[Bitcoin-ml] Proposal for a future blockspace/fee market discovery mechanism

[Gal Buki]

Hi

I was thinking about the issue of the current fee/blockspace market and
how it could be improved in the future given that blocks remain not full.



1. Current situation
Miners have costs that are mainly subsidized by the block reward.
As the block reward decreases every ~4 years and the exchange rate might
not increase at the same pace as the difficulty miners might need
transaction fees to stay profitable.

With full blocks the fee was defined by a blind auction which resulted in
high fees and uncertainty when a transaction was to be included in a block
(if ever).
With non-full blocks we can have 0-fees and there should always be space
in the block but miners might not want to included low fee transaction
because they might need the fees to stay profitable.


2. Proposal
Miners could publish a price for their service. I will call it "blocks to
first confirmation" (bfc) hereafter.

A bfc value of 10satoshi/B signals the price for a guaranteed inclusion in
the next block.
Every miner/pool could define its own bfc value and undercut others. A
market would evolve where the economic equilibrium could be found.

A miner/pool with a higher hashrate could define a higher bcf as the
chance of inclusion in the next block is bigger and thus his service would
have more value compared to other miners/pools.
User who are not in a hurry could opt for a lower fee and get included later.

Assuming we have the following 3 miners/pools on the network.

|------------+----------+-------+
| Miner/Pool | Haspower | bcf   |
+------------+----------+-------+
| A          | 25%      | 10s/B |
| B          | 35%      | 15s/B |
| C          | 40%      | 20s/B |
+------------+----------+-------|

A user who wants his transaction to be included in the next block could
pay 20s/B.
A user who does not care about confirmation time could pay 15s/B and still
be included with a 60% chance in the next block.

A miner is of course free to include any tx, even if the fee is below his
bcf value. Although if he does this in the first block he undermines his
own price thus he might be incentivized to artificially include such
transactions only in later blocks.


3. additional complexity
Miners could have multiple bcf values to signal how much would be needed
for immediate inclusion and how much for inclusion after x of their
blocks.

This could be compared to the postal service where you can chose regular,
express or snail mail.

This could look like this

|------------+----------+-------+---------+----------+
| Miner/Pool | Haspower | bcf   | bcf + 5 | bcf + 10 |
+------------+----------+-------+---------+----------|
| A          | 25%      | 10s/B | 5s/B    | 0s/B     |
| B          | 35%      | 15s/B | 5s/B    | 0s/B     |
| C          | 40%      | 20s/B | 5s/B    | 0s/B     |
+------------+----------+-------|---------+----------+

In the above example a fee of 5s/B would guarantee an inclusion within the
next 5 blocks.
And a 0 fee transaction an inclusion within 10 blocks.


5. Users (everyone sending a tx)

Users can either take the published fee prices as a given or create
transactions with fees below them.
If many user do this the sum of all the fees could become attractive to
the miners and thus they would be included at a certain block.



I'm looking forward receiving your comments.


--
Gal Buki