To test the effect of rating changes on share prices as outlined in our first four hypotheses, we use an econometric-based event study method outlined by Campbell, Lo, and MacKinlay (1997). The idea is to compare firm specific actual returns to normal expected returns surrounding a predefined event. The event in testing our first two hypotheses is simply the announcement of a rating (from any rating agency) for a firm as described in the previous section. The event in testing our third and fourth hypotheses is the announcement of a rating regarding a threshold as described above.