Preparing to kick off shielded set rewards

Mainnet Ops
1

Existing blockchain networks protect digital assets and the ability to transact, but they don’t protect their data. Namada’s primary purpose is to provide data protection for the existing multichain (ie. many chains), beginning with the IBC ecosystem, (which is bigger than just Cosmos!), then Ethereum and beyond.

Further to Mainnet parameter discussions: Knowable's proposal, we’ll need to bootstrap Namada’s data protection by attracting early deposits for each kind of asset, and we can incentivize deposits by reallocating ownership (allocating new NAM token issuance, or “inflation”) to depositors.

From this discussion I think we’ll want to get three questions answered.

Three key questions

  1. How many assets should we incentivize, initially? and which ones?

  2. What size deposit (in tokens) should we target for each asset?

  3. What maximum inflation rate should we use to incentivize each asset?

Things that I think we should care about

a) maximizing social good while minimizing social bad, by starting small and expanding in a considered, controlled way, eg. setting the IBC limit for tokens per epoch in-flow
b) safety of deposits: the Namada network and protocol will be responsible for any deposited assets
c) testing before doing it on mainnet
d) experimenting by observing and updating often

Things I think we’ll need

Goals: protection for what sized transactions, generally? what kinds of metrics do we want to track and achieve?
For example, I’d like for there to be an ATOM pool that protects users’ data for a 1 500 ATOM transaction, which is much more important to me than protecting the data of a 15 000 ATOM transaction. To provide protection, my fairly uneducated estimate is that we would want 50 to 100 deposits, which we can’t properly incentivize, and an ATOM pool between 8 500 and 17 000 ATOM. I’d love some feedback from folks that have a sense for how to reason around quantifying data protection, maybe @cwgoes @brentstone @adrian

Governing
a) mission: perhaps this could start simple and informal, but should get clearer over time… some starting questions may be: which kinds of assets would benefit from being supported by Namada, and why? why incentivize one asset and not another?
b) process: a simple, extensible off-chain process for governing (ie. signalling and deciding support for) new & existing shielded set reward parameters values.

Metrics: tracking and a dashboard for observations

I’d love to hear thoughts from folks like @Daniel @sirouk @preto @cwgoes @awa @brentstone Gian @adrian @Rigorous and any others who may be interested and capable of weighing in on governing, tracking, and displaying metrics in meaningful ways.

8 Likes
2

Eventually, the bigger the pool, the more effective its protection, as explained in Namada’s mission: The Namada mission.

First, we can measure the percentage of cryptocurrency tokens (by relative market capitalization) that are shielded at rest. This includes, for example, shielded assets on Zcash, Namada, Penumbra, Aleo , Anoma , Aztec , etc. - any well-designed shielded pool counts. Privacy is created by assets shielded at rest, so by and large, more assets shielded at rest means more privacy available for users. We should also measure technical stability: are assets that are shielded remaining secure, or are there incidents of them being drained in exploits?

Conversely, a small pool offers less protection.

  • The transaction size could be capped at a certain percentage of the pool size.
  • Also, transaction sizes should not be idiosyncratic, but a few preset amounts.
2 Likes
3

The corollary is for the best data protection, the number of different assets should be kept to a minimum. Better to have big ETH and ATOM pools than to support 500 different assets with many having little to no liquidity.

Exotic assets are hard to protect anyway because of trivial chain analysis.

An added benefit of focusing on a few primary assets is that it reduces the complexity and thus the potential breakage of the infrastructure. Imagine keeping 500 IBC channels alive.

Support for other than primary assets can be done implicitly with intents. Protect TIA? Automatically swap TIA for ATOM through the Osmosis Hub, then process it on Namada.

1 Like
4

@brentstone is there a tool we can use to model max_reward_rate for each shielded set reward? i think it would be helpful to be able to adjust parameters and visually see the effective reward rate for deposits

5

@Gavin the closest thing to a tool (and what I use to calibrate values) is this test in the codebase called test_masp_inflation_playground .

We typically sought to find suitable values that allowed the inflation mechanism to sweep the full range of the reward rate in some pre-desired amount of time. For example, we might say we desire the inflation mechanism to be able to move the rate from 0 to max_reward_rate and then back down from max_reward_rate to 0 all within 30 days and then find parameters that can allow this under the right circumstances. You would also need to make some other assumptions to enforce this as well.

Feel free to let me know if there is some desired inflation behavior similar to what I mentioned - and if it seems like a popular idea I can calibrate the values to something of the sort.

1 Like
6

regarding rate limits, @ValarDragon issued an important reminder

@brentstone have you seen this? thinking it would be good for Namada to have more dynamic flow limits GitHub - Stride-Labs/ibc-rate-limiting

7

I hadn’t seen this, thanks for sharing, @Gavin!

At the moment I believe we just have a constant mint limit and a throughput per epoch limit in raw token amount for each IBC asset. I think a dynamic rate limit could be a nice upgrade post-mainnet.

Do you envision just implementing the Stride rate limits or any other kind of dynamic rate limit?

1 Like
8

the Stride rate limits look compelling, but it would be good to get an opinion from @cwgoes and @ValarDragon

9

@cwgoes @Gavin it seems there is already a collaboration with Stride, because the LSTs of Stride would be more efficient for the MASP since their hurdle rate is zero unlike unstaked tokens. Taking this further, a collaboration with Lido seems also a must to use stETH in the MASP. In general, all the major liquid staking providers in the Ethereum/IBC ecosystem should partner with Namada to include their LSTs in the MASP? It is a win-win: Namada needs assets in the MASP otherwise the project fails, and the LSTs benefit from a new and additional yield with the shielded set rewards

Also, Babylon is offering potential staking yield to BTC holders, couldn’t Namada allow BTC holders to put BTC in the shielded pool with a large reward allocated for the BTC asset? The BTC wouldn’t have the locking like in staking and could be even more attractive to BTC holders than Babylon?

3 Likes
10

we should deeply look at this . I think its important

11

just adding that there’s probably overlap here with key metrics: Key Namada metrics

12

Stride rate limits are indeed a very cool boost to validators, if the token allows for it. works great with Atom

13

Hi all!
Since our epoch is 6h but not 24h I propose to bring down following params as well:

  • rewards_gain_p = “0.25” (make it 0.1)
  • rewards_gain_d = “0.25” (make it 0.1)

It will slow down inflation rate changes and may prevent panic within stakers/validators. Please apply it for both PoS and Shielded set inflation rates.

PGF inflation rate, as Brent suggested, should not be high initially, since we do not know for sure what to do with it right after mainnet launch. It can be increased later when our governance mature enough.

3 Likes
14

I also checked COSMOS chain, and seems like only 0.7% of all transactions had more than 1500 ATOM amount. So I think let us plan for that amount as maximum amount we traget to protect?

For Osmosis I was only able to find that average transaction is about 170 OSMO tokens. But if we just want to use a Cosmos as benchmark for $ equivalent, we can aim at protecting approx. 14000 OSMO transaction.

Regarding shielding rewards, for Cosmos I can see that “inflation rate increases by 13% per annum until 20% if there is less than 66% of all ATOM staked, and if there is more than 66% staked then the inflation rate will decrease at a pace of 13% p.a down to 7%”. So quite high numbers as an alternative investment.

1 Like
15

This max inflation parameter was recently changed to 10%. But yes for stakers there is a hurdle rate, why put assets in the Namada shielded pool if the rewards are higher for staking? That’s why Stride did a collaboration with Namada to allow Stride’s LSTs to benefit also from the shielded pool rewards

1 Like