[WBTC] - WBTC Collateral Request For Comment

This thread is to discuss the addition of WBTC as a collateral type on an accelerated timescale, in light of the State of the peg discussions. Three of the points which should be discussed are:

  1. Is WBTC, in principle, an asset that should one day be added to MCD, with some reasonable choices of risk parameters?
  2. Will adding WBTC as collateral in the very near future, with some reasonable choice of risk parameters, lead to DAI issuance against it which would increase the supply of DAI on the marketplace and help materially to ease the premium over $1.00.
  3. If yes, then which risk parameters should be chosen to avoid exposing the system to excessive risk while still ensuring sufficient DAI issuance?

For (3), the risk parameters that should be discussed are:

  • Debt ceiling
  • Liquidation rato
  • Stability fee
  • Auction parameters
18 Likes

To kick this off by giving my view on my own questions:

  1. BTC is the largest cryptocurrency by market capitalisation and trading volume, and has the most robust financial ecosystem. Custodial BTC tokens like WBTC are one way of securing BTC collateral on the Ethereum blockchain for use with MCD. Accessing BTC collateral via such a token allows MakerDAO to tap into a previously inaccessible market with massive size, even when compared to ETH. In particular, the size of the BTC-collateralised lending market is many times that of ETH, and open interest on BTC futures and swaps products is currently at almost $2.5b (while ETH futures and swaps total around $450mm), which evidences the demand for financing and leverage from BTC holders. Custodial assets like WBTC carry with them a new set of risks, which should be addressed by analysis of the custodian and the creation/redemption processes, and by setting a conservative debt ceiling to minimise exposure to any single custodian (or to a certain class of custodians). BitGo is one of the most reputable custodians in the cryptocurrency industry.

  2. I believe that it will. My evidence for this is:

    • the sizes of the BTC lending, margin trading, and derivatives markets relative to those of our current largest collateral type, ETH. In principle, this means that there could be 5 BTC leverage-seekers for each ETH leverage-seeker (in $ terms). No other asset, that is currently under emergency consideration or otherwise, comes close to BTC and ETH on this scale
    • if WBTC is added with a low stability fee (aiming to be level with or lower than the prevailing market borrowing rates for USD against BTC collateral), an attractive carry trade opens up where someone can lend USD against BTC collateral on the lending markets, financed by DAI issued from a WBTC CDP at a lower interest rate. This carry trade puts sell pressure on DAI/USD. There is an analogous basis trade using futures or perpetual swaps (or even the upcoming dYdX BTC perpetual swap).
    • speaking to traders and lenders who believe that with WBTC added as collateral, there will be a compelling case for BTC leverage seekers to mint WBTC for use in CDPs to take advantage of lower rates.
  3. I’d like to initially propose the following risk parameters for discussion:

    • 150% Liquidation ratio
      comment: our best heuristic for the volatility and liquidity properties of BTC is ETH
    • 1% Stability fee
      comment: the stability fee must match or be lower than market borrowing rates in order to incentivise BTC holders to convert to WBTC and use CDPs
    • 10mm DAI debt ceiling
      comment: our main control for mitigating custodian and redemption risk is the debt ceiling. The debt ceiling must be sufficiently high to allow DAI supply to increase to an extent where it would materially affect the market price. I think even with 5mm DAI minted it would serve as a valuable policy tool in this situation, so I would enthusiastically support a proposal where the debt ceiling was set lower as well.
    • All auction parameters/systems borrowed from the current parameters for WETH (including post-black thursday changes and in particular including the liquidation circuit breaker). BTC/USD oracles should be used for the price of WBTC. There is already a data model in setzer for BTC/USD.
      comment: same as with liquidation ratio, our best estimate for the BTC market is the ETH market
13 Likes

Nice, I think this is really well reasoned and I agree with all the points.

When it comes to the risk parameters I think we should try to onboard this through the full process - meaning they would have to be derived from the risk teams’ ratified general risk model in some form, in addition to community input - if the outcome is similar to these proposed risk parameters I would be very positive about it

5 Likes

wBTC is, to me, probably the most obvious addition that will contribute to the long-term success of this project. Consider this my support if the community is willing to bear the centralization risk. As with USDC, this risk can be mitigated somewhat by adding in other BTC-backed and tracked collateral types later. wBTC has been around for a while, and is generally less volatile than ETH or some of the other collateral types that have been discussed. I get the sense that people are looking for something safer than ETH in the current climate, and BTC has held up remarkably well through the downturn.

A quick squint at the smart contract reveals it’s a pretty standard ERC-20, so there shouldn’t be any roadblocks to the technical implementation. Despite current low liquidity of wBTC on global markets, demand from the protocol likely will cause more BTC to be wrapped to satisfy demand.

6 Likes

I think this market on deversifi (ethfinex/bitfinex) is quite relevant https://blog.deversifi.com/deversifi-lists-wbtc/ It would be interesting if someone could map out whether their claims are true, since that should make it quite a lot easier to deal with WBTC liquidations

3 Likes

Thank you for posting Lev–I really feel even more confident now that CoinList is adding wBTC and making it easier for folks to participate. I reckon that CoinList will be a major competitor to Coinbase and others in the long-term–they got the right stuff.

Regarding 1: I do agree that long term benefits of BTC on Maker are very significant. BTC doesn’t just bring new collateral from the existing user base of Maker, but it opens the door for new users that previously were not participants, such as BTC miners. Not to mention the size of BTC market vs ETH which Lev nicely explained.

Regarding 2: I do get the same sense from market participants that I talk to.
I think CoinList and Deversifi have been mentioned here before, but I want to echo that wrapping and unwrapping of WBTC has become significantly more convenient and scalable over the past couple months because of these initiatives. This has been and will continue to be a snowball effect.

8 Likes

Hi everyone, Ross from DeversiFi here.

Our wBTC/USDT market is mapped 1:1 against the Bitfinex BTC/USD market. We hold significant collateral both on Bitfinex and on chain to do facilitate this and manage any wBTC or BTC collateral imbalances in the background.

We think this is a missing bridge for DeFi and are pleased that we are in a unique position to offer this service to the community.

7 Likes

So, I am also in favour of this, and with Lev’s suggested parameters. Especially after the gov call. You can find more interesting info on wBTC here.

Just a summarising note: Mariano and Cyrus make the pertinent point that we need to make sure there are a viable and sustainable ways to run keepers for wBTC. This mostly means recognising that keepers always need DAI-in-the-hand in order to participate in the next round of collateral auctions. While this is easy with ETH (Oasis, Uniswap etc), it is more complicated with wBTC. This is exacerbated by the potential need to go through something like Coinbase on the BTC chain (if all liquidity and financing have dried up), which can increase confirmation times drastically and push the 6hr auction time uncomfortably close.

That said, I think a 10mm DC is reasonable enough, even given the above. Please feel free to correct/add to my summary if need be.

4 Likes

You wrote some solid points there Andy–agree with everything you posted.

With Regards to Keepers having DAI-in-hand, perhaps there should be a requirement for Keepers to have “Margin/Minimum” requirements of DAI-in-hand to be able to participate in wBTC liquidations? Realistically the business of “Keepers” in the long-term will fall into the hands of Institutional players, or those that have the means to participate.

Re process for Keepers to liquidate CDPs using WBTC, this is what they could do today:

  • a WBTC-collateralized CDP goes under 150% ratio.
  • a Keeper wins the auction, sends DAI to receive the WBTC locked in that CDP
  • Once WBTC is obtained, Keeper can send the WBTC to his CoinList wallet. Unwrapping is done instantaneously (we wait for 15 ETH blocks to credit the WBTC, so ~ 2minutes there, unwrapping is instantaneous) and BTC is credited in Keeper’s CoinList BTC wallet.
  • Keeper can immediately sell BTC for USDC on CoinList.
  • Keeper withdraws USDC and trade it for DAI. CoinList does not support DAI (yet) but many venues offer that pair.
2 Likes

It’s a good point you are making which is that it’s actually possible to avoid waiting for BTC confirmation times at any point, since it’s possible to trade BTC for USDC directly on Coinlist, and withdraw back onto Ethereum.

1 Like

Just to be clear here (and not hating, as above I am in favour) - the keeper would incur at 25bips fee when doing this, correct? Or is that fee only applied elsewhere for retail users?

Fees depend on 14-day volume:
â—Ź Up to 25 BTC: 0.20%
â—Ź 25 - 50 BTC: 0.18%
â—Ź 50 - 75 BTC: 0.16%
â—Ź 75 - 150 BTC: 0.15%
â—Ź > 150 BTC: 0.12%

So if Keeper does a 100 WBTC liquidation, assuming it’s the first trade they do on CoinList, they would be charged 0.15% for unwrapping WBTC into BTC.
The fee is applied on the notional: Keeper would send CoinList 100 WBTC and receive 99.85 BTC in exchange.

1 Like

The system is very much scalable, both in terms of fees coming down, and number of liquidity providers increasing as volumes grow…

Encouraging too much rates and basis trading is not without risks:

  1. If the BTC collateralized lending rates and/or swaps/futures funding rates decline due to lower market confidence, these vault holders would rush to exit the trade.
  2. Any break of the DAI peg upwards would make these positions enormously unstable.

While I agree that the DC is ultimately the most important risk factor to consider, I believe we should avoid setting the SF too low. Doing so could increase our overall risk of credit crises if wBTC vaults closed out their debt all at once.

Overall I think we should try to target vaults as much as possible towards long term holders over arbitrage traders.

3 Likes

This is a good point, thanks. I agree that there are stability implications from unwinding basis trades, which we ought to consider, and are inherent to any lending market. These concerns apply to the existing ETH collateral type, though presumably anyone who has wanted to put such trades on has had plenty of time to, so we could assume that this potential is close to exhausted until market conditions change. I think having the DAI supply increase now due to borrowing against WBTC, only to be contracted later if conditions change, probably puts us in not much worse a position than if we hadn’t issued the WBTC DAI in the first place, provided the debt ceiling was under control, don’t you think?

In general I think we should treat the WBTC SF similarly to the one for ETH.

It’s not important what keeper could do today, but what keepers are actually able to do today.
There are many steps in your recipe and are not trivial. Keepers performance was abysmal during last market crash.

Just a bit of context: on the governance call it was alluded that a BTC transaction to Coinbase is needed to complete the cycle and that such a transaction takes an hour (6 cfs). I think Matt wanted to point out that CoinList already has BTC/USDC pairs which would make this much faster than previously assumed.

Fair enough. Will check the call. Still, i wanted to point out that community seemed to forgot what happens when market quickly crashes. Imo 10m debt ceiling would be mistake (too high). Operating the keeper under normal market conditions is (relatively) easy, not so much when network is fully utilized and fees skyrocket.

2 Likes