I must confess some ignorance on the exact nature of the stress tests being conducted in Europe. When we conducted ours, the criteria were laid out in a fairly robust and comprehensive manner. Everyone could see what assumptions were being used and, though many scoffed at the scenarios as being too rosy, it would appear that there was some measure of success.
From what I have read, such as this post from Naked Capitalism, the tests do not include possible losses that the banks would incur from sovereign debt defaults. In other words, say the fictional Bismarck Financial had $50 billion in equity, but $75 billion in holdings of sovereign debt from Greece, Portugal, and Spain, they would be wiped out in a rolling series of debt defaults. However, under the stress tests (as I understand them), these exposures would be excluded. That would have been similar to if in our own stress tests we had excluded the possibility of further house price declines and rising mortgage defaults. As it turned out, the estimates for house price declines in our own stress tests were actually much more pessimistic than what panned out.
This brings me to a point here for those looking at Europe right now. Our own stress tests, despite being widely mocked by the more pessimistic analysts and observers, actually did a good job of quantifying the unquantifiable. The most dangerous aspect of early 2009 was that no one knew the exposure the banks were facing and financial markets froze under the uncertainty. It was nearly impossible to raise capital to replenish the equity positions of these companies. The stress tests provided a benchmark from which investors could use differing assumptions on whether or not a bank could be saved and decide then whether to supply it with more capital. As it turned out, as I recall anyway, not one of the major banks tested has since failed or had to be swallowed up by another in an emergency basis in the way that National City and Wachovia were in late 2008.
This was all possible because the standards being used were transparent and they were actually fairly sensible. I liken it to if a company forecasts that based on 4% GDP growth they expect to earn $4.50 a share. You can say that's nonsense and you think the economy will only grow 1% so they will earn less, but it's all out in the open and the market can decide, based on good information, what price capital can be raised at. From this perspective, the European stress tests do no seem adequate and I think it would be wrong to apply the relative success of our own experience to Europe.