Skip to main content

Risk Layers

Canon runs six independent risk models in parallel against real-time market data. Each model analyzes a different dimension of trade risk.

1. Order Book Slippage​

Walks the live L2 order book to estimate the actual fill price for your trade size. This accounts for thin books where large orders would eat through multiple price levels.

What it measures:

  • Expected slippage in basis points
  • Liquidity depth at 50bps and 100bps from mid price
  • Order book imbalance (bid vs ask side weight)

Why it matters: A $200K market buy on a book with only $50K of resting asks within 50bps will fill at a significantly worse price than the quoted mid.

2. Liquidation Proximity​

Calculates how far the mark price would need to move before the position (or account) is liquidated. Accounts for existing positions, margin usage, and the exchange's tiered margin schedule.

What it measures:

  • Liquidation price
  • Distance from mark price to liquidation (percentage)
  • Margin usage as percentage of available margin
  • Effective leverage including existing positions

Why it matters: A position with 3% distance to liquidation on a volatile altcoin can be wiped out by normal price action.

3. Funding Rate Analysis​

Analyzes current funding rates and projects annualized costs. Compares Hyperliquid funding to CEX rates to detect divergence that may signal crowded positioning.

What it measures:

  • Current 8-hour funding rate
  • Projected annualized cost
  • Divergence between Hyperliquid and CEX rates
  • Who pays: longs or shorts

Why it matters: An annualized funding rate of 50%+ quietly drains PnL. Divergence from CEX funding can indicate Hyperliquid-specific positioning imbalances.

4. Market Impact​

Estimates how much the trade itself would move the market price, based on trade size relative to recent volume.

What it measures:

  • Expected price impact in basis points
  • Trade size as a percentage of recent volume (participation rate)

Why it matters: Large trades on low-volume assets can move the price against you before your order is fully filled.

5. Value-at-Risk (VaR)​

Calculates the statistical worst-case loss over a 24-hour horizon at 95% and 99% confidence levels. Uses historical return distributions adjusted for skewness and fat tails.

What it measures:

  • 24h VaR at 95% and 99% confidence (percentage and USD)
  • Expected Shortfall (CVaR) - average loss in the worst cases
  • Leverage-adjusted VaR accounting for margin

Why it matters: VaR gives a probabilistic bound on downside. A leveraged position on a high-volatility asset can have a 99% VaR exceeding the margin posted.

6. Open Interest​

Checks how close the asset's open interest is to the exchange's capacity cap. When OI is at or near cap, new positions may be rejected or existing positions face increased liquidation risk.

What it measures:

  • Current open interest in USD
  • Whether the asset is at the exchange's OI cap
  • OI utilization as a percentage of estimated cap

Why it matters: Assets at OI cap cannot accept new positions. Assets near the cap face heightened liquidation cascade risk as the exchange may reduce leverage limits.

Portfolio-Level Analysis​

In addition to per-trade analysis, Canon evaluates portfolio-level risk when the wallet has existing positions:

  • Concentration risk - Whether a single asset dominates the portfolio
  • Correlation exposure - Whether multiple positions move together, amplifying drawdowns
  • Portfolio VaR - Combined Value-at-Risk across all positions accounting for correlations
  • Stress testing - Portfolio performance under historical crash scenarios