Explain the key differences between PoS and PoW economic models: mining, staking and decentralization

24-07-16 12:34

Read this article in 20 Minutes

AI summary

View the summary

Original title: "What are we talking about when we talk about economic models"

Original source: Xiaozhu Web3

Let's start with the Bitcoin shutdown price

Recently, as Mt. Gox began to pay Bitcoin and the German government frequently sold Bitcoin, the price of Bitcoin once fell below $54,000 (now it has risen back above $60,000), hitting the "shutdown price" of some Bitcoin mining machines.

According to a survey agency, if Bitcoin reaches 54,000, only ASIC mining machines with an efficiency of more than 23W/T can make a profit, and only 5 models of mining machines can support it. This means that if the price of Bitcoin falls below the shutdown price, some miners with smaller risk resistance will seek to exit and stop losses. When these miners quit, they often sell Bitcoin for cash and sell mining machines at a lower price, causing the price of Bitcoin to fall further. This phenomenon is called Miner Capitulation.

The so-called shutdown price is actually the cost price of Bitcoin mining machines. So how is this cost price calculated? In order to answer this question, we must first understand Bitcoin's economic model and PoW mechanism.

Bitcoin is pre-programmed with a total supply of 21 million, and a block is mined approximately every 10 minutes, rewarding miners with a number of Bitcoins. The number of rewards was 50 Bitcoins per block at the beginning of Bitcoin, and then the reward will be halved every 210,000 blocks (approximately every four years). The most recent halving event occurred on April 23, 2024. The block height was 840,000 when the halving occurred, and the reward dropped to 3.125 Bitcoins per block. In addition to block rewards, miners also charge a fee for packaging transactions, which is generally between 0.0001 and 0.0005 bitcoins per transaction. The fee is regulated by the market. The more users use Bitcoin to transfer money, the busier the miners will be. If the transaction fee is set too low, the transaction will be ignored by the miners.

When transactions occur in the Bitcoin network, these transactions are placed in the memory pool (mempool). Then, the miner selects a group of transactions from the memory pool and tries to form a new block. To do this, the miner needs to find a specific value in the random number and combine this specific value with the block data to generate a hash value that meets the network difficulty target. This process is "mining". Whoever calculates the qualified hash value first obtains the right to record the account, that is, the mining is successful. The difficulty target is a dynamic value that is adjusted every 2016 blocks (about every two weeks) to keep the average block time of Bitcoin at about 10 minutes. Therefore, the greater the computing power of the entire network, the greater the difficulty target will be.

The computing power mentioned above is the mining capacity of Bitcoin mining machines, that is, how many hash collisions can be made per second. At present, the unit of computing power is generally TH/s, that is, 10^12 hashes per second. The computing power of the entire network is about 630 EH/s, that is, 6.3*10^20 hashes per second. Therefore, each T computing power can theoretically mine 8*10^(-7) Bitcoins per day. For miners, in addition to the purchase of mining machines and the operation and management costs of mining farms, the other expenses are mainly the electricity costs for mining. Taking the Ant S19 pro mining machine as an example, the rated computing power is 110 T and the rated power consumption is 3250 W. It can be inferred that the daily power consumption of each T computing power is 0.709 kW. The electricity costs in different countries and regions vary greatly. According to 0.055 u/kw, the cost of one Bitcoin is about 50,000 US dollars. The figure below is the Bitcoin mining data of F2Pool, which is basically consistent with the author's estimate.

The above premise is based on the total network computing power of 630 EH/s. Once "miners surrender", the total network computing power will decrease, and the cost of mining a bitcoin will also decrease. Similarly, if the price of Bitcoin rises, miners will be profitable, the total network computing power will increase, and the cost of mining a Bitcoin will also increase.

Therefore, the "shutdown price" of Bitcoin is actually the result of market regulation and miners' game, and all of this is based on Bitcoin's simple and effective economic model.

Economic model under PoS

In the economic model of PoW public chains represented by Bitcoin, miners are the most important participants, but in PoS public chains (such as Ethereum and Solana), there is no role of miners, so what is their economic model like?

First of all, we need to know that the biggest difference between the PoS mechanism and the PoW mechanism is that under PoS, there is an access mechanism for nodes participating in consensus blocks, which is usually implemented by staking. In this mechanism, nodes need to stake a certain number of platform tokens to be eligible to participate in network consensus; at the same time, the platform will issue platform tokens to these nodes as block rewards to encourage them to contribute to network stability. Under PoS, nodes that participate in network consensus through staking are generally called validators.

Secondly, if the platform tokens are issued indefinitely (such as Ethereum and Solana), the inflation of the platform tokens also needs to be considered. The issuance of platform tokens is usually achieved by validator block rewards, and the destruction is generally carried out in the form of transaction fees for liquidity recovery, such as recovering them to the project party’s treasury, burning them within the protocol, etc. There needs to be a balance between issuance and recovery, allowing inflation or deflation for some time, but not long-term inflation or deflation to maintain economic stability.

Finally, there is the function of platform tokens. Unlike Bitcoin, which can only be used as transaction fees, PoS platform tokens have the function of generating interest due to the existence of staked block rewards. Therefore, some platforms will also have a design of delegated staking, which can reduce the circulation of platform tokens in the market and maintain economic stability. The liquidity staking we often talk about is usually a third-party protocol designed based on delegated staking, and the APR comes from the staked block rewards (and MEV).

Ethereum

The initial supply of the Ethereum network was 72 million, of which 60 million were allocated to people who purchased ETH in the crowdfunding activities held in July and August 2014 (the average selling price was about $0.3 per coin), and the remaining 12 million were half distributed to 83 people who contributed to the protocol early when the network was launched in 2015, and the other half was left to the Ethereum Foundation. The current total supply of the Ethereum network is about 120 million.

In September 2022, Ethereum switched from PoW to PoS (The Merge) and launched the beacon chain. The inflation design of the Ethereum network is divided into two stages based on this boundary: before Ethereum switched to PoS, about 4.84 million ETH were issued each year, with an inflation rate of about 4%; after Ethereum switched to PoS, about 3.01 million ETH were issued each year, with an inflation rate of about 2.5%. In fact, since Ethereum switched to POS, since EIP-1559 stipulates that each transaction will burn a part of ETH as the basic fee of the network, ETH has been deflationary most of the time, with an average deflation rate of 1.4%.

In the Ethereum network, if a node wants to become a validator of the beacon chain, it needs to stake 32 ETH. More than 32 ETH will not increase the weight of the validator on the network. Each epoch in the beacon chain has 32 slots, each of which is about 12 seconds and generates a block. Ethereum issues rewards by epoch, and the amount is calculated from the base reward, which represents the average reward of each validator under the best conditions in each epoch. The proposer of the block can take 1/8 of the base reward, and the rest of the rewards will be distributed to voters (provided that they vote in line with the majority of other validators) and participants in the synchronization committee. The distribution of rewards is related to the validator's effective balance and the total number of active validators. For details on validator rewards, readers can learn about Ethereum's Gaper consensus, which is also one of the most complex designs of the Ethereum protocol.

Since Ethereum staking requires at least 32 ETH, it does not support entrusting ETH to other validators for staking, and the ETH proposed for staking has a 27-hour unlocking period. These rules are certain obstacles for stakers. Therefore, in order to provide users with a more convenient staking environment, the Liquid Staking Token (LST) protocol has emerged in the market. The principle is to pool ETH together to bypass the minimum requirement of 32 ETH. Each user does not need to operate his own validator. The staking pool will handle the corresponding operations and provide users with corresponding staking certificates to participate in other DeFi applications and improve capital utilization.

StETH, launched by Lido, the industry leader in Ethereum liquidity staking, has occupied most of the market share of the Ethereum LST track. Lido allows ordinary users to stake any amount of ETH through the Lido platform. The staked ETH becomes stETH and can be exchanged for ETH at any time, solving the pain points of native staking. Currently, the staked ETH in the Ethereum network is 32.54 million, accounting for 27% of the total supply, of which Lido contributed 9.8 million, and stETH accounted for 30% of the staked ETH.

Solana

The initial supply of the Solana network is 500 million, of which 38% is given to the community reserve fund, 12.5% to team members, 12.5% to the Solana Foundation, and the remaining 37% to investors. The current total supply of the Solana network is about 580 million, with 460 million in circulation, and a circulation rate of about 80%. The remaining 20% of SOL is locked in the hands of investors and the team, and the most significant unlocking will occur in March 2025, about 45 million.

Solana's initial inflation rate is 8%, with an annual decline rate of -15%, and a long-term inflation rate of 1.5%.

The Solana network has no minimum staking requirement for validators, but the voting power and staking rewards of validators will be distributed in proportion to their staking amount. The Solana network supports delegate staking, where users stake their SOL to existing validators to share the benefits through delegate staking. Delegated staking does not mean delegating SOL to a validator, SOL still remains in the user's wallet, which makes it as safe as holding them. There are currently 1500 validator nodes, and the APR is around 7% on average.

Validators perform the work of verifying transactions and proposing blocks: every time a validator submits a correct and successful vote (which is itself a transaction, and the validator pays the transaction fee), they will receive points; there are no additional points for proposing blocks, the block reward only includes the transaction fees included in the block, and only 50% of the fees flow to the validator as a block reward, and the other 50% will be destroyed. During a period, validators accumulate these points, and then they can "exchange" a certain proportion of SOL rewards at the end of the period. The "exchange" of points to rewards is calculated by equity weighting, that is, the percentage of the total points (the sum of the points of all validators) that the validator accounts for obtains the corresponding SOL.

The current situation of LST on the Solana network is significantly different from that of Ethereum. The proportion of SOL in circulation in the Solana network is staked, which is much higher than Ethereum's 27%. However, LST only accounts for 6% of the staked supply (compared to over 40% on Ethereum). The main reason is that the Solana network natively supports delegated staking, and the DeFi protocol ecosystem is still in its early stages. The problems that Lido and its peers are trying to solve on Ethereum do not exist on Solana. Jito is the leader of LST on the Solana network. Jito entrusts users' SOL to validator nodes that support MEV (Jito-Solana validator client) to become JitoSOL, where MEV income is distributed to stakers as additional income. Therefore, the APR of the Jito platform is higher than that of delegated staking, which can currently reach 7.92%, and JitoSOL accounts for 3% of the staked SOL.

Summary

The economic model is the most important design of a blockchain that aims to run long-term. Compared with the simple and effective economic model of the PoW public chain represented by Bitcoin, the economic model design of the PoS public chain represented by Ethereum and Solana is usually very complex - it is necessary to consider the pledge mechanism, incentive mechanism, inflation parameters, and token functions.

From the perspective of the economic model of the new public chain, most of them use the PoS rather than the PoW consensus mechanism. The reason is that in addition to PoS being more energy-efficient, it also has better throughput and transaction confirmation time, and can process more transactions per second. Performance is the cornerstone of blockchain's large-scale adoption.

At the same cost, PoS is also more secure and easier to recover from attacks. Because validators are stakeholders, honest validators will be rewarded and malicious validators will be punished - of course, the largest stakeholders will get the most rewards, which will also cause wealth concentration problems.