ISDA PRESS REPORT

ISDA PRESS REPORT - MARCH 22, 2001
<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office"
/<
Interest rate swap market opens up - Financial Times
Swedish PM not sure of EU financial reform accord - Reuters
Credit derivatives pose unexpected risks, says Moody's head of OTC -
Risknet.com
Watch derivatives, warns bank - Euromoney
Accounting changes spell $270m charge for Sears - Financial Times
Impatience and optimism on clearing - Euromoney
Basel gives banks the whip hand - Euromoney

Interest rate swap market opens up
Financial Times - March 22, 2001
By Janaki Ghatpande

Till about 12 to 15 months ago, the long-term interest rate swap (IRS)
market was dormant for the lack of sufficient benchmarks despite the Reserve
Bank of India (RBI) permitting the instrument as far back as June 1999.
Ironically, in the last few weeks, there has been a spurt in IRS deals,
indicating a development of the market with increased sophistication in the
design of the products.

While swaps did take place in the past, most of them were short-term swaps
for durations between one to three months. Indeed, swaps of 12 months were
the longest available in the market. Bankers had earlier cursed the lack of
satisfactory benchmarks as holding back the market. The problem still
remains, what is different is the bankers have found innovative ways to
price longer term swaps. Now, the long-term interest swaps have gone up to
even five years.

In an IRS, two parties do a swap for a fixed duration wherein an exchange
takes place which is mainly notional. When bank rates fall, a player
reverses his position and books his profits. Thus, they receive a profit
when rates actually fall and vice-versa.

One party agrees to pay a fixed interest rate for the tenor of the swap
while the other party agrees to pay a floating interest rate. This floating
interest is usually fixed on a benchmark from the debt markets which could
vary from the Mumbai Interbank Overnight Rate (Mibor) or a synthetic rate
derived usually based on a poll of market dealers.

"The hindrance with these swaps is that the benchmarks for them are few and
far between. While some benchmark is on one-year interest rate taken from
the Reuters page (poll of dealers), some use the one-year T-bill average
yield against which the fixed rate is determined. Therefore, the increase in
the number of interest rate swaps provides more bench marks than were
available initially," said a senior dealer with a foreign bank.

Says Tarun Mehrotri, treasurer, HSBC, "As a part of its broader strategy of
providing customised and risk management and funding solutions to its
clients, we provide interest rate swap prices benchmarked on both short-term
rates such as Mibor, CP rates and long-term rates such as one-year T-Bills,
government security yields. This provides clients the flexibility of
choosing the benchmark that best suits their balance sheet needs and
addresses their particular asset/liability management requirements."

"It is primarily a customer-driven and a customised market which is designed
for the corporate keeping in mind their asset liability management point of
view. It is still not monitored. Therefore it is difficult to come up with
an estimate about the volumes," said an analyst with a foreign bank.

"A swap can be important tools for banks and corporates who want to hedge
themselves against conflicting interest rate movements," he added. In the
market place currently, the average size of swaps is between Rs 25 crore and
Rs 50 crore.

Even the Fixed Income Money Market and Derivatives Association of India
(Fimmda) is now shifting its monthly curve to a daily curve allowing for
more benchmarks. "The market has a few market makers right now but as the
number of benchmarks increase and get validity, the market will see more
deals in the IRS market," said Aashish Pitale vice-president research, JP
Morgan Securities India Private Limited.

Swedish PM not sure of EU financial reform accord.
Reuters - March 22, 2001
By Jan Strupczewski

STOCKHOLM, March 22 (Reuters) - Swedish Prime Minister Goran Persson said on
Thursday he doubted whether reforms to create a single European financial
market would be approved at this week's EU summit in Stockholm.

The reform is key to meeting the 15-nation bloc's self-imposed deadline of
integrating its capital markets by 2005 to boost the competitiveness of
European firms through the same easy, cheap access to funding enjoyed by
U.S. rivals.

"I am sorry to say I am not confident at all. But we will try hard," Persson
told a news conference on the eve of the summit, which will review progress
made by the Union towards its goal of becoming the world's most competitive
region by 2010.

"We want to go ahead with it... I hope we will be able to gather not only a
majority but also unanimity for the reform," he said.

EU finance ministers are due to meet later on Thursday to discuss the
reform, prepared by a team led by former central banker Alexandre
Lamfalussy.

The Lamfalussy team proposed earlier this month that the EU should adopt a
"fast tack" procedure under which framework laws could be swiftly enacted,
leaving details to be filled in later by a new European Securities
Commission.

But the proposal has sparked power struggles between the Union's
institutions and met with resistance from Germany, which fears that its
financial centres could be subjected to looser regulation modelled on
London.

Swedish Finance Minister Bosse Rinholm, who will chair the Thursday evening
meeting devoted to the Lamfalussy proposals said agreement on the reform had
reached 99 percent.

The meeting will start at 8 p.m. (1900 GMT) and be followed by a news
conference, the Swedish EU presidency said.

COMMITMENT IN DOUBT

Failure to back the proposals would cast doubt on the EU's public commitment
to an integrated capital market and could further hit the already battered
common currency.

The euro's launch in 1999 created a European capital market with a common
denomination, but the market remains divided in many areas by national
securities laws and accounting practices.

Britain, Denmark and Sweden have opted to remain outside the euro for the
time being.

The financial plan forms part of a wider agenda, laid down by EU leaders
last year, to make the EU the world's most competitive place to do business
within a decade. The March 23-24 summit is due to review progress towards
that goal.

"It is of particular importance that the creation of a common market for
risk capital and financial services will be implemented according to the
timetables agreed," Persson wrote in a column in tabloid Expressen on
Thursday.

The task facing the finance ministers is to stop the issue from snarling up
the summit.

The debate turns on the seemingly arcane point of how a new European
Securities Committee would be able to block detailed legislation in areas
such as issue prospectuses and insider trading, proposed by the European
Commission.

Credit derivatives pose unexpected risks, says Moody's head of OTC
Risknet.com - March 21, 2001
By John Ferry

Credit deterioration rather than actual credit default could trigger credit
default swap contracts under current International Swaps and Derivatives
Association (ISDA) definitions, according to a new report by Moody's
Investor Service.

Moody's contends that ISDA'S current definition of a "credit event" is
broader than the general market understanding of default.

"ISDA'S current definitions... could, in some cases, increase the risk of
loss by triggering payouts for events that are not actually defaults,"
contended Jeffrey Tolk in his report, 'Understanding the Risks in Credit
Default Swaps'.

"Under a credit default swap, losses to investors are determined
synthetically, based on credit events occurring in a reference credit. Thus,
investors' risk is driven largely by the definition of credit events in the
swap," he added.

Credit default swaps are bought as a form of insurance against the risk of a
bond-issuing institution defaulting.

The report states that a restructuring of debt to defer principle payments
may not technically be a default if the lender has been properly
compensated, although ISDA would define this as a credit event.

Under ISDA'S definition of "obligation acceleration", a credit default swap
investor might suffer a loss although an investor in the underlying cash
position would not.

ISDA, which is the leading market association for the derivatives industry,
has set up a committee to look at credit derivative trading practices, in
particular credit event definitions.

Moody's said it is difficult to value defaulted credits to determine the
amount of loss to investors under a cash-settled credit default swap.

"Calculated losses may vary based on liquidity, market conditions and the
identity of the parties supplying bids," the report said.

There is also a moral hazard problem associated with credit default swaps,
which arises as it is the buyer of protection that determines when a loss
event has occurred and how much loss is imposed on the counterparty. The
buyer has an incentive to be over-generous in loss calculation and to define
a credit event widely, Moody's said.

The expanding credit derivatives market is increasingly gaining the
attention of regulators. In a speech made two weeks ago, David Clementi,
deputy governor of the Bank of England, said he was concerned that market
participants may not fully understand the transactions they have entered.

Watch derivatives, warns bank
Euromoney - March 2001

David Clementi, deputy governor at the Bank of England, says the use of
credit derivatives in securitization poses a threat to capital market
stability.

In the closing address to the International Bond Congress in London,
Clementi surprised borrowers, investors and bankers with his concerns at the
rapid growth of credit derivatives and ABS markets, which many banks hope
compensate for the shrinking returns in bond trading and underwriting.

Clementi said there was a danger of a lack of transparency in the market,
because of the increase in the use of synthetic CLOs, where the risk
attached to loans and bonds is sold on -without a sale of the assets.

"These markets mean that a bank need no longer remain exposed to its main
customers but can rapidly take on large exposures to other credits without
any new borrowing by the underlying entities," he said.

Clementi said it was a major challenge for authorities such as the Bank of
England or the
Federal Reserve to keep pace with the rapidly changing world of
securitization and derivatives in collecting financial statistics.

He also fears not all investors understand securitization. "Some
participants in this market may not fully understand, or may have differing
understandings of, the transactions into which they have entered," he said,
adding that uncertainties remain about how courts in different countries I
will treat covenants and agreements in credit derivative deals.

The case of LTY Steel illustrates his concern. LTV is battling Abbey
National in the US over ownership of assets securitized in a deal. The judge
appears to favour LTV though Abbey National says it will appeal if it loses.

Investors, because of this risk, have to make sure they have conducted due
diligence, said Clementi.

The merging of the lending, securities and insurance markets raised new
risks, he said, adding this risk was particularly evident now due to the
debate over whether restructuring of debt constitutes a credit event,
thereby triggering a credit default swap. This row was sparked when Conseco
restructured its debt late last year. The International Swaps and
Derivatives Association has convened meetings to try and find agreement on
this question.

The British Banking Association has predicted that the credit derivatives
business will triple from $586 billion in 1999 to $1/581 billion in 2002.
This item first appeared at <http://www.euromoney.com/bonds<;
www.euromoney.com/bonds.

Accounting changes spell $270m charge for Sears
Financial Times - March 21, 2001
By Christopher Bowe

Sears, Roebuck & Company, the second largest US retailer, said on Wednesday
it plans about $270m in after-tax charges this year due to accounting rules
changes.

The charges will stem from the retailer's adoption of a new Financial
Accounting Standards Board rule on companies~ accounting for use of
derivatives and hedging instruments. Sears detailed the move in its annual
report filing with the US Securities and Exchange Commission.

FASB amended rule No. 133 requires that all of a company's derivatives be
marked to market, valued and recorded on the balance sheet as an asset or
liability. Under the rules, the fair value of the derivatives, which could
range from futures contracts to complex swaps transactions, may be
calculated according to earnings or in "other comprehensive income" provided
certain conditions are met.

In that case, derivatives' gains or losses would be measured against the
value of the item that is subject of the hedge on the income statement.

The new accounting rule became effective for all fiscal years starting after
June 15 last year. Sears said it plans to use the new rule starting in the
first quarter this year. As a result it expects to take a cumulative $270m
charge this year. The transition to the new accounting standard includes
recording a $56m cash flow hedging relationship on the balance sheet and
reclassifying deferred losses from terminated interest rate swaps which
amount to $389m. Combined, they are estimated before taxes at about $175m.

The company uses debt and interest rate derivatives and currency hedges to
manage its risk. Much of Sears' profit comes from its credit card arm.

Sears also said in the filing that it expects about 5 per cent operating
income growth and at least 10 per cent earnings per share growth this year.

Impatience and optimism on clearing
Euromoney - March 2001
By Jonathan Brown

London hosted the first worldwide conference of central counterparties last
month to discuss developing a system of seamless global clearing.

The conference was initiated by the Depository Trust and Clearing
Corporation of the US.

Sir David Walker, senior adviser to Morgan Stanley International,
underscored the importance large securities firms attach to this: "There is,
I believe, a general recognition that the capacity to clear and settle
trades has not kept pace with the capacity of trading platforms to attract
liquidity and handle trades," he said. Without serious thought to improving
clearing and settlement, the present infrastructure will not be able to cope
with large increases in volume and trading.

Representatives of central counterparties from the US, Asia and Europe
discussed procedures for operations and risk management, the benefits of
greater cooperation and legal ramifications. All believe there are
advantages to be gained from the more effective use of collateral in the
systems as well as netting which reduces the amount of settlement that needs
to be handled. Referring to efforts to integrate regional clearing and
settlement systems in Europe, Walker said: "Far greater benefits will accrue
if the CCP and netting arrangements of these new vertical structures are
designed in an inter-operable,
open architecture way.

Jill Considine, chairman of DTCC, was pleased with the progress. "We've
covered the landscape and gained some very important perspectives on
enhancing CCP capabilities, including the views of regulators, exchanges, as
well as the views of our colleagues at other central counterparties around
the world," she remarked. "And we heard from our customers, telling us how
impatient they have grown with what they view as slow and piecemeal changes
in the industry's infrastructure. We must not lose sight of what the
customers value most - strong risk management, low costs and the stability
and reliability of a seamless infrastructure."

Basel gives banks the whip hand
Euromoney - March 2001
By Michael Peterson

It has been a long and exhausting labour of love. Between zoo and 300 bank
regulators from the wealthy G10 countries have flown many thousands of miles
and held countless hours of meetings. The Basel Committee on Banking
Supervision, which meets under the auspices of the Bank for International
Settlements (BIS), has spent the best part of two years rewriting the rules
on bank capital.

The committee, chaired by New York Federal Reserve president Bill McDonough,
steered the discussions. But the bulk of the work was done by four major
sub-committees. Each sub-committee set up working groups to look into some
detail of the proposed rules. The biggest sub-committee, the capital task
force, spawned no less than half a dozen subcommittees of its own.

The proposal document, released on January r6, is suitably massive. Its 541
pages outline a new world order for bank capital which looks, at first
glance, like a vast improvement
on the 1988 capital accord. Where the old agreement sketched out crude
guidelines for calculating the amount of capital a bank needs, the new
proposals delve into the gritty intricacies of collateral haircuts, clean
break' securitization, materiality thresholds and granularity scaling
factors.

Given the Herculean efforts of this dedicated band of officials, it seems
churlish to suggest that the proposals are flawed. But serious doubts are
emerging about the general direction the committee has taken.

The central element of the new rules is that big banks will be allowed to
set their own capital using internal calculations of risk. The regulators
expect a good number - perhaps 60 to 80 institutions - to be doing this soon
after the new accord is introduced in 2004. But Avinash Persaud, State
Street's head of global research, reckons the trend towards greater reliance
on bank risk management systems is a dangerous one. "It is a dereliction of
duty by the regulators," he says. "There is no industry with a longer
history of being unable to control its own excesses.

Flawed accord
The need to update the present rules is pressing. Simplicity was the great
virtue of the 1988 capital accord, but it was also the big weakness. The
accord established guidelines for calculating bank capital, with a basic
minimum requirement of 8% and weighting adjustments reflecting the riskiness
of assets. Members of the BIS then agreed to introduce these as the basis
for national law. The rules have quickly become the standard for other
regulators too, driving up levels of bank capital in many emerging markets,
for example.

The present rules apply a series of weights to the basic 8% capital number
for different assets on banks' balance sheets. Supposedly. risky assets,
such as loans to corporates and asset-backed bonds, are weighted at 100%.
Debts to countries that are members of the Organization for Economic
Cooperation and Development (OECD) are considered risk-free: they are
weighted at 0%. Exposures to banks based in OECD countries count as low risk
assets and are weighted at 20%.

The bank then has to hold capital equivalent to at least 8% of its
risk-weighted assets. And at least half of that capital has to be in the
form of so-called tier-one capital - chiefly equity and reserves. The 8%
figure is a mini-mum. National regulators are free to impose higher levels.
The Federal Reserve, for example, prefers big US banks to hold 10% capital.

The accord's crude system of asset weightings creates some perverse effects.
It means banks based in OECD countries can get much cheaper funding than
those outside the OECD club. This is not necessarily a good thing for
countries that join the OECD. State Street's Persaud points out that the old
Basel system produces regular "OECD membership crises". After they were
admitted to the organization during the course of the 1990's Mexico, the
Czech Republic, Poland and Korea all experienced economic booms fuelled by
suddenly cheaper credit. A year or so later, each boom promptly turned to
bust.

Under the 1988 rules, a junk bond portfolio with a heavy concentration of
telecom names is weighted at 100%. Fair enough. But a diversified loan book
of high-quality corporate names is rated at exactly the same level. Banks
know the Basel weightings produce a distorted picture of their true risk,
and many have become adept at finding ways to reduce their regulatory
capital, using tools such as securstization and credit derivatives. Others
have played the system for all it is worth, stocking up on risky, but
low-weighted assets.

"The old regime had broken down because of the complexity of what the large
and sophisticated banks were doing," says Oliver Page, director of complex
groups at the UK Financial Services Authority and a member of the Basel
Committee. "It was leading to capital arbitrage and we could no longer be
confident that the resulting capital matched the risks they were actually
taking."

Dynamite in the draft
The Basel Committee's first thoughts on revising the capital accord were
mostly about re-jigging the problematic risk weightings. A discussion
document released in June 1999 proposed a series of refinements to the
original risk bucket method. The number of risk weighting bands would be
increased and the OECD link would be scrapped. But what better measure of
credit risk could be used? The committee proposed a controversial solution:
weighting assets according to their credit ratings from the private sector
rating agencies.

The idea of incorporating credit ratings into the accord created a few
misgivings. For one thing many borrowers, especially in Europe, are not
rated. Some also worried that borrowers would cherry-pick the most
favourable ratings. The agencies themselves were not overjoyed at the
proposal. They quibbled over what they saw as inconsistent application of
their ratings from one asset class to another.

The 1999 proposal didn't stop at refining the weightings. It introduced a
couple of new ideas. One was that banks should be charged for operational
risk - the risk of things like system failures and rogue traders.

But the real dynamite came half-way through the 62-page document. The
committee conceded that a small number of very big, sophisticated banks
might be allowed to bypass the whole standard weightings process altogether.
They would be able to set their own capital requirements - in consultation
with their regulator using their own models of risk.

Risk modelling has become a growth industry in recent years. Big banks argue
that they can now model and manage their risk in a highly sophisticated way.
Their models give them a detailed picture of the risks inherent in their
business and the amount of capital they need to set aside for those risks.
Usually, the models say banks need to carry less capital than they would
using the standard risk-weighting method.

Between June 1999 and the end of 2000, as the committees and sub-committees
met to thrash out the details of the new system, the internal-ratings based
approach grew into the biggest part of the proposals. And no longer was it
only the very biggest banks that would be allowed to use internal models.
Now, any decent-size bank could aspire to calculating its own capital needs.

"The big change from our June 1999 paper is that we have designed an
internal-ratings based approach that a good number of banks should be able
to apply," says the FSA's Page. "In June 1999 we just talked about applying
it to some sophisticated banks. But over time we expect a major share of the
banking system will be on the internal-ratings based approach."
It is not completely clear why the committee decided to apply the
internal-ratings based approach more widely. The official line is that a
large number of banks have demonstrated their capability in risk modelling.
But some close observers reckon the committee was under strong pressure from
some European authorities - particularly the German representatives - to
allow a lower threshold for the internal models approach.

The new proposals talk about three approaches to setting capital. The
simplest, the standardized approach, is an extension of the present system,
but with risk buckets now linked to credit ratings. A triple-A sovereign
carries a 0% risk weighting. Single-A sovereigns, double-A corporates and
triple-A supranationals are weighted at 20%. Most loans with a B in their
rating fall into the 100% bracket.

There are also two new weightings. Single-A corporates and triple-B
sovereigns fall into a 50% band. A few risky assets, such as single-B
corporates, carry a 150% weighting. Where a borrower has two ratings, the
lower rating will be used to prevent cherry-picking.

The internal-ratings based approach divides into two different methods:
foundation and advanced. Both approaches allow banks to calculate their own
risks. In the foundation method, the regulator supplies some of the
parameters for the model. In the grown-up version, banks get to do all the
calculations.

The internal-ratings based approach should have several big advantages over
the simpler method. It should dramatically reduce the scope for regulatory
arbitrage, since the differences between a bank's economic capital - the
capital that reflects its true risks - and its regulatory capital will be
too small to bother about.

The internal-ratings based method also has much steeper charges for the very
riskiest assets. Under the standardized approach, the highest risk weighting
is 150%. But under the internal models approach, an asset could carry a 625%
weighting. (In other words, for every dollar at risk, the bank would set
aside 50 cents of capital, assuming it was aiming for 8% risk-weighted
capital.)

But how can we trust banks to set their own capital? The authors of the
proposals give two reasons. First, the regulators will not simply sit back
and let the banks do their own thing. In fact, they will need to step up
their efforts to supervise banks, recruiting teams of savvy analysts to pore
over the banks' calculations.

Second, the market will police the system. One lengthy section of the
proposal - known, for no particularly good reason, as pillar three requires
banks to disclose to the market information on their risk exposures,
risk-management strategies and credit risk mitigation techniques.

"We are moving towards a world in which the regulators step back and allow
the market to police itself," says Mark Intrater, London based managing
director at consultancy Oliver Wyman. "If a bank reports its risks and the
market doesn't like it, that bank will have a hard time getting funding on
favourable terms and its stock price will drop. The world is not ready for
that yet, but this is a step towards it. Markets will always be able to move
faster than any regulations that could be drafted."

Disclosure - or what one analyst, in attempt to catch the mood of the
moment, calls the "Big Brother" part of the proposals -is designed to
prevent banks abusing the internal-ratings based approach. "If we can get
market discipline to work then it won't be the capital regime that drives
the amount of capital banks have, it will be market pressure," says Page.
"We see this already. Because of market pressures, banks already hold
materially more capital than we set as their minimum.

Using banks' internal models to calculate capital may be an attractive
solution to the tricky problem of how to regulate large and sophisticated
banks. But the approach is open to several lines of criticism.

Some say the banks' own models are not all they are cracked up to be. Stare
Street's Persaud says that banks do need risk-management systems, but the
ones they use today are imperfect. "These marker-sensitive risk-management
systems all have the same philosophical flaw," he says. "They assume that
every bank is acting independently: they assume that when they sell, there
is a marker to sell into.

In other words, they understate the real riskiness of panicky, illiquid
markets. "The use of banks' internal models can be a problem," agrees
Standard & Poor's analyst Tanya Azarchs, "to the extent that everyone starts
to use the same kind of models to determine the credit risk of a loan. If
that model is false-signalling it could cause real problems when everyone
heads for the exits at the same time.

What's more, greater disclosure might exacerbate the problem of what Persaud
calls herding. "We are going down the wrong path in insisting on more
disclosure," he says. "Every crisis follows a boom. This tells us that the
crisis is not about a lack of information. It is not about some domestic ill
that the right risk management system could have told us about. The cause of
the crisis lies somewhere in people's collective loss of sense during the
original boom."

Persaud believes the right approach to preventing crises is to encourage
contrarian investing. "We need heterogeneity," he says. "We need lots of
different types of investors playing lots of different types of markets with
lots of different strategies and risk management systems. This emphasis on
transparency is encouraging more herdlike and less contrarian behaviour.
This is why markets are becoming ever bigger in terms of turnover but ever
thinner in terms of liquidity."

The proposals may discourage contrarianism in other ways too. Use of
internal credit models may exacerbate natural economic cycles by encouraging
banks to lend heavily in the good times and cut back on credit when the
economy turns down depressing ratings.
For example, as the US economy has stalled in recent months, bankruptcies
have risen and credit ratings have fallen. Banks' internal credit models
presumably show the same decline in asset quality. Under the
internal-ratings based approach, banks' risk-weighted assets would now be
increasing and their regulatory capital requirement would be going up. That
would encourage them to freeze new credits. In times of rising credit
quality, the reverse would apply, perhaps fuelling a boom.

Another criticism of the banks' internal models is that they are not
conservative enough. S&P's Azarchs thinks banks' own capital models award
them too much benefit for the diversity of their portfolios. 'Banks'
internal portfolio models suggest they could run on as little as half the
capital they have now," she says. 'We don't believe that is the case. We
don't think there is excess capital in the system. If this proposal winds up
decapitalizing the system, that would be very concerning.

For now, banks would not be entitled to a capital reduction purely on the
grounds of credit diversity, although this is something that the Federal
Reserve has hinted could be introduced later. As a result, the banks'
calculations of regulatory capital will still typically come out higher than
their calculations of economic capital under the internal models approach.

Unrealistically short maturities are another problem with banks' own models.
"Most banks' internal models assume a hold rate of one year or less," says
Azarchs. "That is one reason their internal models show a capital surplus."

She believes the advanced internal-ratings based approach falls into the
same trap of underestimating maturities. The foundation version of the
internal-ratings based approach assumes that all loans have a maturity of
three years a reasonably cautious assumption. But on the advanced approach,
banks will be able to plug the legal maturity of loans into their capital
calculations.

Azarchs thinks this would be a mistake. "We don't think this is the right
way to think about it," she says. "A loan has nothing to do with its stated
term, it is representative of a relationship. As long as things are going
well, one expects the loan to be rolled over."
If the capital models give recognition for short legal maturities, banks may
simply extend short loans and roll them over. Banks would employ the same
kind of capital arbitrage they do today when they extend 364-day credit
facilities: such loans are designed to benefit from the zero risk weighting
of undrawn loans with a maturity of less than one year.

National pride
Another worrying aspect of the new models-based approach to bank capital is
that it will mean different things to different regulators. In a way this is
no different from the situation with the current capital accord. But at
least the 5988 version has the virtue of being simple enough for the most
backward regulator to understand.

"Were not sure that all national regulators are equipped to supervise the
internal-ratings based approach," says Azarchs at S&P. "There would
probably be an incentive for all of them to vet a few of their largest banks
right out of the gate - on the assumption that these big sophisticated banks
must know what they are doing."

National pride might also sway some regulators to treat their charges
leniently. "There is still scope for significant discrepancies in the way
these proposals are implemented, particularly in pillar two [supervision],"
says John Tattersall, partner at PricewaterhouseCoopers in London. "Very few
regulators will want the embarrassment of deciding that their major banks
are short of capital."

So what is to stop a regulator letting a few of its banks graduate to the
internal-ratings based approach even when their risk management systems are
nor up to scratch? "The idea of a completely level playing field is an
unreachable goal," says Page at the FSA. 'But we will have processes to try
to ensure there is a reasonably even-handed treatment across the G10. This
will not be some kind of central policing. It will involve being open about
what is going on in each others' countries and feeling free to comment.
Market forces act as another constraint. If a supervisor is known to be a
soft touch, the banks under its supervision will be penalized in the market
place."

Some large banks are privately worried that their own regulators will apply
the new rules more harshly than supervisors in other countries. The
disclosure requirements in the proposal are another big source of concern
for some of them. "The committee seems to have taken a kitchen-sink approach
to disclosure," says Tattersall at PricewaterhouseCoopers. "Everything that
might be desirable seems to have been included."

But Page at the FSA appears to offer some hope that the disclosure
requirements will be watered down. "You need a fair amount of information to
make the internal-ratings based approach work," he says. "The disclosure
requirements we have proposed are desirable, but we also need to be sure
they are cost effective for banks. It may be there are better ways of
arriving at the same results. I certainly hope banks will give us good
feedback on this point.'

The Basel accord is only supposed to apply in the Gin countries, but some
observers have begun to think about the impact of the proposals on the rest
of the world. "Basel was never designed for emerging-market banks," says Ian
Linnell, an analyst at Fitch in London. "But in reality it is adopted by any
banking system that wants to be taken seriously. And the idea that riskier
entities should hold more capital is definitely one that should apply to
banks in emerging markets."

He believes the proposals could have some unintended consequences in less
developed banking systems. ln practice, most emerging market regulators will
adopt the standardized version with some modifications. Others will stick
with the current version of the accord -something they are perfectly
entitled to do. "We will end up with four different methodologies for
calculating bank capital - the 1988 version of Basel, the standardized
approach, the foundation internal-ratings based approach and the advanced
internal-ratings based approach," says Linnell. "Add the salt and pepper of
charges for things like operational risk and you have a recipe for
confusion. The whole concept of a minimum level of tier-one capital could
become relatively meaningless."

Time to collect data
One key uncertainty in all this is how many banks will use each of the three
new ways of calculating capital. The consensus at the moment is that a
minority of banks - but a
majority of banking assets - will quickly move to using internal models.

"My sense is that in 2004 about 40 banks around the world will adopt the
foundation approach immediately and that maybe two or three will go straight
for the advanced approach," says Azarchs at S&P. Within half a decade, the
best part of too institutions could be setting their own capital.

That means banks need to start crunching a lot more data. "The timing is
tight," says Tattersall at PricewaterhouseCoopers. "Banks need to start
collecting the data next year in order to have the necessary three years of
historical data by 2004."

In order to use the foundation version of the internal-ratings approach,
banks need to input reliable data for every asset's probability of default.
If they want to graduate to advanced, they should also be able to make
assumptions about loss severity or recovery rates. "The internal-ratings
based approach looks great in theory," says Linnell at Fitch. "But there is
a real problem with the quality of data. Banks may have a reasonably good
handle on probability of default, but it is very hard to get hold of data on
recovery rates.

Up until now, banks have been surprisingly poor at keeping consistent
records on how many of their borrowers have defaulted and how much of a hit
they took as a result. Even the most rigorous institutions are hazy about
some areas of their business. "Most big, sophisticated banks have rating
systems for some loans that would pass regulatory muster," says Intrater at
Oliver Wyman. "But they also have categories of special loans that are
managed in a less sophisticated manner. It is simply the difference between
using a system for internal management purposes and using it for external
reporting purposes."
Clearly, banks will need to improve their record keeping. It has been widely
suggested that this scramble to put new systems in place will increase the
pace of banking consolidation. Smaller banks, it is assumed, will not be
able to make the necessary investment in technology.

On the other hand, bank mergers have often made it more difficult to get
systems in place. According to Azarchs at S&P, Canadian banks have some of
the best historical data in the world. This is because Canadian institutions
have been prevented from merging. US banks, by contrast, often find it
difficult to reconcile data from a number of legacy systems.

One widespread assumption so far has been that smaller banks will ignore the
internal-ratings based methods and stick with the standardized approach. "A
lot of small, specialist retail operations will - quite appropriately
-ignore everything beyond the standardized approach," says Intrater.
But the internal models approach could yet gain a momentum of its own. If
internal models are successfully used by as many as 6o or 8o banks to reduce
their capital requirements, smaller banks may want to join the party too.
This will certainly be the case if the technology and expertise needed to
use the approach become relatively cheap commodities.

"Perhaps one or two banks might develop a certain way of doing
internal-ratings based measurement and start to franchise that procedure out
to smaller banks," suggests Raj Malhotra, credit analyst at Goldman Sachs in
London. "Smaller banks might be willing to pay for that."

Just as the original Basel accord was soon adopted much more widely than it
was originally intended, so the internal-ratings based approach could become
the standard for any bank which aspires to international respectability.
Once the floodgates of internal capital calculation have been opened, they
may not be easily closed.

Power to the regulator
If that happens, a lot of regulators will need to learn a whole new set of
skills. There are broadly two approaches to bank regulation. The first
involves laying down the rules and checking that banks comply with them. The
second, often called bank supervision, is more intensive: it involves making
individual judgements about each bank's risks and appropriate levels of
capital and continually checking that banks are playing it safe.

The FSA is proud of its record in bank supervision. "In the UK we have
recognized for a long time that there is no way to collect all the
information you need, put it through some number cruncher and come out with
an answer for the amount of capital a bank needs," says Page. "There are
areas where judgement is required in forming the answer.

The Basel Committee proposals have quite a lot to say about banking
supervision - or, in Basel-speak, pillar two. The proposal exhorts
regulators to review banks' internal capital adequacy assessments and to
intervene early to stop a bank's capital falling below the level appropriate
to that bank's risk characteristics.

But this level of supervision will be much harder for some regulators to put
in practice than others. And it is not simply that some regulators are smart
and others are dullards. "Regulators such as the Federal Reserve have the
advantage of a relatively small number of large banks to manage," says
Intrater at Oliver Wyman. "Others, such as the Office of the Comptroller of
the Currency [which regulates smaller US institutions], supervise a large
number of small banks. They will need a much larger staff of supervisors.

Some observers have the feeling that the more clued-up regulators, notably
the Federal Reserve and the FSA, are getting too far ahead of their less
sophisticated colleagues. "A lot of national regulators lack the powers to
go in and do this kind of pillar-two supervisory control," says Azarchs at
S&P. "In addition, they may not have the experience or the staff to do it."

Is it realistic to expect all regulators to supervise their banks in the way
set out in the proposals? "Not all supervisors will suddenly set
differentiated requirements for all banks," says Page at the FSA. "But I
would expect that they will look at some of the bigger ones and at some of
the outliers."

But the Basel Committee can't force countries to invest in creating powerful
and expert regulators, even though this is exactly what the internal-ratings
based approach requires.
"These proposals appear to give greater powers to the market but in some
ways they do the opposite," says Linnell at Fitch. "Regulators will need to
set key data inputs and supervise the entire internal-ratings based
approach. That will increase the burden on regulators. Many are not going to
have the people or expertise to cope."

The big uncertainty of the Basel proposals lies with the capabilities of the
world's bank regulators. Much of the epic work they have written makes good
sense. But when the new accord is implemented in the real world, it will -
like the 1988 accord before it - take on a life of its own.

Administrator for Policy & Media Relations
ISDA
600 Fifth Avenue
Rockefeller Center - 27th floor
New York, NY 10020
Phone: (212) 332-2578
Fax: (212) 332-1212
Email: smarra@isda.org

- Ivy.gif