MSTAR Desktop
The implementation of Exposure, Occurrence and Impact Method for structured risk scenario assessment.
Working with traders exposes a firm to rogue trading; selling
products exposes a firm to mis-selling and to anti-competitive
practices; operating in buildings exposes a firm to natural disasters or
terrorist attacks. From this point of view, operational risk is exposure based.
The design of a loss generation mechanism for a given scenario starts by identifying the Exposure for a scenario, i.e. the resources of the firm exposed to the event considered.
The design of a loss generation mechanism for a given scenario starts by identifying the Exposure for a scenario, i.e. the resources of the firm exposed to the event considered.
A scenario is then fully defined by the
resource (Exposure), the event (Occurrence), and the consequences
(Impact).
MSTAR Desktop implements the XOI
method and allows to you to design and quantify your own scenarios.
Design and quantification are kept independent in order to allow
independent quantification in different legal entities or business
units.
The distribution of potential losses are
calculated using simulation. Scenarios can be tested for what-if
analyses, and aggregated with various correlation methods (causal
dependencies, copulas, etc.)
The main features of MSTAR are as follows:
- Design: create the causal graph that links risk drivers to exposure, occurrence and impact.
- Quantification: evaluate the distribution of each driver, based on data or expert judgment.
- Simulation: sample all drivers in the scenario to assess the distribution of the scenario potential impact.
- What-if analysis: test alternatives on key drivers - stress-testing or risk mitigation.
- Aggregation: assess interdependencies between scenarios and simulate enterprise-wide risk.
MSTAR can also generate simulation scripts in Python, R or Matlab to integrate scenarios into external applications.