In its VaR calculations, JPMorgan Chase uses an expected tail-loss methodology which approximates losses at...

After VaR is computed at the 97% confidence level, the expected tail loss in excess of that confidence level is determined, which is then compared with the VaR estimate at the 99% confidence level.

After VaR is computed at the 99% confidence level, the expected tail loss in excess of that confidence level is determined, which is then compared with the VaR estimate at the 98% confidence level.

After VaR is computed at the 99% confidence level, the expected tail loss in excess of that confidence level is determined, which is then compared with the VaR estimate at the 99% confidence level.

After VaR is computed at the 1% confidence level, the expected tail loss in excess of that confidence level is determined, which and is then compared with the VaR estimate at the 98% confidence level.