How Sentinel Works
🚀 In One Minute​
Sentinel provides risk insights through a straightforward simulation approach:
- We run a what-if exercise on future prices for supported perpetual contracts.
- We use a standard baseline model (called GBM) that replays "typical" up-and-down moves from the last 180 days.
- We simulate holding a position for 10 days (240 hours).
- We support Linear (USD/USDT-margined) and Inverse (coin-margined) contracts.
- We include funding costs by taking the latest rate posted by the exchange in the most recent funding interval and applying it during the simulation.
caution
Bottom line: It's a scenario tool, NOT a prediction. It helps you see a range of possible outcomes, NOT the future.
Monte Carlo Simulations Explained​
Sentinel uses Monte Carlo simulations to provide risk insights. Here's how it works:
- Multiple Price Paths: We generate thousands of randomized price paths based on historical volatility.
- Statistical Analysis: We analyze these paths to calculate probabilities of different outcomes.
- Comprehensive Results: The results are summarized to show liquidation probabilities and potential PnL ranges.
We run 7,500 simulations for each analysis, which provides a robust statistical foundation for our risk metrics while maintaining computational efficiency.
Technical Foundation​
Our simulations are built on established financial risk models with specific parameters:
- Model: Geometric Brownian Motion (GBM) - a standard model for simulating asset price movements
- Calibration: Parameters are calibrated using the last 180 days of market data
- Time Horizon: All simulations project outcomes over a 10-day (240-hour) period
- Funding Rates: We incorporate the latest published funding rate from the exchange
The model captures the typical volatility patterns seen in crypto markets while maintaining computational efficiency.