[Voting Period] Signaling Proposal: Add the Fee Abstraction Module to the Cosmos Hub

I think this is reasonable.

We would love to have a similar module when Duality is live but until it is, I think if it would drive more activity to the Hub it’s worth an integration.

Also worth doing a similar thing with Astroport.

If this is deemed a valuable feature overtime we can do some more analysis on how much value each chain gives in fees and MEV and customize the routing algorithm accordingly.


with a mean time of 4 blocks by IBC transfer,
with 4 transfer to be done in the example
with a block time of >5sec

=> we need >20 seconds to get the 0.00001 ATOM of alternative asset to return to the wallet

This will cost 3 Tx per transfert => 12 Tx
In definitive, this will cost more to the relayers than the real value of the swaps done


Hi @David_Crosnest ! Thank you for sharing this concern.

Fees are batched in the module and only sent over once per hour to be swapped to prevent this exact issue. The value of the transactions involved to the relayer is negligible in comparison with the amount of fees that accumulate from an hour’s worth of transactions on the Hub.

Under this mechanism, fees in non-ATOM assets would be swapped to ATOM and then paid to Hub validators / stakers hourly. fees paid in ATOM that don’t require these transfers will continue to be paid out as normal.

Do you have a rough value of the numbers of asset accumulated per hours ?

What’s about the ICS retributions.
We “asked” the security to be provided against a given % of inflation/fees/…
Will this asset will be swapped as well ?
In the case of Stride, this will create a 50% of inflation selling pressure on the token and push the price to zero.

1 Like

Re: your second point, definitely not!

ONLY fees paid out by users as Cosmos Hub transactions will be subject to the parameters of this module. Inflation or fees from other sources (such as ICS) will not be subject to this. Users would be free to use their ICS rewards to pay fees on the Hub (which is one of the benefits of this module), but only if they wanted to. It would not amount to a forced sale of any ICS revenues.

As far as the numbers, I don’t have this yet because the module isn’t live, but I think we could probably make reasonable estimates by taking the number of transactions that occur hourly on the Cosmos Hub, assigning a reasonable percentage of those transations that would be paid in other assets after the module gets full traction (10-25%?), and estimating the average ATOM those tx fees would cost to get the aggregate numbers.


10-25% ? I think you are very generous. in my opinion you can divide by 10 minimum

To have user using alternate currency for transaction, the UI/Wallet will need to push in the use of this assets.
The default currency will remain the main one by far.

knowing the amount of alternate tokens used right now must be part of the decision and I understand you don’t have this value.

1 Like

I agree that ATOM will remain the primary fee asset. Am unsure of the exact percentage of alt-assets that will be used, but I also think it’s inaccurate to compare this to the number of alternate assets that are used now.

Thanks to this module, wallets and front ends will be much more amenable to having front end options to pay in other fees (especially Keplr, which already has this option available for other chains). Once a wallet opens this capability up, fees paid in non-ATOM assets will increase substantially.

Just seeking tax clarification here: The alternate asset swap is happening on the chains side not the consumers side correct? So that the disposal of the token is not a double (or possibly triple) tax reporting event for the consumer?

I.e. if spending Neutron on an Atom fee swaps the Neutron in my possession for Atom then pays the fees I will have a disposal an acquisition and a disposal per transaction for the IRS. Whereas if the Neutron is accepted as payment for the fee from the consumer there is only one taxable event, the disposal of the Neutron. Most Crypto-IRS software charges per event (usually 100 events, 1000 events etc, as pricing break points). If paying a fee utilizes 3 events for tax purposes this will add up fast for the consumer. And in effect becomes quite expensive. The tax calculation charge will likely greatly exceed the chains transaction cost, but is ultimately just as imperative to the consumer. Consideration of consumer costs should always be at the forefront of these types of discussion.


I’m not qualified to give you assurances or advice on tax status, but I can tell you that from the user / fee payor end it looks like this:

  • User pays fee asset to the Cosmos Hub. At this point, the fee asset is 100% entirely out of the user’s ownership and control and will never be in that user’s ownership and control again, just as it happens with ATOM fee payments now. Accordingly, any subsequent swapping will also not be reflected as a transaction in that user’s wallet.
  • The fee asset sits in the module for up to an hour, along with everyone else’s accumulated fee assets that have been paid to the chain before being sent to Osmosis
  • Any path unwinding / fee swapping occurs and the resulting ATOM are sent back to the Cosmos Hub
  • The ATOM is paid out to Cosmos Hub validators and their delegates as normal.

Based on that flow, I think it’s pretty clear what the tax results (or lack thereof) would be, but if this is a huge concern for you I would always advise consulting with a tax expert in your jurisdiction to ensure you have 100% clarity.

Thank you for the reply and clarification. As long as the swapping is occurring after the payment has been processed it should work as you stated as far as I understand it. Tax costs and tax calculation costs are always a huge concern for me, and should be for anyone engaging the chain. The onus of the extra taxable events should be on the chain then and not on the consumer as far as I can interpret the process as you’ve described it. That is ideal and significantly reduces the costs associated with engaging the chain for the consumer.

1 Like

Yeah no problem!

I tend to agree with you that the tax issue here is pretty clear cut, but I don’t want to be on the hook for giving tax advice that ends up being wrong :sweat_smile:

That’s why i always tell people to consult with experts on these topics.

Glad I could help clarify though!

I think this proposal is the most useless proposal ever, even among those from notional…


  • unnecessary complexity of design
  • risk associated with codebase on Osmosis, which cannot be controlled by Hub
  • slippage risk on minor tokens
  • credible neutrality of Cosmos Hub

very simple alternative

  • periodic auction of non-atom gas fee basket

I have no idea why people are wasting time on this…

1 Like

Always appreciate the colorful feedback from the B-Harvest / Crescent Dex team :sweat_smile:

Hey, do you have a design for this simple alternative? This auction stuff could be interesting.