Estadísticas › Backtesting y evidencia › VaR in practice
Bank risk models were violated zero times before the crisis and ten per cent of days during it
| Bank | Before the crisis | During the crisis | Crisis coverage p value |
|---|---|---|---|
| 1 | 0.4% | 4.0% | 0.000 |
| 2 | 0.0% | 10.0% | 0.000 |
| 3 | 0.1% | 3.4% | 0.000 |
| 4 | 0.0% | 5.1% | 0.000 |
| 5 | no informado | 2.7% | 0.005 |
Daily exception rates against a nominal one per cent one day measure, five large US banks mostly using historical simulation, with the crisis period starting June 2007. Exact pre crisis start dates and the crisis end date are withheld for confidentiality. Over the full sample all but one bank's exception rate was not significantly different from one per cent, so the failure is entirely a within crisis clustering phenomenon.
Five large banks' one day risk measures at a 99 per cent level produced exception rates of zero to 0.4 per cent before June 2007 and 2.7 to 10.0 per cent during the crisis, with coverage rejected for every one of them.
Solo para miembros
El resto de esta página es para miembros
Below this point there are 5 sections, 1 table and 4 named sources, roughly 900 words of it. Every figure carries the source it came from and the date the data is from.
Puedes seguir consultando todas las estadísticas de la biblioteca gratis. La introducción y el resumen están siempre abiertos.