Jared Diamond, Gabriel Bianconi and Alvin Hsia
Hi All, despite GIP-11 passing, market conditions have led to a change in the quoted terms from FRAX. The APR increased from 5% to 6%. Furthermore, given market volatility, we have suggested decreasing the starting TVL figure to (i) provide time for price discovery and (ii) limit potential downside from impermanent loss.
The two updates to the proposal:
- An increase in FRAX APR: 5% to 6%
- Decrease in TVL from $5MM to $1.5MM
As these are material economic changes, we have updated the original proposal to seek community feedback and an updated vote.
- This post is an update from GIP-11 to reflect (i) an updated view on the market (deploying $1.5MM vs $5MM) and (ii) FRAX’s APR moving from 5% to 6%
- Goldfinch enters a partnership with Ondo and Frax to bootstrap DEX liquidity for the GFI token
- Currently, the only DEX liquidity for GFI is GFI/WETH on UniswapV2, which has ~$118k in total liquidity, which is insufficient for any meaningful trade size (~20% price impact on a $10k swap)
- Goldfinch to provide $750k GFI to be paired with $750k FRAX at a fixed 6% APR on Uniswap V2 for 90 days, totaling $1.5MM of GFI/FRAX liquidity
- This partnership costs significantly less than Goldfinch incentivizing liquidity through token emissions
- Goldfinch will earn trading fees (typically 30bps on Uniswap) in exchange for taking on impermanent loss risk
Today, the GFI token has very limited liquidity on DEXs. The GFI/WETH pair on Uniswap only has $118k in total liquidity, which results in very high price slippage for users (~20% price impact on a $10k swap).
This means GFI is basically untradeable on DEXs. We want to bootstrap DEX liquidity for GFI in a cost-effective manner (e.g. without conducting a highly inflationary GFI liquidity mining program).
Ondo Finance has a product called Liquidity-as-a-Service (LaaS) that enables DAOs to take out a low interest rate loan from a stablecoin protocol (e.g. FRAX, FEI) to deposit into one side of a liquidity pair, along with their native token (which is provided by the DAO). This allows DAOs to bootstrap liquidity on DEXs without needing to conduct highly inflationary liquidity mining programs.
Ondo has successfully conducted several LaaS programs with other DeFi protocols such as SYN, NEAR, FOX, UMA, GRO, and MTA over the last couple of months.
Note: The diagram above shows FEI as the stablecoin used for LaaS, but Ondo has entered partnerships with other stablecoin protocols (i.e., FRAX) as well since the post was published.
The basic mechanics of LaaS are:
- Goldfinch takes a 90 day loan from Frax of FRAX (a stablecoin pegged to $1 USD) at a fixed 6% APR
- Frax sends the FRAX loan amount to an Ondo vault, forming one side of the liquidity pair
- Goldfinch sends an equivalent amount of GFI to an Ondo vault, forming the other side of the liquidity pair
- Ondo deposits GFI and FRAX into UniswapV2 as a liquidity pair for 90 days
- At the end of the 90 days, Goldfinch pays back the FRAX loan with interest
- Goldfinch keeps all trading fees on the LP, and assumes any impermanent loss that occurs
- To account for impermanent loss:
If the price of GFI goes up during the 90 day period
1. Goldfinch will end up with more FRAX and less GFI 2. Goldfinch can buy back GFI with excess FRAX
If the price of GFI goes down during the 90 day period
1. Goldfinch will end up with more GFI and less FRAX 2. Goldfinch can sell GFI to pay back the FRAX loan
With the proposed $1.5MM in total GFI-FRAX liquidity, we would expect price slippage on a $10k swap to improve from the current ~20% price impact to a much more acceptable ~1.3% price impact.
If a user would like to swap in and out of GFI with a token other than FRAX (e.g. USDC or ETH), Uniswap and DEX aggregators will auto-route through FRAX pairs with other assets.
Below we examine two extreme price scenarios for GFI over the 90 day LaaS period.
Example: GFI price doubles from $1.80 to $3.60 during the 90 day period and assuming avg. daily utilization of ~3% @ 30bps swap fees
- Total value: $1.5MM → $2.14MM
- Avg Daily Swap Rev - $135 >> $158
- Impermanent loss: -5.7% (~$114k); total value would have been $2.25MM if no LP
- GFI balance Δ: -119,789 (-29%)
- FRAX balance Δ: +316,735 (+42%)
- Frax Assets/Frax Liabilities (I & P): 140%
- Frax Surplus / GFI Buyback power: 305,882 FRAX | $305,882
- Ending GFI Balance: 296,878 GFI | $1,068,760
- Net Ending Asset Value: $1,374,642
- Junior Tranche ROI 83%
Example: GFI price halves from $1.80 to $.90 during 90 day time period and assuming avg. daily utilization of ~3% @ 30bps swap fees
- Total value: $1.5MM → $1.07MM
- Avg Daily Swap Rev - $135 >> $113
- Impermanent loss: -5.7% ($54k); total value would have been $1.13MM if no LP
- GFI balance Δ: +177,089 (+43%)
- FRAX balance Δ: -213,595 (-28%)
- Frax Assets/FRAX Liabilities: 71%
- FRAX Shortfall/Negative Loan Balance: 224,448 FRAX | $224,448
- Gross GFI Balance: 593,756 GFI | $534,380
- FRAX buyback with GFI: 249,387 GFI I | $224,448
- Net Ending GFI balance Δ: 344,368 GFI | $309,932
- Junior Tranche ROI: -59%
These numbers reflect the hypothetical performance of the GFI-FRAX LP position. At maturity, Goldfinch will be expected to assist with rebalancing the assets as needed. The buyback of FRAX with GFI in the above example does not take into account price impact or the availability of third-party on-chain sources of liquidity to execute the trade. This may take an OTC transaction to settle.
- Goldfinch bootstraps DEX liquidity for GFI in a cost-effective and low-commitment manner
- After this 90 day period, Goldfinch can evaluate whether to continue rolling over the LaaS vaults or explore other liquidity solutions for the GFI token
- Goldfinch is required to pay $10,854 of loan interest in FRAX over the 90-day period regardless of price action in LP
- Goldfinch may incur some impermanent loss, although these costs are still relatively small (compared to a liquidity mining program) as seen in the extreme scenarios illustrated above
YES – Enter the LaaS program with Ondo and Frax, with terms outlined in this proposal.
NO – Do nothing.