Ethereum and Bitcoin compared

Bitcoin and Ethereum both run on a blockchain, yet they solve different jobs. The Bitcoin whitepaper of 31 October 2008 is titled “Bitcoin: A Peer-to-Peer Electronic Cash System”, and the first block was created on 3 January 2009. Ethereum started on 30 July 2015 at 15:26 UTC with block 0 and sees itself as a platform on which programs run. Almost every other difference follows from that difference in purpose.

Two networks, two jobs

A smart contract, according to the Ethereum developer documentation, is “a program that runs on the Ethereum blockchain”, meaning a collection of code and data at a fixed address. It is usually written in Solidity or Vyper. Published contracts cannot be deleted by default, and calls to them cannot be reversed. The code is open to inspection. Stablecoins, exchange protocols and staking contracts are all built on this basis.

Bitcoin pursues a narrower goal. The whitepaper describes payments that go “directly from one party to another” without a financial institution in between. Anyone sending bitcoin needs confirmations, not contract logic. According to the project site a new block is added roughly every 10 minutes on average; when the first confirmation arrives is a matter of chance, and there is no guaranteed upper or lower bound. Offer too small a fee and you wait longer. How blocks, hashes and confirmations fit together is set out in our piece on how a blockchain works.

Proof of work against proof of stake

Bitcoin secures its chain with computing work. The whitepaper calls the method “one-CPU-one-vote”: the longest chain is the one with the most work behind it. Miners search for a value whose SHA-256 hash begins with a set number of zeros. On average this produces a block every 10 minutes, and the difficulty adjusts continuously.

Ethereum dropped that method on 15 September 2022 at 06:42:42 UTC at block 15,537,394. Since then it is not computing power that counts but capital put up as a deposit. The project site puts the fall in the network's electricity use at an estimated 99.95 per cent. Time now runs in units of 12 seconds, called slots, and 32 slots make an epoch. In each slot a randomly chosen validator proposes a block.

The arithmetic of an attack changes too. On Bitcoin an attacker needs the majority of computing power. On Ethereum it would take the majority of the ether staked, and anyone wanting to reverse a block that is already finalised would, according to the documentation, have to put up and burn more than one third of all staked ether.

A hard cap against open issuance

Bitcoin has a hard limit: there will never be more than 21 million units. The block reward halves every 210,000 blocks, which is roughly every four years. Since the fourth halving on 20 April 2024 at block 840,000 it has stood at 3.125 BTC. One unit divides into 8 decimal places, so 1 bitcoin equals 100,000,000 satoshi.

Ethereum has no such limit. Before the switch, according to the project site, around 4,930,000 ETH a year were issued on the execution layer and about 620,500 ETH a year on the consensus layer, together about 4.61 per cent a year. The execution layer share then fell away. For the remainder the project site works with an assumed 14 million ETH staked and arrives at about 0.52 per cent a year, a drop in issuance of roughly 88.7 per cent. The 0.52 per cent is therefore the result of that assumption and not a measured value for today; the project site says itself that issuance moves with the amount staked.

On top of that comes burning. Since the London upgrade on 5 August 2021 at block 12,965,000, the base fee of every transaction is destroyed instead of paid out. The project site sets out a worked example: with issuance of about 1,700 ETH a day and an assumed target size of 15 million gas per block, an average gas price of 16 gwei would arithmetically offset that day's issuance. Both inputs are assumptions of that calculation, not measurements from today. The supply can therefore grow or shrink depending on use. That is not a promise.

What gas is and why fees move

Gas measures the computing effort of an action on Ethereum. A simple ETH transfer costs 21,000 gas. Charges are settled in gwei, one billionth of an ETH, that is 0.000000001 ETH. The total fee is the gas used multiplied by the sum of base fee and tip. The protocol sets the base fee itself and changes it by at most 12.5 per cent per block, depending on how full the previous block was. Each block aims at half the applicable gas limit and may grow to twice that target, that is up to the limit itself. No fixed figure sits behind this: the limit moves through validator signalling and through network upgrades.

One point matters for expectations: the move to proof of stake was, in the words of the project site, “a change of consensus mechanism, not an expansion of capacity”, and was never meant to lower fees. Costs do fall on second-layer networks, known as rollups, which post their data to Ethereum and borrow its security. On Bitcoin users pay a fee per unit of data, with no base fee and no burning.

Ten features side by side

Staking, validators and the second layer

Anyone who wants to validate on Ethereum themselves puts up 32 ETH. With the right withdrawal credentials a single validator can carry up to 2,048 ETH. As a rule of thumb the project site names an NVMe SSD with 4 TB, 64 GB of memory and a line with about 50 Mbit/s downstream. The network recognises a deposit in roughly 13 minutes, after which a queue runs until activation.

The risk is called slashing. On a double proposal or contradictory votes, a 32 ETH validator immediately has 0.0078125 ETH burned, proportionately more with a larger stake, after which a 36 day removal period begins during which the stake keeps melting away. On day 18 an additional penalty follows, its size depending on how many other validators were penalised within the same 36 days. In the worst case the entire stake is gone. Withdrawals have only been possible since the Shapella upgrade on 12 April 2023 at block 17,034,870. What this means for returns and tax is set out in our piece on staking, yield and risk; you can run the compounding arithmetic yourself in the staking compound calculator.

Bitcoin has no staking. Anyone wanting to earn there runs mining hardware and competes for the block reward, currently 3.125 BTC plus fees.

What to settle before your first purchase

Start by working out which of the two networks fits your question. Anyone looking for a scarce, unchangeable money supply is talking about Bitcoin. Anyone meaning programs, stablecoins or staking is talking about Ethereum. Both are separate assets with their own risk of total loss, not complements to each other.

• Decide whether you hold the units yourself or leave them with a provider. If a custodian becomes insolvent, everything hangs on how cleanly client holdings were separated.
• Expect fees in two layers: once at the trading venue, once in the network itself. On Ethereum the base fee comes on top, and it does not flow back to you.
• On Ethereum, check whether you really want to tie up 32 ETH. Smaller amounts run through pools, and therefore through a third party with its own risk of failure.
• Record the purchase date, quantity and price for every tranche. For the holding period each acquisition counts separately.
• Look up unfamiliar terms in the glossary before you trigger a transaction. Mistakes in the network cannot be undone.

How two such different assets behave in one portfolio is something you can work through with the portfolio mixer. It shows weights and swings, it gives no recommendation.

Frequently asked questions

Sources