Financial Risk Manager Part 2
Get started today
Ultimate access to all questions.
In the context provided, where a bank’s derivative trading desk currently employs a Value at Risk (VaR) model that utilizes historical simulation with a 3-year look-back period and applies equal weighting to past returns, a new VaR model has been developed. This new model employs the delta-normal method, where volatilities and correlations are estimated using the RiskMetrics Exponentially Weighted Moving Average (EWMA) method over the past 4 years. During a 6-week parallel run, the new model reported no exceedances and consistently estimated lower VaR values compared to the old model. Following this, the model evaluation team, after a brief assessment led by a junior analyst, quickly decided to adopt the new model. Given this scenario, what is the correct conclusion to draw regarding the model replacement?
In the context provided, where a bank’s derivative trading desk currently employs a Value at Risk (VaR) model that utilizes historical simulation with a 3-year look-back period and applies equal weighting to past returns, a new VaR model has been developed. This new model employs the delta-normal method, where volatilities and correlations are estimated using the RiskMetrics Exponentially Weighted Moving Average (EWMA) method over the past 4 years. During a 6-week parallel run, the new model reported no exceedances and consistently estimated lower VaR values compared to the old model. Following this, the model evaluation team, after a brief assessment led by a junior analyst, quickly decided to adopt the new model. Given this scenario, what is the correct conclusion to draw regarding the model replacement?
Explanation:
The correct conclusion for the replacement of the VaR model is option C: "Overnight examination by the junior analyst increased the desk's exposure to model risk due to the potential for incorrect calibration and programming errors." This is because the new model was quickly implemented and underwent an insufficient testing period, which raises concerns about the accuracy and reliability of the model. The lack of exceedances in 6 weeks does not necessarily mean the new model is unbiased; it could also indicate that the model is underestimating risk. The hurried evaluation by a junior analyst, instead of a thorough review by a senior team member, further increases the risk of model errors. This situation is reminiscent of the JP Morgan London Whale case in 2012, where a new VaR model was hastily introduced without adequate testing, leading to significant losses. The key takeaway is that proper model validation and testing are crucial to ensure the accuracy and reliability of risk management models.