Raven MarketDOCS

Arctan AMM

2 min read

The pricing engine behind Raven — a bounded, self-balancing automated market maker built specifically for digital options.

The Problem

In any digital options protocol, if all traders bet on the same side, the protocol becomes insolvent — there isn't enough capital on the losing side to pay winners. Traditional solutions require centralized market makers, position caps, or outright trade rejection. None of these work for a decentralized, permissionless protocol.

The Solution — Arctan Pricing

Raven's AMM uses the arctangent function to price payout odds dynamically based on real-time imbalance between CALL and PUT volume in the pool. As one side gets crowded, that side's odds decrease and the opposing side's odds increase — automatically, algorithmically, with no human intervention.

P(x) = ½ + (1/π) · arctan(α · x)
P(x) = implied probability  ·  x = signed CALL–PUT imbalance  ·  α = sensitivity

The arctan function is naturally bounded between 0 and 1, making it directly interpretable as a probability. At equilibrium (equal CALL and PUT volume), P = 0.5, reflecting maximum uncertainty. CALL and PUT odds always sum to 1 for any given strike and expiry.

What This Means for Traders

The α Sensitivity Parameter

α controls how aggressively odds shift in response to pool imbalance. Higher α means premiums move more sharply as one side gets crowded; lower α produces more stable pricing. α is set per market by the protocol and may evolve as liquidity matures.

ℹ️ For Traders
The OI Balance indicator in the options chain shows the current CALL vs. PUT skew for each strike. Heavy one-sided skew means the market is already leaning that way — and odds have adjusted accordingly. Contrarian positions often carry better expected value in crowded markets.