69929 Tomas I. Magnusson, World Bank December 2005 Formalizing a Debt Management Strategy In short, central government debt management can be defined as the process of establishing and executing a strategy in order to meet the debt management objectives. Undoubtedly, the development of the strategy is the most important debt management decision. Given the market constraints, it is the strategy document that decides on issues such as the level of exposure to foreign currency risk, desired maturity structure of the debt, level of interest rate sensitivity, whether and how much of the debt should be indexed to inflation, and the plan for development of the domestic debt markets. If the government has chosen a strategy that turns out to be too risky, or, at the other end of the spectrum, too costly in order to avoid any risk, it will affect the budget outcome much harder than any misprized and/or badly timed debt management transaction. This paper will discuss the appropriate organizational arrangement, internal procedures and regulatory framework for a successful and sustainable development of debt management strategies. Debt Management Objectives The debt management strategy is the plan on how to achieve the debt management objectives. Thus, the starting point is to decide what these objectives should be. The objectives should not be mixed with the strategy. They are two different decisions with different time horizons, decision-makings and evaluations. However, they are 2 directly linked as the objectives are a prerequisite for development of a strategy; before developing a strategy one has to know what the strategy will be aiming at. Based on country practices, the debt management objectives could be formulated as follows: “Management of the central government debt shall have the following objectives: 1. The central government’s funding needs are always met. 2. The costs of the debt are minimized from a long-term perspective. 3. The risks in the debt portfolio are kept at acceptable levels. 4. Development of the domestic debt markets is promoted.� As with any policy objectives, also the achievement to meet the debt management objectives should be evaluated. Any policy constraint on the debt manager in meeting these objectives should therefore be addressed up front. One common policy constraint is associated with the central bank’s implementation of the monetary policy objectives, which in general is achievement of price stability. For example, the debt manager may prefer in meeting the risk objective to reduce the foreign currency debt faster than the central bank would like to see, the latter worrying on the effect such fast reduction will have on the value of the domestic currency. In other circumstances the central bank might prefer that more treasury bills (T-bills) would be issued than was originally planned by the debt manager, arguing that a better supply of T-bills will assist the central bank in steering the liquidity in the market. One way of addressing this potential conflict is to qualify the above objectives by adding the following paragraph: “In achieving the debt management objectives, the debt management activities shall be consistent with (alternatively, “shall consider�) the aims of monetary policy.� 3 It is important to carefully consider the drafting of the debt management objectives. The objectives should be long term and must have certain robustness in order to serve as anchor for the debt management strategies. Who should finally determine the debt management objectives? The first choice for this decision is the Parliament/Congress, as part of its financial power. It is the Parliament/Congress that approves government tax and spending measures, and, stemmed from this financial power, it normally also has the ultimate legal authority to borrow on behalf of the central government. However, this borrowing power, as a rule and for practical reasons, is delegated to the executive branch of government, normally with some overall restrictions such as debt ceilings. Given that the management of the central government debt may have significant repercussions for future tax and spending levels, extending beyond the term of the current Government, it would be logical for the Parliament/Congress to determine the debt management objectives in legislation. To include these objectives in legislation will give them particular prominence and avoid ad hoc and frequent changes to this important aspect of debt management. This also is in line with regulation of monetary policy, where the objectives of monetary policy normally are determined by the Parliament/Congress through a provision in the Central Bank Act. Another advantage with this structure is that the objectives will form a natural base for the Parliament/Congress to require the executive branch to regularly (at least yearly) report to the Parliament/Congress on how it has used its borrowing mandate, and to what degree it has respected the set objectives. This reporting will be much more interesting for the Parliament/Congress, and the public debate, than sheets of statistics on the different debt management transactions made during the previous year, which is common to see today. This kind of reporting will put pressure on the executive branch to develop a high-quality strategy document. Finally, it probably also will prevent the Parliament/Congress to interfere too much in the execution of the debt management policy, which it normally does not have the right capacity for. For instance, still in many 4 developing countries the Parliament/Congress requires to approve single borrowing transactions, which is a very slow and cumbersome process. Here are some examples on how the debt management objectives have been formulated in the primary legislation of some countries. Macedonia (Public Debt Law of June 2005) “Public debt management objectives shall be the following: - undertaking measures and activities by the Ministry of Finance to the end of ensuring financing of the needs of the state with the lowest possible cost, in the medium and long run, and with sustainable level of risk, and - undertaking measures and activities by the Ministry of Finance to the end of development and maintenance of efficient domestic financial markets.� Portugal (Law on General System Governing the Issue and Management of Public Debt of February 1998) “The management of direct public debt shall be guided by principles of discipline and efficiency, ensuring the availability of the financing required by each budgetary period, in line with the following objectives: 1. Minimisation of direct and indirect cost from a long-term perspective; 2. Guarantee of an even distribution of costs by the several annual budgets; 3. Prevention of an excessively concentrated time-profile of repayments; 4. Non-exposure to exaggerated risks; 5. Promotion of a smooth and effective operation of financial markets.� Serbia (Public Debt Law of June 2005) “The goal of public debt management is to decrease borrowing expenses for the Republic in accordance with a prudent risk level.� Sweden (Act on Central Government Borrowing and Debt Management of December 1988) 5 “The central government debt, with the exception of the debts that are managed by central government enterprises in accordance with Section 3, shall be managed so that the costs of the debt from a long-term perspective are minimised with due regard to the risks associated with debt management. The management shall take place within the framework imposed by monetary policy requirements.� Formalizing the Debt Management Strategy Once the long-term debt management objectives are set, preparation of the strategy document can start. Who is best equipped to prepare this document? To answer this question, we must first discuss what the strategy document normally includes. Again we have to go back to the objectives. Using the example at the beginning of this paper, we can conclude the following. Taking the long-term objective that the central government’s funding needs always are met, the strategy should be to ensure that the government has access to a number of different sources of funding to reduce the risk of relying on a narrow funding base, i.e. to develop a broad funding and investor base. This is particularly important in case of increasing borrowing requirement. Meeting this objective also might include the building up of a certain cash reserve, and smoothing out of the redemption profile of the debt to reduce the roll-over risk. Looking at the cost/risk objectives, we can conclude that trade-offs have to be made. For example, it might be cheaper to borrow in low-coupon foreign currencies than domestic currency. But, on the other hand, the risk in the portfolio normally will be higher if the borrowing is made in foreign currencies than in the local, domestic currency. Also, in most cases it is cheaper to borrow in the short end of the yield curve comparing to long- term borrowing. But in the same time it will be riskier to borrow short term than long term as the short-term interest is more volatile. Thus, one important function of the 6 strategy document is to make these trade-offs, which basically decides the risk tolerance of the government. Finally, development of a strategy to promote well-functioning domestic debt markets probably will include a plan to build up a yield curve through a progressive extension of the maturity of the government securities, to regularly supply government securities to the market, and to consolidate the number of debt issues into a few liquid market benchmarks. Again trade-offs have to be made, e.g., between increasing the roll-over risk in building up liquid benchmark loans and achieving lower-cost funding, in addition to promotion of a well-functioning domestic market. From the above it can be concluded that the unit which prepares the strategy documents ought to have financial skill and, not the least, close contacts with the debt markets. In order to prepare a feasible plan on how much the central government should borrow in different market segments, which instruments to be used to change the interest rate sensitivity of the debt portfolio, and how much the government can extend the maturities in its domestic market, it is necessary to have a good knowledge and understanding of the investor base and how the markets work, and a sense of the market appetite for certain instruments. When this precondition is not met, the strategy document most likely will be just a paper product with little relevance, and will fail in guiding the daily debt management decisions. Due to the importance of having these financial skills and a good ear to the market, as a rule it is the debt management unit that is mandated to prepare the debt management strategy proposal. Indeed, in countries which have an institutionalized process to determine the debt management strategy, the preparation of the strategy is one of the main functions of the debt management unit. Once the strategy has been decided, the debt management unit also will be responsible for drafting the yearly borrowing plan, based on the strategy, and for independently executing this plan. 7 Many countries, particularly in the OECD area, have consolidated the central government debt management functions into one debt management unit, also called the debt management office (DMO). Some countries have chosen to set up the DMO within the Ministry of Finance, while others have established the DMO as a separate agency outside the Ministry of Finance. There are pros and cons with both these alternatives, and the choice also depends on the institutional culture of the country. Other countries, particularly among the developing countries, have kept the fragmented structure of debt management, and instead are trying to coordinate the activities among the different debt management units. For instance, it is common in these countries to find one unit responsible for foreign borrowing from the international financial institutions and bilateral donors, and another unit, normally the central bank, to be in charge of the domestic borrowing. In other countries the fragmentation is even greater, with a third unit responsible for domestic retail borrowing and a fourth unit for borrowing in the international capital markets. In order to move from basic funding to a more cost-and-risk-based debt management, it is essential for these countries to establish one of the ‘debt managers’ as the lead debt manager with the mandate to prepare a debt management strategy for the entire debt. An alternative is to set up a separate debt management coordination unit for this purpose. This unit (either the lead debt manager or the special coordination unit) will by definition then become the middle office for the entire debt. However, as discussed above, it is essential for this middle-office unit to have close contacts with the market, which can be a challenge in these cases as a substantial part, if not all, of the front- and back-office activities will be performed outside this middle-office unit. In addition, this function must have access to accurate and timely debt data, which also can be a challenge as it is common to also find several debt databases in countries with a fragmented organizational structure. 8 However, here it is important to point out that even in countries that have consolidated the debt management functions into a DMO, it is not uncommon that the central bank runs the auctions on behalf of the DMO. Also other functions can be outsourced. But in these cases it is at least clear that the DMO is the principal and that the other units only act on instructions from DMO. So in order for a special middle-office unit to function well in a fragmented organizational framework, it is essential that this unit as a minimum requirement is given overall responsibility and necessary tools to achieve the strategic goals, always contains a critical mass of expertise, is able to prepare comprehensive debt management strategies based on cost/risk trade-offs, has an ear to the market, monitors the debt management activities performed by other units, and that any debt management activity outside this middle-office unit is clearly regulated. This can be regulated in agency agreements between the Minister, the middle-office unit and the other debt management units, or by a Cabinet or Ministerial instruction. In reality, the other units will then act as agents for the middle-office unit. Once the debt management unit has prepared a debt management proposal, the question then is who should decide the strategy. Considering that the strategy includes many trade- offs and basically determines the preferred risk tolerance of the country, the strategy is commonly decided by the Cabinet/Council of Ministers or by the Minister of Finance.1 The advantages of this approach are that it leaves major decisions as to the overall volume of indebtedness and the acceptable risks in the debt portfolio - in terms of their impact on the budget, taxes, government spending programs, or other such fiscal indicators - with political decision-makers, while allowing technical professionals to seek the optimum risk-adjusted outcome within those parameters. This is important because, among other things, it provides clear guidance to the debt manager concerning the acceptable scope of their activities, while giving politicians and the public comfort that 1 In most cases, there is no text-book optimal debt portfolio. Instead every country has to make its own decision on its risk tolerance, based on scenario analyses in the context of local circumstances. This is basically a political issue, closely linked to fiscal policy. 9 the activities of government financial specialists will produce outcomes within an acceptable range. Decision-making Process for Determining the Strategy After the debt management unit has prepared a strategy proposal document, it is not uncommon to find an advisory board mandated to comment on the proposal. This will function as a quality assurance before the Cabinet/Council of Ministers or the Minister takes the final decision. This advisory board normally comprises senior officials with long market practice, with addition of some senior academics in financial economics. Involvement of an advisory board in the decision-making process is also important for the evaluation process, as more described below. As mentioned earlier, it is not uncommon that conflicts might arise between implementation of the objectives of debt management and monetary policy implementation. Any conflict of this sort should be resolved, as far as possible, during the decision-making process of the strategy. A practical way to include the central bank is to ask them to send written comments on the proposal before the final decision is taken. These comments should basically be limited to those aspects of the proposal that might conflict with monetary policy implementation and not on the debt management per se, as this should be the sole responsibility of the debt management unit and the executive branch. Any conflict would then be resolved by the Cabinet/Council of Ministers or the Minister. And, as previously mentioned, if one of the policy constraints is that debt management activities must consider the aims of monetary policy, the central bank views will in most cases be decisive. This procedure also will assist in separating central government debt management from monetary policy implementation. Particularly in the developing countries it is common to 10 find the central bank running the auctions of T-bills and treasury bonds more as a principal than agent of the executive branch. In these cases domestic borrowing is more determined by monetary policy considerations than cost/risk trade-offs. In addition, it dilutes the responsibility of debt management and may generate market perception that debt management benefits from inside information on the future path on domestic interest rates. Another advantage with this separation is that the central bank then solely can focus on monetary policy, which is its main function. The strategy document should have a medium-term horizon. One obvious starting point is a time horizon that covers the same period as the medium-term expenditure framework (MTEF), i.e. often three years. The strategy then will cover the period when budget forecasts and assumptions about developments affecting the debt and borrowing requirements are easily available. However, to make the process robust and considering economical and financial changes, the strategy should be updated on a yearly basis, following the procedure described above. This process also will encourage a transparent debate on the debt management activities at least once a year.2 In a highly volatile environment, the strategy might need to be changed during the budget year. In these cases the same procedure for the regular yearly updates should be followed. In addition to the debt management objectives, some countries have chosen to also prescribe the basic process for strategy development in the primary legislation. For example, in the Swedish Act on Central Government Borrowing and Debt Management it is stated: 2 Considering that the cost-minimization is a long-term (5-10 years) objective, a strategy horizon of only three years might seem to be short. For example, management of refinancing risk may look well beyond three years. On the other hand, in an environment with fast and far-reaching changes it will be difficult to prepare a trustworthy strategy much beyond three years. One way to solve this and to give the debt manager incentives to look beyond three years is that the Cabinet/Minister requires the debt manager to add a rough strategy plan for the remaining 4-10 years. In addition, the Cabinet/Minister can decide that the performance of the debt manager in meeting the long-term cost-minimization goal shall be evaluated on a 5-6 years basis. 11 “Not later than the 15th November each year, the Cabinet shall decide upon the central government debt management strategy. The Cabinet shall invite strategy proposal from the Swedish National Debt Office, and shall allow the Central Bank to express its opinion in respect of the Debt Office’s proposal.� In the secondary legislation (Ordinance Prescribing Instructions for the Swedish National Debt Office) it is further stated that the Debt Office not later than October 1 each year shall submit to the Cabinet a proposed debt management strategy, and that the Board of the Debt Office shall decide on the proposal before it is submitted to the Cabinet. With this legislation in place, the strategy development process has become very robust and substantially diminishes the risk of ad hoc political pressure, which is likely to increase the risk level in the debt portfolio and consequently the country vulnerability. “What Gets Measured Tends to Improve� Once the strategy has been determined, the debt management unit is responsible for execution of the strategy. The first step for the debt management unit is to make the strategy operational through a borrowing plan, and to set up a system for monitoring that the debt management operations will be in line with the strategy. This task normally is handled by a risk control group of the debt management unit. Within this framework, the debt management unit should be allowed to independently execute the strategy, but also be responsible for the result of its activities. With the structure recommended in this paper, the reporting lines will be logical: - The debt management unit/DMO sends a written report (at least yearly) to the Cabinet/Council of Ministers or the Minister of Finance, including an internal evaluation on how its activities have moved the cost/risk of the debt portfolio in 12 line with the requirements determined by the strategy, and what actions it has taken to promote the development of the domestic debt markets, and - The Cabinet/Council of Ministers sends a report to the Parliament/Congress, informing about the chosen strategy and the rationale behind it, and explaining in what way the strategy decision has assisted in achieving the debt management objectives. Though it is rather easy to evaluate how the daily debt management operations have affected the cost/risk in the debt portfolio and consequently whether these activities have contributed to the fulfillment of the strategy goals, it is much harder to evaluate if the chosen strategy was the best choice for the country. With the benefit of hindsight, it is of course easy to find out that another strategy would have been better. But what is interesting and essential to know is if the decision making was founded on sound analysis and that all available information was taken into account. One way of assessing the strategy is to give the decision maker the opportunity to choose between a number of stylized and differentiated debt portfolios whose characteristics have been analyzed in the strategy proposal. The costs and risks of the debt portfolio selected could then be compared with the costs and risks of the other portfolios. However, in practice, to monitor the results of the different portfolios will involve substantial system support and robust quantitative models that probably are beyond the reach of most of the countries. The alternative is to assure that skilled staff is recruited to the debt management unit and that the unit has the necessary tools, such as adequate system support, to enable it to properly analyze the cost/risk structure of the debt under different scenarios. It also makes sense, as described above, to build in a quality control by asking an advisory board to give its opinion on the proposal. In addition, it is helpful to have transparent process, which will encourage an open debate on the proposed strategy and increase the likelihood that all aspects of the strategy have been considered. 13 Finally, transparency is also essential in assisting the markets to have a clear understanding what the debt management activities aim at, which will make the borrowing more cost effective by reducing the risk premiums. Unclear debt management objectives and strategies create uncertainty within the financial community. Investors incur costs in attempting to monitor and interpret the government’s objectives and policy framework, and this uncertainty premium is quickly reflected in reduced demand for the government’s securities, or in higher debt-servicing costs, or both.