Uncertainty & Risk
Optimal decision making requires not only forecasting expected future market conditions, but assessing the risk associated with uncertain and volatile market drivers. EnCompass employs advanced risk modeling that can be applied to any time series value, including but not limited to commodity prices, interest rates, production costs, demand, and renewable generation.
One of the primary options is a two-factor mean-reverting model, with short-term market volatility that tends toward a long-term fundamental outlook, which can also be uncertain.
- Probable distributions can be lognormal, suitable for prices which are positive and skewed higher, as well as normal or uniform.
- Historical distributions that draw from actual conditions that occurred in the past may be used for weather-related variables such as demand and renewable generation.
- Correlation can also be defined among time series to represent the inter-dependence on market drivers.
Sampling & Reporting
EnCompass uses Monte Carlo sampling to assess market and portfolio impacts due to uncertainty. This sampling can be completely random, or utilize Latin Hypercube Sampling, which takes points equally across a distribution and requires fewer draws for convergence.
EnCompass also automates traditional scenario analysis, which simulates all combinations of variables with a defined number of outcomes, such as high/base/low.
The Position report displays a complete picture of all possible outcomes and calculates and displays relevant statistics including mean, deviation, confidence intervals, and correlation factors. Scatter plots and box-and-whiskers charts provide a visual indication of relationships and distributions. Financial contracts such as fuel cost hedges can be added after simulations to assess the impacts on expected values and volatility.
EnCompass combines stochastic details with parallel processing and risk reporting to quickly and easily assess risk and confidence intervals.