The fee market is one of the most hotly debated components to Bitcoins design as questions remain whether it will be sufficient to sustain security after the block subsidy is exhausted. This is likely to become an issue sooner rather than latert as after the next two halvings, we will be down to 1.5625 BTC per block which is a further 75% reduction from today.

There is a popular quote from Satoshi in response t

o this problem:

Satoshi was a smart dude, and he is probably right. Either Bitcoin breaks escape velocity, or it doesn’t.

Ethereum currently also utilises Proof-of-Work consensus and has a similar fee market design. In fact, Ethereum transaction fees on a daily basis are becoming quite competitive with Bitcoins as shown in the daily USD fees paid chart below. This is bullish for both coins.

Despite having crossed the hurdle of developing a strong fee market, a feat that BTC and ETH are the only two projects to achieve, the Ethereum project is now looking to change the dynamics of the fee market via a proposal called EIP 1559. Lets look at why they are proposing a change and what we can expect on the other side if it gets implemented.

This aims to distil some of the content written in this blog post which is a good reference for anyone looking deeper into EIP1559. https://insights.deribit.com/market-research/analysis-of-eip-1559/

The Current Fee Market

As many of you have noticed, Ethereum fees are very high right now with simple smart contract transactions costing a few dollars. This is due to high demand from stable-coins, in particular Tether. In order to get your Transaction mined first, you need to place a higher bid in terms of gas price than everyone else. This auction mechanism makes sense but also is not a great user experience and quickly prices out smaller holders (aka us.).

This is what is otherwise known as a fee auction, highest bidder wins.

Given demand for block-space is an unknown variable, and changes dynamically, this means no fee stability can be achieved. Users and miners both have no idea what the fee will be tomorrow or next year, so how do we budget and factor this into our living and operating budgets and miner incomes?

For both BTC and ETH, the main reason for fee spikes (outside demand) is that block-space is limited. Bitcoin retains 1MB blocks and Ethereum miners set a Gas Cap of 20M which makes getting into a block competitive (only so many transactions can fit in the allocated space). Whilst this promotes a healthy fee market, it clearly has issues with usability and user experience.

Rethinking Transaction Fees

The EIP1559 proposal aims to break down the fee paid by users into two pieces, the BASEFEE and the TIP.

Everybody pays the same BASEFEE which is burned rather than paid to miners. This BASEFEE is dynamic and moves up and down depending on congestion in the chain, targeting a specified block size on average.

The TIP is an extra priority fee, similar to the gas price today, where it enables the same auction mechansm for inclusion. The TIP is the only part of the fee that goes to miners/validators.

Now the major change is in the definition of a FEECAP = BASEFEE + TIP.

By defining a FEECAP you specify the absolute maximum you are willing to pay for your transaction to be mined. As the BASEFEE moves around as set by the protocol with congestion, your transaction will sit in the mempool until congestion drops enough that your total fee gets to the front of the queue.

Now this sounds kind of like the same fee auction market we have today. Where EIP1559 proposes a solution is in flexible block sizes. Miners can change the block size temporarily to deal with large transaction demand and congestion.

Fundamentally, large blocks have two negative impacts:

Blocks take longer to propagate over the network which means miners centralise around those with the most advanced communication channels that can handle large blocks downloads. This aims to avoid wasting PoW on an orphaned chain due to network lag.

Blockchain size increases, requiring data-centres to run which also leads to miner and node centralisation. Few people are willing to run 100TB SSD high performance computers for no reward.

If the block size (the amount of bytes and hence transactions per block) was to increase infinitely, it would centralise both nodes and miners. What EIP1559 does is it allows miners to increase block size temporarily, only when there is provable demand to do so. An example would be major events like the Black Thursday (Mar-12) selloff that liquidated $3M worth of ETH held in MakerDAO.

The incentive for miners to keep blocks small is that since the BASEFEE is burned, they do not get paid much of a TIP. Remember, the user specified FEECAP which contains BASEFEE and TIP. If miners increase blocks for long periods of time, BASEFEE increases consuming the fee that would otherwise be paid to miners as a TIP.

When blocks are small, the TIP is a larger portion of the FEECAP and people can compete to get transactions mined via the fee auction.

Think about the BASEFEE as equivalent to Difficulty (PoW) or ticket price (DCR-PoS) where it adjusts in order to maintain some average network parameter. The protocol will adjust the BASEFEE in order to target the desired block size on average.

Tomorrow we’ll finish up this topic.