When we don't really know how the object of regulation looks like, it might be useful to research it first or in parallel, and here CBs do have an advantage.
Almost every CB in Europe or USA now is dying to hire 'macro-prudential' specialists. The problem is - no one really knows how these specialists would look like...
There was a WSJ feature several days ago on introducing financial considerations (fin markets, banking industry, housing, frictions of different kinds) into DSGE (mostly) models. This is an extremely active area of research right now (couldn't get enough students to engage), shows some promise.
