What could Bitcoin do if the fees are not high enough to provide security?
This is a rough draft in progress…
The Security Problem
In the future, block subsidy goes down. The hope is that fees will be high enough to make a 51% attack impractical and very expensive, but what if this does not happen? What if fees cannot buy enough security?
Bitcoin’s Credibility
Currently relying on inflation. Bitcoin cannot break the promise of 21 million maximum without destroying it’s credibility and value proposition. Demurrage seems like a cheat and much of the community would not go along. Questionable morality about imposing demurrage mid-flight.
Optional Security, for a price
What about an merge-mined side-chain that continues paying a block subsidy via inflation? The side-chain would have it’s own tokens, call those Tokens. Ever block on the side-chain would issue Tokens to the miners, subsidizing the hash rate. The exchange rate in and out would be pegged to the ratio between the quantity of Bitcoin and the quantity of Tokens.
The quantity of Tokens issued with each side-chain block could be set a few different ways:
- Set it to a constant inflation rate. (e.g. Tokens issued per year = 0.5% of existing tokens.)
- Vote on the issuance rate. Every transaction could set a value indicating that the inflation rate should be Higher, Lower or No Change and be weighted by days destroyed (e.g. Tokens in the utxo * (block height at destruction - block height at creation)) Voters have an incentive to maximize the protection of their transactions by encouraging people to peg in (keep inflation low) and by paying miners as much as possible (keep inflation high). They also have an incentive to pay as little inflation as necessary (keep inflation low.) Some aggregated opinion would likely keep inflation at a good level. If people are unhappy with the inflation level they are of course free to peg out to the main chain.
- Market based issuance adjustment. Over the last 2 weeks, did the number of Tokens (not counting issued Tokens) grow? Raise the issuance rate! Otherwise, lower the issuance rate.
Discourage free-riding
Does merge mining ruin the incentive to join the side-chain? After all, if it’s merge mined, the original chain benefits from the hashrate of the side-chain, sort of. How does this work with Namecoin? Could we have a situation where the side-chain is only allowed to merge-mine with the main chain every, say, 100 side-chain blocks? The side-chain (which pays for most of the hashrate) would then have the final word on which candidate main chain to extend. In this way, the full benefits of stability of being on the side-chain are withheld from those who stay on the main-chain, and they cannot get the same level of stability on the main-chain that they can get on the side-chain. Nevertheless, the main-chain is reinforced at least once per day with an endorsement from the side-chain miners.
This is good for Bitcoin?
Is this scheme more harm than good? The downside is that it coaxes transaction fees off of the main chain onto the side-chain, reducing the ability of the main chain to pay for hashrate even farther. Is this scheme incentive-compatible and inevitable, as individuals defect from the main chain in search of more stable settlement of lightning channels?
