Adjusting Axelar Network Incentives: A Proposal to Scale to Thousands of Chains [Call for Comments]

Governance and Grants
1

Preparing Axelar network’s technical infrastructure for an explosion in new-chain connections: short-term [reducing inflation], medium-term [external validation via AVM], and long-term [light clients via AVM] solutions

The Axelar network currently offers inflationary rewards to motivate validators to support external chain connections. This approach has served the community effectively, enabling the network to handle over $6 billion in cross-chain volume and hundreds of thousands of requests. This article proposes for consideration short-term, medium-term and long-term scalability plans in furtherance of maintaining robust and decentralized technical infrastructure. For further context, please read “Streamlining L2 Integrations,” recently posted to this forum.

Key plan highlights

  • Short-term: Modify the network’s inflation parameters. In parallel, introduce a gas-burning mechanism to counterbalance inflation, in furtherance of maintaining robust decentralization on the network.
  • Medium-term: Transition to establishing connections within Virtual Machine smart contracts and replace inflationary rewards with explicit reward pools for each new chain.
  • Long-term: Expand the use of more light-client connections where suitable.

A. Short-term: Incentivizing decentralized security by adapting network inflation parameters & implementing a gas-burning mechanism

The current inflation rate stands at 0.75% per externally verified chain, culminating in a total inflation of 11.5% [1% base inflation, + 14 externally supported chains * 0.75% = 11.5%; see the previous inflation adjustment post] and an AXL staking APR of approximately 14.5%.

To accommodate more Ethereum Virtual Machine-compatible [EVM] chains [estimated at five to seven chains for this plan; medium- & long-term plans below will aim to accommodate many more chains] on the network without compromising incentives, current parameters may need adjustments. Taking into account the approximate technical costs associated with maintaining an external remote procedure call (RPC) node for a standard EVM stack are ~$100-$300/month [the costs vary greatly depending on a chain / deployment model] and the need for the Axelar network to incentivize a minimum of 75 validators to maintain robust technical infrastructure and security in furtherance of decentralization, an added network reward of ~260K-780K AXL [~$90K-$270K at the current exchange rate] per year for each chain should be distributed to such validators in furtherance of such aims. Given Axelar network validators set an average of ~10% commission rate, this suggests a minimum inflation of ~0.26% for each external chain added to the Axelar network to continue incentivizing validators to support external chains [the current total supply is ~1.1B AXL, so 0.26% inflation adds ~2.86M AXL in rewards per chain per year].

We recommend reducing the inflation to 0.3% per external chain. This should bring the current total inflation down to 5.2% and, with the addition of another five chains, result in a 6.7% inflation. This helps ensure that validator incentives are covered while also decreasing the network’s supply.

This short-term modification to the network’s inflationary rewards would help the Axelar network ensure it maintains a high-quality validator set, which would help with the ecosystem’s transitions to the medium- and long-term items proposed below. The medium- & long-term plans would allow Axelar network to onboard many more chains without inflationary rewards.

The inflation-reward adjustment can be initiated by anyone through submission of a governance proposal on the network, and would subsequently be applied automatically by the protocol once a community vote to move forward takes place.

If more chains need to be added to the network, it can transition to a model where a certain amount of rewards goes solely to validators that support an external chain [with no commission], and a smaller inflation reward is applied on top to continue incentivizing users to stake with validators that support more chains.

Implementing a gas-burning mechanism

Currently, when users conduct a cross-chain transaction, they pay gas to the Axelar network, which then redistributes it to stakers. It costs about 0.2 AXL [parametrized] in gas to process a message via Axelar today [plus gas on the source and destination chains]; we suggest removing these gas fees from the supply.

Over time, this mechanism should counterbalance inflation. For example, with a fixed gas fee of the AXL equivalent of $0.5 and 5 bps for selected asset volume transfers, assuming the network processes 100K transactions/day and a daily volume of $100M [a 20-fold increase from current figures], around ~104M AXL [~$36.5M at the current exchange rate] in gas fees would be collected and removed from the supply annually. This would effectively counteract an inflation rate of up to 10%, leading to a deflationary state of the network.

Switching to a gas-burning mechanism as described above would necessitate implementation work from the community, a community governance vote and a network upgrade.

There is a base 1% inflation on the network [for all validators that just run the Axelar consensus], which we do not believe needs adjustment.

axelar-inflation-params-table
axelar-inflation-params-table762×836 39.5 KB

B. Medium-term: Transitioning to Axelar Virtual Machine connections

The Axelar Virtual Machine [AVM], a fusion of Cosmwasm consensus logic and Axelar network interoperability, is being developed to help transform the Axelar network’s connectivity approach. Interchain Amplifier, a service on top of the AVM, is a set of smart-contract templates that developers will be able to instantiate to plug into the Axelar network and route their packets to all other interconnected chains. Multiple validation methods and customization features will be supported. Its key features include:

  • Permissionless network connection setup.
  • Customizable connections for application-specific preferences.
  • No added inflation for new connections. Validator rewards will come from dedicated treasuries, projects or gas fees collected across the connection paths.

To integrate a new chain, a developer would need to do the following:

  • Instantiate an Axelar Gateway on the newly connected chain.
  • Instantiate a Gateway on the Axelar network, specifying the validator set they want to utilize [which would be outside the core consensus validator set].
  • Get a reward pool funded with AXL to incentivize the validators to support the new connection during the bootstrapping phase. [If the number of transactions is high along the new connection, the developer may choose to channel some part of those to the validators versus explicitly funding a reward pool.]
  • Submit a governance proposal for a community vote to add the new chain to the Axelar Router.
  • Upon approval, the traffic from the new chain can be routed to all other Axelar interconnected chains.

To elucidate, Interchain Amplifier would cease adding inflation and would instead use AXL rewards that teams or treasuries deposit into chain-connection pools. The AVM and Interchain Amplifier service are presently being developed and are in the devnet phase, with plans to finalize development in the upcoming quarters.

Existing connections would remain following the existing model; selected new connections should still be considered via the existing model, as they will rely on the full Axelar validator set. Some existing connections [e.g., the connection to Ethereum] could potentially be augmented with additional light-client validation logic for potentially stronger security.

C. Long-term: Use of light clients

Several initiatives in the ecosystem are in progress to employ light clients [including those based on zero-knowledge proofs (ZK)] for connections with the Amplifier. This is particularly apt for L2s as validity proofs gain traction. With light clients, external validator incentives should be unnecessary – only the relayer would need incentives. This is because a light client of chain A gives a compact program P that can be run on chain B to verify transactions from chain A. No external validation support is needed in this model and packets simply need to be relayed by someone. [The Axelar network supports a permissionless relay model, so anyone can transmit these packets.]

The result is that the Axelar network would morph into an overlay featuring a blend of connections, some supported by external validators [at consensus level as today, or at Interchain Amplifier smart-contracts level], others by light clients.

Summary

As the Axelar network continues expanding, making more connections and supporting more applications, its compounding effects increase exponentially. To support more chains in the short-term, the Axelar network should reduce the inflation parameters and add a gas-burning mechanism. With the instantiation of the Axelar Virtual Machine and Interchain Amplifier services, new types of connections would be possible. Those connections could help eliminate the need to add external inflation to support more chains. Instead, AXL would obtain further utility: it can be used to explicitly fund the reward pools for validators that require external validator support and to fund relayers. A combination of the inflation reduction and increased utility for AXL to fund reward pools would help ensure that the community can help scale the Axelar network to support all the connectivity needs of Web3 ecosystems.

6 Likes
2

I support the idea of making the AXL token deflationary.

I think core inflation of 0.75% needs to be gradually reduced. (Like the Fed rate, only in the other direction :smile:. . This is to avoid causing a collapse in validator yields so that they can cover their server expenses.

I think it’s the right decision to hold a vote and lower the inflation rate from 0.75, for example, to 0.6, and continue decreasing it to 0.3%.

8 Likes
3

I like the idea of reducing inflation and making the AXL token deflationary. I think perhaps we need to bring inflation down more gradually from 0.7%5 to 0.5-0.6% and gradually go to a rate of 0.3%.

6 Likes
4

Love the long-term vision and well thought out tokenomic model. Becoming deflationary based on usage is a win-win for all stakeholders. If this can be done while ensuring robust infrastructure then it is the obvious path forward.

Depending on the timeline for adding additional chains I would agree with others and propose a step-laddered approach to decreasing external chain rewards to coincide with each additional chain. This will avoid an abrupt change in validator projected operating costs. If they are all to be added at once I see no issue with making the jump right to .3%.

One question we do have as a prominent relayer in Cosmos is has any thought be given to how relayers will be incentivized? It’s mentioned that the long-term use of light-clients will not require validator incentives, but only relayer incentives. At this time relayers aren’t really incentivized all that much in the Cosmos. A lot of us do it mostly as public goods and because we’ve been given substantial foundation delegations. I think long-term we as an ecosystem (Cosmos wide) need to think about the relayer model as well. Will Axelar be implementing ICS29 and the relayer fee module, or does it already?

2 Likes
5

This is a broad and far reaching proposal for adjusting the tokenomics to allow Axelar to scale as the number of chains increases. Such a fundamental change needs to be understood and considered carefully, especially by validators.

We completely agree that inflation needs to be significantly reduced, and the short-term proposal of moving to 0.3% per chain makes sense. We also think implementing a gas-burning mechanism would be a significant positive change for the network, and welcome discussion on how parameters such as a fixed gas fee would be set.

If there is agreement that this broad direction of travel is correct, is the next priority to have a focused discussion on reducing the inflation rate? This could cover the target per chain and whether to implement in a single move or gradually as the number of chains increase. A governance vote on this specific aspect could then follow.

After this we could move to a focused discussion on implementing gas-burning, again concluded with a vote.

1 Like
6

We support these changes whole heartedly, making it easier for adding new chain and making the AXL token less inflationary will benefit the ecosystem

We would like to know whats the plan for non-evm chains, what can we expect in that front in future.

2 Likes
7

As a short-term, governance proposal can be made to adjust inflation per external chain. Medium / long-term initiatives are being worked on by various ecosystem teams and will come to governance votes later. But all are in-line to continue improving programmability, security, and scalability of the Axelar network.

3 Likes
8

The first line indicates that the current inflationary parameter is 0.75% per supported chain, with a proposal to reduce it to 0.3% per chain. This suggests a reduction of inflation, but not necessarily its complete cessation.

But with what you are proposing, if the inflation is put into effect and if the burning rate of 104M AXL per year is maintained, equilibrium could be reached in just over a year. However, for burning to consistently exceed creation, it would be necessary either to increase the burning rate, to reduce further the rate of new token creation, or a combination of both.

Based on the calculations with the provided data:

The current total inflation, at the rate of 0.75% for the 14 supported chains, is 11,025,000 AXL per year.
With the reduced proposal to 0.3%, the total inflation would be 1,764,000 AXL per year.
The annual surplus or deficit of tokens, once the proposal is implemented and assuming that 104M AXL are burned each year, would be 102,236,000 AXL.
This means that, each year, the network would burn 102,236,000 AXL more than what is created by the proposed inflation.

Starting from the current supply of 1.116 billion AXL and aiming to reach a maximum supply of 1 billion AXL, we would need to burn 116 million AXL.

Taking into account the annual surplus of burned tokens (102,236,000 AXL), it would take approximately 1.135 years to burn enough tokens to reach the desired maximum supply of 1 billion AXL, assuming all other parameters remain constant and that the rate of token creation and burning does not change over time.

This is a theoretical scenario where we consider limiting the maximum total supply of AXL to 1 billion units.

If the target maximum supply of AXL tokens were reduced to 800 million or less, it would require at least approximately 3 years to burn enough tokens to achieve that goal. That’s why I chose to set it at 1 billion, as it seems reasonable in terms of duration, rather than wasting too much time.

9

It looks like this proposal was made over a year ago and we have yet to even vote on the ‘short term’ plan.

When can we expect to see an official governance proposal? Would that be submitted by the Axelar Foundation or somebody within the Axelar internal team? Surely it does not fall on the peripheral community.