3. Central government borrowing and debt management 3.1 Objectives and comprehensive strategy
The National Debt Office raises and manages loans for the financing of the central
government debt, deficits arising in the central government budget, and other expenditures decided upon by parliament. The financing takes place through nominal loans, inflation-linked loans, and loans in foreign currency. The bulk of the borrowing is done by means of
instruments designed for trade in large quantities on securities markets. In addition, the Debt Office uses instruments especially suited to borrowing directly from the household sector.
The objective of the National Debt Office is to manage the central government debt so as to achieve cost minimisation in the long run while at the same time taking risk in the
management into consideration (the Law (1998:1387) on central government borrowing and debt management). According to the annual regulatory communication for 1999, the Debt Office shall achieve lower absolute costs for the debt through measures to enhance market efficiency. Moreover, the Debt Office shall, over a five-year period and with due
consideration of risk, attain a lower cost for the government debt in relation to the cost of the benchmark portfolio established by the Board of Commissioners.
In May 1998, parliament decided on a new decision-making structure and a new formulation of the objective of the central government debt policy. The new rules took effect at the end of the year. Accordingly, the 1999 business year was the first year covered by the new set of rules. The changes primarily imply that existing practices are clarified and codified and that a more consistent system of management, evaluation and delegating is created. This meant that the government provided guidelines for the composition and interest fixing period of the entire government debt instead of as before only giving guidelines for the volume of foreign currency borrowing.
The Board of the National Debt Office has the responsibility for putting the decisions of the government into operation. Accordingly, the board has established benchmark portfolios for debt in nominal kronor and foreign currency, respectively. It has also established permissible discrepancies with regard to maturity (interest fixing period) and foreign currency
composition, respectively. In this manner, the Board has laid down limits for the risk-taking in relation to the benchmark portfolios. By estimating the cost of borrowing in accordance with the benchmark portfolio and comparing with the actual cost of borrowing, a measure of the result of the activity of the Debt Office is obtained.
However, comparisons with benchmark portfolios only provide a partial picture of how the cost of the debt is influenced by the management of the debt. They do not, for example, catch the effects of measures which influence the cost of both the actual debt and the benchmark.
Measures to enhance market efficiency, among others, belong here. If the Debt Office through its choice and design of instruments or through initiatives to improve the functioning of the bond market or in other ways contributes to a lowering of borrowing costs, the costs as measured by the benchmark portfolio decline to the same extent. The work of the Debt Office to improve market efficiency is guided by the objective of attaining the lowest possible cost, but the effects have to be assessed by other methods than benchmark portfolios. In its regulatory communication for 1999, the government stated that measures to enhance market
efficiency are to be described and analysed. If possible, the effects of these measures on the interest rate costs are to be quantified. This account is to be found in Section 3.2.4.
The management of the central government debt is influenced as well by more all-embracing risk considerations than those measured by discrepancies in relation to the benchmark
portfolio. By borrowing via instruments with different characteristics, for example nominal and inflation-linked loans, and in different markets, for example in kronor and in international capital markets, and from both institutional and household investors, the dependence on the borrowing terms prevailing on any given market is reduced. Like the handling of a well- managed asset portfolio, the government debt policy is guided by a diversification approach aiming to minimise the risks of large fluctuations in borrowing cost. This diversification approach found expression in the Debt Office´s proposed guidelines and also characterizes the decision, adopted by the government, on the guidelines for the management of the debt in 1999.
Evaluation by means of benchmark portfolios can be described as giving a measure of the relative costs for some parts of the debt while the objective refers to the absolute costs for the debt in its entirety. For a more complete evaluation it is therefore essential to try to consider whether the debt as a whole is composed and managed in such a manner that the absolute costs become as low as possible given the constraints that have to be taken into account. This means that the evaluation has to include other aspects than benchmark portfolios. The
complexity and difficulty of finding suitable yardsticks for absolute costs, the effects of measures to improve market efficiency etc. makes it necessary to take qualitative assessments into account as well.
3.2 Borrowing in the securities markets
3.2.1. Strategic considerations
In order to attain the goal of long-term cost minimisation, the Debt Office has laid down a number of strategic points of departure for the borrowing. They are enumerated below.
Predictability and transparency. In the Swedish market, where the central government is the dominant issuer, it is important that the Debt Office acts in a long-term and predictable manner. The borrowing strategy must be known and measures announced in a timely fashion.
Clear rules of the game for investors and dealers help keep down the yield requirements and thereby reduce the borrowing costs. In the past, foreign currency borrowing has been
characterized by predictability to a much lesser extent. The increased integration of different markets, which for example has manifested itself in the introduction of parallell benchmark loans in kronor and in euro, makes the difference between domestic and foreign borrowing smaller. Predictability and transparency must therefore nowadays be heeded in the foreign currency borrowing as well.
Diversification. By diversifying the borrowing, i.e. exploiting several different markets, the Debt Office can achieve a lower cost on each market. The Debt Office borrows in both domestic and foreign currency, through nominal as well as Inflation-linked Bonds, and in different maturities. Diversification also reduces the risk, since the Debt Office becomes less exposed to changes in the situation on different markets.
Market maintenance. Being the incomparably largest borrower on the domestic market, the Debt Office can lower the central government borrowing costs by promoting liquid and effective markets. This is done for example by concentrating the borrowing on a limited number of instruments. Moreover, constant efforts are being made to improve the functioning of the markets. The increased integration between domestic and foreign borrowing makes market maintenance more important in the foreign currency borrowing as well.
Limited refinancing risk. The borrowing is to be conducted in such a manner that the maturity profile is smoothed. This reduces the risk that the borrowing costs are run up in the event that borrowing terms are particularly unfavourable at a point in time when a large part of the debt is to be refinanced.
3.2.2 The structure of the market
The start of the EMU has brought large changes in the European capital markets. To a greater extent, European investors have diversified into new categories of credits. This can be
assumed to have helped increase the interest in markets for government paper outside the EMU. Moreover, liquidity has been channelled to the largest markets, in particular the German market. The emphasis on liquidity has entailed that smaller markets for government paper within the EMU has found it difficult to attract new investors from other EMU
countries. Borrowing costs have therefore risen in the smaller countries in comparison with Germany. In the Swedish market for government paper, turnover fell in 1999 by about 15 percent compared with the previous year.
The amalgamation of eleven markets into one implied that the agents in the market intensified their efforts to improve the liquidity, transparency and efficiency of the market. This has occurred through the establishment of a number of electronic platforms for trade in the bond market. Euro-MTS, which trades in government paper in euro, is an example of such a system. At this time only the most liquid government papers in the EMU are traded in this system, but in several of the smaller countries, plans are well advanced to participate by establishing local MTS-systems. Holland launched such a system already in 1999. This is an example of how electronic trading can increase access for government paper from smaller countries.
Trade in futures contracts
In order to increase interest in Swedish government paper among foreign investors, the Swedish market place is subject to the same demands for liquidity, transparency and
efficiency. An important step towards increased transparency and trade in an electronic, order- driven system was the introduction of two futures contracts in the fall of 1998. The contracts were intended to function as the main instrument for trade in interest rate risk. However, the trade became excessively dependent on active market makers. These were to support liquidity in the initial phase whereupon the underlying order flow was meant to take over. However, the underlying order flow never became sufficiently large; in practice the contracts stopped
functioning when the market makers – during the summer of 1999 – stopped quoting binding two-way prices. Instead the old structure based on trade by telephone was restored.
A new electronic system for trade in government paper
As trade in the futures contracts came to an end, the need to make the market more efficient appeared more urgent and consequently work was started on the establishment of an electronic
platform for trade in government bonds. The dealers and the Debt Office participate in this work. The main purpose of an electronic trading platform is to attract a larger number of participants and thereby improve liquidity and increase the availability of government paper.
Changes in the circle of dealers
During the course of the year JP Bank ceased being a dealer and Salomon Smith Barney began.
3.2.3 Market maintenance in the domestic market
In its regulatory communication the government has described how the Debt Office´s market maintenance measures are to be described and analysed. If possible, the effects of these measures shall be quantified.
The concept of market maintenance can include a number of different measures and it is difficult to give an unambiguous definition. The broadest approach would be to assume that practically everything except a mechanical borrowing to cover government funding
requirements is to be considered as market maintenance. In such a case, the term market maintenance would include for example information for investors, actions in matters of market structure and other activities not directly related to the actual borrowing transactions.
A more narrow definition would be to endeavour to specify transactions which the Debt Office enters into or might enter into and which have a market maintenance character. They would be defined on the basis of whether they could be assumed to lower the absolute interest rate costs for the government as distinguished from transactions that aim to reduce the relative costs and are captured when costs are measured relative to the benchmark.
Even though a large part of the activities of the Debt Office can be described as measures aiming to lower the absolute borrowing costs for the government, the broad definition does appear too sweeping. It would be more reasonable in this context to discuss primarily different kinds of activities in the market which aim to lower the absolute costs. But then one needs to keep in mind that this is a relatively narrow definition and that other definitions are possible.
Below is a classification of different market maintenance measures together with a description of their purpose. In addition, a certain quantification of possible costs is given.
Exchanges and buybacks as a means to improve liquidity and transparency
Creating loans with large outstanding volumes improves liquidity and thereby the pricing of these loans. A liquid secondary market and transparent pricing reduces liquidity risk
premiums which reduces the government´s borrowing cost.
The cost of creating large loans arises because of transaction costs when older loans are exchanged for new ones as a result of the difference between the buying interest rate and the selling interest rate. In the market for nominal bonds in 1999 these transaction costs are estimated at 50 million kronor in connection with exchanges amounting to 28 billion kronor in total. This cost has to be seen in relation to the gains in the form of a more liquid market.
Estimated on the basis of the issue volumes of the most recent years, about 70 billion
annually, a lowering of the borrowing cost by one basis point corresponds to savings of about 50 million kronor. Evidence from other markets, for example the American market for
government paper, indicates that the cost difference between liquid and less liquid loans can
amount to 10-15 basis points. Thus, the cost of reduced liquidity can be expected to exceed the transaction costs arising as a result of exchanges made for the purpose of creating larger loans.
On the market for Inflation-linked Bonds, the uncertain pricing picture does not allow a quantification of the transaction costs in connection with exchanges.
Exchanges and buybacks for the purpose of reducing risk in connection with refinancing Benchmark loans with less than a year to maturity are typically exchanged for packages of Treasury Bills that are neutral with respect to duration. The exchanges serve two purposes:
first, the refinancing risk is reduced when the maturing of several large bond loans is spread on a number of Treasury Bill loans (which is not to be viewed as market maintenance but is warranted by risk considerations); second, investors are offered the opportunity to exchange short bond loans which risk becoming less liquid for more liquid Treasury Bills. This promotes liquidity in the Treasury Bill market as well. The transaction cost for these
exchanges is nil since the buying interest rate is established on the basis of a weighted average of the interest rate on the Treasury Bills being issued.
On-tap exchanges in the market for inflation-linked bonds as means to improve the secondary market
On-tap exchanges offered to dealers in Inflation-linked Bonds increase their opportunities to trade in the secondary market. This improves liquidity. There is, however, a risk that the Debt Office assumes the role normally performed by intermediaries and other participants in the secondary market. The volumes transacted by the Debt Office therefore have to be adapted to the total activity in the market for Inflation-linked Bonds.
When the Debt Office receives transaction compensation in the form of a difference between the buying and selling interest rate, there should be no direct cost involved. Afterwards, however, it can turn out that a cost arises if the price relation between the two loans is unfavourable. But since the exchange activity is rather limited in relation to the outstanding stock, gains or losses arising from on-tap exchanges tend to be small in relation to the total cost of the borrowing in Inflation-linked Bonds.
Activities in the repo market to improve the functioning of the market
In addition to government paper supplied in repos when there is a borrowing need, the Debt Office also offers repos in the secondary market designed to improve the functioning of the market. The purpose is to maintain good liquidity in the market for government paper. In this case as well, it is important that the Debt Office does not become too dominant a participant, since this could obstruct the general development of the market. However, the total repo activity of the Debt Office, i.e. short-term financing transactions and transactions to improve the functioning of the market, amount only to about two percent of the turnover on the market for repos.
In newly issued government paper where the outstanding stock is small, a certain minimum outstanding volume is guaranteed. This arrangement is designed to facilitate trade in newly issued loans with small outstanding volumes. Without this facility there would be a risk of locking-in effects in case a large part of the outstanding volume is held by a small number of investors. These transactions are conducted at a rate 15 basis points below the repo rate of the Riksbank and thus involve a small saving for the Debt Office.
For Iinflation-linked Bonds each of the Debt Office´s dealers is offered 120 million kronor in guaranteed repo volume. The repo facility is used regularly and facilitates trade since dealers can sell bonds not held in their portfolios. These repos are also conducted at a rate 15 basis points below the Riksbank´s repo rate and thus involve a small saving for the Debt Office.
In order to help the situation on the market for government paper around the turn of the year, the Debt Office introduced an additional facility during the period December 15 - January 10.
The purpose was to minimise anxiety about disturbances in the system and reduce the risk of shortages in the market. Since many participants suspended their repo activities in the period before and after the transition to the new year, the facility was widely used by the dealers. It is likely that as a result, the market functioned better around the turn of the year.
To sum up, the activity in the repo market contributes to a lowering of the domestic interest rate level and thus to a reduction of the financing costs. The size of this saving is difficult to estimate, however. Apart from this positive effect on the absolute interest rate level, there is also an effect that can be measured. This arises because repos are more cost effective than other short-term borrowing since repos are conducted at a lower interest rate than the overnight rate. In 1999, this entailed a saving of 4.2 million kronor.
3.2.4 Borrowing in nominal Swedish kronor
The main part of the National Debt Office´s borrowing in nominal Swedish kronor is
conducted through nominal Treasury Bonds and Treasury Bills. Bonds are used primarily to cover long-term borrowing needs while Treasury Bills are used to offset short-term
fluctuations in the borrowing requirement. The distribution is governed by the benchmark for the nominal debt. In order to balance daily surpluses or deficits in central government
disbursements, the Debt Office also conducts liquidity management through overnight borrowing, deposit transactions, and repos.
In spite of a surplus in the borrowing requirement in 1999, a normal issue activity in nominal Treasury Bonds and Treasury Bills has been maintained. This is due to the continued net amortization of the foreign currency debt (see definition in footnote 1 on page 8) in an amount equivalent to about 24 billion kronor and to the Debt Office´s high activity in kronor/foreign currency swaps (an increase by about 40 billion kronor). In addition, borrowing in Inflation- linked Bonds has been rather limited (about 3 billion kronor).
Nominal treasury bonds
At the end of 1999, the outstanding stock of nominal Treasury Bonds was 667 billion kronor (excluding repos)1, a decline during the year of about 14.4 billion kronor. During the year auctions were conducted every other week on altogether 25 occasions. The volume issued amounted to 3 billion kronor on each occasion, except for the first auction of the year which was for an amount of 4 billion kronor (see figure 3.1). The average maturity was 7.3 years, implying that bonds issued were on average 1.5 year shorter than in 1998. This shortening is explained primarily by the fact that the Debt Office mainly conducted issues in new seven- year benchmark loan, while issues in 1998 to a great extent were conducted in an 11-year benchmark loan.
1 Treasury Bonds including repos amounted to 678.9 billion kronor.
Figure 3.1 Volume of nominal Treasury Bonds offered
0 1 2 3 4 5 6 7 8
SEK Billion
1998 1997
1996 SEK Billion 83
SEK Billion 50
SEK Billion 81 SEK Bilion 76
1999 SEK Billion 169
1995
Each column represents one auction. The total volume offered per year is stated above the columns
During the first half of the year interest in the auctions was on par with that of the previous year. However, the auctions during the second half were not quite as well received by the market. Both the tail spread (the difference between the average interest rate and the highest accepted rate ) and the interest rate increment (the difference between the average interest rate and the market rate) rose during the second half, while at the same time the subscription ratio (total bids in relation to the volume offered) declined, which is without exception
unfavourable for the auctions (see table 3.1). A possible explanation of this development could be a generally less ample liquidity – the turnover was about 20 percent lower during the second half than during the first half in both bonds and bills. The reduced issue volumes amounted to only 300 million kronor. One explanation could be the declining borrowing requirement.
Table 3.1 Average figures for tail spread, yield mark-up in basispoints and cover ratio för nominal Treasury Bonds 1995-1999 (Standard deviations in brackets)
Tail spread1
1,93 1,42 1,02 1,05 0,55 0,53 0,59 0,60 0,61 0,91 (0,78) (0,74) (1,10) (0,56) (0,67) (0,30) (0,39) (0,23) (0,36) (0,58) Yield
mark-up2
1,95 0,34 -0,04 1,71 0,39 0,56 0,90 0,61 0,74 1,28 (1,85) (1,61) (2,58) (1,78) (1,58) (1,20) (1,71) (1,74) (1,66) (1,86) Cover
ratio3
2,01 2,46 2,27 2,00 2,46 2,58 2,33 2,50 3,08 2,62 (0,91) (1,30) (1,38) 0,85) (1,12) (1,02) (0,83) (0,96) (2,17) (1,18)
1 Difference between average yield and highest accepted rate offered.
2 Difference between average issue yield and market yields immediately prior to the auction. When exact market yields have not been readily obtainable, these have been reconstructed in some cases, and therefore the figures should be interpreted with some caution.
3 Volume of bids submitted in relation to volume offered.
Benchmark Loan 1044 (3.5 percent, April 2006) was introduced during the year. The loan is a parallel to Loan 1044E, which is denominated in euro. In the event of a future Swedish
membership in the EMU the two loans will merge into one loan.
The only scheduled loan maturing in 1999 was Loan 1028 with an outstanding amount of 18 billion kronor. The main part of the loan had been exchanged for Treasury Bills during the previous year. Apart from regularly maturing loans the Debt Office also manages buy-backs and exchanges4. During 1999 altogether 40 billion kronor were offered under exchange auctions between nominal bonds, of which 34.2 billion were subscribed.
In May, the National Debt Office launched a facility for stripping of Swedish Treasury Bonds.
Such facilities have existed for a long time in the USA and France and have recently been introduced in several countries in Europe. Stripping implies that the Debt Office offers exchange transactions whereby coupon bonds are bought and at the same time a set of zero- coupon bonds – which together give exactly the same cash flow as the coupon bonds – are issued. The exchange can also be made in the opposite direction, i.e. loans that have already been stripped can be put together as the original bond. The purpose of the stripping facility is to offer the market a new investment vehicle in the form of nominal zero-coupon loans.
However, like in other countries where stripping has been introduced, the start has been sluggish. So far, only 6 million kronor of Loan 1041 have been stripped. The poor interest probably reflects concern about inadequate liquidity in the newly created instruments, which
4 All volumes in this paragraph are exclusive of non-comp sales.
1995 1996 1997 1998 1999
95:1 95:2 96:1 96:2 97:1 97:2 98:1 98:2 99:1 99:2
initially risk having limited volumes. It is probably felt that this risk is greater in a situation where the government debt is expected to decline.
Since 1998, the Debt Office offers a so called non-competitive facility in bond auctions. This facility implies that the dealers, on the basis of the allocation during the four most recent auctions, can request an additional allocation at the same average price as at the auction. The purpose is to encourage the dealers to submit aggressive bids in the auctions and make it more attractive to sign dealer agreements. . Theoretically, the measurable cost that the facility entails for the Debt Office should be weighed against a lower interest rate cost in the auctions.
In principle, the dealers are offered an option to acquire a certain volume of bonds, something that should be mirrored in more aggressive bidding. The facility comprises altogether 10 percent of the issue volume at the respective auctions. During 1999, the use of the facility at regular auctions has amounted to just under 50 percent. Non-comp is also offered in exchange auctions. Here the degree of usage has been slightly higher, or just under 70 percent. The additional revenue of the dealers since the facility was introduced at the end of 1998, is estimated at about 20 million kronor, including exchange transactions.
Treasury Bills
The total offered volume of Treasury Bills in 1999 was a nominal amount of 351.5 billion kronor, of which 327.5 billion was issued. The average maturity was 0.6 year. On December 31, 1999, the outstanding stock of Treasury Bills amounted to 246.4 billion kronor (paid amount, excluding repos)5, equivalent to an increase of 22.7 billion during the year.
Figure 3.2 Volume of Treasury Bills offered
0 4 8 12 16 20 24
28 SEK Billion
12-months 6-months 3-months
1996 1997 1998
SEK Billion 233 SEK Billion 144
SEK Billion 320
1999
SEK Billion 352
SEK Billion 227
1995
Each column relates to one auction. The total volume issued each year is stated above the columns
5 Treasury Bills including a reverse repo reduced the debt to 243.6 billion kronor.
Like with nominal Treasury Bonds, Treasury Bills are issued through auctions every alternate week. Apart from managing the maturity composition of the total debt, issues of Treasury Bills also serve the purpose of offsetting short-term fluctuations in the borrowing requirement.
This implies that the issue volumes, like the size of the stock of outstanding Treasury Bills, are subject to relatively large fluctuations from month to month. Figure 3.2 shows how the issues have been distributed by maturity at the auctions during the last four years.
The interest rate spread (the difference between the average yield and the highest accepted rate) has displayed a similar development as in the case of bonds, i.e. the first half of the year resembled the previous year, while in the second half conditions were less favourable (se table 3.2). The interest rate spread almost doubled compared with the same period the year before, especially in auctions of longer maturities. One reason could be that investors chose to invest in the shorter end of the curve to a greater extent. During the latter part of 1999, monetary policy was changed in a more restrictive direction, which led to increased uncertainty concerning changes in short-term interest rates. However, the interest rate increment (the difference between the average interest rate and the market rate) has more than halved which is a positive sign. The subscription ratio (bids in relation to amount offered) is largely
unchanged.
Table 3.2 Average figures for tail spread, yield mark-up i basispoints and cover ratio for Treasury Bills 1995-1999
(Standard deviation in brackets)
1995 1996 1997 1998 1999
95:1 95:2 96:1 96:2 97:1 97:2 98:1 98:2 99:1 99:2 Tail-
spread1
0,96 0,96 0,98 0,78 0,88 0,76 0,66 0,71 0,76 1,30 (0,81) (0,63) (0,79) (0,57) (0,71) (0,52) (0,61) (0,65) (0,64) (0,86) Yield-
mark-up2
0,74 0,67 0,40 0,76 1,23 1,37 1,23 1,06 0,52 0,35 (1,45) (2,99) (1,40) (1,56) (1,45) (1,10) (1,06) (0,86) (1,28) (1,47) Cover-
ratio3
3,39 2,97 3,61 2,95 2,60 1,85 2,70 2,32 2,36 2,29 (2,09) (1,75) (2,04) (1,77) (1,78) (1,22) (1,76) (1,28) (1,17) (1,60)
1 Difference between average yield and highest accepted rate when offered.
2 Difference between average issue yield and market yields immediately prior to the auction. When exact market yields have not readily been obtainable, these have been reconstructed in some cases, and therefore the figures should be interpretated with some caution.
3 Volume of bids submitted in relation to volume offered.
3.2.5 Liquidity management
The National Debt Office conducts liquidity management for the purpose of offsetting the daily fluctuations in the central government´s payment flows. Liquidity deficits and liquidity surpluses are actively financed and invested, respectively, on the Swedish money market.
During most days in 1999 (about 80 percent), the payment flows resulted in deficits. Surpluses arose only during a few days in the middle of each month. During 1999, the Debt Office borrowed on average 16.2 billion kronor per day on short term. Of this amount, 13.4 billion was financed on the deposit market and 2.2 billion by means of repos. The remainder was financed through issues of on-tap bills and through participation in the Riksbank´s weekly repos.
Borrowing on the deposit market normally takes place at an interest rate that corresponds to the repo rate of the Riksbank. Compared to borrowing in the deposit market, the Debt Office can obtain cheaper financing through participation in the repo market. In order not to assume a dominant role on the repo market, the Debt Office limits its participation in this market.
Result and evaluation
In 1999, financing in the repo market entailed savings of 4.2 million kronor compared with financing in the deposit market. In 1998, the savings amounted to about 7 million kronor, in 1997 to about 3.5 million kronor and during the 1995/96 budget year, which comprised of 18 months, to 10.5 million kronor. The decline from 1998 to 1999 was mainly a consequence of a reduced volume of repos. The reason was that the Debt Office chose not to influence the volume in the market through its pricing of repos. The Debt Office chose a price that generally was higher than the pricing by the market. Minor incidents of shortages in the market for government paper may also have been a reason why the volume declined.
Repo and deposit transactions are also used when the Debt Office invests surplus liquidity.
When the Debt Office had large investment needs, buy-backs were also conducted to some extent, mainly of Treasury Bills approaching maturity. Investments in the overnight market are usually made at an interest rate level corresponding to the Riksbank´s repo rate. Buybacks were made at more or less the same rate as the Riksbank´s repo rate. Accordingly, there has been no savings or additional costs as a consequence of the buybacks.
On-tap bills
Issues of Treasury Bills on tap were introduced in May 1999 as a supplement to the other instruments used for liquidity management. The Treasury Bills are issued at maturities ranging up to six weeks and permit the Debt Office to raise longer loans instead of rolling over
overnight loans. On-tap bills were issued totalling the amount of 56 billion kronor with an average maturity of just over two weeks, and the interest rate level at which these were issued was marginally lower than the Riksbank´s repo rate.
3.2.6 Benchmark results for the nominal kronor debt
For the fiscal year 1999, the Board of the National Debt Office adopted a new benchmark for debt in kronor. This benchmark differs from that of previous years in many respects. Firstly, the benchmark is constructed as a transaction benchmark. Briefly, this means that the benchmark is defined as a series of transactions, which for most of the debt is equivalent to issues. Secondly, the benchmark is arranged so that a certain average maturity of the debt rather than a certain duration is to be maintained. Thirdly, the new benchmark is not based on exactly the same definition of debt as the benchmark of past years; borrowing from and lending to central government agencies has been excluded. As in past years the benchmark only applies to the nominal debt. It was decided that the new benchmark was to apply as of April 1999. During the first quarter of the year the debt was considered to be in line with the
benchmark, i.e. the result for these months is equal to zero by definition. For Treasury Bonds and Treasury Bills it is natural to define the benchmark in terms of issues in existing and planned loans. For the swaps between fixed interest rates in kronor and floating interest rates in foreign currency, which the Debt Office uses for some of its foreign currency borrowing, the benchmark is based on the assumption that the planned volume is distributed in a certain manner over the year and is raised in a certain combination of 5-, 7-, and 10-year swaps.
Finally, the benchmark for the retail market debt is expressed as monthly changes in principal and in maturity.
For the nominal debt in kronor, the result is calculated as the difference between the cost of the actual debt and the cost of the benchmark. The costs are measured as changes in market value. Since Treasury Bonds represent the largest single part of the debt and the rest of the debt also in general has a maturity in excess of one year, a large part of the cost will not be realised. However, the result has not been divided into a realised and an unrealised part.
Activities which have led to discrepancies in relation to the benchmark
In 1999, the Debt Office has not issued exactly in accordance with the established benchmark and by definition has therefore deviated from the benchmark. However, the deviant issues have in principle been warranted by changes in the borrowing requirement, market
maintenance, an ambition to reduce refinancing risks etc. In no case has an active decision been taken to issue a different bond other than the one in the benchmark on the basis of a certain interest rate assumption. In any event, the differences between the actual issues and the benchmark have been rather modest and have not created any major differences between the two portfolios, whether in maturity or result.
The difference in respect of maturity between the benchmark and the outcome is due instead to the Debt Office having entered into larger volumes of so called kronor/foreign currency swaps than allowed for in the benchmark. The Debt Office obtains a fixed interest rate in the swaps and since the floating rate it pays in the kronor swap exactly corresponds to the floating rate it receives in the basis swap6, the swaps influence the duration of the overall kronor debt in the same manner as a purchased bond. Thus, a larger volume of swaps implies a shortening of the maturity of the portfolio. In addition, the swaps that were entered into during the year had longer average maturities than indicated in the benchmark, which further increased the discrepancy relative to the benchmark. This discrepancy is justified by the cost savings implied by the greater difference between the longer swaps and the government bond rates.
Development of interest rates and result
The interest rate development during the year was characterized by two trends. First, interest rates generally rose during the year in the case of both Treasury Bonds and Treasury Bills.
From the beginning of April, when the benchmark was adopted, until the end of December, the ten-year rate increased by 130 basis points to about 5.7 percent. The second trend consisted in the global economic recovery leading to stronger expectations of quicker and larger increases in short-term interest rates. Diagrammatically this trend is indicated by the so called yield curve becoming much steeper at the short end than at the longer end. In figure 3.1, showing yield curves for the Swedish market in government paper, both of these trends are illustrated.
6 In a basis swap, a variable rate is paid in one foreign currency and a variable rate is received in a different foreign currency. A kronor/foreign currency swap consists of a regular interest rate swap and a basis swap.
Figure 3.3 Yield curves beginning and end of 1999
Interest rate (%)
0%
1%
2%
3%
4%
5%
6%
0 2 4 6 8 10 12 14 16
Maturity (years) 12/31/98 12/31/99
Yield curves have been estimated from market rates for treasury bills och benchmark bonds in the Debt Office´s model for estimation of zero-coupon curves.
The result for the debt in kronor is measured, as mentioned above, as the difference in cost, based on market value, between the benchmark and the actual portfolio. Since higher interest rates result in lower market value, cost as measured in this manner will fall when interest rates rise. The market value of a portfolio with a longer maturity will fall more than one with a shorter maturity.7
The fact that the maturity has been consistently somewhat shorter in the outcome than in the benchmark, has implied a negative result for the nominal kronor debt. Viewed over the year as a whole the cost of the outcome has been about 200 million kronor higher than the cost of the benchmark. Although this is a large amount, the difference is very small in relation to the total cost of the debt. In percentage terms the amount corresponds to about two basis points
(hundreds of a percentage points). It should also be emphasized that the model for the
calculation of the result is based on a few simplifications, so that the result must be considered to be within a reasonable margin of error.
Comparison of results for the most recent five-year period
Table 3.3 below shows the results for the nominal kronor debt during the most recent five years. The estimate of the cost of the private market debt for 1995-98 has been adjusted in accordance with the calculation principle adopted for 1999. Even with this adjustment, however, it is not altogether easy to compare the result with that of earlier years. On the one
7 Since the difference in maturity between the actual portfolio and the benchmark is relatively small, this relationship is not always valid: in some months the kronor debt as a whole has had a shorter maturity but the cost has nevertheless fallen when interest rates have risen. The reason has to do with the distribution of the interest rate risk along the yield curve in the outcome and in the benchmark, respectively. The fact that maturity is much less correlated with changes in market value than is duration, further reduces the possibility to analyse the result on the basis of changes in maturity.
hand, the construction of the benchmark has been changed and on the other, the calculations for earlier years used to include the cost of borrowing from and lending to central government authorities. However, the effects on the overall result of the borrowing and lending activities are not of the order of magnitude that they can explain the large difference between last year´s result and that of 1995-98.
Table 3.3 Result for the nominal krona debt in relation to benchmark 1995-99
1994/95 1995/96 1997 1998 1999 Sum (18 mths)
Result (millon kr) -325 -16 250 661 1 178 -202 -14 938 Result (i % of debt) -0,04 -1,1 0,06 0,11 -0,02 -0,27
Judging from the size of the results, the Debt Office has been considerably closer to the benchmark than in past years. This is largely due to the difference in the construction of the benchmarks. During 1995-98, the benchmark was a duration figure and the benchmark portfolio included exactly those issues needed to arrive at this duration. The transactions under the benchmark portfolio were often such that they could not be replicated in the actual portfolio. Thus, the transactions under the actual portfolio could end up being very different from the benchmark, with consequent effects on the result.
Of the five years, it is in particular 1995/96 that shows a very large figure for the result. This is explained by the fact that the Debt Office during this time, following upon the large increase in debt during the early 1990´s, chose to increase the average maturity of the debt in order to reduce the refinancing risk. However, the benchmark for the kronor debt happened to be changed only after the maturity of the actual debt had been changed. As a result, a
discrepancy arose in relation to the mid-point of the original benchmark and this had a relatively large effect on the result. In retrospect, it would have been more appropriate to change the benchmark in line with the average maturity of the actual debt. Then there had been no discrepancy and consequently no effect on the result; the actual cost of the debt had been the same in any event.
The benchmark for 1999 was constructed so that it would be possible to replicate to a very high degree. This is what the National Debt Office has also done in all essentials, which on the whole, has yielded a result close to zero.
3.2.7 Borrowing in Iinflation-linked Bonds
The Debt Office has issued Inflation-linked Bonds since April 1994. The main purpose is to achieve a further diversification of the composition of the debt and also to offer investors diversification opportunities by means of a new class of assets. The long-term requirement posed on borrowing via Inflation-linked Bonds is that this form of borrowing not be more expensive than nominal borrowing. In the guidelines for 1999, the government stated that the Debt Office should aim not to reduce the outstanding stock. In order to meet the long-term
requirement, different kinds of market maintenance measures are needed, which may in the short run lead to somewhat higher costs or to a decline in the outstanding stock.
Measures taken during 1999
In April, the Debt Office returned to auctions and introduced two new loans – of which one 30-year loan – with a deflation guarantee, all of which in line with international examples. At the same time a possibility was introduced for dealers to conduct running exchanges between different inflation-linked bonds with the Debt Office. The purpose was to facilitate the role of the dealers as market guarantors and thus to contribute to market efficiency. The return to auctions was, apart from the adjustment to international practice, a way of transferring the pricing of the loans to the market and to increase transparency with respect to the supply during the year. For 1999, it was announced that issue volumes would be up to 5 billion kronor, depending on demand.
In December the Debt Office announced further changes to be introduced during the course of 2000. To contribute actively to a broadening of the investor base – which is the key to a bigger and more efficient market in Inflation-linked Bonds – and to work to familiarise and increase knowledge of Inflation-linked Bonds as attractive additions to well-managed portfolios. These activities will be carried out in co-operation with dealers and primarily aim at smaller
institutional investors, but in the longer term will target households as well. In order to increase dealer incentives to cultivate potential customers, the dealer contracts were
terminated at the turn of the year and will be replaced by one-year contracts with active and committed market participants who view Inflation-linked Bonds as an interesting product to offer their customers.
Market developments
Auction activities were begun in April with three auctions, in which almost the entire offered volume of 2500 million kronor was placed on the market. Three more auctions were planned for August but were cancelled. The auction activities were concluded with three auctions in October, which had the same design as the auctions that had been cancelled in August. This time a little more than 70 percent of the total offered volume of 3000 million kronor was placed.
Altogether, the outstanding stock (measured as accumulated, discounted amount, which is how Inflation-linked Bonds are measured in the accounts of the government debt) rose in 1999 from 93.7 billion kronor to 96.5 billion kronor8, corresponding to an increase from 6.5 percent of total government debt to 7.0 percent.
8 Including repos, the amounts are 93.8 and 96.7 billion, respectively
Figure 3.4 Outstanding stock of Inflation-linked Bonds 1994-1999
0 10 20 30 40 50 60 70 80 90 100
Coupon bonds
Zero coupon bonds Billion Kronor
1994 1995 1996 1997 1998 1999
As shown in figure 3.4, there is a continuing trend towards coupon bonds at the expense of zero-coupon bonds. The reason is on the one hand that the refinancing risk that arises when zero-coupon bonds mature is reduced, and on the other, that investors more and more have showed a preference for coupon bonds.
The ten-year real rate of interest began and ended the year at about 4 percent, but at its lowest, it stood at 3.3 percent (in May) and at its highest it was 4.5 percent (in August), see figure 3.5.
During the year, the ten-year nominal rate of interest rose from 4.1 percent to a high of 6.1 percent (in October), standing at 5.7 percent at the end of the year. This means that the break- even inflation, i.e. the difference between the nominal and real rates of interest, for this maturity rose from 0.3 percent to about 1.6 percent. At the same time the actual rate of inflation has increased from -0.4 in January to 1.2 percent in November. The break-even inflation is the highest rate that the average future rate of inflation can amount to in order for borrowing via Inflation-linked Bonds to be cost neutral as compared with borrowing via nominal bonds.9
9 Strictly speaking it is the relationship between break-even inflation and future expected inflation that determines whether borrowing via Inflation-linked Bonds is preferable to nominal borrowing.
Figure 3.5 Ten year nominal and real interest rate, break-even inflation and actual inflation 1999
-1,00 0,00 1,00 2,00 3,00 4,00 5,00 6,00 7,00
qr 1 qr 2 qr 3 qr 4
SGB1043 (5%, Jan. 2009)
IL 3101 (4%, Dec. 2008)
Break-even inflation
actual inflation
The break-even inflation amounted to 0.9 percent, 1.1 percent, and 1.6 percent at the three auction periods in April, August, and October, respectively. This means that the issue conditions for Inflation-linked Bonds became more and more favourable during the year as compared with the actual inflation on these occasions. That the Debt Office nevertheless decided to cancel the auctions in August was due to the fact that the price formation in the secondary market showed evidence of being excessively influenced by the Debt Office´s auction. Real interest rates rose noticeably prior to the auction while at the same time, on the contrary, nominal rates of interest fell. A consequence of which break-even inflation declined by 0.6 percent. The Debt Office considered it appropriate to arrange a series of meetings with investors where the background to the cancelled auctions and the Debt Office´s views of the market were explained. Following the cancellation of the auction, the price formation has normalised.
Results
The calculated result of total borrowing in Inflation-linked Bonds since the start in 1994 increased from 5.1 billion kronor in 1998 to 6.8 billion in 1999, see figure 6.1. The improved result in 1999 has to do with the lower rates of inflation that have been recorded during the period. The result is estimated as the cost difference up till the present between borrowing in nominal and in inflation-linked bonds, which consists of the difference between expected inflation at issue and realised inflation until the time of evaluation. A rising inflation can have adverse effects on the actual result when the bonds are redeemed.
Figure 3.6 Calculated results of total borrowing in Inflation-linked Bonds accumulated for period 1994-1999
0 1 2 3 4 5 6 7
1994 1995 1996 1997 1998 1999
Million kr
Commissions
The Debt Office´s dealers receive commission for their activities in both the primary and the secondary market. During 1999, these commissions amounted to 2.8 million kronor.
3.2.8 Borrowing in foreign currency
The amortization of the foreign currency debt, begun in 1998,continued during 1999. The size of the amortization, as laid down by the government, was 25 billion kronor. The amortization in combination with smaller volumes of maturing loans and buy-backs of outstanding loans resulted in a decline in gross borrowing in foreign currency.
Like in previous years, borrowing in foreign currency was directed towards long-term borrowing in euro and swaps between kronor and foreign currency, so called kronor/foreign currency swaps. The amortization of the outstanding volume of commercial paper continued.
The break-down between the various kinds of borrowing is based on a balance between achieving low costs in the long and in the short run. Kronor/foreign currency swaps offer the lowest measured cost, but do not to the same extent provide the positive effects in terms of a broader base of international investors as does long-term borrowing in euro. Long-term borrowing in euro also contributes to an increased interest in Swedish government paper in kronor, which should contribute to lower borrowing costs on the domestic market.
Borrowing in foreign currency via the swap market
Borrowing via swaps between kronor and foreign currency, so called kronor/foreign currency swaps, constitutes a cost efficient substitute for traditional borrowing directly in foreign currency. This form of foreign currency borrowing can be seen as consisting of two parts. One part involves borrowing in kronor through regular bond issues, i.e. the Debt Office incurs a debt in kronor. The other part involves the kronor/foreign currency swap, concluded with
banks as counterparts. In the kronor/foreign currency swap, the Debt Office receives an asset in kronor – equally large as the debt in kronor incurred through the bond – and a debt in foreign currency. Thus, the debt in kronor has been transformed into a debt in foreign currency. The savings arise because the Debt Office receives a higher interest rate from the bank on the swap than the interest it pays on the bond. This difference is called swap spread.
Since swap transactions are conducted with banks they involve a credit risk. In order to handle this risk the Debt Office has continued to conclude so called CSA Agreements.. These
agreements mean that the Debt Office´s risk exposure to a counterpart is almost fully
eliminated (see also section 7.4: Credit risks in the derivatives and investment activity). The CSA agreements are therefore a prerequisite for the Debt Office to be able conduct, and increase the scope of, kronor/foreign currency swaps without the credit risk increasing at the same time.
Market developments in 1999
At the beginning of the year, the swap spread, i.e. the difference between the interest rates on government bonds and corresponding swap rates, declined from about 60 basis points to about 40 basis points. The combination of positive signals in both Sweden and England concerning EMU membership and a global reduction of swap spreads influenced the development in Sweden. During the summer, the swap spread widened and in the period July-September it reached levels in excess of 80 basis points. At most it stood at about 110 basis points. Many factors contributed to this development. An increased focus on computer problems in connection with the transition to the new year led to a reduced willingness to incur credit exposure and this tended to widen the swap spread.
Figure 3.7 Difference between Swedish 10-year swap interest rate and 10-year Treasury Bond interest rate
Basispoints
Swap interest minus bond interest 10 year in SEK
99 Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
20 30 40 50 60 70 80 90 100 110 120 130
During 1999, the Debt Office conducted kronor/foreign currency swaps to the tune of 40 billion kronor, an increase by 9 billion kronor compared with 1998. In 1999, kronor/foreign currency swaps accounted for about 65 percent of the borrowing in foreign currency. At the end of the year, the outstanding volume of swaps amounted to 112 billion kronor,
corresponding to about 26 percent of the debt in foreign currency (including the swap transactions). During the second quarter, the Debt Office chose to increase its borrowing via the swap market at the expense of other borrowing options since the swap spread remained at high levels.
Long-term foreign currency borrowing (borrowing in capital markets)
The Debt Office has continued to pursue the strategy of issuing loans in euro according to the concept of parallel loans. This means that identical loans are issued both in euro and kronor.
On the occasion of a possible Swedish membership in the currency union the two loans merge into one, denominated in euro. The reason for this strategy is that the Debt Office, in view of the increased competition between central government borrowers, needs to widen the investor base for Swedish Central Government Bonds and promote an internationalisation of the Swedish market.
At the beginning of the year, a seven-year loan in euro (1044E) was introduced. It was later augmented to a total outstanding amount of 2.5 billion euro. This loan supplements the existing loan in the ten-year segment (1043E). The purpose of the seven-year loan was to widen the investor base for Swedish Government Bonds further and to establish a point of reference in the euro market with a linkage to the Swedish market. The loan was placed mainly in Europe with a focus on German investors. The Debt Office wanted to reach out to investors which had not earlier bought euro loans to a great extent.
Figure 3.8 Interest rate spread for Swedish euro denominated Government Bond vs German Government Bond 1999, basispoints
0 5 10 15 20 25 30 35 40 45 50
Jan-99 Feb-99 Mar-99 Apr-99 May-99 Jun-99 Jul-99 Aug-99 Sep-99 Oct-99 Nov-99 Dec-99 Basispoints
1044E 1043E
Evaluation
The seven-year loan was issued at 12 basis points above German Central Government Bonds with a corresponding maturity. At the same time the ten-year loan was trading in the
secondary market at an interest spread of about 30 basis points above German bonds. The relatively large difference is explained partly by the limited supply in the seven-year segment, partly by the unique position held in the ten-year segment by German Central Government Bonds, resulting in generally larger interest rate spreads in relation to other borrowers in this maturity. The same result, i.e. that the seven-year loan is trading at a more competitive interest rate spread than the ten-year loan, also holds in a comparison with other government
borrowers. The seven-year loan is generally trading at a lower interest rate than loans issued by comparable european countries, while the reverse is true for the ten-year loan. The seven- year loan has, for example, been trading at an interest rate 0-5 basis points lower than the corresponding Finnish loan, at the same time as the ten-year loan has been trading 5-10 basis points higher.
Table 3.4 Interest rate spread vs German Bonds for som european countries for 7 year resp. 10 year maturities, basispoints December 1999
7 year 10 year
Sweden 13 29
France -19 9
Italy 29 25
Belgium 17 24
Finland 15 24
Portugal 5 32
Souce: Reuters
The analysis can be supplemented by the actual borrowing cost, which in addition is
influenced by the level of swap interest rates of corresponding maturity. A comparison of the actual borrowing cost for the two euro loans at their respective time of issue shows that the cost of the seven-year loan was about 5 basis points less than the cost of the ten-year loan. In sum, the more favourable development for the seven-year loan, in relation to other borrowers as well as in terms of actual borrowing cost, can be largely explained by the decision to issue at a maturity where competition from other borrowers was relatively limited at the same time as demand was strong.
Short-term foreign currency borrowing (commercial paper)
The decision to amortize 25 billion kronor of the foreign currency debt in combination with a concentration on borrowing in other markets led to a strong reduction of short-term
borrowing. At the turn of the year the reduction amounted to 26 billion kronor, which more or less eliminated the outstanding stock. The limited gross borrowing of 10 billion kronor in 1999 had mainly to do with margin payments on security for the CSA contracts that the Debt Office has with counterparts. In 1998, 51 billion kronor was issued.
Liquidity management in foreign currency
Since the project of introducing liquidity management in foreign currency was started in the spring of 1999, circumstances have changed. Borrowing in foreign currency declined and at the same time a shift towards kronor/foreign currency swaps instead of traditional foreign currency borrowing took place. The Debt Office judged that the gross borrowing need in foreign currency would be limited in the coming years as well, and no special resources were allocated to liquidity management. In order to quickly be able to manage foreign currency liquidity without special arrangements, it is planned that the use by the Debt Office of foreign currency accounts in dollars and euro in the Riksbank be expanded to comprise all borrowing and repayment of foreign currency loans, borrowing and repayment of kronor/foreign currency swaps, and large interest rate payments. The highest permissible balance on the accounts was set at the equivalent of 20 billion kronor on the euro account and 14 billion kronor on the dollar account, but with a combined upper limit of 20 billion kronor.
Result
The average borrowing cost declined in 1999. This is a consequence mainly of lower
borrowing costs and larger volumes of kronor/foreign currency swaps. The distribution of the borrowing by different types of borrowing and their respective cost are summarized in the tables below.
Table 3.5 Volume for various types of borrowing 1995-1999 in billion kronor
1995 1996 1997 1998 1999
Krona/currency swaps - 20,8 20,6 30,7 40,0
Public loans 98,4 57,0 34,1 34,2 22,3
Private placements 20,1 15,0 15,8 1,9 -
Commercial paper1 0,8 -3,5 -37,0 -44,0 -25,7
1 Borrowing via commercial paper show nett borrowing while other posts show gross.
Table 3.6 Costs for various types of borrowing 1995-1999 in basispoints under USD Libor
1995 1996 1997 1998 1999
Krona/currency swaps - -30,8 -31,8 -52,3 -60,0
Public loans -7,8 -14,7 -21,3 -6,7 -3,0
Private placements -6,0 -23,2 -23,5 -24,8 -
Commercial papers -12,5 -12,5 -19,0 -17,5 -18,8
Average costs 1 -7,5 -19,6 -24,7 -27,9 -39,6
1 Average costs exclusive commercial paper.
In 1999, the average borrowing cost for borrowing in kronor/foreign currency swaps was USD Libor minus 60 basis points. This is the lowest level recorded since this activity was started in 1996. Thus, the cost of borrowing in the swap market is about 55 basis points less than the cost of borrowing by means of the capital market. In present value terms this is equivalent to a cost saving of about 1200 million kronor compared with borrowing on the capital market.
During the last five-year period, the average cost of foreign borrowing has improved from USD Libor minus 7.5 basis points in 1995 to minus 39.6 basis points in 1999. The
improvement is due mainly to the increased use of kronor/foreign currency swaps. During the last year, the average borrowing cost has declined by about 12 basis points, which in present value terms is equivalent to an improvement of about 400 million kronor compared with a situation in which the borrowing cost had been the same as in 1998.
3.2.9 The management of the debt in foreign currency Results for fiscal years 1994/95 – 1999
The distribution between foreign currencies and the level of interest rate risk in the foreign currency debt is laid down by the Board of the National Debt Office. Since the Board has provided room for deviations in relation to the benchmark, albeit within well defined limits, there is scope for an active management of the debt. During the last five fiscal years, the active management of the debt in foreign currency has produced a positive result of 3.6 billion kronor, i.e. the interest cost of the central government has been reduced by this amount.
Positions based on expected developments in interest rates during the period 1994/95 – 1999 have contributed 3.7 billion kronor. With some simplification, a large part of the positive
result can be explained by the application of a consistent strategy during period 1994/95 – 1998.
The plans for the introduction of a common European currency (euro) at the turn of the year 1998/99 entailed that interest rates were expected to converge to the low German interest rate.
Thus, the strategy of the Debt Office was based on the assumption of a fall in interest rates in the so called high-interest rate countries Italy, Spain, and also Great Britain. The fall in interest rates in these countries was heavy and this gave positive results.
In total, positions based on expectations of the exchange rate development during the period 1994/95 – 1999 have produced only a minor effect on the result.
Since the 1992/93 fiscal year, the Debt Office has made use of external portfolio managers.
These manage a smaller part – less than 12 billion kronor in 2000 – of the foreign currency debt on the same terms as the Office. Thus, the results of the Debt Office can be compared with that of the managers. Viewed over the period of the most recent five budget years, the National Debt Office has turned in a better result than the average results of the managers.
The result in relation to the extent of risk-taking can be analyzed by comparing the information ratio. A high information ratio indicates a good result in relation to the risk- taking, i.e. the higher the value the better. During the five-year period 1994/95 – 1999, the Debt Office has attained a higher information ratio than the average of the external managers.
This means that the result of the Debt Office has not been achieved as a consequence of a higher risk taking than that of the external managers.
