This is the Cato Daily Podcast for Friday, June 5th, 2015. I'm Caleb Brown.

Central Bank stress tests of the banking system aren't credible because no central banks want to say that the banking system is less than solid.

And the models used to test the banks are routinely gamed.

That's the conclusion of Kevin Dowd, author of alchemists of Loss.

He's also a Cato Institute adjunct scholar. Endowed, author of Alchemists of Loss.

He's also a Cato Institute adjunct scholar,

we spoke today.

The idea of splitting a sandwich with somebody

involves quite often this simple set of rules which is I cut you choose and that gives me

a clear incentive to deal as fairly as possible because I know that you're going to choose the bigger half.

It seems that when you make those kind of models more complicated, maybe they offer more opportunities for people to game them.

And that's essentially the point I think that you're making when it comes to stress testing of banks undertaken by Central Banks around the world.

So what are the problems that you see with the stress tests that have been undertaken here in the United States by the

Federal Reserve and I assume the ECB and other Central Banks?

Well the first point is that models are very, very gameable.

They're very easy to game, and it's very easy to hide the gaming from the client who might

be essentially the taxpayer. So for example if you had a bunch of mortgage junk you

can easily fiddle the probability of default in the in the portfolio you can fiddle with the sample that you choose.

It's very easy for an expert to calibrate the model in a way that appears to be scientific

and objective but really fulfills the ulterior objective

of ensuring that junk looks like AAA.

Yeah, so in one specific example, and this goes, this requires a lot of unpacking.

...