ChristoffersenBacktestForUnconditionalCoverage function

Christoffersen Backtest for Unconditional Coverage

Carries out the Christiffersen backtest for unconditional coverage for a VaR risk measurement model, for specified VaR confidence level.

ChristoffersenBacktestForUnconditionalCoverage(Ra, Rb, cl)

Arguments

• Ra: Vector of portfolio profit and loss observations
• Rb: Vector of VaR forecasts corresponding to PandL observations
• cl: Confidence level for VaR

Returns

Probability, given the data set, that the null hypothesis (i.e. a correct model) is correct.

Examples

# Has to be modified with appropriate data:
   # Christoffersen Backtest For Unconditional Coverage for given parameters
   a <- rnorm(1*100)
   b <- abs(rnorm(1*100))+2
   ChristoffersenBacktestForUnconditionalCoverage(a, b, 0.95)

Author(s)

Dinesh Acharya

References

Dowd, K. Measuring Market Risk, Wiley, 2007.

Christoffersen, P. Evaluating interval forecasts. International Economic Review, 39(4), 1998, 841-862.