Chapter 36: Key Takeaways

DeFi Integration and Liquidity Mining

Composability is DeFi's defining property: any smart contract can call any other without permission. Prediction market outcome tokens flow through DEXes, lending protocols, yield aggregators, and derivatives platforms as first-class financial assets.
The three levels of composability are morphological (standardized token interfaces), atomic (multiple protocol interactions in one transaction), and syntactic (well-documented callable functions).
Outcome tokens can be used as collateral in lending protocols, traded on DEXes, deposited in yield aggregators, and bundled into structured products.

Prediction market AMMs (constant product, LMSR, etc.) require liquidity providers who deposit token pairs (YES/NO or YES/collateral) into pools.
LPs earn trading fees proportional to their pool share and the trading volume. Fee APR = (daily volume * fee rate / TVL) * 365.
The implied probability in a constant product YES/NO pool is: $p_{YES} = x_{NO} / (x_{YES} + x_{NO})$.
Capital efficiency is lower than in standard DeFi pools because prediction market tokens have bounded value ranges (0 to 1).

Impermanent loss (IL) is the difference between holding tokens in a pool versus holding them in a wallet. It occurs whenever the price ratio changes from the entry point.
In prediction markets, IL is maximized at resolution because one token goes to 1 and the other to 0 --- the maximum possible price divergence.
For a constant product YES/NO pool entered at price $p_0$, the IL at price $p_1$ is: $IL = 2\sqrt{p_1(1-p_1)} - 1$.
IL is zero only when the price returns to the entry level. It increases monotonically as the price moves away from the entry point in either direction.
LPs must exit before resolution or accept catastrophic IL. A 5-7 day buffer before expected resolution is a reasonable guideline.

Trading fees: The base yield source. Depends on volume/TVL ratio and fee rate.
Liquidity mining: Protocol-distributed governance tokens provide additional yield (often 10-50% APR) but carry token price risk.
Lending yield: Outcome tokens deposited as collateral or lent directly earn interest.
Stacked yield: Layering multiple yield sources (fees + mining + lending) amplifies returns but also amplifies risk.
The break-even condition for an LP: total fee income + mining rewards > impermanent loss. This is more likely in high-volume markets with higher fee rates.

Impermanent loss risk: The primary risk for prediction market LPs. Worse than standard DeFi IL because resolution creates guaranteed maximum divergence.
Smart contract risk: Composing multiple protocols multiplies the attack surface. A bug in any layer can drain funds.
Governance token risk: Mining rewards denominated in volatile tokens may lose value before the LP exits.
Oracle risk: If the underlying prediction market's oracle fails or is manipulated, all LP positions are affected.
Resolution timing risk: Unexpected early resolution can catch LPs who planned to exit later.

Flash loans enable uncollateralized borrowing within a single transaction. They can be used for arbitrage (buying tokens on one platform and selling on another) or for attacks (manipulating oracle prices).
MEV (Maximal Extractable Value) affects prediction market traders through front-running, back-running, and sandwich attacks.
Sandwich attacks: A searcher front-runs a large trade (buying before the victim), then back-runs (selling after), profiting from the price impact.
Defenses include private mempools (Flashbots), slippage limits, TWAP oracles, and off-chain order matching (as used by Polymarket).

Range tokens: Pay out if a prediction market price stays within a specified range.
Options on outcomes: Call/put options on outcome tokens. Black-Scholes is inappropriate due to the bounded, non-lognormal distribution of outcome token prices.
Leveraged exposure: Split collateral, sell one side, deposit the other as collateral in a lending protocol, and repeat. This creates synthetic leverage.
Insurance products: Protocols like Nexus Mutual can insure against smart contract failure in prediction market protocols.

Formula	Description
$p_{YES} = \frac{x_{NO}}{x_{YES} + x_{NO}}$	Implied probability in constant product pool
$IL = 2\sqrt{p_1(1-p_1)} - 1$	Impermanent loss for YES/NO pool
$\text{Fee APR} = \frac{V \times f}{TVL} \times 365$	Annualized fee yield
$\text{Net APR} = \text{Fee APR} + \text{Mining APR} - \text{IL rate}$	Net LP return
$\text{Sandwich profit} = \Delta p \times q - 2 \times \text{gas}$	MEV sandwich attack profit