Warp Devlog #18
There have been some important changes since the last update in the Warp Magazine #05, namely to staffing as seen below. This sprint, work continued on the elements of Chisel audit, deploying new pairs, and updates to frontend and backend. Please see below for more details.
As always, here are some links relating to the Warp universe;
- Warp finance website
- v2 dApp platform
- Deconstructing the Warp vision (the three pillars)
- $warp magazine edition #01 (WARP, the ultimate incentive mechanism)
- $warp magazine edition #02 (Looping, the ultimate gains multiplier)
- $warp magazine edition #03 (Chisel, the ultimate DeFi yield aggregator)
- $warp magazine edition #04 (Lending versus minting)
- $warp magazine edition #05 (Chisel; B2B bribeable TVL)
- Old v1 platform
We are Hiring!
Unfortunately, since our last dev-log update (Magazine #05), our lead developer had reduced their hours to half-time, and have reduced their hours further from this week to pursue a new opportunity. This means, progress on all contract work will be hampered until we can find replacement developer resources. We are therefore actively looking for additional solidity developers resources, and the team has been actively contacting developers through our channels and networks. We are also open to recommendations and self-promotion, specifically of developers with experience in DeFi protocols; proficient in concepts such as pricing oracles, borrowing and lending, vaults, flash loans, etc.
If you have a recommendation, or are looking to submit a CV for yourself — please contact Mr Stiive on twitter (@MStiive) or telegram (@MrStiive). Please note that to prevent bots, generic developer spam in the main TG channel will not be tolerated, and will be deleted and may incur a permanent ban.
As stated above, our lead solidity developer reduced their hours to part time since the last update, however we were still able to make progress on the audit process. Our dev made a plan for updating the Chisel contracts in response to the audit, and has implemented about half of the body of work thus far, including the ‘critical’ bug requiring a transfer hook on the rewardSnapshot (this allows us to maintain composability whilst denying any exploit reward potential). The remaining items are mainly low priority, for things like gas optimizations.
As previously mentioned, we are actively looking for more developer resources to speed up the process of the remaining items — developers, please refer to the section above to get in contact.
Yearn Vault and Oracle Adapter
[As some of you might have guessed from the leaked asset icons in the TG…]
In the past fortnight, our team also worked on creating a new vault and oracle adapter to be able to implement Yearn Finance assets as collateral on Warp! Yearn Finance has recently implemented their own v2 vault system, including the new st-yCrv and and lp-yCrv boosted yield assets. Yearn assets are audited and well trusted, and now that we have compatibility the Warp team intends to list a number of high APY Yearn assets in preparation for the launch of Chisel.
As previously discussed at length in Warp Magazine #03, Warp platform intends to have a range of differing high APY pairs listed, where the yield exceeds the borrow interest rate. In this way, Warp borrowers can leverage yield farming, while providing industry leading high-APY interest in trusted tokens (USDC / ETH / etc) to suppliers with minimal and isolated risk.
For example, one freshly deployed pair currently under testing is “st-yCrv”, being Yearn’s new staked v2 derivative of $CRV. This vault is clearly trusted as it has already attracted $6.8M in TVL, and has managed to maintain an appealing high APY.
This asset currently has 36.11% APY, and we have listed the pair with a maximum borrow limit of 70% LTV, with the liquidation threshold at 75% LTV.
We also deployed a new high-APY interest rate model to accommodate new high-APY collateral pairs. The interest rate model looks as follows;
Let’s assume a user purchases $10k worth of st-yCrv, and borrows against it at 70% LTV on Warp at an interest rate of ~10% (~84% utilization as per above model), they user;
- Purchases $10k of st-yCrv earning 36.11% APY
- Adds the $10k of st-yCrv to Warp as collateral
- Borrows $7k USDC at 70% LTV
This means the user now has $7k liquid, and $3k locked up earning 36.11% APY on $10k ($3,611). However the user is paying 10% interest on their $7k loan, therefore incurring $700 in interest. So the user’s net profit is $3,611 — $700 = $2,911 on their $3k locked up, meaning a net APY of $2,911/$3,000 = 97.03% APY
If the user wished to increase their estimated net yield from $3,611, they could reinvest their $7k liquid in to st-yCrv to earn ~$6,138 per year, and then re-add to Warp to either lower their LTV (from 70% to ~41%) and reduce their chance of liquidation of the non-stable collateral, or to use it as collateral to borrow a further $4,900 from the Warp Vault (aka loop), thereby maintaining the same high leveraged APY. As previously mentioned, the team are working on automating this looping process.
You may note from the interest rate model a significant ‘kink’ at 80% utilization. This is done to incentivise borrowing <80% utilization, and to incentive adding more supply >80%.
For example, if the pool was just ~20% utilized, the interest rate would be closer to 1.5%. In this case, the leverage borrower would earn ~118% APY.
Likewise if the pool was near full utilization, the interest rate charged would be around ~33%, and thus the borrower would achieve just ~43% APY (not much better than just holding st-yCRV raw), while the suppliers would get a whopping 33% APY on their supplied USDC, paid in USDC with minimal and isolated risk (ie no liquidation risk, large liquidation buffer), therefore incentivising further suppliers to join. In all cases, Warp users are incentivsed to add more TVL to the platform.
Note that the collateral yield is self compounding, so the value of the collateral should increase over time (e.g. by 36% per annum if price is stable), and thus the LTV for the borrower should decrease over time. The loan can therefore be viewed as a ‘self-paying loan’.
Further to the yield, the borrower can also profit from the value of $CRV increasing, so could become highly profitable if/when “bull market” returns and the leveraged position increases in value (e.g. effectively a leveraged long on $CRV)
However, on the other side, if the value of $CRV goes down in value, or the peg of $yCRV is lost, the borrower could be adversely affected, and liquidated.
Therefore, a borrower may prefer an option for a ‘stable’ collateral and ‘stable’ debt, as per our next pair;
Another asset we have on boarded is yvCurve-MIM, the stablecoin of Abracadabra SPELL. This is currently yielding 7.16% APY on yearn, and we have coupled this with a high-leverage maximum borrow limit of 94% LTV, with 95% LTV being the liquidation threshold.
In this case, if a user added $10k of yvCurve-MIM to Warp and borrowed the maximum $9.4k at an example 5% APY, the user would net 41% APY with reasonably low risk of liquidation (MIM would have to drop to 95c, which it has only done a few times in its existence, if at all).
Considering that this represents a stable-stable loan (unlikely to be liquidated), and that MIM is overcollateralized (similar to Maker DAI) with a long performance history, we believe that the risk of long-term depegging is low, and the pair thus can provide a good risk versus return yield for Warp borrowers.
On the other hand, in this example, suppliers to this pair will be earning a healthy 5% on their USDC, with no risk of liquidation, and explicit counter-asset risk exposure.
More to come
All yearn vaults are audited, and therefore we believe these assets to represent safe high-APY self-compounding collateral assets for leveraging on Warp. However, although the vaults are deemed safe, contagion potential within the underlying assets still need to be assessed before pairs can be listed.
Eventually veWARP users would be able to decide whether the 5% APY on stablecoin MIM or the 10% APY on st-yCrv provides better risk-versus-reward, and assign capital accordingly. At the same time, borrowers will determine the utilization rates and thus the actual interest rates for suppliers. It is impossible to predict where the system will find equilibrium, and will instead rely on the free-market to dynamically find this.
Like all pairs, in the final state of v2, veWARP holders will firstly need to assign a max percentage of capital to the pair, and then borrowers need to validate this by accepting the risk-vs-reward and borrowing from the pair (they take the bulk of the risk), before any funds are exposed. Therefore, the risk-vs-reward is always monitored in real-time by the ‘decentralized brain’ of the aggregate of the free-market participants.
Furthermore, Warp team will be monitoring assets, and have admin controls to pause pair borrowing and lending if we believe any pairs pose risk to our users, thereby providing a quick feedback loop response.
Another interesting point to note is that Yearn recently launched their v2 vaults, and will soon implement their new tokenomic reward structure, $veYFI. This should bring additional attention to the yearn ecosystem, and will revamp the fee structure which may improve yields.
Further, new Yearn vaults are being listed all the time post Yearnv2 and in preparation for the new veYFI fee and reward structure. For example, the newly listed 1:1 pegged USDPax vault at 72% APY looks interesting as a high LTV high-APY self-compounding leverage farm 👀
For example, the 72.76% APY on USDPax with a 94% LTV borrow would yield a whopping 429% APY on a stablecoin farm for the borrower, whilst providing 50% APY low-risk USDC yield for suppliers/Chisel. Considering USDPax is redeemable 1:1 for USD, this could also be a recent risk-vs-reward pair to list, especially considering all Warp pairs on v2 are isolated.
Which yearn vaults would you like to see implemented? Please discuss in our community TG or discord.
With all these new high-APY pair options, when Chisel audits are finally complete, suppliers of Chisel will be able to experience low-maintenance high-APY, with governance evaluating the best risk versus reward exposure, by using the dynamic liquidity aggregation to optimize each of the utilization rates. Meanwhile, the borrowers in aggregate actively manage the exposed positions by taking on the bulk of the risk of the farming opportunity, in exchange for a cut of the yield (yield APY — borrow interest). In Magazine #03 we describe how this ecosystem mimics an actively managed mutual fund, with the Chisel suppliers being the investors, and the borrowers in aggregate (decentralized brain compute) being the money managers.
Frontend and backend improvements
We deployed some changes to differentiate our staging environment with the backend such that we can test new pairs on mainnet (not all are available on Goerli) without them appearing live on the main website. For example, the assets above are already deployed and being tested on mainnet in our staging environment. Further, we continued work on the dynamic APY calculations which estimate the APY post position (e.g. taking into account post borrow utilization), which has proved more difficult than we originally thought due to the provided mathematical model being incorrect. We also tried some gas optimizations for the borrow process, but work on this is ongoing.
We will continue to provide updates to the community, and the next update is scheduled for the first week of December (5th+).
Join one or more of our social platforms to follow our journey! We are creating DeFi’s first stablecoin lending protocol that enables and optimizes LP token deposits as collateral, and much more!