Launch plan for SILK + Lend

I’d like to share the current launch plan for Silk and ShadeLend with the community, open the floor for feedback, and provide some transparency about how the backing of Silk will be handled going forward.

First, please make sure you are intimately familiar with the ShadeLend documentation. CDP-backed tokens like Silk are fairly complex, and we have iterated considerably on existing CDP models to vastly increase the reflexivity of Silk – the ability for Silk to respond to changes in supply and demand. The documentation can be found here: Lend - Shade Protocol

In addition, Silk has comprehensive documentation available here, which also provides necessary background knowledge for the topics discussed in this post: Silk - Shade Protocol

At launch, ShadeLend will provide the following vaults to the market:

  1. stkd-SCRT - 40% max LTV
  2. stATOM - 60% max LTV
  3. stOSMO - 45% max LTV
  4. stLUNA - 45% max LTV (if Stride makes it available in time for launch)
  5. IST - 85% max LTV
  6. DAI - 85% max LTV
  7. FRAX - 85% max LTV
  8. USDC - 85% max LTV

All vaults will offer introductory rates of 0% interest and 0% base borrowing fee. Borrowing fees have a fixed component, which is the minimum borrowing fee, and a variable component, which is based on redemption volume (more redemptions implies excess supply = higher variable borrowing fee to discourage further minting of Silk), so you may still be charged a borrowing fee if there is any redemption activity. These introductory rates will persist for two weeks, at which point fees will begin to increase once a week until they are at their final rates. These target rates will depend on vault utilization, so we are unable to provide that information at this time. For example, a vault that is not seeing a lot of borrowing activity will not have its interest raised as high as one that is seeing a lot of borrowing activity. In addition, a riskier collateral will have a higher target interest rate than a lower risk collateral.

Please note that if you borrow during the introductory period, you will not pay any fees upfront (other than the potential variable borrowing fee), but once fees begin to increase after the introductory period, your loan position will begin to accrue interest. You will not be subject to the additional borrowing fee unless you take out a new loan after the borrowing fees are increased.

The approximate range for interest rates will be 3-5% for volatile collateral, and 0.5%-2% for stable collateral. Per the documentation, interest rates can only change once every 7 days, and can only change by a maximum of 1% per 7 days. Any fee changes will be communicated on Discord, Twitter, and the Shade Forums.

Over time, the LTVs of these vaults may change as well. LTVs can be adjusted by the governance admin, but they can only be increased, not decreased. The LTVs are intentionally conservative at launch, but as liquidity on ShadeSwap and the Stability Pool deepens, we will raise the LTVs in accordance with the relative risk of the collateral to provide more capital efficient borrowing opportunities for users.

Going forward, we would like to add the following collateral to ShadeLend in a second phase post-launch:

  1. XMR
  2. ETH
  3. BTC
  4. stJUNO
  5. stSTARS
  6. ShadeSwap LPs

For XMR, ETH, and BTC, ShadeSwap does not support these pools at launch, so it does not make much sense to support them as collateral on ShadeLend until we have pools available on ShadeSwap.

For the other Stride staking derivatives, these are lower liquidity tokens, and fundamentally staking derivatives are higher risk than non-derivative tokens. It would be preferable to put off incorporating these derivatives until liquidity on ShadeSwap deepens and the market shows demand for these tokens.

For ShadeSwap LPs, LP vaults are slightly more experimental and require a little more development work to fully flesh out. In addition, there are some other risks associated with using LP tokens as collateral that we haven’t fully modeled yet. Using LP tokens as collateral is also dependent on pools having deep liquidity from a variety of LP providers to prevent large swings in LP value. As a result, we have to see how liquidity develops on ShadeSwap before we commit to accepting ShadeSwap LPs as collateral.

This is the current plan for the two-phase rollout of ShadeLend vaults. Beyond this, we would like to begin the governance handoff by empowering the community to make further adjustments beyond these initial conditions. The first phase of this would be signaling proposals where SHD holders can vote to change collateral parameters and launch new vaults. Fully automated governance functionality is considerably more development effort, which is why the first phase will be signaling proposals to introduce the ability for the community to impact the protocol directly. This initial signaling proposal functionality is expected to be rolled out in a couple of months after the launch of Swap and Lend, and after they have had some time to settle.

4 Likes

Very clear explanation and I like the two phased rollout.

Personally I’m most excited for the BTC & Eth collateral options. I’m glad you’re waiting for these pools to launch on swap and that you’ve made adjustments with depth of liquidity in mind.

Is there a timeline for when this second phase could start?

3 Likes

Creating new vaults and reconfiguring them is trivial, so from a technical perspective the second phase can be rolled out immediately (except for the LP vaults, which I will touch on.)

For the derivatives, the biggest thing we’d wanna pay attention to is how the staking derivative pools perform on ShadeSwap. We don’t want to accept any collateral on Lend that can’t also be traded on Swap, and we don’t want to oversaturate the market with too many derivative pools right away, so whenever those pools are live on Swap, we can add the vaults on Lend at the same time.

For XMR, ETH, and BTC, again it’s just dependent on when they’re live on swap. This is also a case of us not wanting to throw too many pools in everyone’s face at once, so once the initial set of pools have had their time to settle and we’ve acquired some baseline liquidity, those pools would be rolled out and then the vaults could be added at the same time.

For the LP vaults, the primary thing we need to do is finish some work on the dev side (not much, mostly shadeswap specific oracles and testing). From there, we’d be able to launch LP vaults very easily. The two big questions for LP vaults are:

  1. Can we safely support SILK LP? SILK LP is self referential (about half of the LP is SILK, so SILK ends up backing SILK), but that also may not be a problem since that liquidity is locked for the term of the loan. That SILK can still enter the market if people swap heavily in that pool and pull SILK out, but that just may mean we have to limit how much SILK is backed by SILK-containing LP. This is an open question we haven’t investigated too much.
  2. Can we ensure LPs still receive staking rewards when their money is used as collateral on Lend? We believe the answer is yes as we implemented this functionality on the Swap side, but we never fully tested it end-to-end. While this wouldn’t be a show stopper for supporting LP collateral on Lend, it would obviously not be good to force people to choose between using their LP as collateral or for staking rewards.
4 Likes

Looking good. Thanks for the explaination.

That part about silk backing silk is a interesting question. I am glad it is already on your radar.

2 Likes

I’m fully in support for the launch plan for Silk and ShadeLend, and am very pleased to see introductory rates offered for all vaults. It makes sense to have a multi-phase rollout as more complex collateral, functionalities, and generalized volatility and stability are introduced/observed. I have 2 main questions regarding the proposed launch plan:

  • How are starting Max LTV’s calculated?

  • Who/what is the Gov Admin?

1 Like

LTVs were chosen based on historical volatility of the assets weighted by market cap, with a big “risk premium” allowance which makes them much more conservative than that metric would suggest. this considerable risk premium is mostly because if the chain is down and we can’t liquidate, or if the stability pool empties out and we can’t liquidate positions fast enough, it gives more time for liquidations to happen safely if a coin is on a downward spiral. Gov admin is just the admin on the contract, it would eventually be a DAO contract that allows voting outcomes to automatically trigger contract executions, but that is the “fully automated governance” i mentioned in the OP which is quite a bit of work, so for now it would just be a secure multisig admin wallet owned by humans which has admin privilege on the contract.

2 Likes