Examination of Witnesses (Questions 1080
TUESDAY 13 NOVEMBER 2007
Q1080 Mr Dunne: If I can quote to
you from this article, it comes from a Bloomberg's market report
in July which said that corporate bonds rated BAA, the lowest
Moody's investment grade rating, had now reached 2.2% default
rate over five-year periods from 1983 to 2005, according to Moody's,
but from 1993 to 2005 CDOs, which have only been going that long,
with the same BAA grade, suffered five-year default rates of 24%.
Are you going to suggest that that CDOs have much higher default
rate with the same rating than corporate bonds over their life?
Mr Drevon: No, it does that in
general. It does it in specific rating levels, and I think you
commented on one rating level for a specific horizon, but even
within the CDO categories there are a number of different types
which will have different performances.
Q1081 Mr Dunne: Can we turn for a
moment then to how you decide at a certain point that credit is
deteriorating? We have touched on what happened in Northern Rock,
where you did not decide it was deteriorating until the Bank of
England had stepped in. Is it the case that you tend to reduce
grades of debt that you see heading towards default shortly before
the final default in order to improve these performance statistics
that we have just been talking about?
Mr Hancock: Certainly not. Indeed,
the way the statistics are actually published, you can look through
that, so the investor is quite able to look at what the rating
was one year, two years, five years before the default, so there
is absolutely zero incentive to do that, and it will be seen through
by the users, who have access to all of this information.
Mr Madelain: I think what is important
to note is the performance of the rating. There is a very high
degree of transparency about that. I think all agencies publish
a huge volume of statistics, either at the aggregate level or
at the asset class level, actually tracking the performance of
our ratings. So, it is certainly an area where transparency is
Q1082 Mr Dunne: Have you read Peter
Warburton's report? This is language which you may disagree with,
I expect. It says, "The final trick that rating agencies
pull is to post-validate their assignment of a rating by making
sure that very few bonds actually default from a high rating.
Hence, by heavily downgrading a nearly bust bond a few weeks before
its final demise, the agencies can claim that it defaulted as
a C-rated bond rather than a DD-rated bond which it was when the
bad news hit?
Mr Hancock: Can I just reinforce
that the user of all our ratings has all of the data available
to identify if that behaviour is prominent.
Mr Bell: I would also say, if
you look, for example, at the table that we include in our submission
of the default rates in the US RMBS, they are from initial rating
not from final rating.
Q1083 Mr Dunne: I think your best
defence to that charge actually is what happened with Northern
Rock, because you failed to downgrade them and perhaps you should
have seen the warning signs a bit earlier. A couple more questions,
if I may, Chairman. One is in relation to the information that
is available to you as a rating agency in the US compared to Europe.
You routinely receive information not generally available to the
public markets from issuers, but in the US information on underlying
collateral is generally available to investors, whereas it is
not in Europe, and a charge has been made that during the summer
crisis prices of securities began to show investors perceiving
risk well ahead of the rating agencies in the US. That did not
happen in Europe because the information was not available. Would
anybody like to comment on that?
Mr Taylor: I do not think we have
any problem at all with greater transparency in the markets. We
have no problem whatsoever with the market seeing absolutely what
data we get.
Q1084 Mr Dunne: You would be quite
happy to see information made available to investors in Europe
in the same way as it is in the US? That does not happen at the
Mr Bell: Yes, absolutely, in fact
we welcome it and we have been, in some cases, urging, particularly
in this crisis, our clients to make that information available,
because we think it is good for the market that it should be available.
Q1085 Mr Dunne: Do you see any parallels
with what is happening in the US banking sector: losses being
generated by investors, particularly in these CDOs, between other
market failures such as the Lloyd's insurance market? If you take
the example of the Piper Alpha loss of one billion dollars, because
the way the reinsurance arrangements worked, that translated into
a 16 billion dollar Lloyd's reinsurance loss. Do you see investment
in CDOs by CDO funds as creating a spiral in a similar way to
that, or potential risk of a spiral?
Mr Bell: I think you need to distinguish
between a credit loss and a mark to market loss. The credit risk
never disappears, but neither is it necessarily magnified by being
repackaged, it is just moved around. In terms of credit losses,
the credit losses suffered so far in the global market as a result
of the events of the summer have been actually very small. The
losses you are looking at which are being announced by all the
banks are mark to market losses. Undoubtedly, if you have many
transactions, including in this synthetic area, then you have
much greater value of issues out there, and on a mark to market
basis you clearly have a greater chance that the losses will be
magnified. In terms of credit losses, there is no magnification
because the loss does not get magnified as it gets moved around.
Q1086 Mr Dunne: Until somebody defaults,
and then it does get magnified.
Mr Bell: Sure, but there is no
magnification. The ultimate default, the borrower in the sub-prime
who borrowed 25,000 or 100,000 dollars, can only default to the
100,000 dollar tune even if that loss has been repackaged in an
RMBS or a CDO. It has been moved around, but he cannot default
more than 100,000 dollars. The losses that you are now witnessing
in the system are mark to market losses as these securities' values
have been marked down, and I think this is one thing that is worth
bearing in mind. In terms of the magnitude of those mark-downs,
we, for example, looked at one triple-A prime UK RMBS bond that
traded at 80 cents in the dollar or 80 pence in the pound. That
loss, taking into account the credit enhancement already in the
bond, assumed that the person who sold it at 80 pence in the pound
was selling it on the assumption that in a UK prime residential
mortgage backed security 80% of the pool would default and the
price of properties in the UK would fall by 40%. I do not think
that on any valuation theory, other than Armageddon, anybody believes
that eight out of ten UK borrowers are going to default on their
mortgage and the price of houses in the UK is going to fall by
half. What you are seeing in the market today, all those enormous
mark to market losses, does not reflect credit deterioration,
they also reflect a clear element of panic.
Q1087 Mr Dunne: Are any of you considering
liquidity and introducing a measure of liquidity as part of your
Mr Prescott: We are setting up
a working party to look at liquidity in financial institutions.
Mr Hancock: It is certainly something
we will be looking at.
Q1088 Chairman: Why do you need a
working party in the light of Northern Rock? Why do you not just
go ahead and ensure that you assess liquidity? For God's sake,
you have all given Northern Rock a really good rating, the disaster
happened and now you are saying, "We are going to set up
a working party because we never assessed liquidity." Why
do you not just say here and now you are going to assess liquidity?
Mr Hancock: Certainly within the
rating of Northern Rock we did assess liquidity. Now clearly our
assessment of liquidity did not withstand the repercussions
Q1089 Chairman: So they failed.
Mr Hancock: I think what Philip
was referring totell me if I am wrongwas some sort
of separate indicator for liquidity in addition to the probability
of loss indicators.
Q1090 Mr Dunne: My question is how
do you do that if you are not participating in the market?
Mr Taylor: We are looking at it.
We are investigating it because market participants are asking
if we can provide that kind of service. We will investigate and
do the best we can to look into it and see if we can put something
together. Maybe we need to buy in new expertise, new tools and
new data. There is no guarantee we can come up with that kind
of product, but it is work in progress.
Mr Bell: The decision to do such
a process is fairly easy to take; the creation of such a scale
is actually quite complex.
Q1091 Jim Cousins: Looking at the
events of this summer and, indeed, the autumn as well, would you
expect a bank approaching the Bank of England's credit facility
to inform you?
Mr Hancock: We would not be at
all surprised if they did not, given the hugely sensitive nature
of that discussion.
Q1092 Jim Cousins: You seemed to
imply earlier that you would have such an expectation?
Mr Madelain: What I said is that.
I said earlier that
Q1093 Jim Cousins: A lot earlier.
My memory is quite good.
Mr Madelain: I said earlier that
we were expecting systemic support to be made available to the
Q1094 Jim Cousins: No, you said that
Northern Rock had told you that they had approached the bank.
Mr Madelain: No, I did not say
that, or if I said it I should have
Q1095 Jim Cousins: If you did say
Mr Madelain: I do not think I
meant to say that.
Q1096 Jim Cousins: You did not mean
to say it?
Mr Madelain: No.
Q1097 Jim Cousins: Would you expect
a bank approaching the credit facilities of the Bank of England
to tell you?
Mr Prescott: I think they would
only do that at the very last minute.
Q1098 Jim Cousins: They would only
do it at the very last minute?
Mr Prescott: Yes.
Q1099 Jim Cousins: How many banks
have in fact told you that they have approached the credit facility
of the Bank of England? You have just said to this Committee that
you want very high standards of transparency. I have asked you
a rather simple and obvious question and you are dumbstruck?
Mr Madelain: No, I think the answer
to your question is