Blockchain Analysis Series Part 3: Examining the NEAR Protocol

Authors: Lukas Bruell, Chris Smalley, River Fields

In our third installment of a four-part series, we analyze the NEAR protocol, focusing on its network design and implementation of token supply and validator incentives. Parts one and two, which similarly analyzed Ethereum and Avalanche, can be found here and here, respectively.

If you are already familiar with NEAR’s background, feel free to skip through to your desired topic.

An Overview of the NEAR Network

NEAR is a decentralized application (dApp) development platform running on the NEAR Protocol blockchain, a Proof-of-Stake (PoS) based Layer-1 blockchain network. The NEAR network was developed to address the limitations of earlier blockchain networks that prevented them from realizing mainstream adoption, namely seamless usability for both developers and end-users in addition to network scalability and security. To achieve this, NEAR designed its chain with a focus on the following pillars, which also influence its economic design:

• Usability — Acknowledging the importance of usability for both end-users and developers, the NEAR protocol aims to provide a more user-friendly blockchain solution by eliminating many of the obstacles commonly associated with blockchain applications such as onboarding complexities for end-users and limited compatibility with more widely-used programming languages for developers. NEAR employs familiar programming languages and offers a comprehensive suite of developer tools, empowering developers to more readily launch dApps on the network
• Scalability — NEAR offers a dynamic sharding approach, wherein network nodes are divided into multiple shards during periods of high usage, allowing for parallelized computation across the shards
• Simplicity — NEAR is dedicated to minimize complexity, making the protocol easier to understand and work with
• Sustainable Decentralization — The protocol facilitates efficient governance to ensure long-term sustainability and encourage innovation

For consensus, validators are chosen based on the Thresholded Proof of Stake (TPoS) model which selects validators based on a dynamic auction process that determines the allocation of “seats” to potential validators. The primary objective of this auction mechanism is to ensure equitable distribution and foster broad participation in the network’s validation process, thereby promoting decentralization and security. The dynamic auction process is based on the concept of a “seat price” which is the minimum amount a validator must stake to be eligible for participating in the network’s validation activities. Validators express their interest in participating in the process by submitting a signed transaction containing pertinent information, such as the amount they wish to stake and a new public key that will be used for signing blocks.

To secure a seat in the validation process, a validator’s stake must exceed the current seat price. Validators who successfully meet this criterion will then be assigned specific roles within the network, such as block or chunk producers, responsible for creating new blocks or chunks of data, or hidden validators, who participate in the validation process without producing blocks or chunks. Validators are assigned two epochs ahead of time (T+1) so that validators can be shuffled between shards and perform the appropriate shard state sync.

Lastly, similar to Avalanche, NEAR enables delegation so that users can participate in the protocol’s rewards without running their own validating node. Validators can attract delegators by setting up a staking pool contract, which defines commission fees and reward distribution split. A list of current validators and staking pool contracts can be found here.

The following two charts provide key metrics regarding the NEAR blockchain network and the greater NEAR ecosystem from Q2 2022 to Q2 2023.

Source: Messari

Source: Messari

NEAR Economics

Network Supply — validators

NEAR employs a dynamic sharding system called “Nightshade” to address issues associated with static sharding, such as the concentration of applications on single popular shards. In this system, validators are divided into block producers and hidden validators. Block producers are responsible for receiving state transitions, generating chunks, exchanging chunks with others, and maintaining data — chunks are a fraction of a block and are a result of sharding. Hidden validators, on the other hand, are distributed across all shards (unlike in a static system where they are assigned to single shards) to enhance the blockchain’s security and ensure accurate block production and data availability.

Since hidden validators do not produce blocks, it becomes challenging to reward them at the time of block production. As a result, the NEAR protocol distributes rewards over the course of a time period referred to as an epoch, addressing this issue effectively. At the end of every epoch, rewards are distributed between validators, developers, and the protocol treasury. The total epoch reward is calculated as:

Validators as a group are paid a fixed 90% of the epochReward with the remaining 10% going to the protocol treasury. The allocation to the protocol treasury, which is maintained by the NEAR Foundation, will ideally provide a continuous stream of revenue for ecosystem development. The ultimate goal is for the protocol treasury to be governed in a decentralized manner.

In this model, a validator’s reward is determined proportionally to their stake as a percentage of the total amount staked by all validators. To reduce complexity, the individual validator reward is the same if they were a block producer or a hidden validator. Validators will fail to receive the reward if they do not meet a certain online participation threshold, known as the online threshold. For example, if a validator does not produce enough blocks in a given epoch then they will not receive the reward and will be removed from the validator pool for the subsequent epoch.

On the penalty side, we see that NEAR follows an approach similar to Ethereum by seizing a portion of the validator’s stake (slashing) rather than Avalanche’s implementation that does not penalize malicious or lazy validators. However, unlike Ethereum, NEAR does not remove any of a validator’s stake for going offline (Ethereum removes the equivalent staking reward for missed target and source votes). NEAR’s approach for lazy behavior is to have the validator kicked out automatically at the end of the epoch which means they will automatically lose the potential to earn rewards for the next two epochs. For truly malicious behavior, NEAR will slash validators for double signing a block at the same height or for signing a block with an invalid post-state root.

NEAR Token Issuance

The NEAR token serves as the foundational native asset within the NEAR ecosystem and is inherently integrated into all accounts. As the primary facilitator of economic coordination among network participants, the NEAR token plays a crucial role in governing the operations of the network. NEAR is utilized to cover all transaction fees on the network, to participate in the validation process via staking, and to have an influence on governance processes. Additionally, it fosters the development of innovative behaviors and functionalities in applications built on the NEAR platform.

NEAR tokens are minted (aka coinbase) on an epoch basis, which lasts approximately half a day, with the overall inflation of the system determined by the size of the epoch reward for running a validating node. The maximum rate of minting new NEAR tokens is capped at 5% per year. The 5% inflation rate can be adjusted per epoch based on the volume of transaction fees collected — as the protocol captures more fees, the inflation rate decreases, potentially leading to zero reward from coinbase and a deflationary environment. This mechanic is similar to Ethereum’s EIP-1559 implementation; however, the base fee on Ethereum is burned in a transaction and users can incorporate a priority fee. For reference, an illustrative NEAR inflation rate per annum given differing assumptions can be seen below.

Source: NEAR Blog

Minting new tokens is effectively an indirect tax on token holders, especially token holders that are not validators, so it is generally preferred to minimize new token issuance. However, too small a coinbase combined with insufficient transaction fees can result in reduced interest to become a validator which in turn reduces the compute required for security and increases centralization of validators. To solve this tradeoff, NEAR’s approach is to set a ceiling for the maximum coinbase that dynamically decreases depending on the total fees in the system. Of course, the necessary ingredient to reduce inflation is the number of transactions on the network — NEAR currently does about $1,973 in daily fees. Of these $1,973 daily fees, 70% is burned by the protocol (discussed more below), or about $1,381 fees burned daily. Over the past year, we see that NEAR has had 144,525 daily active users which translates to an average fee burnt per user of $0.0096. At the current token supply of 987.2 million, NEAR is issuing approximately 49.4 million tokens per year or 135,233 tokens per day. This means that for NEAR to reduce coinbase reward to 0 and instead rely on transaction fees to reward validators, the daily active users would need to increase from 144,525 to 14,718,242 or about a 300x increase. As previously stated in the Avalanche article, the transaction fees per user would likely increase as there is increased adoption of the NEAR protocol. More users would perform more complex transactions and more applications would build on NEAR which would increase the users’ number of transactions per day. Noting these caveats, these calculations do however highlight that NEAR is currently unsustainable and likely to remain at the inflationary cap of 5% in the near future. Additionally, NEAR’s core strength of sharding improves its throughput and inherently reduces the fees paid by users which makes it more challenging for users to cover the money needed to reward validators and remove inflation on the network. This problem is further compounded by the 30% smart contract rebate which is discussed further below. NEAR could potentially address some of these issues by either increasing daily activity or by increasing the complexity of the transactions. For example, on-chain games could help address this issue by dramatically increasing the number of transactions on-chain. Another is complex DeFi protocols that provide high gas transactions. Let’s now explore how validators are rewarded on NEAR.

Network Demand

Fees

NEAR’s approach to gas distinguishes itself from Ethereum and Avalanche, which employ a single upfront transaction fee. NEAR separates the pricing of blockchain-specific resources into three main categories: compute, bandwidth, and storage. The pricing of these resources undergoes gradual adjustments depending on usage, with a smoothing effect applied. This departs from the auction-based pricing mechanisms of Ethereum and Avalanche, offering developers more predictability in transaction and storage costs.

Compute is a temporary resource consumed during transaction execution. The cost of each CPU instruction is denominated in gas units, with its price determined by the gradually adjusted gas price. Bandwidth, typically measured in bytes, is converted into gas units on the NEAR platform. On NEAR, storage is a long-term scarce resource, requiring a minimum account balance for applications or users to utilize it. The minimum balance scales linearly with the storage amount used by the account. The necessary amount of NEAR tokens per byte is fixed and subject to change only through major governance decisions.

NEAR’s gas prices are constructed to be predictable, but not fixed. If the previous block is more than half full, the gas price increases by a parameter called “alpha.” Conversely, if less than half full, it decreases by “alpha.” Additionally, because of NEAR’s dynamic sharding design, pricing for transactions within a shard and for cross-shard transactions can remain the same.

In general, the total gas of a given transaction can be computed by:

• CPU Instructions is the cost of processing with complicated state transitions costing more
• α is the relation between a unit of computation and unit of bandwidth
• Size of transaction is the size in bytes

These predictable pricing mechanisms allow block producers to estimate the total amount of gas that will be used to execute a block. Block producers can overspend the gas needed as they will receive the difference upon execution.

Unique to NEAR, 70% of gas used in transactions is burned while the remaining 30% is rebated to contracts that are interacted with in the transaction. Fees are distributed based on gas usage. This unique dynamic allows developers to receive a flow of funds from client applications; however, it remains to be seen if it promotes the right incentives on the network — developers might actually be encouraged to create contracts that expend more gas. We will discuss this concept more in Part IV of the series.

Key Economic Takeaway — Infrastructure Funding

One of the major aspects that the NEAR protocol has attempted to solve is the problem of public goods funding — a complex issue that societies have grappled with for centuries. This raises the question of the optimal way to pay for public goods. Typically, public goods are commodities or services that benefit all members of society, and which are often provided for free through public taxation. Public goods have two main criteria: they are non-rivalrous which means they do not dwindle in supply as more people consume them and they are non-excludable which means that the good is available to all citizens. A significant consideration that is related to public goods is the problem of free-riders, which are members that use the public good but refuse to pay for it. Non-tax paying citizens who enjoy the benefits of a nation’s public goods exemplify this issue.

In web3, permissionless blockchains can be considered public goods because they are open to anyone with a wallet and they benefit all members. Illia Polosukhin, Co-Founder of NEAR, proposed that Uniswap is an example of a public good on Ethereum as these protocols serve an integral role in the ecosystem. Although, one could argue Uniswap is a quasi-public good as it shares characteristics of both private and public goods. Uniswap is only partially non-excludable as it will charge higher fees to users via gas during periods of congestion, just as toll booths can charge drivers on the freeway during rush hour. This raises the question of which blockchain-based public and quasi-public goods should be universally funded and the most efficient funding method.

NEAR’s approach to solving this public goods funding conundrum involves transferring fees to the smart contract developers. As mentioned previously, 70% of a user’s transaction fees will be burned while 30% go to the smart contract developers of the contract that they interacted with. On the surface, this is a novel way to pay smart contract developers that create infrastructure for the benefit of the community, such as a DEX providing liquidity to all like Uniswap. Overall, public decentralized goods funding can be debated and may come with positives and negatives highlighted below.

The negative take on the forced attempt to reroute fees towards smart contracts and public goods is that it may produce negative externalities such as black markets, network congestion, collusion, and others. The economic definition of “rent” is important to think about when analyzing the 30% paid to smart contract developers. In economics, rent is an amount of money earned that exceeds that which is economically or socially necessary. Rent typically arises from market inefficiencies or information asymmetries. Economic rent is considered unearned income because it is not based on any increase in productivity or the addition of any new goods or services. Economic rent often arises because authorities create a patent or license which prevents competitors from entering the market and allows for the patent or license holder to extract rent from individuals that demand the goods or services. For example, companies that have rights to extract natural resources, such as oil, gas, or minerals, can earn economic rent if the price they can sell their resource for is significantly higher than the cost of extraction. This is especially true if the rights are acquired cheaply or if the resources become more valuable due to increased global demand or reduced supply.

The funding mechanism for NEAR’s smart contracts could be seen as a form of economic rent extraction, which could lead to network inefficiencies over time. Developers, for example, might entice users to engage with their smart contracts even if they impose higher gas fees, promising future transaction fee refunds via airdrops. This could result in unnecessary network congestion and unhealthy competition. Moreover, developers might collude to block the deployment of more efficient smart contracts, thereby maintaining their economic rent. This could stifle innovation and hamper the protocol’s future growth.

Viewed from another perspective, this financing model can be interpreted as a robust competition to develop a highly sought-after smart contract. Moreover, the entire NEAR ecosystem would benefit from its popularity, as 70% of transaction fees processed through the smart contract would be burned, thereby increasing the scarcity and potential value of the NEAR token.

Consider a scenario where a developer is in the process of creating an application with Uniswap-level popularity. They might opt to deploy it on the NEAR network, as they are motivated by the platform’s user-friendly development tools and the prospect of retaining 30% of transaction fees. Such an application could act as a magnet, attracting a user base that might not typically engage with the platform.

For example, NEAR could attract a project like GMX which has brought thousands of users over to Arbitrum and where it is the number one protocol with $460 million in TVL. The relationship between NEAR and this new application would be highly symbiotic as the app would bring in users and fees and the developers would get to keep 30% of transaction fees which they wouldn’t be able to achieve on other platforms. This aspect could drive innovative and unique applications to start their project on the NEAR network because of the novel payout benefits.

The concept of public goods funding or effective infrastructure design is extremely important as it will have far-reaching implications in blockchain configuration going forward, as the most efficient design will win. Notably, Vitalik Buterin has written extensively about public goods funding. In a blog post published in November 2019, he suggested several methods:

• Coin issuance
• Extracting a share of transaction fees at the protocol level (e.g., through mechanisms like EIP 1559)
• Levying a portion of transaction fees from certain layer-2 applications (e.g., Uniswap, or perhaps from some scaling solution, or even state rent within an Ethereum 2.0 execution environment)
• Charging a portion of other types of fees (e.g., ENS registration)

NEAR Protocol has adopted the second approach, implementing its own variant of EIP 1559. Meanwhile, the landscape of decentralized public goods funding has evolved, with methods such as Optimism’s retroactive public goods funding, Gitcoin’s quadratic funding for open-source initiatives, and Moloch DAO’s democratic voting system for project investments.

However, it’s essential to remember that the most effective public goods funding model remains a topic of ongoing exploration. When considering any economic design, it’s also critical to anticipate and manage any unintended consequences. Both users and developers typically act out of self-interest, which can often lead to network inefficiencies and friction. Therefore, designing a system that aligns individual incentives with the collective good is a crucial aspect of optimizing public goods funding in the blockchain space.

Key Economic Takeaway — Data Analysis

In addition to our infrastructure analysis of NEAR, we evaluated the correlation and regression of fundamental metrics on the NEAR network, using the same methodology used in Parts I and II. We observe that protocol Revenue and Trading Volume have the strongest correlation with Price, revealing a correlation coefficient (r) of 0.78 and 0.73, respectively. A comprehensive correlation matrix between Price and other fundamental metrics can be seen below.

The ordinary least squares regression analysis using NEAR Trading Volume, Fees, Daily Active Users, Active Developers, and Code Commits as independent variables yields an r-squared value of 0.808, as shown below.

These results indicate a strong explanatory power of the independent variables on Price.

Next, in Part IV, we will look into the Economic Analysis of the Solana network.