United Kingdom
Financial Sector Assessment Program Technical Notes and Detailed Standards Assessments

This technical note reviews the Financial Sector Assessment Program of the United Kingdom. It examines the United Kingdom’s public debt management practices using the IMF-World Bank Guidelines for Public Debt Management as a framework. It analyzes the government’s Code for Fiscal Stability, transparency, accountability, debt strategy, and risk management framework. It also provides a detailed assessment of the antimoney laundering and combating the financing of terrorism regime and compliance of the Basel Core Principles of the United Kingdom.

Abstract

This technical note reviews the Financial Sector Assessment Program of the United Kingdom. It examines the United Kingdom’s public debt management practices using the IMF-World Bank Guidelines for Public Debt Management as a framework. It analyzes the government’s Code for Fiscal Stability, transparency, accountability, debt strategy, and risk management framework. It also provides a detailed assessment of the antimoney laundering and combating the financing of terrorism regime and compliance of the Basel Core Principles of the United Kingdom.

I. Introduction1

1. The amount of public sector debt outstanding in the U.K. is low. Public sector net debt totaled £303 billion in nominal terms or 30 percent of GDP as of February 2002 (Table 1), the lowest ratio since 1992, and also the lowest among all the G-7 countries (Table 2). Total public sector gross debt (i.e., public sector net debt before short-term financial assets are deducted) consists almost entirely of central government gross debt, since the vast majority of local government and public corporations’ gross debt is borrowed from the central government and is thus netted out when calculating the consolidated figure. In addition, although more than £4 billion of local government debt is not held by the central government, this is offset in the public sector net debt figures by a similar amount of central government debt held by public corporations, such as the Post Office.

Table 1.

Public Sector Net Debt

(£ billions, percentages in italics)

article image
Source: Office for National Statistics.

2002 data are as of end-February.

Table 2.

General Government Net Financial Liabilities for G-7 Countries

(As a percent of GDP)

article image
Source: OECD Economic Outlook No. 70, Government Debt Statistics: Fiscal Balances & Public Indebtedness

2. Gilts (government bonds) are the main component of the stock of central government debt (Table 3). Most consist of conventional instruments, although there is a significant stock of inflation-indexed (index-linked) bonds outstanding. National Savings instruments issued to U.K. households represent the largest non-gilt segment. There is also a small amount of Treasury bills outstanding, which are issued for government cash management purposes. Foreign currency debt is very small, and is hedged by matching foreign currency reserves.

Table 3.

Distribution of Nominal Central Government Gross Debt by Instrument

(£ billions, percentages in italics)

article image

The nominal value of index-linked gilts has been raised by the amount of accrued capital uplift.

Source: Office for National Statistics.

3. The stock of gilts is mainly held by institutional investors. U.K. insurance companies and pension funds hold most of the gilts outstanding (Table 4). The rest is mainly held by other domestic financial institutions and nonresidents.

Table 4.

Distribution of Gilt Holdings as of December 2001 (market values)

(£ billions, percentages in italics)

article image
Source: Office for National Statistics

Net of central government holdings.

4. The U.K. government follows a sustainable investment rule, which states that public sector net debt as a proportion of GDP will be held at a stable and prudent level over the economic cycle. Other things being equal, policy is for net debt to be maintained below 40 percent of GDP over the economic cycle. The government also has a second fiscal rule known as the golden rule, which states that, over the economic cycle, the government will borrow only to invest, and not to fund current spending. The low level of government indebtedness in recent years has enabled the government to respect these rules with comfortable margins to spare.

5. The rest of this chapter examines the UK’s public debt management practices using the IMF-World Bank Guidelines for Public Debt Management as a framework. The focus is on the management of the central government’s domestic wholesale market obligations by the Debt Management Office (DMO), since these account for most of the debt outstanding and are the most relevant from the standpoint of the stability of the U.K. financial system. In light of the low level of borrowing by the U.K. government in recent years, special attention is placed on the measures taken by the authorities to maintain a well-functioning gilt market. The coordination of debt and cash management with monetary policy is also examined given the transfer of sterling cash management from the Bank of England to the DMO in 2000. Section A examines the objectives of debt management; transparency issues are discussed in Section B; and the institutional framework is outlined in Section C. The debt strategy followed by U.K. authorities is reviewed in Section D; the risk management framework in Section E; and the functioning of the gilt market in Section F. An overall assessment of U.K. public debt management practices is provided in Section G.

II. Objectives and Scope of Debt Management

6. The current objective for debt management was set in 1995. It is as follows:

  • To manage over the long-term the cost of meeting the Government’s financing needs, taking into account risk, whilst ensuring that debt management is consistent with the objectives of monetary policy.

7. This objective is to be achieved through the following strategies:

  • Pursuing an issuance policy that is open, predictable, and transparent;

  • Issuing conventional gilts that achieve a benchmark premium;

  • Adjusting the maturity and nature of the government debt portfolio, by means of the maturity and composition of debt issuance and other market operations, including switch auctions, conversion offers, and buy-backs;

  • Developing a liquid and efficient gilts market; and

  • Offering cost-effective retail savings instruments.

8. The objective explicitly takes account of risk and focuses on managing debt service costs over the long-term. The authorities seek to ensure that expected debt service costs are robust against a variety of economic outturns. The main way of doing this is by considering the effect of issuance on the ensuing government debt portfolio. They do not believe that they are able to predict which security will prove to be cheaper than any other, since they do not believe that they are any better informed than the market on key macroeconomic variables. Instead, they prefer to select a debt portfolio that protects the government from as wide a range of shocks as possible.

9. In terms of the operational delivery of the objective, the 1995 review of debt management heralded a move away from a highly discretionary debt management policy. It rejected the thesis that discretion benefited the government in that it could sell appropriate amounts of debt at advantageous prices. It concluded that under such arrangements the government would pay an unnecessary premium, since it would be systematically attempting to beat the market and there would be no certainty or transparency in the path of issuance policy. As a result, the government has shifted in favor of a policy that promotes a more efficient, liquid and informationally more complete market. This includes adhering to a policy of annual published remits, which set out in advance issuance in terms of type and maturity of gilt, a pre-announced auction calendar, and a move from gilt sales by tap to an auction-based issuance system.

Coordinating debt management with fiscal and monetary policy

10. Responsibility for debt and cash management has been formally separated from monetary policy, and there are mechanisms in place to ensure that the conduct of these policies is appropriately coordinated. The DMO officially became operational as an executive agency of HM Treasury in April 1998 and took over responsibility for debt management from the Bank of England as of that date. Full responsibility for cash management was assumed in April 2000. Prior to April 1998, the Bank of England acted as the government’s agent in debt and cash markets. The transfer of debt management responsibilities to HM Treasury helped to mitigate any perception that the Government’s debt and cash operations benefited from inside knowledge over the future path of interest rates. It has also helped to avoid a potential conflict of interest, or perception of a conflict, between the objectives guiding debt management and monetary policy operations. And, it appears to have aligned incentives within HM Treasury to improve cash forecasting, as evidenced by the significant improvement in mean absolute forecasting errors documented in the chapter on systemic liquidity arrangements.

11. The separation of responsibilities allows for the setting of clear and separate objectives for monetary policy and debt management with benefits in terms of reduced market uncertainty and hence lower interest rates. The Bank of England’s Monetary Policy Committee (MPC) is able to raise any issues about the implications of debt management for monetary policy with HM Treasury’s representative at MPC meetings. However, in order to avoid creating any external perceptions that the DMO is privy to monetary deliberations, it has no contact with the MPC (including the non-voting HM Treasury representative) with regard to interest rate decisions or its thinking.

12. The debt management objective still has a reference to monetary policy. It ensures that the Bank of England’s monetary policy responsibilities will not be undermined by the DMO or by the Treasury (e.g., “printing money” to meet the government’s cash requirements, or DMO cash management operations interfering with the Bank’s monetary policy operations).

13. At the operational level, the framework for cash management developed by HM Treasury and the DMO in discussion with the Bank of England is designed to ensure that DMO operations do not cut across the Bank’s monetary policy operations. The DMO avoids holding tenders at times when the Bank is conducting money market operations, and does not hold reverse repo tenders at the 14-day maturity range. It also does not conduct ad hoc tenders on MPC decision days. These restrictions do not apply to bilateral operations conducted by the DMO owing to their relatively low market profile compared to auctions. However, the DMO acts as a price-taker in its market operations.

14. The DMO may also issue Treasury bills to assist the Bank of England in its management of the sterling money market. If so requested by the Bank, the DMO would add an additional amount of bills to a tender and deposit the proceeds at the Bank at an interest rate equal to the average yield at the tender. If the DMO were to do this, it would identify, in its usual announcements, any amounts being issued for the Bank. This facility has not been used.

15. On occasion, sizeable unanticipated cash inflows and outflows may occur too late in the day for their impact to be smoothed by the DMO in the money market. To take account of this, arrangements have been put in place with the Bank of England and settlement banks to cope with late changes to the forecast without disadvantage to the market. In circumstances where there is an Exchequer cash surplus, the surplus is taken into the Bank of England’s 4:20 p.m. overnight repo facility and is allocated at a non-penal interest rate (i.e., the additional refinancing is provided by the Bank to the settlement banks at its two-week repo rate). An Exchequer cash deficit can also be offset by bilateral borrowing from a number of settlement banks through a special end-of-day transfer arrangement.

16. The DMO targets a balance of £200 million at the Bank of England in order to manage the rare circumstances where unanticipated changes to the forecast are not fully accommodated in the usual end of day arrangements. The current structure replaces the previous arrangements through which late changes were absorbed by fluctuations in the size of the government’s Ways and Means advance from the Bank of England. Since the transfer of cash management to the DMO, the £200 million target has, as expected, covered all but two daily cash shortfalls. The two exceptions both occurred in the 2000–01 fiscal year, and only required modest (around £20 million) fluctuations in the size of the Ways and Means advance (which normally stands at £13.4 billion). Both were reversed the following business day.

17. Debt management is coordinated with fiscal policy in that the annual borrowing program is set by HM Treasury on the basis of the fiscal plans contained in the government’s budget. The borrowing program is updated in line with the update to the government’s cash forecast, issued at the time of the Pre-Budget Report.

18. The DMO operates as much as possible on a commercial basis and distances itself from a role in supervising financial markets. For example, it sets its own terms and conditions when managing the risks associated with its operations, such as rules governing eligible collateral in its market operations.2 However, it tries to promote well-functioning markets by operating as a “good market citizen”—e.g., by setting a good example in the conduct of its market operations and responding to requests for information from the Bank of England and the FSA on market developments. It also plays an active role in the market committees that discuss the functioning of the markets in which it operates, and works closely with market participants to introduce new innovations such as electronic trading systems to the interdealer gilt market. It is also prepared to serve as an instrument of the Treasury should there be a need for public intervention to support the markets, provided that such an operation would be conducted in a transparent fashion and in a way that clearly separates it from the DMO’s normal market operations.

III. Transparency and Accountability

19. Public debt management in the U.K. is conducted in an open and transparent fashion in accordance with the government’s Code for Fiscal Stability. Indeed, market participants effusively praised the authorities for the major improvements in transparency that have been introduced since 1997. The DMO and HM Treasury maintain websites that contain a comprehensive set of information and publications that provide data on, and explanations of: the objectives guiding debt and cash management activities; the debt/cash management governance and institutional framework; the government’s finances; and debt/cash management operations. Under statute, the government must publish details of its borrowing activities during the fiscal year, and its plans for the year ahead before the end of the fiscal year. These are contained in the Debt and Reserves Management Report (DRMR) published by the Treasury. The DRMR also explains the debt and foreign exchange reserves management operations of the fiscal year just ended, and for the coming fiscal year, details the financing requirement, the forecast sales of gilts, their breakdown by maturity and instrument type, and the gilt auction calendar, along with planned short-term debt sales including Treasury bills.

20. An auction calendar is issued at the end of each quarter by the DMO, which confirms auction dates for the coming quarter and states which gilts are to be issued on which date. Normally eight calendar days before an auction, the amount of stock to be auctioned is announced (and if it is a new stock, the coupon). At this point, the stock is listed on the London Stock Exchange and ‘when-issued’ trading commences. A quarterly auction calendar is also issued for Treasury bill tenders.

21. Market-makers and end-investor groups are consulted regularly through meetings with DMO staff during the formulation of these plans (and also quarterly before the DMO announces specific auction stocks for the quarter ahead). They are also regularly consulted ahead of time whenever the authorities are considering changes to debt management practices or initiatives to enhance the functioning of the market. Minutes of the consultation meetings are published within a few hours of the meetings.

22. Gilts are now issued by auction. The DMO retains the ability to buy-back or issue gilts in smaller quantities (by tap) and conduct special repos at short notice for market management reasons—i.e., to ensure that the functioning of the market is not severely impaired by technical pricing distortions associated with excess demand for a specific security and the like. Operational transparency is enhanced through regular close consultation with market participants.3

23. The National Audit Office (NAO) annually audits the government’s Debt Management Account (DMA), the account through which all financial transactions entered into by the DMO in pursuit of its debt and cash management objectives pass, and the National Loans Fund (NLF), through which all gilt issuance is recorded. It also audits the administrative (or agency) accounts of the DMO. The results of these audits are published each year. In addition, the DMO’s annual report reports its performance in meeting the objectives set forth in its business plan. This report and the DMO’s annual business plan are published each year on the DMO’s website. Like all U.K. government executive agencies, the DMO’s activities are to be regularly reviewed by the parent department, the results of which are reported to Parliament.

24. One area where there continues to be some opaqueness in the process is the lack of information on the considerations that lie behind the cost-risk trade-offs implicit in the annual remit provided to the DMO. The authorities expect to publish work shortly that examines the linkages between fiscal policy and the debt portfolio. It includes the development of a comprehensive asset and liability risk monitor to aid the quantification of risks faced by the central government in its balance sheet. A preliminary version of the risk monitor was published in the 2002–03 DRMR as a precursor to the publication of the Whole of Government Accounts in 2005-06.

IV. Institutional Framework

25. The institutional framework for debt management is well specified with a clear articulation of roles and responsibilities. Government debt management is principally handled by two executive agencies. The DMO manage the government’s domestic wholesale market borrowings and its domestic cash balances. National Savings and Investments (NS&I) oversees the government’s savings instrument program, which raises funds directly from U.K. households. Foreign currency debt management and the management of the government’s foreign currency reserves are still conducted by the Bank of England as agent for HM Treasury under a formal remit between the two institutions.4 The annual remits provided to the DMO and NS&I and the MOU between the Bank and HM Treasury are published in the DRMR.

26. As with all U.K. government executive agencies, the DMO’s relationship with HM Treasury is outlined in a framework document that sets out the roles and responsibilities for each of the parties involved in debt management.5 The basic structure for debt management is that Treasury Ministers, advised by officials in HM Treasury’s Debt and Reserves Management Team, set the policy framework within which the DMO takes operational decisions within the terms of the annual remit provided to it by Treasury Ministers. The DMO’s business objectives include a requirement for the DMO to advise the Treasury about the appropriate policy framework, but strategic decisions rest with Ministers. The Bank of England acts as the DMO’s agent for gilt settlement and retains responsibility for gilt registration.

Legal framework for borrowing

27. The legal framework underpinning debt management provides HM Treasury with wide discretion as to how to raise money by borrowing. It does so through two statutory funds, the National Loans Fund, and its sub-account for wholesale market borrowings, the Debt Management Account. Its main power to borrow for the National Loans Fund is conferred by section 12 of the National Loans Act 1968. This provides that the Treasury can raise any money that it considers expedient to raise for the purpose of promoting sound monetary conditions in the U.K., and this money may be raised in such manner and on such terms and conditions as the Treasury think fit. Section 12(3) of the same Act makes it clear that the Treasury’s power to raise money extends to raising money either within or outside the U.K., and in other currencies. There are no set limits on the extent to which the Treasury may borrow from outside the United Kingdom.

28. The Treasury’s power to borrow for the Debt Management Account is conferred by paragraph 4 of Schedule 5A of the National Loans Act 1968, and this paragraph, like section 12 of the Act, gives the Treasury a wide discretion as to how to raise money. Paragraph 4(3) is similar in terms to section 12(3) of the Act, and it provides that the Treasury’s power to raise money under paragraph 4 extends to raising money either within or outside the United Kingdom, and in other currencies. Again, there is nothing in Schedule 5 of the Act to limit the amount of money the Treasury may borrow from outside the United Kingdom. In practice, Treasury borrowing takes a wide range of forms and ranges from the issuing of long-term securities (gilts) to the issuing of short-term Treasury Bills (12 months maximum) under the Treasury Bills Act 1877.

Organizational structure within the DMO

29. The DMO’s own organizational framework is clear and well-specified. The Chancellor of the Exchequer, with advice from Treasury officials, determines the policy and financial framework within which the DMO operates, and delegates to its Chief Executive operational decisions on debt and cash management and day-to-day administrative issues. The Chief Executive is appointed by HM Treasury, and variously reports to the Permanent Secretary (on expenditure and related issues), to Treasury Ministers (on policy issues), and to Parliament (in the formal presentation of accounts). In particular, the Chief Executive is responsible to Treasury Ministers for the overall operation of the DMO, and for delivering the remit in a way that he/she judges will involve the least long run cost to the Exchequer, subject to being compatible with other policy considerations.

30. The DMO employs approximately 80 people organized around ten business units, and has a structure of corporate governance in place to assist the Chief Executive in carrying out his responsibilities. There is an informal high-level Advisory Board, advising the Managing Committee, which is the senior decision making body for the office.6 The Managing Committee is in turn supported by a Credit and Risk Committee and strategy groups for each key business area (Debt, Cash, Investment). There are currently two external non-Executive Directors on the Advisory Board, both of whom are also on the office’s Audit Committee, together with a member of HM Treasury. And, there is an appropriate separation of front, middle, and back office activities. If the Advisory Board is expected to become a permanent fixture in the DMO, the authorities may wish to consider formalizing it in the DMO’s governance framework, and introduce some selection criteria to guide its membership.

31. The DMO is financed as part of HM Treasury and operates under arrangements that control its administrative expenses. It is subject to an internal audit function that reviews the systems of internal control, including financial controls, and to external audit by the National Audit Office. The Chief Executive is the Accounting Officer for both the office’s administrative accounts, and the accounts of the Debt Management Account (DMA) through which all its market transactions pass.

32. Appropriate human resource policies and systems are in place. The Chief Executive of the DMO is responsible for setting the DMO’s personnel policies and managing staff. The Office has delegated authority for pay, pay bargaining, training and setting terms and conditions in order to recruit, retain and motivate staff. Nonetheless, personnel policies must be consistent with wider public sector pay policy and the Civil Service Management Code. In practice, this means that while the DMO has some discretion with respect to the setting of staff compensation levels, the overall compensation policy must be approved by the government. The DMO achieved Investors in People accreditation in June 2000—an internationally-recognized ISO accreditation that recognizes the DMO’s commitment to staff training and development, and its efforts to foster a “learning” environment in the organization. DMO staff are bound by conflict-of-interest guidelines and the code-of-conduct applicable to the U.K. civil service.

33. Policies and procedures have been adopted to control operational risk. The Statement of Internal Controls (SIC) in the DMO’s Annual Report & Accounts (ARA) 2000–2001 describes the DMO’s approach to managing its operational risk. The adequacy of the DMO’s management of risk and internal controls is regularly reviewed by the DMO’s Audit Committee, which is chaired by an external non-executive director. In addition, the DMO has established a Business Continuity Plan. This Plan includes a capability that would enable it to conduct its business at its Disaster Recovery Site in the event of its main offices becoming unavailable (see the SIC).7 In the event that the DMO had to invoke this plan, it would post a notice on its web site notifying market counterparties of the event and of further information relevant to the continued operation of its business. A business continuity plan is also being developed in conjunction with key market participants under the stewardship of the Bank of England to mitigate the effect of a severe dislocation to the infrastructure supporting the U.K. markets.

34. The lack of an automated bidding process for Treasury bill and gilt auctions is one area where the DMO is currently exposed to more operational risk than debt managers in some other industrial countries. Bids are submitted by auction participants to the DMO by telephone, and DMO staff call the bidders back to confirm the details of each bid before the auction is processed. While a telephone system is very resilient, such an approach can potentially result in human errors, and may also slow the processing of the auction. As a result, government debt managers in several industrial countries have automated the bid-capture process using Bloomberg or customized systems, and have cut the amount of time spent processing the auction to 15 minutes, compared 30 minutes for gilt auctions. The DMO would also like to automate its bid-capture process, and expects to consult the market later this fiscal year. One hurdle to overcome in developing such a system will be to ensure that it can handle bidding limits for dealers and end-investors; however, some countries have been able to develop systems with such a feature. Inasmuch as this would represent a transfer of operational risk from the DMO to the market, such a proposal may not necessarily be enthusiastically endorsed by the DMO’s counterparties. Nonetheless, the DMO is encouraged to follow through with its desire to automate the bid-capture process for gilt and Treasury bill auctions (although the latter should wait until these instruments are dematerialized). Once these systems are installed, the telephone bidding system could be used as a fall-back, for use in cases where the electronic systems are not available for some reason.

V. Debt Strategy

35. HM Treasury, with input from the DMO, determines the desired structure of new issuance for the year ahead taking into account the financing requirement and the costs and risks associated with various instruments. This remit is published in the annual DRMR, and is expressed in terms of the percentage issuance across each class of gilt and overall financing to be raised through the issuance of Treasury bills.8 After consulting HM Treasury and market participants, the DMO takes further tactical decisions about the features of new debt issues, and the precise timing of new borrowings during the year in line with the overall target. Significant changes to projected borrowing requirements may result in revisions to the remit. The Chancellor’s Pre-Budget Report (typically published in October or November) provides an opportunity to revise the remit, if necessary, in light of changes to HM Treasury’s economic outlook.

36. The risks associated with the government’s debt are prudently managed. The government’s policy has been to issue debt across a variety of instruments with a definite preference in favor of longer-term securities. At 7.5 years (at end-March 2002) the average duration of the gilt stock is longer than that of most OECD governments. This partly reflects a desire to minimize rollover risks, which in the past were significant when indebtedness levels were high. In addition, by issuing longer-term conventional debt, the government has been able to profit from the high institutional demand for long maturity debt from U.K. insurance companies and pension funds, which contributed to an inverted yield curve in recent years (see Section F). Along with a relatively smooth redemption profile, this helps to provide additional certainty to projections of future nominal debt servicing costs. Long duration debt also helps to limit the effect of any supply-side shock on the government’s fiscal position.

37. If the yield curve distortions should recede at some point in the future, the authorities may wish to consider shortening the average duration of the gilt portfolio. In an upwardly sloping yield curve environment, this would result in debt service cost savings at the expense of some added risk to the debt portfolio—risk that the U.K. government should be well-placed to tolerate given its low level of indebtedness. To be fair, the authorities have already begun taking steps along this path. The 2002–03 remit provides for a more balanced distribution of gilt sales across the short, medium, and long tenors compared to the one in 2001–02, which called for the majority of gilt issuance to take place in the long (15+ year) tenor.

38. Almost a quarter of the government’s marketable debt portfolio consists of index-linked gilts and Treasury bills, and the debt issued by NS&I is also short-duration in nature. In the event of a demand shock this should allow the changes in the debt servicing cost relating to these parts of the national debt to mitigate the resulting effect on the government’s fiscal balance. Index-linked securities also provide protection against a “nominal” shock. The expected cost advantages of issuing these securities have declined over time as the spread between nominal and real yields has settled just above the 2½ percent inflation target for monetary policy. The authorities noted that there remains strong demand in the market for them with limited substitutes available from private sector issuers.

39. U.K. governments have not used foreign currency debt to finance domestic borrowing requirements in peacetime, reflecting the belief that foreign currency risk to the government’s balance sheet is neither desirable nor cost-effective. Instead, the issuance of such debt in recent years has been to augment the foreign currency reserves. Issuing liabilities in the currency in which the government wishes to hold foreign currency assets helps to hedge its foreign currency reserves against unexpected exchange rate fluctuations.

40. The development of an active currency swap market has meant that the currency debt is issued in, and the currency in which assets are held, do not necessarily have to be the same. Consequently, the authorities have recently issued additional amounts of domestic currency debt, and swapped the proceeds into another currency, in order to augment the government’s foreign currency reserves and maintain issuance volumes in the domestic market. Value for money is the primary concern when deciding whether to fund the foreign currency reserves from debt issued in sterling swapped into foreign currency, or from the issuance of foreign currency denominated debt, with the comparison being made on a swapped basis. Stringent controls are in place to limit the amount of basis risk and credit risk associated with these swap transactions. Approximately £10 billion of such swaps were outstanding at the end of March 2002.

Cash management

41. The DMO’s management of the government’s cash balances is highly respected by the market. The DMO manages the government’s domestic cash balances primarily through a combination of Treasury bill tenders conducted on a competitive bidding basis and repo or reverse repo transactions. Treasury bill tenders tend to be used to manage longer-term swings in cash flows within the fiscal year, while repo operations with market participants are used to manage shorter-term swings of less than one month. Market participants praised the DMO for its professionalism in executing its cash management transactions. They especially applauded its smooth investment of the much larger than expected receipts from the third generation mobile phone license auction (3G) in fiscal year 2000–01.

42. Although the cash management process is working well with much smaller forecast errors than in past decades, U.K. authorities conceded that the quality of information received from government departments on the magnitude and timing of cash flows could be improved, and that the timing of these flows could be better synchronized so as to minimize the volatility of cash balances. HM Treasury’s Exchequer Funds & Accounts team coordinates the cash forecasting process across government departments and agencies, and a system of financial rewards and penalties has been established to encourage good forecasting behavior at the department/agency level. Although this system is helping to encourage departments and agencies to provide better cash forecasts at the margin, the authorities may wish to consider larger rewards/penalties in order to motivate better performance. They could also consider providing financial managers in government departments/agencies with more information on the timing and magnitude of aggregate government revenues and expenditures, and introduce a system of rewards/penalties to encourage them to try and achieve a better synchronization of the timing of their expenditures with government cash receipts.

VI. Risk Management Framework

43. Work is underway on managing risk in the debt portfolio by assessing the resilience of cost and tax smoothing properties for different debt structures to a range of economic conditions and shocks. HM Treasury and the DMO will consider the “cost-at-risk” and robustness of different issuance strategies to different possible economic outcomes generated using Monte Carlo simulations within a reduced form macroeconomic model. These results will be further subjected to a range of stress tests to allow for the possibility of particular shocks, for example supply shocks.

44. This should help to quantify an optimal debt portfolio against which an issuance strategy and long-term performance could be assessed, and thus could facilitate the identification of a counterfactual portfolio. The two institutions believe that this counterfactual portfolio will provide a useful analytical tool when setting the terms of the DMO’s annual financing remit. This counterfactual would help articulate, in the form of some simple indicators, the government’s preferences for the appropriate structure of the debt. This might help the public understand better the government’s cost-risk preferences, and would help separate accountability for the choice of the optimal debt portfolio (a policy decision) from accountability for the realization of the remit (DMO’s tactical mandate).9 It would also help address comments made by the Treasury Select Committee, which noted that portfolio benchmarks could “help produce a clear published assessment of the costs and risks faced by the DMO.” 10

45. However, HM Treasury and the DMO would have significant concerns if this approach led to an explicit performance benchmark for the DMO. The absence of such a target is a deliberate choice that has been made for a number of reasons. They believe that such an approach would encourage short-term opportunistic behavior (or the perceptions of such behavior) on the part of debt managers as they strive to meet the performance benchmark to the detriment of the long-run cost minimization objective. First, there is no expectation that the DMO can predict movements in the gilt yield curve more accurately that the market as a whole, other than by exploiting inside information about future gilt supply or other relevant policy announcements. As a result, such a target might lead to a market perception that HM Treasury and the DMO would seek to outperform the benchmark target, either by taking views on the future course of interest rates or by having access to privileged information, to the detriment of the separation of debt and cash management from monetary policy. Secondly, in an environment where the government debt manager is the dominant issuer in the domestic currency, it is not possible to derive an objective benchmark that would not be affected by the very actions of the debt manager it is designed to assess. Third, a benchmark approach would curtail the ability of HM Treasury and the DMO to pursue wider debt management objectives, for example to maintain liquidity or build benchmarks at particular tenors. The DMO believes that this conflict could not simply be resolved through the use of interest rate swaps, since the swap counterparties would purchase gilts at the tenor the government was swapping out of to hedge their exposures. This would negate the effect of the initial strategic motivated issuance.

46. Introducing some indicators that simply articulate the cost-risk preferences underpinning the remit would not necessarily induce opportunistic behavior on the part of the DMO. The latter could still be governed by the structure of the current remit, which sets strict guidelines for the amount and tenor of new borrowings. There would not be any changes to the incentives it currently faces. The risk of opportunistic behavior would only arise if the DMO was given a more flexible remit that gave it additional discretion to set issuance strategy or use derivatives to actively manage the risk profile of the debt stock in pursuit of an optimal debt structure. Several countries in the European Monetary Union have given their public debt managers such discretion, which is understandable as these countries are no longer sovereign borrowers and compete against one another in the deep Euro-area bond market. However, there is no need to change the structure of the DMO’s remit at this time.

Scope for active management

47. Since almost all government borrowing is conducted in domestic markets, U.K. authorities do not seek to actively manage the debt portfolio to profit from movements in interest rates and exchange rates. The authorities believe that this would risk financial loss, as well as potentially send adverse signals to the markets and conflict with monetary and fiscal policies. It could also add to market uncertainty, which in turn could be passed on to the government in the form of higher borrowing costs.

Contingent liabilities

48. The government’s contingent liabilities are managed separately from the debt program. Generally, the government tries to avoid taking on such liabilities, but the decision on whether to finance commitments through debt or through guarantees is one that is made at the political level after taking account of information provided by HM Treasury and the DMO on the opportunity cost of financing the commitments in the market. Details of the government’s contingent liabilities are published each year in the Supplementary Statements accompanying the publication of the Consolidated Fund and National Loan Fund accounts. A list of contingent liabilities and their maximum potential loss is produced. (They mainly consist of the cost of nuclear waste cleanup, export credits, and guarantees provided for the construction of the Channel Tunnel.) Additional work is planned to quantify and assess the risks of these liabilities so that their estimated values can be calculated within a probabilistic framework.

VII. Developing and Maintaining an Efficient Market for Government Securities

49. The gilt market is the smallest segment of the sterling capital market, although the gilt repo market is becoming an increasingly important component of the sterling money market. Table 5 compares the size of the gilt market with the sterling money market and markets for other fixed income securities and equities. The relatively small size of the gilt market reflects the government’s modest borrowing requirements in recent years. Although the gilt market is small, these securities are actively traded in the secondary market, while other sterling-denominated bonds are not (Table 6). Gilt trading volumes have held fairly steady despite the shrinking size of the market in recent years. A 1997 BIS study suggests that the secondary market in gilts is less active than comparable markets in France and North America, but compares favorably with many government bond markets in Continental Europe and Japan (Table 7).

Table 5.

Sterling Capital Markets Amounts outstanding

(£ billions, end of period)

article image
Source: Bank of England.

Defined here as amounts outstanding in the interbank, certificate of deposit, gilt repo and stock lending, bill, and commercial paper market.

Measured as market capitalization of the FTSE All-Share index; 1990 data are estimated.

Single currency interest rate swaps, notional principal outstanding. 1990 data are not available so the table uses 1992 data; 2000 data are end-June.

Table 6.

Secondary Market Turnover in Selected Sterling Markets

Average daily amount (£ billions)

article image
Source: London Stock Exchange and London International Futures and Options Exchange

Foreign firms listed for trading on the London Stock Exchange.

Table 7.

1997 Cross-Country Comparison of Government Bond Market Trading Volumes

article image
Source: Bank for International Settlements, Committee on the Global Financial System. 1999. “Market Liquidity: Research Findings and Selected Policy Implications” BIS-CGFS Study No. 11. May.

The range of interdealer bid-offer spreads for benchmark securities, expressed in 1/100 of a currency unit for the face value of 100 currency units. Spreads tend to be narrower for shorter-term securities, and wider for longer-term ones.

Data may include trading other than outright transactions, such as repos or buy/sellbacks.

Annual trading volume divided by the stock of debt outstanding.

50. Most of the trading in gilts consists of over-the-counter trading between market makers and their customers. Market participants suggested that the gilt interdealer market is less active than comparable markets in Europe and North America. Institutional clients expect to be able to transact £25 to £100 million with tight spreads, while trades of £5 million are more common in the interdealer market. Market participants reported that the market is dominated by about 25 insurance and pension fund clients that deal primarily through the five most active gilt-edged market makers (GEMMs). The GEMMs prefer whenever possible to intermediate transactions through their client base, rather than trade with each other, in order to retain as much of the bid-offer spread as possible. The limited number of large investors that are regularly present in the market has helped to limit search costs and increase the viability of such a strategy. Institutional clients are reportedly becoming more adept at trading off immediacy versus price when negotiating trades with the GEMMs; in the past, they tended to insist on immediacy at the expense of price. GEMMs have reportedly become more cautious since 1997 about quoting tight spreads for large orders as their risk management practices tightened up.

Structure of the Gilt Market

Gilt-edged Market Makers (GEMMs)

51. The U.K. Government bond market operates as a primary dealer system. There are 16 firms recognized as primary dealers (GEMMs) by the DMO. Each GEMM is authorized by the Financial Services Authority (FSA), and must be a member of the London Stock Exchange so that its trading practices are subject to supervision by the Exchange. The GEMMs undertake a number of market-making obligations, in return for certain benefits.

52. The obligations of a GEMM are to participate actively in the DMO’s gilt issuance program; to make effective two-way prices on demand in almost all gilts (those issues that have very small amounts outstanding are excluded); and to provide information to the DMO on market conditions. 10 of the 16 recognized GEMMs are also recognized as Index-linked Gilt-edged Market Makers (IG GEMMs), and their market-making obligations extend to cover inflation-indexed gilts.

53. Beginning in 2002, GEMMs are also required to provide firm two-way quotes to other GEMMs in a small set of benchmark gilts.11 The purpose of this new obligation is to ensure there is a minimum amount of liquidity in the intra-GEMM market for the benefit of the entire secondary market for gilts. These quotes are to be made for at least 5 hours during the eight hour London trading day on any of the three recognized interdealer broker (IDB) screens. They ensure that at least £28 million of liquidity is available across the yield curve to any GEMM needing to lay-off an undesired position. Although non-GEMM market participants do not have access to these screens, the transparency of the market is aided by the prices that GEMMs voluntarily make available through various wire services, and by the fact that the DMO publishes a real-time indicative mid-prices for a selection of benchmark gilts, drawing on GEMM prices as published on the wire services. At the retail level, the transparency of the market is enhanced by the fact that all trades of less than £50 thousand (par-value) are published on a real-time basis by the London Stock Exchange on its ticker tape.

54. Although the GEMMs were consulted by the DMO on several occasions prior to the introduction of the mandatory price quoting rules, some of them now regret their introduction. They view the mandatory quoting rules as an annoyance that causes them, at the margin, to pay more attention to servicing the IDB market than they would otherwise prefer. While recognizing that these rules should help to make the IDB market more accessible to potential new market entrants, which could add liquidity to the overall market, they fear that the rules may simply serve to generate artificial trading between themselves, to the distraction of servicing their client base. Smaller GEMMs are reportedly the most active users of the IDB market, since the larger ones are better able to intermediate through their own customer base. On balance, the new rules should prove to be a useful step in developing a deep and liquid interdealer market—a market that could become more important if the underlying investor base broadens out over time and search costs for GEMMs increase accordingly. Although the maximum bid-offer spreads are wider than the typical spreads seen in this market, there is no need to narrow them in. They currently serve a useful backstop function, and the future growth of the IDB market would be best served by leaving it to the participants to determine appropriate spread margins.

55. GEMMs receive a number of benefits in return for their obligations. These include: exclusive rights to competitive telephone bidding at gilt auctions and taps, either for the GEMM’s own account or on behalf of clients;12 exclusive access to a non-competitive bidding facility at outright auctions; the exclusive facility to trade or switch stocks from the DMO’s dealing screens; exclusive facilities to strip and reconstitute gilts; an invitation to a quarterly consultation meeting with the DMO (allowing the GEMMs to advise on the stock(s) to be scheduled for auction in the following quarter, and to discuss other market-related issues);13 and exclusive access to gilt IDB screens. In addition, any transactions undertaken by the DMO for market management purposes are only carried out with or through the GEMMs.

Gilt Inter-Dealer Brokers (IDBs)

56. There are three IDBs operating in the gilt market. Their main purpose is to support liquidity in the secondary markets by enabling the GEMMs to unwind on an anonymous basis any unwanted gilt positions acquired in the course of their market-making activities. Almost all inter-GEMM trades are executed through an IDB. Non index-linked GEMMs do not have access to index-linked screens.

57. IDBs are registered with the London Stock Exchange (LSE) and endorsed by the DMO. The DMO monitors this segment of the market on an ongoing basis to ensure that the IDBs’ services are available to all GEMMs on an equitable basis, and that the market maker structure is effectively supported by the IDB arrangements. The IDBs are also subject to specific conduct of business rules promulgated by the LSE. For example, they are prohibited from taking proprietary positions or from disseminating any market information beyond the GEMM community.

58. Although IDBs cannot take proprietary positions, they serve as the counterparty between the GEMMs involved in a transaction going through them. This is done in order to protect the anonymity of the GEMMs, who only see the name of the IDB when they transact through it. While no IDB has reportedly ever failed to honor its transactions, their central counterparty role in the inter-GEMM market could, in theory, be a source of risk. However, in that event the London Stock Exchange’s default rules would come into effect. Under Rule 15.18 a Stock Exchange official (default official) would determine any payments due to or by the IDB to its counterparties on the transactions in default (on the basis of the current market or “hammer” prices of the bonds in question). Under Rule 15.24 (b) the default official would put the counterparties to unsettled trades in direct contact with one another, and the counterparties is required to complete the trade at the price at which the original order was submitted. In the second half of 2002, the London Clearing House introduced a central counterparty clearing service for cash and gilt repo transactions. This service is currently restricted to conventional gilt transactions.

59. Most IDB transactions are executed through the voice broking systems of the IDBs, rather than through their recently introduced electronic trading platforms. However, the popularity of the latter is growing quickly, especially for straightforward transactions involving benchmark securities. Voice brokers are very useful for other trades, which tend to be more complex involving hedges against futures or links to other transactions, where the price and amounts need to be negotiated with the help of a broker. That said, the number of IDBs is expected to shrink as market preferences settle in favor of one or two electronic dealing systems.

Mechanisms used to issue gilts

60. The DMO uses market-based mechanisms to issue debt and manage the stock of debt outstanding in a transparent and predictable fashion. Auctions are the exclusive means by which the DMO issues gilts as part of its scheduled funding operations. Two different auction formats are used to issue gilts:

  • conventional gilts are issued through a multiple price auction; and

  • index-linked gilts are auctioned on a uniform price basis.

61. Different auction formats are employed because of the different nature of the risks involved to the bidder for the different securities. Conventional gilts are thought to have less primary issuance risk. There are often similar gilts already in the market to allow ease of pricing (or if more of an existing gilt is being issued, there is price information from the existing parent stock); auction positions can be hedged using gilt futures; and the secondary market is relatively liquid. This suggests that participation is not significantly deterred by bidders not knowing the rest of the market’s valuation of the gilts on offer. A multiple price auction format also reduces the risk to the Government of implicit collusion by strategic bidding at auctions.

62. In contrast, a uniform price format is used for index-linked gilt auctions. Positions in index-linked gilts cannot be hedged as easily as conventional gilts. The secondary market for these gilts is also not as liquid as for conventional gilts. Both of these factors increase the uncertainty of index-linked auctions and increase the “Winner’s Curse” for successful bidders—that is the cost of bidding high when the rest of the market bids low. Uniform price auctions reduce this uncertainty for auction participants and encourage participation. In addition there are fewer index-linked bonds than conventionals on issue, so pricing a new index-linked security might be harder than for a new conventional one.

63. GEMMs have access to a non-competitive bidding facility under both formats. They can submit a non-competitive bid for up to 0.5 percent of the amount of stock on offer in a conventional gilt auction. In an indexed-link auction, the proportion of stock available to each IG GEMM is tied to their performance in the previous three auctions and an aggregate ceiling of 10 percent of the total amount of stock on offer.

64. The DMO allots securities at its sole discretion. Consistent with practices in other major industrialized countries, in exceptional circumstances it may choose not to allot all securities of offer—e.g., where the auction would only be covered at a level unacceptably below the prevailing market yield. In addition, in order to protect the integrity of the auction process, it may decline to allot securities to a bidder if it appears that to do so would likely lead to a market distortion. As a guideline, successful bidders, GEMMs or end-investors, should not expect to acquire at auction for their own account more than 25 percent of the amount on offer (net of the GEMM’s own short position in the when-issued market or parent stock) for conventional gilts and 40 percent for index-linked securities. In addition to minimizing the risk that an investor, or group of investors, could exert undue influence over the prices of individual securities, these limits also help to promote a diversified investor base for the government’ securities.

Tap issues

65. The DMO only uses taps of both conventional and index-linked gilts for market management reasons, in conditions of temporary excess demand for a particular stock or sector. The last one was in August 1999.

Conversion offers and switch auctions

66. Market-based mechanisms have also been used to minimize debt stock fragmentation and promote a well-functioning secondary market for gilts. In addition to the above operations, the DMO will occasionally issue stock through a conversion offer or a switch auction, where debt holders are offered the opportunity to convert or switch their holding of one gilt into another at a rate of conversion related to the market prices of each stock. In both cases, the main purposes of these operations are to:

  • build up the size of new benchmark gilts more quickly than can be achieved through auctions alone. This is particularly important in a period of low issuance; and

  • concentrate liquidity across the gilt yield curve by reducing the number of small, high coupon gilts and converting them into larger, current coupon gilts of broadly similar maturity.

67. Conversion candidates will normally have around five years or more to maturity and less than £5.5bn nominal outstanding. In addition, conversion offers will not be made for a stock that is cheapest-to-deliver, or has a reasonable likelihood of becoming cheapest-to-deliver, for any gilt futures contracts (with any outstanding open interest).

68. The price terms of any conversion offer are decided by the DMO, using its own yield curve model to provide a benchmark ratio for the offer. It then adjusts this ratio to take some account of the observed cheap/dear characteristics of the source and destination stocks. Conversion offers remain open for a period of three weeks from the date of the initial announcement of the fixed price ratio in order to allow retail investors to participate. The appropriate amount of accrued interest on both gilts is incorporated into the calculation of the dirty price ratio, for forward settlement. The conversion itself involves no exchange of cash flows. When conducting these operations, the DMO chooses stocks with similar durations in order to limit its exposure to yield curve fluctuations that may arise during the three week life of the conversion offer.

69. Acceptance of such offers is voluntary and debt holders are free to retain their existing stock although this is likely to become less liquid (i.e., traded less widely, with a possible adverse impact on price) if the bulk of the other holders of the gilt choose to convert their holdings. Should the amount outstanding of a gilt be too small to expect a two-way market to exist, the DMO is prepared, when asked by a GEMM, to bid a price of its own choosing for the gilt. In addition, the DMO would relax market-making obligations on GEMMs in this “rump” gilt. The DMO would announce if a gilt took on this “rump” status.

70. In addition to the main purposes identified for conversion offers, switch auctions were introduced in 2000 in order to:

  • facilitate switching by index-tracking funds as a particular stock is about to fall out of a significant maturity bracket, thus contributing to market stability; and

  • allow the DMO to minimize rollover risk by smoothing the maturing gilt redemption profile by offering switches out of large maturing issues into the current 5-year benchmark (or other short-term instruments).

71. Switch auctions are held only for a proportion of a larger stock that is too large to be considered for an outright conversion offer. The DMO ensures that a sufficient amount of the source stock remains for a viable, liquid market to exist following a switch auction. Hence, it will not hold a switch auction for a conventional stock that would reduce the amount in issue to below £4.5 billion (nominal) and are only held where both the respective stocks are within the same maturity bracket, although here the maturity brackets overlap (short and ultra-short 0-7 years; medium 5-15 years; longs 14 years and over). In addition, it will not hold a switch auction out of a stock that is cheapest-to-deliver, or has a reasonable likelihood of becoming cheapest-to-deliver, into any of the ‘active’ gilt futures contracts. It may, however, switch into such a stock.

72. Switch auctions are open to all holders of the source stock, although non-GEMMs must route their bids through a GEMM. They are conducted on a competitive bid price basis, where successful competitive bidders are allotted stock at the prices which they bid. There is no non-competitive facility and the DMO does not set a minimum price.

73. The same principles apply to index-linked switch auctions with the following exceptions. First, these switches will only be held where both the respective stocks have longer than 4½ years to maturity, and where the source stock has not been auctioned in the previous six months. Second, the (nominal) size of any single switch auction is limited to £250 million to £750 million of the source stock, and the DMO will not hold a switch auction that would leave an index-linked stock with a resultant amount outstanding of less than £1.5 billion nominal. Third, the auctions are conducted on a uniform bid price basis, whereby all successful bidders will receive stock at the same price. Where a GEMM’s bids are above this price it will be allotted in the full amount bid, but allotments for bids at the striking price may be scaled. Published results will include the common allotment price; the pro rata rate at this price; the real yield equivalent to that price (and the inflation assumption used in that calculation), and the ratio of bids received to the amount on offer (the cover). Only one index-linked switch auction has been held to date.

Gilt Repo

74. The DMO has the ability to create and repo specific stocks to market-makers, or other counterparties, under a special repo facility if, for example, a particular stock is in exceptionally short supply and distorting the orderly functioning of the market. In response to a previous consultation exercise, the DMO introduced a non-discretionary standing repo facility in June 2000 for the purpose of managing actual or potential dislocations in the gilt market. Any GEMM, or other counterparty, may request the temporary creation of any non-rump stock for repo purposes. The DMO charges an overnight penal rate; and the returned stock is cancelled.14 The facility is cash-neutral, since the operations involve the creation of additional stock in exchange for general collateral securities.

Gilt Strips

75. The U.K. gilt strip market was launched on 8 December 1997. Not all gilts are strippable (see below). The strip market was introduced to permit investors to:

  • closely match the cash flows of their assets (strips) to those of their liabilities (e.g., annuities);

  • to enable different types of investment risk to be taken; and

  • to bring the range of products offered in the U.K. market in line with other large markets, such as the United States, Japan, Germany and France.

76. From the issuer’s perspective, a strips market can result in slightly lower financing costs if the market is willing to pay a premium for strippable bonds. As of March 28, 2002, there were 13 strippable gilts in issue totaling £133.94 billion (nominal). Of these, £1.4 billion of stock was held in stripped form. All issues have aligned coupon payment dates. This means that coupons from different strippable bonds are fungible when traded as strips. However, coupon and principal strips paid on the same day are not fungible in order to protect the overall size of an issue and maintain the integrity of various benchmark bond indices. The first series of strippable stocks were issued with June 7/December 7 coupon dates; however, in 2001, the DMO issued two new conventional stocks with coupon dates aligned on March 7/September 7. These stocks became strippable in April 2002. The second series of coupon dates was introduced to avoid cash-flows becoming too concentrated on just two days in the year.

77. Although anyone can trade or hold strips, only a GEMM, the DMO or the Bank of England can access the service provided by CREST to strip (or reconstitute) a strippable gilt. GEMMs are obliged to make a market for strips. Thus, consistent with practices in other major industrialized countries, end-investors must deal through a GEMM.

78. The market in gilt strips has grown slowly since its inception. Factors that have contributed to this slow take-off have been the need for pension fund trustees to give appropriate authority to fund managers to invest in strips and the inversion of the yield curve over the period since the inception of strips, which makes strips appear expensive relative to conventional gilts. Retail demand for strips has also been hampered by the necessary tax treatment, whereby securities are taxed each year on their accrued capital gain or loss even though no income payment has been made. However, the ability to hold strips within some tax exempt savings products should reduce the tax disincentives to personal investment in strips.

Recent factors shaping the U.K. Gilt Market

79. As indicated above, there have been a number of advances in issuance techniques since the 1995 debt management review. The range of debt instruments has been refined and expanded, and numerous structural changes have taken place in the U.K. government debt markets. The overall aim of the reforms has been to help lower the cost of public financing over the long-term, responding to both endogenous and exogenous factors that have influenced the U.K. debt market during the period. In recent years these factors have included budget surpluses, the rapid rise of the U.K. corporate bond sector, institutional changes (particularly those relating to insurance companies and pension funds), and increasing technological and other advancements, which have enhanced systems, market structures and debt instruments around the world.

Declining levels of government debt

80. As in other currencies, the sterling debt market has seen increased annual private issuance at a time when the U.K. has been running a budget surplus. Thus, the government’s percentage of the overall outstanding sterling debt has been steadily declining (Table 5). Falling government funding requirements have resulted in gilts acquiring a scarcity premium, especially the longer-dated bonds, which in turn led to a reduction in yields and an inverted yield curve that was pronounced in both absolute terms and relative to those observed in other G-7 countries (Figure 1 and Table 8). At the same time, the U.K. has enjoyed a low inflation/low interest rate environment, so the need to enhance returns has led investors to increase their appetite for (credit) risk.

Figure 1.
Figure 1.

Par Gilt Curves

Citation: IMF Staff Country Reports 2003, 208; 10.5089/9781451814200.002.A001

Table 8.

Yield Curve Slopes in G-7 Countries 1/

article image
Source: Bloomberg Financial Services.

Yields on 10-year benchmark bonds minus yields on 2-year benchmark bonds expressed in basis points.

End of period for December 1996 and March 2002. All other years represent averages of end of quarters.

81. As the U.K. Government’s budgetary position improved, gross issuance of gilts declined from a peak of £55 billion in financial year 1993–94 to a trough of £8 billion in 1998–99. However, given that the government’s borrowing needs are cyclical, there is a benefit in maintaining a minimum level of issuance so that market infrastructure is sustained and the market remains sufficiently liquid and retains the capability to absorb future larger gross issuance. Table 9 summarizes the government’s forecast for the Central Government Net Cash Requirement (CGNCR) over the next few years. The medium term forecasts point to an increasing level of net borrowing in response to planned investments in public services that is fully consistent with the fiscal rules.

Table 9.

April 2002 PBR forecasts for the CGNCR (£ billion):

article image

82. In view of the limited amount of gilt issuance in recent years, the DMO adopted a number of strategies to concentrate debt issuance into larger benchmark issues, currently at three maturity points, with a 5-, 10-, and long terms to maturity. These larger issues enable the government to capture a liquidity premium across the yield curve. Examples of some of the initiatives include: running down the stock of NS&I securities, postponing the ramping up of the Treasury bill program, introducing buyback swaps and secondary market purchases of existing stock, and increased holdings of financial assets (such as running larger cash balances). The DMO has also used conversion and switch auctions (see above) to build benchmark issues.

83. The authorities decided to launch a structured gilt buy-back program in FY2000–01 in order to add to gross issuance and thus help to maintain liquidity in the market during a time of strong demand. Following market consultations, reverse auctions were reintroduced in FY2000–01 while the DMO acquired from the secondary market, short-dated index-linked gilts and double-dated gilts. It also purchases near-maturity gilts (with less than six months residual maturity) as part of its regular operations to smooth the cash flow effects of redemptions.

84. The authorities have been pleased with these operations. There was more than 90 percent take-up of the conversion offers, apart from the one conducted in November 1998. The switch auctions have all been covered with a comfortable margin, and the three longer-dated switches have secured very attractive forward-dated funding rates. The rates at which the DMO has repurchased stock in the program of reverse auctions were at yields that were considered “cheap” relative to the DMO’s fitted yield curve.

Effects of pension fund and accounting rule changes

85. In recent years, the pricing of gilts and other sterling-denominated bonds has also been significantly affected by the rules governing pension funds, notably the Minimum Funding Requirement (MFR) and Financial Reporting Standard 17 (FRS 17). The conventional gilt market is especially exposed to the effects of such rule changes because U.K. pension funds and insurance companies are the principal investors in the gilt market.

86. The MFR test was introduced in 1997 as part of the 1995 Pensions Act, and is applied to the assets of defined benefit occupational pension schemes. The MFR test seeks to ensure that a defined benefit pension fund holds enough assets to balance its long-term pension liabilities, discounted over time. In the late 1990s, there was some concern that the MFR had significantly influenced pension funds’ investment decisions. In particular, pension funds may have held more gilts as a hedge against short-term fluctuations in the MFR discount rate (which is based on the prevailing market yield on a basket of gilts with a maturity of 15 years) than would otherwise be the case. Furthermore, this increase in the demand for gilts appeared to have been relatively price-insensitive. Together with the decline in the net issuance of gilts, this excess demand may have contributed to the inversion of the gilt yield curve.

87. Reflecting these concerns, the authorities initiated a review of the MFR. In the March 2001 Budget the Chancellor of the Exchequer announced the Government’s intention to repeal the MFR, and in February 2002 application of the MFR test was relaxed. While there remains some uncertainty as to what will take its pace, institutional investors have begun shifting their demand for fixed-income investments away from gilts in favor of non-gilt bonds in anticipation of the demise of the MFR, and the gilt yield curve has significantly disinverted. However, this relief could be short-lived. U.K. authorities and market participants noted that there is some uncertainty surrounding new European Community pension rules that are in the process of being developed. Some fear that these could prove to be more stringent than the MFR if they require pension funds to be fully funded on a continuous basis. If so, this would add to the incentives for pension funds to prefer fixed-income securities over equities in order to achieve a better matching of their assets and liabilities.

88. Pension funds’ demand for bonds is also being whetted by the introduction of FRS 17, and over time by an aging population, which is leading many pension funds to shift their investments from equities to bonds as their liabilities shorten in duration. Details of FRS 17 and how it compares to International Accounting Standards (IAS) can be found in the chapter dealing with accounting, disclosure, and corporate governance issues. Briefly, it requires that companies’ defined benefit pension assets be measured at fair value, and that liabilities be discounted to present value using the prevailing yield on AA-rated corporate bonds with a similar tenor to the scheme’s liabilities. The net surplus or deficit is recorded in the balance sheet and ongoing service costs (including the basic cost of pension provision) are recorded in the profit and loss statement. Other surpluses and deficits arising from fluctuating market values of fund assets will be recognized in the statement of total recognized gains and losses (STRGL). There is a transition period prior to full adoption of the standard for accounting periods ending on or after June 22, 2003.

89. This accounting change has not had any material effects on the gilt market. However, many pension funds are sharply increasing their holdings of non-government bonds in anticipation of the full implementation of the standard, so that they can achieve a better matching of reported values for pension fund assets and liabilities and limit the effects of changing market values on the sponsor’s financial statements. The Bank of England’s “Markets and Operations” article in the Spring 2002 Bank of England Quarterly Bulletin noted that non-gilt holdings as a proportion of insurance companies’ and pension funds’ asset portfolios have doubled since 1997 to16 percent of total assets, while equity holdings declined by more than 6 percentage points to 61 percent. In mid-2002, the Accounting Standards Board announced that it would delay the full implementation of FRS 17 in order to bring it in line with IAS and EU directives.

90. An obvious puzzle that emerges is why did U.K. insurance companies and pension funds continue to invest in long-term gilt securities when the yield curve became inverted, rather than invest in other sterling debt instruments or offshore on a hedged basis, and why did U.K. companies and foreign borrowers not move more aggressively to tap the sterling market for long-term funding. A couple of reasons were offered by market participants and U.K. authorities.

91. First, it does appear that over time U.K. corporate and nonresident borrowers have in fact sought to access the sterling bond market for funds. In an environment of modest government borrowing requirements, corporate and international issuance of sterling debt has grown rapidly to the point that they now outstrip U.K. government borrowing (Table 5). At times, the European Investment Bank has been a larger borrower in sterling than the DMO. This stands in marked contrast to 1995 when the amount of gilts outstanding was double that of non-gilt sterling bonds. However, U.K. corporate borrowers’ appetite for long-term funds is reportedly very price sensitive. Thus, they appear to be only willing to enter the market when long-term gilt yields are less than 5½ percent. Similarly, foreign borrowers’ willingness to enter the sterling market is very sensitive to currency swap rates. The latter are not always attractive after one takes account of transaction costs and counterparty risk, even though long-term sterling yields may appear low on an unhedged basis.

92. As for U.K. insurance companies and pension funds, they traditionally have not had much of an appetite for non-gilt debt. One reason they have not invested in government bonds in other countries is that it is almost impossible for them to implement the necessary exchange rate hedge because a foreign exchange swap market does not exist for long term investment horizons—three years appears to be the longest tenor available—whereas they would want to hold an investment with a tenor exceeding 15 years. Second, it takes time to change trustee rules and other governance clauses that have limited the discretion of portfolio managers when it came to investing in private debt securities. Pension fund trustees are reportedly unsophisticated and conservative—they have been reluctant to obtain the necessary expertise to supervise more complicated investment strategies. Moreover, a strong equity market in the late 1990s enabled their plans to achieve returns that more than satisfied their actuarial requirements. Thus, they had no incentive to pursue more sophisticated investment strategies because any gains at the margin would have accrued to the plan sponsors, not to the beneficiaries to whom they are accountable. That said, given the weakness in equity markets over the last couple of years, some insurance companies and pension fund portfolio managers have begun taking steps to hire staff that are experienced in investing in corporate debt.

93. The effects of the MFR and FRS 17 on the sterling bond markets illustrate how rules and regulations designed to address public policy concerns can have unexpected consequences on the functioning of financial markets. Authorities need to be especially vigilant of such effects in circumstances where markets, such as the sterling fixed income market, are dominated by a narrow group of investors. However, it is also important to note that these effects are often transitory. The distortions in the gilt yield curve are now beginning to fade as insurance companies and pension funds become more adept at investing in non-gilt bonds, and as other bond issuers enter the sterling bond market to exploit the favorable borrowing opportunities.

Emergence of alternative trading systems

94. Another significant challenge confronting U.K. authorities is the growing popularity of alternative trading systems. Many GEMMs have been active supporters of a number of fixed-income electronic trading systems, including inter-dealer platforms, such as Broker Tec and Euro MTS, and multiple dealer to client systems, such as Trade Web and Bond Click (now part of Euro MTS and Bond Vision).15 In addition, GEMMs have considered adding gilts to their single dealer to customer platforms, where they exist for other government bonds.

95. As part of its continuing commitment to encourage liquidity and transparency of the gilts market, the DMO consulted widely in 2000 about the possible impact of electronic trading systems on the secondary market for gilts and how the DMO’s relationship with the GEMMs might change as a consequence. That work continued during 2001, and led to the introduction in 2002 of an inter-GEMMs electronic market with mandatory quote obligations in the more liquid gilts.

96. This approach is similar to the mandatory liquidity provision common in most European government bond markets. Although this poses an extra burden on the GEMMs, a central committed market should benefit the entire market (including the GEMMs), especially in times of limited issuance. It ensures that the GEMMs have access to a minimum depth of liquidity in certain bonds and that prices in that market are fully efficient, allowing the GEMMs to carry on their wider market-making activities in confidence. This could be particularly valuable in an environment where a number of trading venues exist. The DMO hopes that this model will make it more likely that entry barriers facing prospective GEMMs remain at acceptable levels, maintaining a high degree of competition in market-making services.

97. The DMO sees little need to extend the model to the full electronic dealership model, where the centralized primary dealer core is augmented through direct participation of other financial institutions. However, this model may evolve naturally from the inter-GEMM model.

98. The issue of who should supply the necessary trading systems was delegated to a group of elected representatives of the GEMMs and a representative of the DMO. This group has concluded that the best way is to allow each GEMM to supply their prices to any recognized IDB. Adopting this approach preserves competition in the provision of IDB services, and ensures that brokers will have a continuing incentive to develop their services and encourage further technological innovation. While liquidity would not be concentrated with just one broker, there should not be any market fragmentation, since all GEMMs will have equal access to all recognized IDBs.

99. At the retail level, three of the GEMMs that specialize in servicing retail clients have launched a multiple dealer to client electronic trading service, BondScape, for sterling-denominated fixed income products. This platform provides some basic bond analytics, in addition to supplying firm prices in bonds in small size, adding to the prospective investors’ information set. The service is also provided to the network of retail brokers that deal directly with individuals.

100. Electronic trading systems also open the way for primary auctions to be held electronically. In addition to alleviating operational risk and facilitating faster auction processing (as discussed previously), it would open the possibility for more direct access to auctions by a wider set of market participants. Some settlement safeguards would need to be introduced, however, to protect the government against settlement risk in case some of the new participants failed to honor their auction obligations. That said, it is not clear whether major investors would choose to participate directly in the auction. Some prefer to avoid execution risk by dealing in the when-issued market, while others, especially index-trackers, would likely wait to trade until after the auction so that they can execute at a price close to the one used in the calculation of gilt indices.

VIII. Overall Assessment

101. The management of public debt is not a source of vulnerability for the financial system given the low level of government indebtedness and the conservative approach to managing the risks associated with the debt. The UK’s practices are fully consistent with the IMF/World Bank Guidelines for Public Debt Management. Indeed, the authorities deserve to be commended for the smooth transfer of cash management responsibilities from the Bank of England to the DMO in 2000, which has contributed to a highly efficient process for managing cash balances, and the large number of innovative initiatives undertaken to minimize the fragmentation of the debt stock and maintain issuance volumes in a period of modest borrowing requirements. Going forward, consideration could be given to some minor technical innovations, such as: (i) introducing some indicators to explicitly articulate the government’s preferred cost-risk tradeoff in order to increase public understanding of the decisions underpinning the debt management remit, and reinforce the separation of accountability for debt policy strategy (a policy decision) from accountability for the tactics used to implement the strategy (the DMO’s tactical mandate); (ii) automating the bid-capture process used in debt auctions in order to reduce operational risk and improve auction-processing times; and (iii) formalizing the role of the advisory board in the DMO’s governance framework and introducing some selection criteria for members if this body is expected to become a permanent fixture in the DMO.

102. Despite the U.K. authorities having undertaken a number of steps to maintain the depth and liquidity of the gilt market in an environment of modest borrowing requirements, the gilt market has been distorted by changes to accounting and pension rules. It is important to note that these distortions are gradually being resolved by market participants themselves as insurance companies and pension funds become more adept at investing in non-gilt securities, and as other borrowers enter the market to take advantage of the high demand for sterling-denominated debt securities. Thus, there does not appear to be a need for public intervention beyond the steps already taken by the DMO to maintain gilt issuance volumes and minimize fragmentation in the debt stock. The main lesson to be drawn is that policymakers need to be cognizant of the unexpected consequences for financial markets of public policy actions in other domains. This is especially true for markets such as the gilt market, that are dominated by a narrow group of market participants.

103. Finally, some market makers expressed concern about the new mandatory quotation rules in the interdealer gilt market. While conceding that typical bid-offer spreads in this market are well within the maximum permissible limits, they believe that the need to be seen quoting prices in this market distracts them, at the margin, from focusing their efforts on serving their underlying customer base. In contrast to government bond markets in the U.S. and the largest Continental European countries, most of the trading activity in the U.K. consists of transactions with customers rather than interdealer trades. On the other hand, the new rules should help make the interdealer market more accessible to a broader range of market makers. Over time, a more active and accessible interdealer market could prove invaluable as the overall market develops and the investor base broadens out. Thus, the new rules appear to be appropriate at this stage of the market’s development. In addition, while the maximum bid-offer spreads are wider than the typical spreads seen in this market, there is no need to narrow them in. They currently serve a useful backstop function, and the future growth of this market is probably best served by leaving it to the participants to set appropriate spreads through competition.

1

This paper was prepared by the FSAP mission team as part of the background work for the U.K. FSAP in the summer-fall of 2002. The primary contributor to this paper was Mark Zelmer of the IMF’s Monetary and Financial Systems Department.

2

The range of collateral accepted by the DMO in its market operations is set without recourse to prudential information from the Bank of England or the FSA. It consists of all gilts, Treasury bills, and foreign currency issues of HM Treasury; selected European government €-denominated sovereign issues; selected highest-rated supranational £- and €-denominated debt; selected bank bills, and other high quality short-term debt issued by supranationals and foreign governments. Consistent with market practice, the DMO does not publish specific details regarding the collateral it will accept, but does publish broad outlines of the collateral accepted in repo and reverse repo trades. This enables it to set rules that are free of public policy considerations and devoid of signaling content. Its market counterparties learn the details through experience.

3

Full details of all these instruments and operations are available in the Gilt Operational Notice and Cash Operational Notice on the DMO website.

4

The Exchange Equalization Accounts (the government’s repository for its foreign currency reserves) are published each year. There is also a formal monthly requirement to publish the level of reserves and indicate whether (and if so, the amount) the authorities have intervened in the foreign exchange market.

5

A full description of all the DMO’s responsibilities, objectives and lines of accountability is set out in the current version of its Framework Document (July 2001). It and other relevant documents can be found on the DMO’s website.

6

The Advisory Board does not have any formal standing in the governance framework and its proceedings are not published. The two members were appointed by the Chief Executive Officer through an informal selection process. They provide the DMO’s management team with access to external expertise in audit issues and the formulation of investment management guidelines—the latter has taken on added importance now that the DMO has cash management responsibilities.

7

For example, under one of the arrangements the DMO and the Bank of England can use each other’s trading room as a backup facility in the event of an emergency. This arrangement has been successfully tested in practice. For example, on one occasion the processing of a T-bill tender was successfully transferred from the DMO to the Bank in the middle of an auction.

8

The amount of money to be raised in the wholesale market each year is set after taking into account projections of the net amount of funds that the NS&I expects to raise directly from U.K. households. NS&I projections are based on experience in previous years, and its own sales targets set in consultation with HM Treasury.

9

Recent monetary history offers a useful parallel in this regard. In an inflation targeting framework, one cannot say with precision what the target rate of inflation should be—it is ultimately a question of judgment. However, the adoption of inflation targeting helps to separate accountability for the choice of target (a strategic decision) from accountability for tactical decisions related to the implementation of monetary policy

10

“Government’s Cash and Debt Management” (HC 154) was published on May 22, 2000. It provides a comprehensive review of the Government’s cash and debt management arrangements, as well as records of the oral evidence provided by officials and expert witnesses.

11

Mandatory bid-offer spreads on inter-GEMM quotes are to be 3 basis points or less, and the two-way quotes are expected to be good for dealing amounts of up to £5 million in the case of short and medium-term gilts, and £2 million in the case of long-term gilts. Most IDB trades take place at prices well within these bid-offer spreads. Thus, the maximum allowable spreads are not normally binding in practice. Moreover, in the event of a severe crisis in the gilts market, the DMO would be prepared to consider relaxing the maximum spread obligation, since it may not provide adequate cover for the risk being undertaken. This should ensure that this obligation does not have any significant negative effects on the resiliency of the gilt market in periods of market stress.

12

In order to prevent undue market concentration, GEMMs and their customers are each subject to allocation limits at the auctions. The limits are 25 percent of the amount of securities being auctioned (40 percent in the case of index-linked auctions).

13

The DMO also holds quarterly meetings with the representatives of end investors. Minutes of these meetings are published shortly afterwards on the DMO’s website. Additionally, there are annual meetings with the relevant Treasury Minister for both groups in January as part of the preparations for the annual remit, generally published in March.

14

The penal rate of interest is ten percent of the prevailing Bank of England two-week repo rate. For example, if the Bank of England’s two-week repo rate is 4 percent, then the institution borrowing the stock from the DMO would receive an annualized rate-of-return of 0.40 percent on the funds lodged with the DMO in exchange for the stock.

15

Discussion of this issue draws heavily from A. Holland, 2001, “The Development of Alternative Trading Systems in the U.K. Gilt Market: Lessons and Implications.” This paper was prepared for the Financial Market Structure and Dynamics Conference organized by the Bank of Canada in November 2001.

United Kingdom: Financial Sector Assessment Program Technical Notes and Detailed Standards Assessments
Author: International Monetary Fund