Examination of Witnesses (Questions 1020
TUESDAY 13 NOVEMBER 2007
Q1020 Mr Simon: I know it is possible.
Mr Madelain: And we can disagree
sometimes on things that are very importantthe cut off
point between investment grade and non-investment grade, for example,
or special situations where we may take a view and other raters
may take a very different view.
Q1021 Mr Simon: We have been hanging
this around the Northern Rock situation. Give us an example in
the Northern Rock situation of such a nuanced disagreement and
the positive impact that it had?
Mr Madelain: At the moment, I
Q1022 Mr Simon: No, no, I am talking
about the run up to the run on this bank, which none of you sussed
out in advance. To be fair, nobody else noticed it either and
it is obviously not your faultI am not saying it is your
faultbut give us an example of how this might have worked
in this particular instance where you rated it in some nuanced
different way and thereby sent a subtle signal to the market?
Mr Madelain: We had a higher rating
for Northern Rock than Fitch and S&P. We had a double issue
rating on Northern Rock, and the reason we had a higher rating
was because in our rating methodology we do assign a higher weight
to systemic support to bank ratings.
Q1023 Mr Simon: But that does not
really make any difference, does it? It is a different methodology,
everybody knows that, everybody knows what the methodology is,
so unless you are actually going to be making decisions in a different
way, unless you are going to be forming views differently
Mr Madelain: We formed a view,
which was that effectively we are rating the bank higher.
Q1024 Mr Simon: That is because you
always rate those higher banks higher than they do, and everybody
knows that, so what is anybody learning from this?
Mr Hancock: It comes back to the
users of ratings value a variety of opinions. In this case Moody's
took a different slant on the likelihood of state intervention
in the case of Northern Rock. These are opinions, there is no
right or wrong, and clearly our users value having a variety of
Q1025 Mr Simon: Very briefly, Mr
Bell, two hypotheticals. If one of the three of you did not exist,
would it be a big problem for the market? Secondly, if none of
you existed, would it throw the markets into crisis?
Mr Bell: If one of us did not
exist, it would narrow and reduce the number of opinions that
investors can turn to.
Q1026 Mr Simon: That is called a
Mr Bell: Yes.
Q1027 Mr Simon: If one of you did
not exist, there would be one fewer of you than there are now.
I understand that. I would like a little bit of interpretation,
which is what you do for a living after all.
Mr Bell: I think the more educated,
informed opinions there are, the more
Q1028 Mr Simon: So it would be better
if there were ten of you?
Mr Bell: Absolutely.
Q1029 Mr Simon: Good.
Mr Bell: If none of us existed,
it depends which markets you are looking at, but in markets which
are structured finance, which are global markets with fairly complex
instruments, it is difficult to see how such markets could meaningfully
exist without a series of independent opinions that looked directly
at the transactions.
Q1030 Mr Simon: Why cannot we just
have a computer model? Everybody knows what these criteria are.
Why can we not just have a programme and everybody runs it through
at their desk?
Mr Madelain: They are available
Q1031 Mr Simon: Why do we need you
then? Why do we need to pay you to do that?
Mr Bell: Because the thing about
computer models is they are very inflexible and, therefore, they
are subject to gaming. There are a lot of very highly paid people
in banks whose job it is to try to figure out a better mouse trap,
and if you have an inflexible model with no human element to actually
see how the model is being gamed, you will get yourself into a
lot of trouble fairly quickly.
Q1032 Mr Simon: I am sorry if I am
going on a little bit, but when I ask, "How come you all
agree or you do not quite agree?", nobody says, "We
disagree because we have added a little bit of human element into
this, because we made a slightly different judgment." The
only reason you disagree is, "Oh, we have got a slightly
Mr Madelain: It is not. It is
human judgment. That is exactly what it is.
Q1033 Mr Simon: You could write that
into your model?
Mr Madelain: No, it is not a model,
it is actually a view we form over time for bank ratings. We assign
a higher element.
Q1034 Mr Simon: In all cases; in
Mr Madelain: No.
Q1035 Mr Simon: "In these cases
we assign this rating"that is not a warm, human, touchy
judgment. You could write it as an algorithm.
Mr Madelain: Well, eventually
you can, but how you came to that conclusion is exactly the image,
and that is what you differentiate as.
Q1036 Mr Simon: That is another truism.
These are all human judgments because the algorithms are written
Mr Bell: Let me give you an example
outside of Northern Rock which I think is useful. Both Moody's
and us rate structured finance transactions in emerging markets,
including Russia. They are not, you will be glad to know, rated
triple-A. However, we have formed quite different views about
the nature of the risk of sovereign interference in Russia on
various asset classes and Moody's view of their analysts based
in Russia, knowing the market and knowing the Government, about
what is more likely, an interference with a mortgage transaction
or an interference with a consumer loan transaction, is the exact
opposite of ours. We have our view; they have their view. As a
result we rate those transactions differently and we explain why
we rate them differently. That is a classic example of the human
subjective element based on our staff in Russia and their understanding
of what is happening. Moody's staff in Russia have a different
understanding. I am not saying I believe ours is better because
I am S&P, but I think it is of value to investors to be able
to see those different ratings, to understand why they are differently
assigned, to understand the subjective element behind them and
then, as an investor, to make their own view as to whether they
feel Moody's is right or whether we are right.
Q1037 Chairman: You mentioned about
value to investors to see the different ratings. Would you take
it then that some investors can mistake a good credit rating for
a green light to invest?
Mr Bell: My experience of investors
over the 20 years that I have been in the structured financial
market is that the spectrum of investors in structured finance
is huge. At one end it is composed of extremely sophisticated
funds that have their own
Q1038 Chairman: My question is a
very simple one, Mr Bell: do you think that some investors can
mistake a group credit rating for a green light to invest?
Mr Bell: I think some investors
may well have done that, yes.
Q1039 Chairman: You would all agree
Mr Bell: Yes.