. If a significant number of countries switch some of their reserves to Bitcoin, it becomes in their interest to ensure the network survives. They may therefore run national Bitcoin mining farms to ensure their stored wealth remains extant and usable.

3.Bitcoin mining develops into an international lottery.

Block rewards may eventually be insufficient to encourage mining businesses, but as an individual windfall they are sizeable up to about 2072. With 52,560 blocks being generated every year, assuming a population of 10 billion, an individual has a 1 in 200,000 chance of finding a block if the mining population consisted solely of people running low-power miners in the background on their phones, for example.

Of course, this would require wide-spread adoption of Bitcoin by the world population in general, an appreciation of how to keep it running, and pressure on mobile device manufacturers to include a low-power mining ASIC chip in each phone they make, but with 15 billion mobile devices currently running (or mostly sitting idle) on the planet there’s room for a swarm of personal miners taking over responsibility for the network.

4.Satoshi Nakamoto unexpectedly returns, with the following explanation for their absence: “Sorry for the radio silence. I’ve spent the last three decades meditating in a cave in the depths of Peru, where I have had a break-through working on a new global reserve currency system that is entirely peer-to-peer, without a trusted third party validator. The paper is available at …”

And everything is right again.