I find the following from the Bloomberg News report as published in The Boston Globe article, Treasury official faults credit rating downgrade:

S&P made a $2 trillion calculating mistake and then changed the rationale for its decision, raising “fundamental questions about the credibility and integrity of S&P’s ratings action,’’ John Bellows, acting assistant secretary for economic policy, wrote in a Treasury blog post.

The dispute stems from how S&P used figures from the Congressional Budget Office. The discrepancy didn’t change the downgrade decision, S&P officials said, because Treasury’s $2 trillion figure was derived by calculating government debt over a 10-year period while S&P’s ratings are determined using a three- to five-year timetable.

It appears that S & P may have come to the conclusion that it needed to downgrade the US credit rating and then went searching for the evidence that would justify the foregone conclusion.

In making hypotheses, I suppose there can be circumstances where your gut tells you what the hypothesis should be before you find the evidence to support it.  However, especially in this case where the hypothesis came before a formal search for evidence, you must be extremely diligent to not look only for supporting evidence.  You must be as diligent in searching for any evidence that would contradict your hypothesis.

As pointed out by Nassim Nicholas Taleb, The Black Swan: The Impact of the Highly Improbable, it doesn’t make any difference how many swans that you see that are white, seeing just one black swan is enough to convince you that not all swans are white.  Even before you see the black swan, you ought to be cautious about claiming all swans are white.

Yes, I know,  if I am to follow my own logic, I should look for evidence that S & P did not make its decision to stick by its original conclusion because of  confirmation bias.  I’ll leave that as an exercise for the reader.  After all, I don’t get paid millions of dollars for writing these blog posts.  I would say millions of dollars may be at stake for S & P if they should lose their reputation, what reputation they have left, for making good calls on the creditworthiness of investments.  Remember their investment grade rating of banks’ mortgage backed securities.

August 8, 2011 9:00 AM

As mentioned in the post Resistance To Revenue Increase Causes S & P Downgrade, I was able to read the full S&P report.

I can’t tell if I was reading the original report or the corrected one, but it does look to me as if the quote mentioned above from John Bellows on the Treasury blog may have been very self-serving.

Perhaps S & P didn’t so much as make a mistake as they did use different assumptions from the CBO report from which they took their numbers.  I don’t know if this difference was mentioned in the original report, but it is very clearly stated in the report that I read.

I think this shows that you cannot trust what either political side has to say about the report.  You have to read the report and decide for yourself what you think it says and what you think it means.

Here is the link where you can read the full 8 page S & P Report.