I think in the other thread about this, someone posted something to the effect that the spreadsheet with the flawed formula was not questioned or tested because it provided justification for the kind of reckless, short-sighted risk-taking behavior that the decisionmakers at the bank wanted to promote anyway. Spot-on. In organizations like this, the decisionmakers will always massage models and abuse statistics to support the decision they already know they want to make anyway. That's why it's important that they are forced into a structure where testing and oversight are required.