Readings

Protocol	Participation method	Returns	Risk
GMX	GLP - counterparty for each trade	70% of fees accrued
price of GLP → weighted avg of tokens + trader PnL + fees	Market risk, trader pnl, counterparty risk (default)
Gains	gDAI vault, GNS staking	GNS staking - 40% of market order fee and 15% of limit order fee ($\approx 32.5$% of trade fee)	trader pnl, counterparty risk (default)
Overlay	OVL	burn or mint OVL whenever trader makes negative or positive pnl	trader pnl, counterparty risk (default), potential death spiral
D8X	participation fund	earnings split between AMM fund $a$ and participation fund $p$ s.t. participation fund receives $\min(p/(a+p), 0.5)$	trader pnl, counterparty risk (default)
dydx	two-sided orders	weighted by the sum of (amount provided / spread)
Perp	liquidity pool (also holds impermanent position when long-short imbalance exists)
easy LP (bullish → liquidity in vUSD if price above range else used against shorts and vUSD converted to vBTC)	80% of generated fees	Market risk, trader pnl, counterparty risk (default)
Drift	DAMM liquidity pool
insurance fund	80% of fees, split pro-rata

insurance fund stakers receive 50% of revenue pool pro-rata (other 50% reserved for the protocol) | trader pnl, counterparty risk (default) | | Hubble | liquidity pool (used as countarparty for each trade) | trading fee reflected in favorable open price of impermanent position | trader pnl, counterparty risk (default) |

Todo

Notes

analysis of LP returns

motivation

perp exchanges benefit from having extra liquidity deposited that we can use to cover any risk arising from
1. long-short imbalance (inventory risk) and
2. undercollateralized positions (counterparty risk)
need to make sure that whenever LPs deposit their assets, they have a clear understanding of APY and risk

background

research on how LPs can participate in existing (perp) exchanges and what their rebates are e.g.) GMX, Gains, Overlay, dydx, Perp, Drift, Hubble, D8x etc.
- high-level design (+ on-chain analysis of realized returns)

D8x

Price is based on spot plus insurance premium (GBM assumption risk-neutral price with reduction to digital option pricing so that insurance premium is just probability of AMM default)
- One-period pricing, suggests calibration to typical holding period
Funding based on ewma mark and follows BitMex model
Default (insurance) fund sized via “cover n” which is ewma version of representative position size e.g. 5% of all traders
- assume stress scenario, where if the market is long heavy, price shoots up and $n$ shorts representative position defaults, defined similarly for short heavy market
Optimal AMM target size (collateral vault) based on a target insurance premium, set so default probability is low, and has baseline (profit above is sent to default) + stress size (default sends to AMM fund)