Margin System

Diffusal uses a portfolio margin system inspired by CME SPAN (Standard Portfolio Analysis of Risk). Unlike isolated margin where each position is collateralized independently, portfolio margin allows netting of positions and calculates collateral requirements based on worst-case portfolio loss under stress scenarios.

---

Cross-Margin Architecture

Users do not collateralize individual positions. Instead, all positions share a single collateral pool with deposited USDC and multiple option positions across different series.

This enables capital efficiency: a long call can offset some of the risk from a short put, reducing total margin requirements.

---

Portfolio Equity Calculation

The core health metric is Portfolio Equity:

$$\text{Portfolio Equity} = \text{Deposited Collateral} + \text{Unrealized PnL}$$

Unrealized PnL

For each position, unrealized PnL is calculated against current mark price:

$$\text{Unrealized PnL} = \sum_{i} (\text{current\_mark}_i - \text{entry\_premium}_i) \times \text{position\_size}_i$$

For long positions (size > 0):

• Positive if option value increased (mark > entry)
• Negative if option value decreased

For short positions (size < 0):

• Positive if option value decreased (obligation worth less)
• Negative if option value increased (obligation worth more)

Example

If deposited collateral is $5,000:

$$\text{Equity} = \$5,000 + \$200 = \$5,200$$

---

Stress Testing (SPAN-like)

The margin system uses stress testing to determine required collateral. We test the corners of the risk space to find worst-case portfolio loss.

Stress Scenarios

We apply shocks to both spot price and implied volatility:

This creates 4 corner scenarios:

Why These Values?

• ±30% spot: Captures significant market moves (e.g., crypto flash crashes, major news events)
• +50% IV: Volatility spikes are common during market stress (VIX can double in hours)
• -30% IV: Vol crush after events resolve (post-earnings, post-FOMC)

Stress Loss Calculation

For each scenario:

1. Recalculate all option premiums with stressed spot and IV using Black-Scholes
2. Compute portfolio value change vs current mark
3. Track the maximum loss across all scenarios

$$\text{Stress Loss} = \max_{s \in \text{scenarios}} \left( \sum_{i} (\text{current\_mark}_i - \text{stressed\_mark}_{i,s}) \times \text{position\_size}_i \right)$$

---

Required Margin Calculation

Initial Margin (IM)

$$\text{Initial Margin} = \text{Stress Loss} + \text{Adverse PnL Buffer} + \text{Notional Buffer}$$

Where:

• Stress Loss: Maximum loss across stress scenarios
• Adverse PnL Buffer: $\max(0, -\text{Unrealized PnL}) \times 1.05$ (covers current losses with 5% buffer)
• Notional Buffer: $\text{Total Notional} \times 0.15$ (15% of total position notional)

Maintenance Margin (MM)

$$\text{Maintenance Margin} = \text{Initial Margin} \times 0.80$$

Health States

---

Worked Example

Consider a portfolio with ETH at $3,000 spot and 50% IV:

Positions

Deposited Collateral: $2,000

Step 1: Current Premiums (Black-Scholes)

Using Black-Scholes with spot=$3,000, IV=50%, r=5%, T=30/365:

Step 2: Unrealized PnL

Step 3: Stress Losses

Testing scenario 1 (Spot -30%, IV +50%):

• Stressed spot: $2,100
• Stressed IV: 75%

After testing all 4 scenarios, maximum stress loss: $843.45

Step 4: Required Collateral

$$\text{IM} = \$843.45 + (\$315.55 \times 1.05) + \$750 = \$1,924.78$$ $$\text{MM} = \$1,924.78 \times 0.80 = \$1,539.82$$

Step 5: Health Check

$$\text{Equity} = \$2,000 - \$315.55 = \$1,684.45$$

Since $1,684.45 >$1,539.82 (MM), the position is Healthy.

---

Risk Parameters

---

Mark Price vs Trade Price

The margin system uses mark price (Black-Scholes theoretical) for all calculations, not the trade price:

This prevents manipulation: a user cannot avoid liquidation by submitting fake orders at favorable prices.

Entry Premium Recording

When a trade executes, the actual trade price is recorded as entry premium. For example, if a user buys 10 contracts via RFQ at $51.50 each, the entry premium is$51.50. Later, if the current mark price (from Black-Scholes) is $50.20, the unrealized PnL would be ($50.20 - $51.50) × 10 = -$13.00.

The user paid above theoretical (the spread), so they start with a small unrealized loss.

---

Integration with Trading

Pre-Trade Check

Before any trade, the system verifies:

1. New position margin: Would the post-trade portfolio be healthy?
2. Sufficient collateral: Does user have enough deposited USDC?

The trade execution updates positions for both buyer and seller, then performs a post-trade margin check to ensure both parties remain healthy. If either fails the margin check, the transaction reverts.

Deposit and Withdrawal

• Deposit: Always allowed (increases equity)
• Withdrawal: Only if post-withdrawal Equity ≥ Initial Margin

$$\text{Withdrawable} = \max(0, \text{Equity} - \text{Initial Margin})$$

---

Gas Optimization

The margin calculation is gas-intensive due to Black-Scholes evaluations. Optimizations include:

Target gas cost: ~50,000 gas per margin check on high-throughput EVM chains.

---

Contract Implementation

The margin system is implemented across several contracts:

---

Related

• Margin Calculations — Mathematical formulas and derivations
• Liquidation — What happens when equity falls below MM
• Black-Scholes Model — Pricing used in stress calculations