APPENDIX 4
Supplementary memorandum by BASIC
WHAT IS THE SUBSIDY TO THE ECGD?
1. ECGD PROPOSE
TO USE
A VARIANT
OF VALUE
AT RISK
UNDER TRADING
FUND STATUS
(VAR)
1.1 VAR quantifies losses with a defined
level of confidence (over a defined period).[19]
eg "losses are estimated to be £1 billion or less 99%
of the time over the next year". The higher the level of
confidence used, the greater the maximum loss. The longer the
period over which the value is "at risk", the higher
the loss.
1.2 VAR can be treated as a capital sum,
and a return (cost of capital) on this sum.
3. THE UNCERTAINTIES
ASSOCIATED WITH
THE VAR APPROACH
ARE:
3.1 How are the probabilities of default
calculated?19
3.2 What level of confidence should be used?
3.3 Over what period should the VAR be calculated?
3.4 What should be the cost of capital?
Should it be a commercial rate, or reflect the Government's cost
of debt?
A change in any one of these variables will
result in significantly different estimates of ECGD subsidy.
4. OUR ALTERNATIVE
APPROACH IS
TO LOOK
AT WHAT
THE MARKET
CHARGES FOR
IDENTICAL RISK
4.1 ECGD predominantly insures debts owed
by foreign sovereign governments. Such governments borrow on the
international debt markets. The international debt markets apply
a "risk premium" (over UK sovereign debt) to such foreign
sovereign debt. The ECGD guarantees take the place of this risk
premium.
4.2 Our alternative approach is well established
as a pricing mechanism for credit derivatives, and was first suggested
in the context of the ECGD by KPMG in 1999[20]
5. HYPOTHETICAL
EXAMPLE (IGNORING
TIME VALUE
OF MONEY)
5.1 Assumptions
| million |
|
| Principal guaranteed (assume bullet repayment after 5 years)
| 300 | |
| Annual premium: | 4 |
|
| Losses on the contract | nil
| |
| Borrowing rate of foreign government | 3.5%
| |
| Borrowing rate of UK Gov | 3.2%
| |
| 99% confident losses will be less than |
40 | over a 5 year period |
| Cost of capital, based on riskiness of equivalent market loan portfolio
| 20% | |
5.2 Cash surplus on contract (5 yrs x 4m pa)
5.3 Value at risk approach
| million |
|
| Annual cost capital implied by VAR: 40m x 20%
| 8 | |
| Cost over 5 years (8m x 5 years) | 40
| |
| Less 20 million premia received (40m x 20m)
| 20 | deficit |
5.4 Cost implied by the market (on date contract signed)
| million |
|
| Annual cost of guarantee: (3.5%-3.2%) x 300m
| 9 | |
| Cost over 5 years: (9m x 5 years) | 45
| |
| Less 20 million premia received (45m x 20m)
| 25 | deficit |
6. SUBSIDIES OUR
APPROACH FAILS
TO CAPTURE
6.1 ECGD offers exporters the opportunity (but not the
obligation) to enter into a contract for a significant period.
Exporters also have significant latitude as to when and whether
they go on cover through a contract. These are very valuable financial
options.
6.2 ECGD often guarantees transactions that contain significant
risks in addition to underlying sovereign risk—either because
they are with a special purpose vehicle with limited access to
foreign government guarantees, or because of the nature of the
underlying product.
6.3 The Bank of England acts as "lender of last
resort" to the banking sector. Benchmarking overseas sovereign
debt against commercial rates fails to recognise that commercial
rates are themselves subsidised by central banks via this prudential
system.
6.4 Fixed Rate Export Finance (FREF) has been acknowledged
as a subsidised product since 1999. Subsidies via FREF, quantifiable
via the hedging commonly undertaken by ECGD, are not included
in our total.
7. WHOSE GOVERNMENT
BUDGET DOES
THIS SUBSIDY
CURRENTLY APPEAR
UNDER?
7.1 Using the Government's balance sheet to guarantee
ECGD guarantees increases the Government's borrowing rates.
7.2 The incremental increase for each pound of debt is
very small. However, the total cost to the exchequer equates to
the subsidy above—at least in theory. In practice it is
impossible to measure whether the cost of more or less than indicated
by theory.
7.3 It is for similar reasons local government borrowing
(and borrowing in general) is constrained.
8. WHO BENEFITS
FROM THE
SUBSIDY, AND
BY HOW
MUCH?
8.1 The subsidy is shared between the exporter, the banks
and the overseas government. If the objective of the subsidy is
to preserve jobs, then a "direct" subsidy would be cheaper,
and could be targeted at producing products of clearer benefit
to the UK economy where market failure was a possibility.
9. WHAT INFORMATION
MIGHT THE
COMMITTEE ASK
THE ECGD FOR?
What would be the cost of providing an export guarantee for
a typical deal commercially? A bank (or panel of banks) marketing
credit derivatives could give such a quote (or quotes)[21].
This amount could then be compared with the return on capital
calculated under VAR, and with the premia payable.
10. CONCLUSION
Why does the ECGD not compare its operations with the private
sector? Such a comparison would quantify, in a straightforward
manner; the benefits the ECGD is providing to the exporter and
foreign government.
Is it legitimate to compare the commercial and ECGD values?
Yes. Government has no special ability to lay off risk in the
international debt markets.
Is it legitimate to subsidise exports? Not from an economic
perspective. As the York report shows, more jobs can be created
by acting otherwise. There may be social or diplomatic reasons,
but there are not economic ones.
And if there is a political decision made to subsidise exports
does not plain good governance (let alone the need for political
accountability) demand that the costs of such a policy be known?
19
This might be represented on the graph to the left below by point
A. The area to the left of point A would be 1% of the total area
under the curve below:
Capital market studies indicate capital
market prices are not normally distributed, but have "fat
tails" as indicated in the graph to the right. The likelihood
of large losses is much higher with "fat tails". This
is one facet of the risk of "catastrophic loss". The
comment of KPMG in 1999 in the context is interesting:
"The probability of default . .
. is revised on the basis of a "subjective adjustment"
. . . (which is) significant and the process is opaque, unstructured
and not formalised." Para 3.2.7. Back
20
KPMG `ECGD Risk Management Review' para. 4.7.6
". . . it is possible to . . . hedge through the use of reinsurance
or credit derivatives".
Interestingly KPMG assumed there were
benefits accruing to the Government of "Paris Club wrap-up"
(para 4.7.6). However, such benefits would imply the Government
could itself profitably trade emerging market sovereign debt.
This seems unlikely. Back
21
An investment bank would use the following information, only
some of which would be required from ECGD, and some of which is
available in the market:
1. UK Government and foreign government market yield curves (ie
interest rates), in the various currencies agreed in the loan
document for the period of the deal. These might be implied using
an interest rate parity where the data is not directly available.
2. Profile of loan drawdown and loan
replayment, for each currency, agreed under the contract.
3. Where optionality exists on drawdown
or repayment dates and amounts, the various significant combinations
should be modelled, with associated estimates of probability,
in order to value this optionality.
4. If choices exist as to the currency
drawn down, the variables implied by currency options in the market. Back
| 
