So how are these stress tests different?

It will be interesting to see the results of the stress tests for the banks. Whywould they be any different from the stress tests the banks shld have been conducting as part of their regular risk management process? Or is that they never did it? or ignored the results? Or they just figured the long tail events would never take place?

I am asking a fundamental question here - does the manner of identification and measurement of risk need to be changed? Do the VaR calculations need to be done in a different way? Seems to me - not different but in a more detailed way now. So that all kinds of exposures are identified and captured in the formula.