Risk Measurement
Employing Value at Risk (VaR), Simulation Models, and Balance Sheet Analysis to quantify market exposure.
Menu
Value at Risk (VaR) Engine
VaR estimates the maximum amount a bank could lose over a specific time period, with a given probability (e.g., 99% confidence level). It is a powerful summary figure for trading activity risk.
Institutional Applications:
- Regulatory Capital charges (Basel III/IV Compliance).
- Internal Economic Capital allocation across units.
- Performance-based compensation modeling.
- Establishing and monitoring Risk Appetite Limits.
Hierarchy of VaR Models
Simple VaR (Parametric)
Estimates sensitivity to small price changes (e.g., 1% rate shift) assuming a particular statistical distribution, typically normal distribution.
Probability BasedHistorical Simulation
A direct approach using historical data to revalue the portfolio at past market prices, effectively capturing extreme "tail" observations.
Direct ObservationSubjectivity vs. Objectivity
Friction often exists between objective VaR metrics and intuitive human judgement. A trader may see "Zero Risk" in a position due to central bank interventions, while models flag a limit breach.
The Balanced Solution:
Risk Limits define the boundaries of institutional safety (Objectivity), while Judgment enables selecting the optimal position within those boundaries (Subjectivity).
Asset-Liability Management (ALM)
Strategic management of the fixed/floating rate asset mixture to protect Net Interest Income (NII) against volatile rate shocks.
GAP Analysis
Fundamental identification of maturity mismatches between rate-sensitive assets and liabilities.
Simulation Analysis
Subjecting forecasted cash flows to price shocks, shifts, and non-parallel yield curve rotations.
Embedded Options
Managing liquidity gaps arising from early loan repayments or premature deposit withdrawals.
Contingency Hedging Strategy:
Core Balances
Utilizing stable current account balances as a natural hedge.
Covenant Controls
Stipulating penalties to mitigate embedded option risk.