Proposed Changes to SecretSwap

It’s becoming more clear from the recent price movement that SEFI as a protocol is not set up with the right incentives to benefit the HODLRs. In anticipation of governance roll out I’m putting out some items which I plan to propose:

  1. Cut cashback down from 10% to 1%. Being rewarded for trading is a nice touch, we don’t need it be a very large part of all rewards. Free coins are being dumped, and treated as a discount on their purchases. “Cashback” almost implies it. We need to give people just enough to get a taste, but not enough it moves the needle on their purchase decisions. The rewards are likely all accruing to arb traders, not average joe users making a simple swap every so often.

  2. Replace all SCRT-XYZ LP earn rewards with SEFI-XYZ LP earn rewards. If free coins are being given out, we need to ensure the users receiving them are invested in the protocol. How do we ensure this? By forcing them to HODL and be invested with the rest of the users. Rewarding people with SEFI, when they do not need to hold any is a perverse incentive. These free coins are being dumped.

  3. Exit LP Fees. People threw out some high figures in the chat, but 5-10bps feels right if we must select one number. We could implement a curve 0 to 6 months than scales down from 30bps to 0bps. This is to encourage users to LP for the longer term.

  4. Buyback. In addition to the above Exit LP Fees, we will take 5bps of all LP fees collected and buy back SEFI. Ideally this should be distributed to HODLrs, but I’m not anti-burn.

There was a lot of excitment months ago about building out features to SecetSwap. We should make a list of the things we want to see added to the protocol. Please add below, share thoughts of what you would like to see done when we get Gov.

  1. Order Book
  2. Yield Aggregator
  3. __________________ - Your Answer
2 Likes
  1. I agree with your sentiment regarding cashback. I wasn’t aware that it was 10% but regardless of the level, I would assume that the average person gets very little out of cashback and it’s likely arbitrage traders who are going through the motions to accumulate cashback. Decreasing it might mitigate it to the point where it’s not worth their time. It’s not like people will use SecretSwap just for the cashback so make it a happy coincidence rather than what could be a gimmick.

  2. I’ve not formed a view on this yet. I understand where you’re coming from and see how it will theoretically support SEFI as a coin but there must be a reason why other Dex’s don’t use the native token but use the coin supporting the chain (Uniswap largely uses ETH and Pancakeswap largely uses BNB).

  3. I strongly dislike the idea of exit LP fees. When there are “fees” to selling your coins like with SAFEMOON, the crypto community spoke out against it, this is no different. It’s not encouraging you to hold, it forces you to hold because you will likely make a loss otherwise. That will drive informed people away from SEFI, the majority of the crypto community have probably never even held a coin for six months yet. Do we even have any statistics that can demonstrate the average age of an LP?

  4. I have no problem with this. It doesn’t really do anything to SecretSwap as a product but will in the long-run contribute to some positive price action as total supply reduces.

Something important that I think needs to be considered is that SEFI is a governance token by nature and therefore it is important that if SEFI was to be used in place of SCRT within pairs, there has to be a way to capture this and enable LP holders to vote too. Otherwise, this creates a conflict for token-holders.

Note that I don’t currently hold any SEFI and folded those tokens into my SCRT bag.

4 Likes
  1. I am not sure this would be necessary if options 2 or 3 get implemented. I like the idea to incentivize trade, it generates volume and this attracts more users, etc.

  2. “All” seems a bit too radical for me. We should certainly replace most of them. Other DeFi do the same and seems to work.

  3. This is highly controversial, I have mixed feelings. Many large (20~30M TVL) BSC pools have " Deposit fee: <0.1%, Withdrawal fee: 0.1% " and that’s even for long term holding.

  4. How about we distribute only for those holding for more than X days/weeks/months too ? Or even burn x% and airdrop the rest so we don’t create a SEFI Elite…

On #2 why instead of replacing would you not just add more SEFI pairs? Don’t think you can replace. SCRT is the onramp to the whole ecosystem and needs pairs with large liquidity.

3 Likes
  1. Right now SEFI is being used as a prop for SCRT, not treated like an independent protocol built on top of it. The move to make all SCRT pairs SEFI is to add millions of dollars in buying pressure, stop the dumping by people not invested in the token, and have the protocol benefit the protocol holders. This is not uncommon and is done on other dexes like THOR or Sifchain. People will still be able to LP SCRT-XYZ, but they won’t be rewarded for it. If people want they can go back to staking SCRT to keep yield.

  2. I’m sympathetic to the LP Exit fee pushback. We modeled this off the deposit fee you mentioned, but thought there shouldn’t be friction to enter a pool. The fee can come out on the backend so users think twice if it’s worthwhile. This is to stop mercenary capital, or at least make it benefit the holders.

  3. I’m very strongly against anything that buckets holders rewards by length of time held. The protocol shouldn’t be forever captured by the first users. If new whales want to come in and join they should take same rewards as any other.

1 Like

On 2: You will dump the SCRT price if you cancel those reward farms.

1 Like

Thats a SCRT token issue…

2 Likes

Thank you for taking the time to put this together. My two cents…

  1. Agreed. The current Cashback rewards do indeed seem to be dumped regularly on the market.

  2. I think this is probably a smart move. It may be sensible to wait until SCRT has some more arrows in the quiver as SecretSwap/Bridge currently make up a large part of the Secret Network in the eyes of many. Another possible solution may be to allow including SEFI alongside SCRT in LP pairs. The SEFI portion could potentially act as a reward-multiplier of sorts, providing higher rewards to those that lock SEFI alongside SCRT in their lending pair.

  3. No strong feelings either way…

  4. I think this could work well for keeping the SEFI price stable.

1 Like

This is the proposed distribution for the rewards (post cash-back decrease). Highlighted yellow indicates it’s the only pair right that currently requires SEFI to earn. The others in that section all have SCRT side pair.

I took information from the current earn page, and the last notes I saw on the rewards distribution (though it’s out of date). Taking into account the goal of being cross-chain and adding stable coins. Proposed rewards on the far right and of course assume SEFI on the other side of the pair.

You’ll note that the sSCRT-SEFI rewards are increased as many trades will be routed through it. It’s probably a little stable coin heavy, but that could be where we pull from next to seed a new pair like ENJ.

2 Likes

Could you explain the allocation model you have used? I don’t get why you suggest swBTC should have a much higher than allocation than other pairs.

Regarding switching to SEFI pairs instead of SCRT pairs, aren’t you afraid LP TVL will drop drastically and make SecretSwap less usable?

  1. Any data to support this? If people are using Cashback heavily then that’s an indication that it’s increasing trading volume. Cashback is a unique mechanism that rewards trading volume as opposed to TVL. To me that’s an extremely important metric. All other rewards focus solely on TVL.

  2. Strongly disagree to replacing all. Maybe a better mix is needed though.

  3. This is important, but requires non-trivial development and migration.

  4. Similar, and likely not that useful beyond marketing. Currently, there are ~$6k fees generated from swaps, which would increase the price at ~0.25%. I agree this becomes more interesting if there are other fees collected and the general amount increases.

I’m very much in favor of improving the value capture of both SEFI and SCRT (which is in fact the thing that I think about the most right now), but we need to more sustainable methods to achieve that IMO.

6 Likes

I looked at the current amount of money in millions locked on the earn page, it’s weight in percen terms and the last seen rewards distro. The 15% WBTC seems high, but when you combined both ETH pairs it’s equal weighting. This is similar to what it is now. I was thinking it may even be a bit low and should be bumped up to 20% as the main asset in crypto universe. Open to any suggestions and feedback.

I am not as concerned about a TVL drop due to switching to SEFI. Many people who are LPing own or owned (dumped) SEFI. This is the issue and why it’s tanking. We need to tie the protocol rewards to the holders of the protocol. While the transition may be bumpy, the end point is more resilient. You can’t vampire attack the LP pairs, if the LP are all part owners. I think this is a more sustainable business mode.

2 Likes

Incentivizing non-SEFI pairs is good in the sense that it will lead to an increase in TVL (at least in theory). However, it is a recipe for indefinite, constant sell pressure on SEFI. Perhaps this is a short-term initiative to attract higher TVL, but in my opinion we cannot keep non-SEFI pools incentivized indefinitely as it will undoubtedly have an adverse effect on the price of SEFI.

1 Like

I have no data, but my understanding of the way people value things given to them for free (which is very little). The cashback mechanism may drive trading volume, but it doesn’t generate value for SEFI. The only trading volume which may provide value (esp without fee capture) is arb trading, in which case the trader is rewarded with the arb. Other than that all trading volume is captured by SCRT (gas) or LPers (mostly non-sefi pairs).

The current game theory shows free coins are going to users that haven’t proven any commitment to SEFI. This creates a negative feedback loop where people concerned about short-term profit will dump the asset. The changes I’m suggesting would limit rewards to users who have shown commitment to the protocol itself. This would turn SEFI into an owner-operated type of business where most LPers are truly committed to the long-term success.

I agree on the fees that it’s currently not a meaningful amount, but it’s important to begin behaving for the long term as soon as we can. Exit fees, swap fees, lending fees (??), etc in the long term will replace the SEFI distribution currently taking place. I’m also not against having another item to lean on for marketing. Sienna announced this day one because they understood the appetite for it.

1 Like

Just wanted to chime in here. And I will say that I have no hard data on this, however I have been more or less with the community since the swap was released. Even having experienced the market-wide dip, I really have no reason to run for the hills. I think the community and project are great overall.

However I have noticed that SEFI is not only creating pressure to its own value. Arguably its also making the sSCRT - SEFI pool one of the least attractive pools to LP with at any given time. And I have a feeling that outside of an extremely bullish market (lets just assume a crabbing one for now) the mechanics of people dumping SEFI for even SCRT, which is then more or less being arb’d on other exchanges …

It seems to be creating choppy environments across all pools as it keeps putting SCRT into a constant state of speculation. Now over time SCRT as a token should settle down as additional DAPPS are build on network (not unlike ETH, etc) but right now it feels like circular value logic.

Perhaps someone that’s inside the spreadsheets more than I right now to play with numbers could game this out, but it would seem to me a short term way of stabilizing this push and pull effect (SEFI <> SCRT) are having on each other would be to create a split reward mechanism for:

  • SEFI STAKING
  • sSCRT - SEFI

This alone could encourage people to get into the habit of compounding more medium to longer term to capture more consistent value.

OR … and this could be more practical in the near term … I could see how it could make sense to provide split rewards for strictly:

  • sSCRT - SEFI

… as then holders have the additional option to place SEFI in the SEFI STAKING pool for compounding (and to hedge the inflationary nature) and with staking validators within the SCRT staking mechanism. Whereby SEFI only pays APY out in SEFI and SCRT staking only pays out in SCRT.

Perhaps someone has some thoughts about this?

1 Like

There is no timeline where SEFI is worth billions and SCRT is worth very little/0. So, no. The overall success of SEFI is also contingent on the growth of SCRT.

Suggesting that SCRT pairs should be removed completely implies an overall decrease in TVL and volume, which are two ultimately key metrics for the success of SEFI. Nobody wants to trade in an illiquid DEX, and volume = cashflow when it comes to AMM tokens (considering you are proposing a buyback implementation).

This doesn’t mean things are fine as they are though. Simply saying that turning around and saying it is a SCRT issue is disingenuous. I think 4. is a good point and even if not significant due to current volume, should be implemented.

Strong yes on #4 and strong no on #2/#3. For number 1 I’d say maybe…I haven’t done any math to argue for any particular cashback level. I would say though that the existence of the cashback mechanism justifies raising swap fees by 5bps (to 35 overall) with the 5bps being used for buyback.

I think most people agree with the buyback idea but what I would additionally propose is this: bought-back sefi goes to stakers. If possible I’d suggest that it only go to stakers who lock up for some period (say, 21 days, to match the scrt period and because it’s kind of just long enough to stop people from short term trading to frenetically). Maybe this could be accomplished with a lock-up contract that receives the bought back sefi. You could deposit sefi to the contract for a share in the “pool” and the contract could stake its sefi for the usual staking rewards.

You’re arguing some position you assume I hold that SEFI would be worth more than SCRT. The issue as it stands is that SEFI is being used for the benefit of SCRT and not it’s holders. These are two different things. The value accrual to SCRT should come from the increased usage of the network, not capture of the protocol. The SCRT issue is figuring out how to bring more value to the token beside using SEFI to lock up millions in value. It’s TBD if that would lead to decreased TVL or volume. I assume it will have no effect. SEFI has a captive audience at the moment… do you want to swap or not? The protocol is just not being positioned as such.

1 Like

It’s not TBD if it leads to decresed TVL or Volume. sSCRT pairs are the most liquid besides swbtc-seth according to secret analytics. If there is no incentive to hold SCRT, then a lot of that SCRT would either go back to staking, sold, etc. To say it’s not clear TVL will go down seems wishful thinking.

Lower TVL leads to lower volume, as lack of liquidity makes trading much less attractive due to slippage. That leads to less fees, and less buyback pressure in the mechanism you are proposing. Resulting on SEFI being overall less valuable fundamentally.

1 Like

You’re assuming no one will convert the sSCRT to SEFI to continue earning rewards. The 2nd most liquid pair beyond the sWBTC-SETH is a SEFI pair. It’s the ONLY ONE that gets rewards on its own protocol. Reread the last sentence. This is what I’m highlighting. The protocol is being used as a tool to prop up SCRT, not for its own benefit. We need a serious realignment in the way rewards are distributed to fix the issue.

People will go where the yield is. If they choose not to convert some sSCRT to SEFI (reverse the dump) then they can go back to staking. Others will take advantage of their absence. The returns are what drives the LP. You can see it reflected in the fact that no pair on the earn page and release of new pairs. Almost all rewards are quickly taken advantage of until the yield <100%.

1 Like