Explanation

Historical Simulation Method for VaR:

• Portfolio value: USD 800 million
• Number of observations: 400 days
• Confidence level: 99%

Calculation: $\text{Observation number} = (1 - \text{confidence level}) \times \text{total observations}$ $\text{Observation number} = (1 - 0.99) \times 400 = 4$

This means we need the 4th worst loss from the ordered returns.

From the given returns (ordered from highest to lowest), the worst returns are:

• -1.91% (worst)
• -1.87% (2nd worst)
• -1.84% (3rd worst)
• -1.82% (4th worst)

VaR calculation: $\text{VaR} = \text{Portfolio value} \times \text{4th worst return}$ $\text{VaR} = 800,000,000 \times 0.0182 = 14,560,000$

Therefore, the 99% VaR is USD 14.56 million

Note: The 4th worst return is -1.82%, not -1.76% which appears earlier in the sequence.