Chapter 3. Cash and Debt Management: Interaction, Coordination, and Integration

Abstract

This chapter identifies the main cash and debt management (CDM) functions of a ministry of finance (MoF), explains the importance of interaction and coordination between them, and sets out how this can be best achieved under various institutional structures. Although the chapter discusses the CDM items in combination, it approaches the interactions primarily from the perspective of cash management and the wider treasury. It does not discuss debt management policies in detail, which would require a more extensive literature (Bangura, Kitabire, and Powell, 2000; Shah, 2007).1 The chapter also offers a number of country examples, describes how cash and debt units are organized in Latin America, and discusses the usefulness that integration has on both functions in one unit.

Introduction

This chapter identifies the main cash and debt management (CDM) functions of a ministry of finance (MoF), explains the importance of interaction and coordination between them, and sets out how this can be best achieved under various institutional structures. Although the chapter discusses the CDM items in combination, it approaches the interactions primarily from the perspective of cash management and the wider treasury. It does not discuss debt management policies in detail, which would require a more extensive literature (Bangura, Kitabire, and Powell, 2000; Shah, 2007).1 The chapter also offers a number of country examples, describes how cash and debt units are organized in Latin America, and discusses the usefulness that integration has on both functions in one unit.

International experience has favored more recently the structuring of both functions in one unit because of efficiency, both in developing a consistent policy stance and operationally. It appears, however, that the integrated structure is particularly useful when there is a certain level of market development and when the CDM functions are active and well developed. In any case, a good coordination mechanism is essential.

This chapter is organized in three sections. The first discusses the organization of CDM, the second provides international experience with particular focus on Latin America, and the final discusses the advantages and disadvantages of the integration model. There is a tradition in Latin America of the treasury and debt management functions being separate. It is recognized, however, that an integrated unit can be more efficient when a more active management is envisaged and if there is a good level of market development.

The chapter emphasizes the importance of coordination mechanisms to improve the efficiency, transparency, and value for money of cash and debt operations. The experience regarding cash and debt committees in Latin America is described as a mechanism that favors coordination.

Performance of CDM in institutions throughout the region is assessed. This includes the annual borrowing plan preparation; treasury single account (TSA) management and monitoring; cash surplus investment; cash flow forecast preparation; treasury bond (Tbonds) and treasury bill (Tbills) issuances, as well as corresponding auction calendars; risk management and analysis; market and investor relationships; and accounting and statistical report preparation. Although most of these functions are performed by the CDM offices in the region, active cash management is not prevalent in many countries. This may explain why only three countries (Brazil, Colombia, and Peru) have adopted integrated offices, given that financial markets that are less developed and have less active cash management do not require integration between the two functions.

Throughout this chapter, DMO (debt management office) is used specifically to refer to a semi-autonomous debt management function, often with cash management integrated with it, whereas DMU (debt management unit) is used more generically or where the function is more closely integrated with the rest of the ministry, although it may still have operational or managerial responsibilities delegated to it.

Organization of Cash and Debt Management Functions

Cash Management Functions

The overriding objective in all countries with regard to government cash management is to ensure that cash is available to execute the budget efficiently and to meet government obligations when they fall due. Modern cash management, however, has other objectives (Lienert, 2009; Williams, 2010).

  • Cost-effectiveness: Borrowing only when needed, to minimize government financing costs; and maximizing any returns on surplus idle cash.

  • Risk management: Protecting government short-term assets, for example, by insisting on collateral when cash is invested, and ensuring that there are always sufficient avenues open to secure short-term financing when required.

  • Support other financial policies: Debt management, monetary operations, and financial market development.

These objectives generate a range of functions, which go beyond those of the traditional treasury’s mostly passive role in monitoring cash balances and maintaining a cash buffer to handle volatility and unanticipated outflows. The traditional treasury tended to restrict expenditures or delay bill payments, when necessary, to avoid overdraft or unacceptably expensive borrowing. This has been the reason why cash management in the past often has been considered simply as an extension of the budget execution function. Modern cash management, however, requires planning to ensure the smoothing of daily and weekly cash flows through active borrowing and lending in money markets, as well as having the appropriate tools to cope with financial market volatility. Cutting planned expenditure as a result of a lack of cash equates to cash rationing. Effective cash management avoids the need for cash rationing.

It is important to emphasize in this context that the investment of surplus cash is not only an added value; it is an intrinsic part of modern cash management. Cash flow smoothing implies the investment of surpluses; and smoothing generates benefits by way of a lower cash buffer as a result of reduced cash volatility, which also facilitates monetary policy operations. Well managed, an investment will be—in itself—cost effective and thus reduce the net debt interest bill.

The development of a TSA, including the consolidation of all government cash balances into a single bank account, is the first prerequisite for efficient cash management. Most countries, including those of low income, have a TSA in place—or, at least, have implemented policies and processes to build the TSA, although it may still take time to complete. Such is the premise of this publication. Cash management functions look beyond the TSA and revolve around three main activities:

  • 1. Monitoring and accessing government cash and related short-term assets

    • Identification of assets that are within the scope of cash management. This relates mostly to ensuring that the TSA is complete and is fully accessible, although there may be other assets that are available in certain circumstances. These may include those assets that are held in fiscal stabilization funds or other reserve funds.

    • Arrangements that are in place to allow the cash manager to monitor TSA balances in as close to real time as possible.

    • Integration of policy understandings into the cash management function that determines where temporary surplus assets will be placed; and maintenance of a cash buffer.

  • 2. Cash flow forecasting

    • Net cash flow forecasting of at least three months in anticipation to be made across the TSA. The timing of future peaks and troughs should be predicted to enable decision making with regard to borrowing and lending maturities. It also provides cash managers—and, hence, the spending units—with sufficient time to plan and execute budget changes in the face of potential emerging problems. This implies alerting the budget unit and others involved to assess their priorities, determining in a timely way the appropriation authorities to be withheld, and informing spending units. Giving some notice of these actions is preferable to simply imposing arrears upon suppliers. In any event, the cash manager should not be involved in deciding which suppliers should not be paid at any given period.

    • Daily, weekly, or monthly forecasting of flows, depending on the cash manager’s targets and/or capability.

    • Networking and exchange of information with spending units, revenue administrations, and those responsible for the execution of the budget to enable successful forecasting.

  • 3. Financial market interaction

    • Identification of options to manage, cost-effectively, government cash flow deficits and surpluses. The core of these is likely to be the issuance of Tbills and the use of repurchase (repo) and reverse repurchase (reverse repo) agreements. The options may also include bank borrowing and deposits in the central bank or commercial banks. For deposits and other investments, credit quality and liquidity are crucial; there should be no risk to principal; and buffers, in particular, should be held in highly liquid form (including, for example, term deposits that can be withdrawn with penalty). The maturity of investments also should be geared toward the cash flow profile, so that they mature on days when cash is likely to be required, thus smoothing the net cash flow.

    • Very short-term arrangements in place to meet unanticipated cash flow deficits. These safety nets often include overdraft facilities with the central bank or commercial banks.

    • Development of short-term borrowing and investment plans and their execution in an effective and low-risk manner. This may also extend to the management of a transaction throughout its lifecycle.

    • Associated liaison with the central bank and market participants.

These functions and interactions are summarized in Figure 3.1.

Figure 3.1
Figure 3.1

Main Cash Management Functions and Interactions

Source: Authors’ elaboration.

There are many supporting operations that are associated with these core functions. These include databases to maintain cash flow forecasts, for financial transactions, and for accounting purposes. Information sharing and coordination are crucial with revenue-collecting agencies and spending ministries, as well as for those relevant offices within the MoF. Coordination with debt management is self-evident—and is further discussed below—and equally essential is coordination with the central bank. This involves the flow of information, especially with regard to real-time information on the incoming and outgoing flows of the TSA (assuming that it is held by the central bank and that information is also essential for other accounts held there). Furthermore, there should be clarity with regard to the respective responsibilities and operational interaction in money markets, as well as to the services offered (e.g., as a banker or fiscal agent) by the central bank (Pessoa and Williams, 2012).

The importance of the risk assessment and management responsibilities of the cash manager are often underestimated. They include the following:

  • Liquidity risk: Ensuring that liquid funds are available and avoiding overdrafts or expensive emergency facilities.

  • Funding risk: Securing the ability to raise funds at market yields, when required.

  • Forecasting risk: Making decisions on the basis of imperfect estimates of the borrowing requirement, or with insufficient information of the volatility, or lumpiness of underlying cash flows.

The risk assessment and management responsibilities of the cash manager also include the more familiar risks:

  • Market risk: Associated with the management of cash balances.

  • Credit risk: Relating to investment counterparties.

  • Operational risk: Associated with transactions, payments, and accounts.

Debt Management Functions

The core function of debt management is widely acknowledged. The objective of most countries is similar to the following: “The main objective of public debt management is to ensure that the government’s financing needs and its payment obligations are met at the lowest possible cost over the medium to long run, consistent with a prudent degree of risk” (IMF and World Bank, 2003). Many countries will add a secondary objective that relates to domestic financial market development.

The key roles of debt managers in achieving this objective are to establish and execute a strategy to manage government debt, to raise required funding amounts, and to achieve their cost and risk objectives. It is essential for debt managers to process the transactions and manage associated data, as well as achieve other government goals. Although domestic market development may be a common secondary objective, the skills of debt managers can be put to use in other policy areas, such as managing other government assets and liabilities, including contingent liabilities.

Debt management functions, therefore, fall within three main categories:2

  • 1. Financial market interaction; in particular, the execution of transactions and negotiations with other creditors:

    • Implementing the annual borrowing plan, including related transactions (e.g., derivatives), and contribution to the design of that plan. Transactions may be in domestic or foreign markets, and may involve securities or loans and credits; they also may include guarantees and on-lending of external loans to, for example, state-owned enterprises.

    • Maintaining regular contact with market participants—domestic and, in many cases, external. This includes lenders (private and public), investors and primary dealers, and other market intermediaries. There may need to be understandings with the central bank, either in relation to money or debt market development, or to avoid operational clashes.

  • 2. Debt management strategy (DMS) design:3

    • Design of strategy options and their presentation to policy-makers who will ultimately approve a strategy, based on the government’s trade-off between cost and risk.

    • Associated analytical tools. Models for the analysis of cost and risk trade-offs, assessment of macroeconomic and financial market data, and monitoring of risk indicators.

    • Development of the recommended annual borrowing plan to operationalize the medium-term strategy.

    • Monitoring and reporting on performance against the strategy and its implementation, and reporting on risk indicators and other targets.

  • 3. Transactions processing and recording:4

    • Transactions registration, confirmation, and settlement, as well as associated documentation.

    • Debt servicing, maintenance, and reconciliation of debt transactions.

    • Database management and statistical reporting.

    • Monitoring guarantees and the repayment of on-lending.

Debt managers will often perform other functions, although they may be shared with others in the MoF. These may include:

  • Financial reporting: Includes financial transactions undertaken and the financial and risk characteristics of the debt portfolio. These reports are required for reasons of transparency and accountability, auditing, and policymaking. They may be internal or external, and for the market stakeholders, Parliament or Congress, or the wider public.

  • Risk monitoring and compliance: Debt managers will face a similar range of risks as do cash managers, although often with greater exposure, given the sums involved and the greater impact of poor decisions or process failures. Risk analysis, especially of market risk, is at the heart of DMS, and the management of operational risk (including maintaining business continuity and disaster recovery plans) is particularly relevant.

  • Stakeholder relationship management: As is already clear, debt managers interact with a range of stakeholders. Debt managers, in turn, must develop stakeholder relations policies and a communications strategy. A summary of the contacts are included in Figure 3.2.

  • Policy and advisory services: This function may encompass a wide range of activities, with examples that include advice on domestic debt market development; analysis of public-private partnership proposals; development of policies; and analytical tools to handle contingent liabilities, manage financial assets (other than purely cash), authorize borrowing by subnational entities or state-owned enterprises, buy currency (or hedge currency risks) for other ministries, and manage contractual claims on government.

Figure 3.2
Figure 3.2

Stakeholder Relationships of Debt Managers

Source: Authors’ elaboration.

These functions and interactions are summarized in Figure 3.3. The format is similar to Figure 3.1 above; and the links between cash and debt management, and the common importance of interaction with the financial markets and of risk management, are apparent.

Figure 3.3
Figure 3.3

Main Debt Management Functions and Interactions

Source: Authors’ elaboration.

Cash and Debt Management Interaction

The Importance of Coordination

The importance of close coordination between CDM functions is self-evident.5 In the first place, financing the gross borrowing requirement of government requires choices between instruments: internal or external, short- or long-term, fixed-rate or floating-rate, retail or wholesale, and so on. In particular, these choices will have direct implications on the mix of shorter-term and longer-term instruments; that is, between Tbills and Tbonds.6 Good practice dictates that these choices are made in the context of the medium-term DMS, but they must also take account of market appetite, market volatility, and interest rate prospects. Price considerations are summarized by the yield curve which extends across the full range of maturities, often with anomalies between different market segments, not least between the money and debt markets (below or above one year to maturity, respectively), which have to be well understood by the issuer.

From the supply perspective, government financing choices are made in the context of the profile of financing flows. Most countries have marked quarterly, monthly, and intra-monthly cash flow patterns associated with the timing of tax receipts, payment of wages or salaries, transfers to subnational units, and so on. The pattern may be exacerbated by the in-year timing of debt redemptions. If there is an underdeveloped money market, this pattern has to be reflected in the pattern of Tbond issuance, which also has to be geared to Tbond redemptions.

All these considerations require close interaction between the functions and a common understanding of the market and agreed issuance choices. On a day-to-day basis, there will be further coordination requirements, including:

  • Linkage of issuance dates with redemption dates to maximize the opportunities for investors to rollover into a new issue.

  • Maturity dates, selected to avoid weeks and, especially, days of heavy cash outflow (e.g., salary payments) and, indeed, to target days of cash inflow (the due date for tax payments).

  • Mitigation by debt managers of cash management problems that potentially arise when large Tbonds come to maturity. Debt redemptions tend to be concentrated on only a few days of the year, particularly when successive tranches of Tbonds have been issued to build up their volume sufficiently to support a liquid secondary market. In these circumstances, liability management operations run by the debt manager; for example, debt buy-backs (reverse auctions or bilateral purchases) or Tbond exchanges (switch auctions and conversions) can have the effect of smoothing the redemptions or pushing them forward in time to the benefit of the cash manager.

The potential strain between CDM objectives over whether to issue Tbonds or Tbills when faced with an imminent cash shortage is lessened as the scope for active cash management develops. Debt managers prefer to issue Tbonds with a stable and predictable pattern. Regular issuance reduces market uncertainty and investors can better plan ahead. With a liquid money market, the timing of Tbond sales can be separated from the profile of the government’s net cash flow. It is left to Tbills and other money market instruments to deal with short-term fluctuations. That, in turn, greatly improves the transparency and efficiency of debt management.

As this interaction with the market develops, the coordination of CDM functions becomes especially important. It ensures that the government presents a consistent face to the market. Where two parts of government are interacting with the market, there are risks of giving conflicting signals, adding to uncertainty and potentially distorting the money market. Those staff who directly interact with the market and manage transactions need to build a relationship with individual intermediaries, whether they are selling Tbonds or Tbills, borrowing or investing in the repo or other money markets, or intervening for wider reasons. They may also need to intervene in the money market for debt management reasons, for example to lend Tbonds to unwind a blockage in the repo market, or in the debt market for cash management reasons, as noted above. That requires a single point of contact across a range of CDM operations.

Meeting the Coordination Requirements

The potential institutional structures are outlined below. The requirements outlined above, however, suggest a number of areas where interaction is important.

  • Preparation of DMS: The preferred portfolio structure may need to take account of cash management requirements; for example, for a sufficiently large stock of Tbills to secure sufficient liquidity, allowing cash managers some flexibility in their issuance to accommodate cash flow volatility. In other cases, there may be a need to build a cash buffer. Where domestic market development is an objective, many reform priorities will be common to both debt and money markets. These may include stimulating competition between intermediaries or the development of the repo market (which generates a demand for Tbonds as collateral and provides more options to cash managers).

  • Preparation of annual financing plan: As well as the factors noted above, this has to take account of the in-year profile of cash flows. The timing of debt redemptions or major payments may be particularly relevant. If the debt manager wishes to frontload issuance for prudential reasons, cash managers may be responsible for cost-effectively managing any cash surpluses that arise.

  • Monthly or quarterly issuance plan: Debt managers properly want to announce their issuance plans in advance in the interests of transparency, predictability, and to reduce market uncertainty. Cash managers, on the other hand, want to retain some flexibility to respond to unanticipated cash flows. The trade-off may take the form of a Tbill issuance program that is less specific than the Tbond program when they are announced to the market, with the Tbill program perhaps indicating ranges rather than firm amounts.

  • Short-term responses to market volatility: Tbonds are often viewed as the main instrument for financing the annual borrowing requirement (after allowing for external loans and credits and, possibly, external Tbond issuance), with Tbills essentially used to reduce cash flow volatility. Within the year, there is much more flexibility to respond to changes in market demand, interest rates, or cash flows.

  • Interaction with the central bank: Operational understandings with the central bank—the timing and type of market interventions, instruments used, use of intermediaries—are important to CDM, and should normally be discussed jointly with the central bank. There are more specific requirements for cash managers (access to the TSA, scope for managing temporary surpluses with the central bank, interest on accounts) and the information flows needed will be different. In the supply of services, the central bank may be fiscal agent for Tbill and Tbond auctions, although its role as banker may be more relevant for the cash manager.

This last point hints at the scope for administrative savings that effective coordination can generate. Both functions have some skill requirements in common—in particular, an understanding of financial markets (i.e., transactions in securities; structure of demand; and market infrastructure, which includes legal and taxation frameworks, trading conventions, and settlement and custodial systems). Many of these skills are often scarce in the public sector and are expensive to recruit and retain. It makes sense, from a cost-effectiveness perspective, to avoid duplication of the capability.

There is a similar potential for administrative savings in the data management and operational risk management tasks. Both functions manage important data and should have high regard to operational risk. It is not necessary to use the same database to manage the transactions and stock data of Tbonds and Tbills, although many countries have found it sensible to do so. The database must ideally be capable of managing a range of liabilities (i.e., external and domestic securities, loans, and credits) and, at a minimum, it should be capable of storing liability data securely; capturing selected market data for valuation purposes; making data available for analysis; projecting debt servicing payments; and triggering the payments in due course. In practice, however, many countries lack a single database that covers all these functions. Both the main propriety debt management databases in wide use (CS-DRMS of the Commonwealth Secretariat and Debt Management and Financial Analysis System (DMFAS) of the United Nations Conference on Trade and Development (UNCTAD)—cover a range of database functions in their latest versions). They are, however, still primarily recording systems and not transactions processing systems. Moreover, different countries may not have upgraded their databases, and although some use a mixture of systems with potential additional capabilities, they are not always successfully realized and may still be inadequate for more sophisticated debt management purposes, let alone cash management.

Cash management potentially involves many more daily transactions than debt management.7 That, in turn, argues for a more sophisticated transactions processing system that integrates the front-, middle-, and back-office capabilities, and ideally is electronically connected also to the relevant settlement systems so that individual transactions are processed seamlessly, thus reducing the risks associated with multiple interfaces. A highly schematic summary of the key CDM system requirements is presented in Figure 3.4, indicating the shared and separate requirements. All these systems require careful management and skilled IT staff. Back-up protocols and disaster recovery sites are essential, along with a business continuity plan that extends to all processes and systems (Storkey, 2009). The sharing of such facilities is sensible financially and, indeed, in many cases disaster recovery facilities will extend to the whole of the ministry’s operations.

Figure 3.4
Figure 3.4

Shared Information technology Systems

Source: Authors’ elaboration.

These considerations are among those that have led to the integration of CDM functions in many developed countries. This also applies increasingly in middle-income countries.

Organizational Structures

Integrated Cash and Debt Management Unit

The organizational structure of a modern DMU is based on the familiar separation of responsibilities between the front, middle, and back office. A summary of the key functions is shown in Box 3.1.

The analysis and negotiation of loans and credits from multilateral and bilateral lenders is mostly a front office task. Once a loan has been agreed, however, the information is passed to the back office to manage the data and to ensure that disbursements are recorded and the debt is serviced by the due date.8 There is not always much choice for the terms and conditions of loans and credits but, where possible, they should be negotiated consistently with the DMS and within a framework set by the middle office. For these reasons, the teams handling external loans and credits should be part of the integrated DMU, although it often makes sense for them to be a separate management unit within the front office; the type and flow of work is very different from that involving securities markets.

Key Functions of a Modern Debt Management Unit

  • Senior management (supported by internal audit and compliance).

  • Front Office: Primary issuance and execution, internal and external, and all other funding operations, including loans and credits and secondary market transactions (debt and cash).

  • Middle office (1): Policy and portfolio strategy development and accountability reporting.

  • Middle office (2): Internal risk management (policies, processes, and controls).

  • Back office: Transaction recording, reconciliation, confirmation, and settlement; maintenance of financial records and database management; debt servicing.

This structure still allows for the contracting out of some functions (e.g., to the central bank as fiscal agent for the handling of auctions). Some countries contract out debt registration and debt servicing functions, although the debt office retains policy control.

Cash management functions integrate comfortably with this basic structure, although there may be some differences in practice, depending on the nature of the cash management responsibilities. The three basic functions identified above are:

  • Cash flow forecasting: Cash flow forecasting may be shared with others in the MoF or treasury. There may be a separation between those compiling the “above the line” (i.e., revenue and expenditure) forecasts, which may fall to those officials monitoring execution of the budget, who will often also be the ones drawing on information from the revenue and spending departments; and those projecting “below the line” transactions (debt and other financing operations), where the cash and debt managers may be better placed. Integrating the forecasts would normally be a middle-office activity in the DMU. The middle office would pull together the different sources of information and present forecast sensitivities or what-if scenarios to senior management for decisions on Tbill issuance or other borrowing and lending. The more active the cash management, however, the more important it will be to have forecast support close to the front office. Thus, although the weekly decision parameters may be driven by the middle office, if the front office’s task is to take account of the emerging flows within the day, it will need an up-to-date picture of what is happening to government transactions. Someone in the front office will probably need to be monitoring the position and receiving in-day updates.

  • Monitoring and accessing the TSA: Real-time monitoring of the TSA is probably part of the front office function noted above. There will be other roles, however, within the DMU. Daily reconciliation may be necessary, which can sometimes be done within the transaction processing system, but otherwise manually—usually by the back office or the accounting support team. Underlying analysis of the required cash buffer, as well as retrospective analysis of forecasting performance, would be a task for the middle office. If there was any question of drawing on other reserve funds, that policy question would also fall to the middle office, which will tend to have the links with the macroeconomic and fiscal policy departments within the ministry.

  • Financial market interaction: In this case, the location of the function is unambiguous. As already stressed, it is important that there is only one interface with the financial market. There may still be cash or debt market specialists within the front office, but the management should be unified at that level.

The slightly different systems requirements of debt and cash managers have already been noted. They should be managed, however, as an integrated function by the IT department, whether that is in the DMU or the wider ministry. Of the other roles, however, that cut across the front, middle, and back offices within the DMU, it is worth noting that there will be new capabilities required of the governance and risk management frameworks. Cash management may require more frequent decisions, with less time for analysis and consultation than does debt management; and cash management objectives are qualitatively different, with the focus not so much on the cost-risk trade-off, but also on ensuring that cash is available when needed—whatever else happens. The decision-making structure must support this; that means that the DMU—and, in turn, its front office—must have the delegated authority to make short-term borrowing and lending decisions without prior approval of a minister or senior official(s) elsewhere in the ministry (although they periodically may want, of course, to set or review the parameters of the delegation). The potentially wider range of counterparties—particularly when investing surplus cash—adds a major dimension to credit risk analysis, and the additional operational risk management requirements have already been mentioned.

The functions of the integrated DMU are summarized in Figure 3.5. The figure distinguishes between those that fall into the front-, middle-, back-office split and those that are “horizontal” in nature.

Figure 3.5
Figure 3.5

Functions of an Integrated Debt Management Office

Source: Authors’ elaboration.

Separate Cash and Debt Management Functions

There are many countries that maintain separate functions, although the boundaries between them differ from place to place. The cash management function will often be fully integrated within the treasury. The management of cash will have developed alongside the management of government payments, and the emphasis will have been on monitoring cash balances—whether held in the TSA or elsewhere—ahead of releasing budget authority or processing payments. That naturally also becomes the location for cash flow forecasting, although as the possibility of more active cash management comes into focus, the coordination issues addressed in this note also become more apparent. How onerous they are in practice may depend on where the debt management function has developed. In some cases, it is part of the treasury, and coordination is relatively straight forward. In others, however, the policy importance of debt management has been emphasized, perhaps distinguishing it from the more “executive” role of the treasury, and it has developed closer to the macro-fiscal policy functions at the heart of the ministry.

As already argued, at a minimum:

  • There should be a single interface with the market for market transactions, selling Tbonds and Tbills, and investing short-term cash surpluses. That also means sharing front-office systems, such as those from Bloomberg and other data providers. There have been unfortunate examples where a treasury is building a front office to manage Tbills with an entirely separate part of the ministry issuing debt; this is wasteful and damaging. The single front office may be part of the DMU (perhaps the most usual) or the treasury, or both may use the central bank as agent.

  • Support services should be shared, particularly those where there are substantial economies of scale (e.g., in relation to disaster recovery).

  • The operational risk framework should take account of the interaction between the functions, particularly if the institutional separation means that there are more manual or semi-manual interfaces.

  • There should be frequent and systematic exchanges of information between the debt managers and cash managers with regard to the profile of cash flows, issuance plans, and options.

Coordination of management decision making is also of great importance. There are two dimensions to this: strategic and tactical.

The DMS, typically prepared initially within the middle office of the DMU, will usually take account of the role of Tbills in the debt portfolio. If short-term cash assets are significant, the DMS should consider the structure of assets, as well as liabilities, in an asset and liability management framework. In particular, by matching explicitly the risk characteristics of various financial assets and liabilities on the government’s balance sheet, the exposure to economic shocks can be greatly reduced. The DMS should be endorsed by a Public Debt Committee or similar body, comprising all the main macro-fiscal and financial policy functions. Its main roles are the formulation of strategic debt management policy objectives; the mandating of those responsible for strategy execution; and the setting of targets and objectives and subsequent monitoring of performance.9

Decisions on short-term cash management must also take account of the wider policy framework. Interaction with the debt management issuance program is an obvious aspect of that, although there may be other decisions to be made; for example, transferring cash between accounts, bringing forward asset sales, or—as a fallback—restraining expenditure appropriations. For these reasons, some form of Cash Coordinating Committee (CCC) is usually recommended. Meeting frequently, its main role would be to review recent developments and the latest cash forecasts, and to decide on action in the period ahead. As stressed above, in an active cash management environment, its role is not to approve individual transactions (although it may decide details of forthcoming Tbill auctions), but to agree to the parameters and policy framework within which those transactions are made. In doing so, the CCC would bring together relevant functional responsibilities. These might include the central bank, partly for its views on market developments, but the CCC can also be the vehicle for passing the latest cash flow forecasts to the central bank which, in turn, are an important input into its liquidity forecasts. Larger spending ministries and the revenue authority could attend if judged useful. Summary details are presented in Box 3.2. If it is necessary to hold back appropriations or cut approved budgets, the detailed decisions should not normally be made by the CCC; that is a role for the budget directorate.

The functions of the split organization and the associated coordination structures are summarized in Figure 3.6. It assumes that the DMU is not part of the treasury.

Figure 3.6
Figure 3.6

Cash and Debt Management Coordination

Source: Authors’ elaboration.

Cash Coordinating Committee

Members

  • Head of the Treasury (chair)

  • Heads of (or senior representatives from):

    • Debt Management Directorate

    • Budget Directorate

    • Fiscal Policy Directorate

  • As required: representatives of

    • Central Bank

    • Revenue Administration

    • Larger line ministries

  • Secretary from Cash Management Unit

Responsibilities

  • Review cash flow out-turns and the comparison with forecasts

  • Review cash flow forecasts for the period ahead

  • Decide on the action needed to ensure cash adequacy over the period ahead, making recommendations accordingly.

Meetings

  • Weekly

International Experience

The variety of institutional arrangements for DMUs that countries have adopted has expanded considerably in the last two decades. They can be categorized according to their degree of institutional independence:

  • Offices within the government: Departments operating within the Ministry of Finance (or equivalent) in which the DMU reports directly to the minister: Argentina, Belarus, Brazil, Bulgaria, Canada, Colombia, Czech Republic, Indonesia, Italy, Japan, Korea, Latvia, Lithuania, Macedonia, Mexico, New Zealand, Peru, Poland, South Africa, Turkey, United States, Uruguay, and Vietnam.

  • Established as an agency by the government: Established by the government as an agency, often formalized as a DMO, with substantive operational autonomy in relation to the MoF (or equivalent): Australia, Belgium, France, Netherlands, Nigeria, and United Kingdom.

  • Established by a specific law: Agencies whose functions and operational frameworks are defined by a specific law: Austria, Finland, Ireland, Portugal, Slovakia, Sweden, and Thailand.

  • Established under the General Corporation Law as a public company. A public company established under the general law of companies with no specific provisions that distinguish it from other companies in the private sector (e.g., Germany and Hungary), while operating within the framework of policies set by the MoF (or equivalent).

In all these cases, “independence” has to be qualified. Many DMOs, including some that are fully integrated in the ministry, have substantial operational and managerial independence. They should, however, all be executing strategies that have been agreed at senior political levels, potentially by or on the recommendation of a Public Debt Committee as noted above. The creation of the DMO helps to support this distinction between the strategy, which is properly approved by ministers, and its implementation—which is a more technical process and can suffer from undue political influence and the market uncertainty that it can generate.

The creation of DMOs dates back to the eighteenth and nineteenth centuries,10 but it was during the 1990s and 2000s that a wave of organizational reforms took place in many countries. Countries realized that there is a performance gain from the integration and the possibility to use the reform as a way to improve professionalism, develop a more strategic approach, and generally enhance the power and tools available for managing government resources. A DMO also supported the separation between debt management and fiscal policy from monetary policy—an important consideration, especially in the Eurozone countries.

There are counter arguments to the creation of an agency. The more institutional distance there is between the MoF and the DMO, the stronger the governance framework must be. There is otherwise a problem of the DMO developing its own agenda that is not fully aligned with that of the ministry: the “principal-agent” problem. A strong governance structure, however, has to be managed; in effect, the ministry would need to retain a competent “intelligent customer” function to ensure that the DMO was given clear objectives and targets, and to monitor its performance against them. That in turn, however, would risk spreading skilled resources yet more thinly. For these reasons, it is usually recommended that, although the agency structure has many advantages, it should preferably remain institutionally part of the MoF or close to it, rather than entirely separate. Table 3.1 shows a sample of DMOs with respective dates of creation or major structural reforms.

Table 3.1

Countries with Debt Management Offices: Date of Creation or Restructuring

article image
Source: Authors’ elaboration.

It is true that no DMO is exactly the same, but all of them do have clearly identified divisions that take care of front-, middle-, and back-office functions. In general, they also have additional divisions that respond to specific issues, such as a division that deals with loans from international financial institutions as donors. The structure of the French Treasury, for example, follows to a large extent the typical DMO model. It has a very compact structure and a staff of only 38 people (Figure 3.7).

Figure 3.7
Figure 3.7

Organizational Structure of the French Treasury

Source: French Treasury. Available at www.aft.gouv.fr.

The internal governance structures of the DMO are also important. Arrangements are needed for internal policy making, delegation, and communication. Risk management policies and procedures should be established, consistent with objectives and best market practice. The vertical front-, middle- and back-office structure needs to be supported by horizontal mechanisms for business planning, the operational risk management framework, and capacity building, including IT, training and staff conduct. Many DMOs establish internal decision making and reporting structures that cover risk policies and audit, as well as the main operational areas of CDM.

Other circumstances that seem to influence the structure of the CDM units are related to the development of the financial market and a more active cash management. There are three main stages of development to consider: (i) no domestic government debt markets; (ii) domestic bond markets that are not deep/liquid; and (iii) domestic markets that are deep/liquid. In the first case—and to an extent the second—countries rely mostly on donations and loans provided by bilateral and multilateral institutions. In these situations, the need to have both functions in the same units is less clear-cut. When debt management is related to the preparation of project loans, knowledge of the financial market is not a precondition. In the third case, the government has more debt options; should develop a DMS and implement an annual financing plan that is consistent with the DMS; and build a yield curve to give to market participants a benchmark for pricing other instruments. On active cash management, the situation is similar. Unless the domestic financial market is robust and active, there are few options to investment by the treasury. In addition, all these circumstances require additional technical skills not always available in the public sector. It is in this active environment that close coordination is essential, and the benefits of an integrated unit become especially evident.

Functions and Organizational Structures in Latin American Countries

This chapter applies the framework discussed above to categorize the way CDM functions are organized in 17 LAC countries.11 We applied a questionnaire to collect information on the main functions, mechanisms, and structures of coordination; reports and systems available for recording transactions; characteristics of the issuance of Tbills and Tbonds; size and qualification of the workforce; provision for legislative authorization to issue debt; and advantages and disadvantages of an integrated structure.

Distribution of Functions

There are only three countries in Latin America that operate under an integrated DMO model (Brazil, Colombia, and Peru); however, in none of them have the reforms yet been fully implemented. In the other countries, the Treasury and DMOs operate under independent structures and rely on coordination mechanisms to assure consistency of policies and procedures. Table 3.2 shows how typical CDM functions are distributed in each country. In both models, coordination is very relevant, but it is even more relevant when functions are separated. Some of the issues relevant for coordination are related to the relationship with the market (as emphasized above, this can be complicated and risky when the market has to relate to multiple entities dealing with Tbills and Tbond issuance), a definition of the most efficient volume and frequency of auctions (less optimal decisions may be made because of insufficient coordination), and the reporting of transactions (debt amounts and cash balances, or debt market and money market transactions, may be found only in separate reports, making it difficult for investors and other external analysts to assess, comprehensively, government finances).

Table 3.2

Main Functions of Cash and Debt Management in Latin American Countries

article image
article image
article image
Sources: Treasury and debt management units of each country.

Six out of 14 major functions are usually performed by treasuries and 8 by DMUs.12 Typical treasury functions are: (i) cash flow forecasting and planning; (ii) monitoring and management of the TSA; (iii) investment of excess cash; (iv) issuance of Tbills; (v) payment of debt service; and (vi) payments flowing from execution of the budget (and, in some cases, management of the execution authority). Typical DMU functions are (i) development of the annual financing plan; (ii) review and execution of the annual financing plan; (iii) issuance of Tbonds in the domestic market; (iv) issuance of other securities in the domestic market; (v) market risk management and analysis; (vi) relations with the market; (vii) recording of financial operations; and (viii) issuance of accounting and statistical reports.

Annual Financial Plan

Annual financial plans are important to define the strategy to issue debt. It is important that these plans are well integrated with the medium-term fiscal framework and the budget process. In 15 countries the DMU is responsible for developing the annual financial plan (ARG, BOL, BRA, CHI, COL, DR, ECU, HON, MEX, NIC, PAN, PER, PRY, SLV, URU). In HON, it is developed with some degree of coordination with its Treasury and in CHI and GUA, in coordination with the Budget Directorate, Dirección de Presupuesto (DIPRES). In CRI, the Treasury prepares this plan. The DMU is also in charge of the in-year review and execution of the financing plan in the same 15 countries and, in 3 of them, in coordination with the treasury (ARG, MEX, and NIC). In CHI, the Budget Unit—jointly with the treasury—performs these tasks.

Issuance of Tbills

Issuance of Tbills is very important for short-term liquidity purposes. In a few countries, it is not the same unit that is responsible for issuing Tbills and Tbonds in the domestic market. Furthermore, in a few countries, the treasury has been assigned to undertake both tasks. Tbonds are issued in the domestic market by the DMU in 14 countries (ARG and PRY, jointly with the treasury; BOL, BRA, CHI, COL, DR, ECU, GUA, HON, MEX, PAN, PER, URU) and in 3 countries, they are issued by the treasury (CRI, NIC, SLV). In SLV, this task is performed by the Treasury, jointly with the Central Bank. In 10 countries, Tbills are issued in the domestic market by the DMO (BOL, BRA, COL, DR, ECU, HON, MEX, PAN, PER, URU), in 5 countries by the treasury (ARG, CRI, GUA, NIC, SLV), and in 2 countries (DR, HON) by both units. In PRY, they were authorized in 2014 but have not been issued yet. The split responsibilities for Tbill and Tbond issuance in the countries mentioned can potentially cause confusion in the market in terms of government strategies and policies.

In 9 countries, the DMO issues or transacts other securities (including repos) in the domestic market (ARG, BOL, BRA,13 CHI, COL, DR, PAN, PER, URU). In 3 countries (HON, CRI, NIC), the treasury does this; in 2 (ECU, MEX), however, it is the Central Bank and, in CHI, it is the Budget Unit.

Cash Flow Forecasting and Cash Planning

Cash flow forecasting is an essential activity of the treasuries and is important to anticipate cash needs during the upcoming months. In 16 countries, cash flow forecasting and planning and the monitoring and management of the TSA are the responsibility of the respective treasury (ARG, BOL, BRA, COL, CRI, DR, ECU, GUA, HON, MEX, NIC, PAN, PER, PRY, SLV, URU). The only exception is CHI, where these activities are undertaken by DIPRES.

Investment of Cash Surplus

Investment of cash surplus is a way in which to obtain some financial benefits from available resources. Many countries invest their short-term cash surpluses only at the central bank or at the public commercial bank where the TSA is maintained, although a few do not make investments at all. In some countries, there are provisions that prohibit the treasury from performing short-term investments. Sacrificing the capacity of treasuries to invest available financial resources might seem to undermine the objectives of a TSA, and it can prove difficult to attract autonomous entities to deposit their resources in the TSA.

Cash surpluses are invested by the treasury in 12 countries (ARG; BOL; CHI, jointly with the Budget Unit; COL; CRI; DR; HON; MEX; NIC; PAN; PRY, in progress; PER), and in ECU by its Central Bank. In GUA, investment is prohibited by law, and in URU and SLV, there is no active investment. In BRA and ECU, all TSA balances remain invested at the Central Bank.

Risk Analysis and Management

A good risk analysis is fundamental to identify potential risks and manage the exposure of treasury and DMUs in terms of investing liquidity or issuing debt.

The function of risk analysis and management has been allocated in most countries to the DMU. Risks are analyzed and managed by the DMU in 10 countries (ARG, BRA, COL, GUA, MEX—in some cases, in coordination with the treasury—DR, NIC, PAN, PER, URU); in BOL and HON by the DMU and the treasury; by the Central Bank in ECU; and by the Budget Unit in CHI. In PRY, it is not performed.

Market Relations

It is essential to a modern treasury and DMU to define a transparent and close contact with the financial market. Some of the units in Latin America are considered very efficient and transparent.

The relationship with the market is, in 13 countries, the responsibility of the DMU or the integrated unit (ARG, BRA, CHI, COL, DR, GUA, HON, MEX, NIC, PAN, PER, PRY, URU). In MEX and NIC, this task is performed in coordination with the treasury. In CRI and BOL, the task has been assigned to the treasury and to the Central Bank in ECU.

The Institute of International Finance (IIF) assesses investment relations according to two main indicators: (i) overall assessment of investor relations and data transparency practices; and (ii) assessment of data dissemination practices. The situation in 11 countries of the region is shown in Table 3.3.

Table 3.3

Investor Relations and Data Dissemination Practices

article image
Source: Institute of International Finance (2014).

Recording of Financial Operations

Recording financial operations is usually the responsibility of the DMU. This activity is important in terms of accountability and transparency. The recording of financial operations is a function assigned to the DMU in 11 countries (ARG, BRA, COL, ECU, GUA, HON, NIC, PAN, PER, PRY, while in MEX it is also by the Treasury). In another 3 countries (BOL, CRI, DR), the treasury bears this responsibility; in CHI, it lies with the Treasury and Budget Unit; in URU, with the Central Bank; and in GUA, with the Accounting Unit.

Accounting and Statistical Reports

The maintenance of accounting and statistical reports is also important, but these functions are performed in Latin America by differing units. They are assigned to the DMU in 12 countries (ARG, BRA, CHI, COL, CRI, ECU, GUA, MEX, NIC, PAN, PER, URU); to the treasury and the DMU in 6 countries (ARG, HON, MEX, NIC, PRY, URU); in 4 countries to the Accounting Unit (BOL, DR, GUA, SLV); and in PAN to the Office of the Comptroller.

Payment of Debt Service

The same diversity is encountered in relation to the payment of the debt service. The payment of debt service is assigned to the treasury in 7 countries (CHI, CRI, ECU, MEX, PAN, SLV, URU); in 5 countries to the treasury and DMU (DR, GUA, HON, NIC, PRY); in 5 countries to the DMU (ARG, BOL, BRA, COL, PER); in URU to the Central Bank; and in PAN to the Accounting Unit and Office of the Comptroller.

Budget Execution

The budget is executed by the treasury in 7 countries (BOL, CHI, HON, GUA, MEX, URU, while in SLV also by the Budget Unit); by the DMU in 7 countries (ARG, BRA, COL, CRI, ECU, PAN, PRY); by the budget unit in 4 countries (DR, GUA, PER, SLV); by the executing units in NIC; and in GUA by the Accounting Unit.

Organizational Structure

The size and complexity of the organizational structures of the treasury and DMUs are quite variable in Latin America (Table 3.4). Integrated DMUs have a workforce of between 110 and 150 staff; treasury units between 50 and 100 staff; and the smallest offices from 8 to 31 staff. Large DMUs have between 50 and 75 staff, and the smallest from 8 to 34 staff.

Table 3.4

Number of Staff Working on Treasury and Debt Functions in Some Countries in Latin America

article image
Notes:

Treasury and debt functions are carried out in an integrated unit.

Does not include staff that perform other functions, such as accounting and oversight of subnational governments. NA – not available.

Coordination Arrangements

Coordination between CDM units is very important considering the need to finance budget execution in all units of the government. International experience favors formal organization of committees with clear functions and mandates, periodic meetings, and participation of the main areas responsible for the definition of cash plans, debt strategy, Tbond issuance, exchanges of information, and results assessment.

There are formal and informal coordination committees in almost all countries in LAC. In 13 countries (BOL, BRA, CHI, COL, CRI, DR, GUA, HON, MEX, NIC, PAN, PER, PRY), there are formal committees to coordinate decisions and policies relating to CDM. The rest of the countries rely on informal mechanisms and meet when needed (ARG, URU). Some of the committees deal with other issues, such as monetary policy and budget. The main functions of such committees are discussed in Box 3.3.

Functions of the Main Coordination Committees

Twelve countries have formal coordination committees and two have informal committees (Argentina and Uruguay), relating to treasury and debt functions.

ARG: Committee of informal coordination of the Secretariat of Hacienda and the Secretariat of Finance, which is responsible for managing the financial management system.

BOL: Committee which advises the Deputy Minister of the Treasury and Public Credit in analyzing cash flow forecast, financing needs, and the debt situation. Participation at weekly meetings by the Department of Planning and Operations of the Treasury (Dirección General de Programación y Operaciones del Tesoro) and the Public Credit Department (Dirección General de Crédito Público).

BRA: Has a Debt Management Committee (as well as a Budget Execution Committee (Junta de Programação Orçamentária). The Debt Management Committee has monthly meetings and is responsible for the cash forecast of the public debt, taking into account the debt strategic plan (annual borrowing plan, other budgetary revenues for debt payments, indicators of composition and maturity of the debt, and others), monitoring the debt strategy plan, and defining the auction calendar. The Budget Execution Committee defines bimonthly budgetary and financial ceilings for the central government.

CHI: The Coordination Committee between the Treasury, Budget Department, and MoF, which advises the Minister of Finance, discusses at monthly meetings budget, Treasury, and debt-related issues.

COL: The Treasury Committee, which holds monthly meetings, discusses cash management; debt financing and market analysis; revenue collection; budget execution; among other related functions of the Public Credit and National Treasury Department (Dirección General de Crédito Público y Tesoro Nacional). There is another inter-institutional Treasury Committee between the MoF and the Central Bank, where debt and cash matters are involved; its main objective is monetary coordination and issues related to the management of Treasury balances, operations affecting domestic debt market liquidity, and the issuance of Tbonds for monetary policy purposes.

CRI: The Auction Committee has weekly meetings and brings together representatives of the Office of Public Credit, Treasury, and the Office of the Minister of Finance.

DR: The Council of Public Debt is a high-level committee involving the Secretary of Finance, Governor of the Central Bank, and Secretary of Economy, Planning and Development. It meets twice a year. The functions of the Council are to propose the debt policy and strategy, propose the debt ceiling, recommend the financial characteristics for new debt, and recommend the maximum amount of collateral and guarantees.

GUA: The Programming and Execution Committee advises the Minister of Finance on matters of budget, accounting, Treasury, public debt, and fiscal analysis. It meets twice a month and defines the allocation of payment authorization for entities that implement the budget.

HON: Committee to advise the Deputy Minister of Finance on cash and debt management. It holds monthly or bimonthly meetings.

MEX: The Technical Investment Committee and Working Committee advise the Secretary of the Treasury on issues related to debt, budget, revenue collection, and cash planning. The Technical Investment Committee meets annually and the Working Committee monthly.

NIC: The Financial Operations Committee is composed of the Minister of Finance and the General Directorates of the Treasury, Public Credit, and Legal Unit, and it meets weekly. It is responsible for approving the annual borrowing plan; defining Tbond issuance and the auction calendar; approving the results of the auctions, policies on financial investment, and the bids for financial investment; coordinating with the Central Bank’s Tbond issuance; and submitting quarterly reports.

PAN: The Treasury Committee advises the Minister of Finance and meets weekly. Participation at the meetings includes the units responsible for Treasury functions, debt, revenue collection, and budget.

PER: The Assets and Liabilities Committee (ALC) and the Cash Committee (CC). The ALC reviews and evaluates the guidelines, policies, principles, strategies, and methodologies to optimize the overall management of financial assets and liabilities of the government (the operating rules of ALC are in the process of approval). The CC establishes the conditions for a balanced budget for monthly income and expenses of the Treasury; approves the cash budget; coordinates and agrees on the actions necessary for the implementation of the cash budget; performs a comprehensive monthly monitoring of the implementation of public expenditure; and assesses the implementation of the budget.

PRY: An Interagency Advisory Team analyzes the current macroeconomic conditions to define the economic, monetary, and fiscal policy and short- and medium-term projections; analyzes market conditions (maturity, interest rates, issuance calendar)—domestic and international—to consider the issue of Tbonds or guarantees provided by the Treasury; and advises the Minister of Finance on other financial and technical issues. It meets before a Tbond issuance is scheduled.

SLV: Cash and debt management are handled separately. Short-term debt is coordinated, involving the Central Bank, Treasury, and Directorate General of Public Credit.

URU: There is no formal committee. Coordination is managed through an ad hoc group that meets periodically, which brings the heads of the units of Debt Management, Treasury, and the Macro-Fiscal Advisor to discuss issues of common interest.

In 5 countries the committee is a cash or treasury committee. In other countries, the committee is an “auction” committee (2 countries), an Investment Committee (1), a Budget Execution Committee (1)—with participation of the Presidency and the Ministers of Finance and of Planning—or a Debt Committee (2). In PER, there are two committees: Cash Committee and Assets and Liabilities Committee (ALC); as there are also two in MEX—the Technical Investment Committee and Working Committee.

In 9 countries, the committee has advisory and decision-making functions; in 6, it has only decision-making functions; and in one (PER), it has only advisory functions. When it has advisory functions, it provides advice mostly to the Minister, and in 2 cases, also to the Vice Minister (Viceministro); in one country, only to the Vice Minister (BOL); and in CHI to the Treasury and DIPRES.

In one country, the committee is chaired by the Treasurer and in another by the Deputy Treasurer; in 6 countries, it is chaired by the Minister of Finance, and in 3, by a Vice Minister. In PER, the Minister chairs the ALC and the Vice Minister the CC; and in MEX, the Minister chairs the Technical Investment Committee and the Vice Minister chairs the Working Committee. The Secretariat of the committees is supplied by the treasury in four countries, while in PER by the Directorate of Analysis and Strategy for the ALC and the Vice Minister of Finance for the CC. In NIC, it is chaired by the Legal Advisory Directorate.

The frequency of meetings of the committees varies from weekly to biannually. The committees meet on a monthly basis in 7 countries (BRA, CHI, COL, HON, PAN, PER, while in MEX, the Working Committee); weekly or biweekly in 4 countries (BOL, CRI, DR, GUA); every six months in 2 countries (DR, URU); and in SLV and PAR, only when issuance of Tbills is required.

The committees in most cases do not include some of the key stakeholders, such as the Tax Administration Agency, the Central Bank or the Budget Directorate. Regarding the main units/functions represented in the committees, besides the Treasury and DMUs (in 9 countries), the Tax Administration Office is represented in only 2 countries (COL, MEX—but in COL, it participates as a guest); the Central Bank in 2 countries (COL, DR—but in COL, in a separate committee where only the Treasury and DMU meet with the Central Bank); and the Budget Unit participates in 4 countries (CHI, COL, GUA, MEX).

Frequency of Issuance of Tbills and Tbonds

Seven countries do not issue Tbills to meet temporary liquidity needs. These are (i) BOL, BRA (issuances only for reasons of monetary policy and preference of the market); (ii) CHI, COL (only for reasons of monetary policy); (iii) DR uses credit lines to finance short-term liquidity needs; and (iv) NIC and PRY do not issue. PRY has recently approved legal authorization to issue Tbills, but no operation has been executed so far.

The frequency of issuance of Tbills varies from weekly to yearly. It is weekly in BRA, MEX, and SLV; fortnightly in PER and CRI; and quarterly in COL. In HON and ECU, frequency is not predetermined and, in GUA, in general, it is only once a year.

The duration of the Tbills varies from 1 to 12 months. They are 6 months and 12 months in BRA,14 and one year in COL, GUA, ECU, and HON. In COL, the Tbills are initially 364 days and reopened in four auctions when they come to 90 and 180 days of remaining maturity. In CRI, Tbills are offered twice a month with durations of no less than nine months. Tbills are from 1 to 12 months in PAN; 3, 6, 9, and 12 months in PER; and 28, 91, 182, and 364 days in MEX. In SLV, the terms are from 30 days up to 360 days.

In most of the countries, the timing of the issuance of Tbonds does not necessarily coincide with periods of temporary cash shortages; it has other drivers.

  • BOL: Weekly funding needs are assessed to cover temporary shortages, followed by a decision on the amounts of Tbills to be issued. As there have been no liquidity problems, Tbonds ranging from 2 to 50 years have been issued to finance the budget.

  • BRA: The Treasury issues securities according to the medium- and long-term strategy DMS. The resources from these issuances are integrated in a subaccount of the Public Debt Office, which constitutes the debt cushion. Current policy is to have a cushion equal to at least three months of debt payments (principal and interest). Auction dates are predefined in the annual borrowing plan at regular intervals (specifically, auctions of securities with fixed rates occur weekly, irrespective of the dates of debt payments). As liquidity in the TSA is very high, there has been no need to issue short-term liquidity Tbills to support budget execution.

  • CHI: Dates of Tbonds are defined to match with temporary shortages of cash in relation to the issuance of Tbonds denominated in local currency. In the case of external issuance, the strategy is to have benchmarks at low cost.

  • COL: Short-term Tbills are issued for monetary contraction operations, whose issuance depends on the needs of monetary policy. The market is informed beforehand of the calendar of the auctions and amounts. These Tbills can also be issued for temporary operations, according to the liquidity needs of the Treasury. Although there is currently no outstanding balance, when such auctions are required, a calendar is issued beforehand. For Tbonds, there is a predetermined schedule, which follows a calendar previously published to the market. The securities issued are benchmarks of the yield curve for the medium and long term (i.e., those with maturities of 5, 10, 15, or 20 years).

  • CRI: Issuance of Tbonds matches refinancing needs, and the shaping of the portfolio makes use of other instruments (swaps, reverse repos).

  • DR: Has used credit lines to finance short-term cash needs.

  • ECU: The primary objective of Tbond issuance is to cover the financial gap identified in the annual budget and, in executing the financing plan, to adjust it to the needs of budget execution based on the cash forecast.

  • GUA: The strategy is to place all Tbonds when there is adequate demand and the interest rate is adequate with the understanding that all required resources will be raised during the fiscal year.

  • HON: The issuance dates coincide with the periods in which a cash shortage is expected.

  • MEX: Tbills are issued to cover temporary cash shortages. A flexible approach has been adopted in relation to the weekly amounts issued, while observing the quarterly target.

  • NIC: Tbond issuance dates match the maturities of the Tbonds to reduce refinancing risk on the domestic public debt.

  • PAN: Tbonds are issued in different amounts, taking into account the temporary cash needs.

  • PER: The main strategy is to perform regular issuance of Tbonds in local currency (referred to as “regular auction”), to provide predictability to market participants, ensure stability of the sovereign curve yields, and preserve financial stability, including during times of uncertainty. To complement this strategy, the Treasury can carry out “special auctions,” which depend on market conditions and should be in line with overall strategic objectives. These auctions can also reflect the specific liquidity needs of the Treasury to cover eventual mismatches between revenues and payments.

  • PRY: Long-term bonds are issued only to finance capital investments (golden rule).

  • URU: The government has adopted a 12 month, prefinancing policy covering the debt service for at least one year.

Treasuries in Latin America have aimed to ensure predictability on the issuance of Tbills and Tbonds. With the exception of ECU, PAN, HON, and countries that do not issue Tbills, the other countries adopt a predefined issuance calendar. Predictability has been achieved because (i) issuance follows a debt strategy that covers at least one year; (ii) the financing program aims to give security to the investors; (iii) issuances are close or coincide with the expiration dates of other bonds to minimize refinancing risks; (iv) some countries have predefined their debt to minimize market volatility; and (v) issuance is done when demand is large and interest rates low.

Legislative Authorization

The ideal situation is when legal authority is given to the MoF to issue debt, according to a lump sum approved in the budget. In many countries, however, the framework is not flexible and, for each operation, it is necessary to have specific legislative authorization. The lack of proper legal authorization to manage debt issuance can be a risk in terms of adequate availability of resources during the year and it can be an impediment to manage the debt more efficiently.

In most countries, the legislature has to authorize the issuance of debt, although in more than half of the countries, this authorization is not required for the issuance of Tbills. In most countries, debt issuance is authorized by the budget law in a lump sum (ARG; BOL, only the external debt; BRA; CHI; COL; CRI; DR; GUA, only for short term; HON; MEX; NIC; PER; PRY, through the annual debt law; SLV). In COL, the issuance of internal or external debt requires various approvals, including by the legislature.15 In URU, the law authorizes an annual increase in the net public debt on a preestablished indexed amount. In other countries, individual and specific authorizations are required for each operation, particularly on external debt (in CRI and PRY for bilateral external borrowing and GUA, HON, SLV, and URU for all operations of external debt). In some countries, legislative authorization for issuance of Tbills for liquidity purposes is not required if redeemed before the end of the fiscal year (ARG; BOL; CHI; COL; CRI; HON, except when beyond the presidential term; DR; NIC). In ECU, prior legislative authorization to issue Tbonds or Tbills is not required. In BOL, domestic debt is authorized through presidential decree. In PAN, issuance of domestic and external; and short-, medium-, and long-term debt requires only that the Executive informs the Assembly on the amounts authorized by the President.

Publication of Reports and Systems for Recording Transactions

Publication of plans and reports is essential for fiscal transparency and accountability. Usual reports that are published include (i) annual borrowing plan; (ii) in-year cash flow forecasting; (iii) Tbond and Tbill issuance calendars; (iv) results of Tbond and Tbill auctions; (v) debt position and cash balances; (vi) CDM performance reports; (vii) annual financial statements; and (viii) financial statistics.

In LAC, Treasuries and DMOs are mostly transparent. They prepare and publish a substantial number of reports, although the content and depth vary.

Fourteen countries prepare annual borrowing plans (BOL, BRA, CHI, COL, CRI, DR, ECU, GUA, HON, MEX, PAN, PER, PRY, SLV, URU,) and 10 publish them (BRA, CHI, COL, CRI, ECU, GUA, HON, PAN, PER, URU). Fifteen countries prepare cash flow forecasts (ARG, BOL, BRA, CHI, COL, CRI, DR, ECU, GUA, HON, MEX, PER, PRY, SLV, URU), although none publishes them. In 10 countries, a Tbond issuance calendar is published (BRA, CHI, COL, CRI, DR, GUA, HON, MEX, PER, URU), and in 8 countries, a calendar of Tbills is published (BRA, COL, CRI, MEX, PAN, PER, SLV, URU). Results of auctions are publicly available in 15 countries (ARG, BOL, BRA, CHI, COL, CRI, DR, GUA, HON, MEX, PAN, PER, PRY, SLV, URU).

Annual financial statements are published in 13 countries (ARG, BOL, BRA, COL, CRI, DR, GUA, HON, MEX, PAN, PER, SLV, URU) and financial statistics are published in 14 (ARG, BOL, BRA, CHI, COL, CRI, DR, GUA, HON, PAN, PER, PRY, SLV, URU).

Six countries publish a report on debt management performance, comparing plans with objectives established. Only in CRI is a report that assesses the cash management performance in comparison with the objectives publicly available; and information on daily or monthly cash positions is published only in 5 countries (ARG, CHI, COL, GUA, SLV). Table 3.5 presents the preparation and public availability of cash and debt reports for 10 types of information.

Table 3.5

Main Cash and Debt Reports

article image
article image
article image
article image
article image
Source: Treasury and debt management offices of each country.

Many information systems are used in the region to record the issuance and repayment of Tbonds and Tbills. DMFAS (the UNCTAD system) is used in 6 countries (ARG, ECU, HON, NIC, PAN, PRY) for recording the issuance and repayment of Tbonds and Tbills, and in these countries, they are also recorded in their financial management information systems (in a few cases, both systems interface). In 5 countries, they are recorded in their own debt system (BRA, CHI, COL, MEX, PER). In COL (internal debt) and MEX (all debt), data are also recorded in central bank systems. In 2 countries, the relevant data are recorded only in central bank systems (CRI, URU) and in one country (SLV), in the stock exchange system.

Advantages and Disadvantages of Integration

The CDM functions operate well in some respects in LAC, despite the use of different structures or organizations. It appears that historical tradition has prevailed instead of adopting a modern DMU approach. Even where the potential advantages of integration are recognized (see Box 3.4), the need to change current legislation and administrative—and perhaps congressional—practices creates barriers to change. The three Treasuries in which both functions are integrated considered that coordination has been facilitated, that there is tendency for a more comprehensive financial policy, and that the consolidation of financial operations is easier. In none of them, however, is a fully integrated operation in place.

In the countries with two separate units, the more effective ones are those with better coordination mechanisms and support of better integrated accounting and financial information systems. Specialization and autonomy are considered advantages of the model with separated offices. It is recognized, however, that consolidation of financial information and risk management analysis is more difficult and that there may be some difficulties in coordinating issuance and management of short-term liquidity. It has also been noted that sometimes the sensitivity of the DMU to the short-term liquidity problems identified by the treasury is not optimal. In some cases, the treasury has had to ration cash to cope with localized liquidity problems because the DMU has not been quick enough to react. In this regard, the DMU has preferred to keep the debt issuance calendar unchanged.16 Most of the countries recognize that coordination is the main bottleneck regarding the separated model.

Advantages of The Integration and Separation of Cash and Debt Management According to Country Authorities

Summarized below are the opinions of country officials on the issue of integration of cash and debt management functions. Authorities point out that there are more advantages than disadvantages toward integration.

Advantages of integration:

  • Allows for a comprehensive overview of financial issues related to budget financing and cash needs and, therefore, better liquidity management. Allows for an integrated strategy on the financial requirements of debt and treasury management.

  • If the two are not integrated, financial programming may be affected, since the debt office may not have the same sense of urgency in terms of timing and amounts—”if issued, it is fine; if not, it is fine, too.”

  • If not integrated, the risk of concentration on a particular date is greater, which implies large cash withdrawals on the same date (and higher liquidity risk). Thus, integration can help to smooth the maturity profile and improve flexibility on the corresponding interest rate risk that is generated by a higher volume of refinancing (for financial markets).

  • Defines comprehensive strategies for the implementation of investment operations (active) and funding (passive), taking into account the different opportunity costs.

  • Represents an advantage in relation to the impact of strategy planning, given that with coordination, the execution times and transmission are lower.

  • Legal and risk policies are centralized, which are considered transversal policies and similar in nature in terms of debt and treasury management.

  • When they are not integrated, registration of transactions is done separately, which can create delays and difficulties in consolidating information and reporting. When integrated, statistics are prepared more comprehensively.

  • When not integrated, it is dependent on another unit (Public Debt), because it lacks direct information (debt to cash flow). In addition, units may have conflicting views, putting at risk the coordination of the cash and debt units.

  • Staffing requirements are reduced.

  • It places more relevance on the economic evaluation of the operations. A financially preferable operation in the short term, and analyzed individually, may threaten strategic objectives in the long run and against the balance.

Advantages of keeping separate units:

  • Align incentives either because who defines the benchmark is different from who is responsible for managing the cash. In this case, the Ministry of Finance sets the benchmark, the budget office decides its implementation, and the treasury executes and controls.

  • A risk of being integrated is the concentration of responsibilities on debt and treasury operations under only one manager, implying an ethical impediment when making decisions that impact debt, foreign exchange, and monetary markets (“Chinese Wall”). However, this issue is mitigated by the creation of strategic committees, such as the Treasury Committee or Asset-Liability Management Committee, to approve policies and guidelines.

  • Because they are different functions, specialization is advisable.

  • The areas of competence of each one are respected.

CRI is the only country that had an integrated unit in the past, but now it has two separate units. Their assessment, however, is that the integrated unit worked better than the current arrangement.

As countries are becoming more sophisticated in terms of cash management, it appears that the separation of functions imposes more disadvantages than advantages. With the (i) accumulation of large cash surpluses; (ii) tendency of the treasury also to invest not only in the short term, but also in the medium and longer term (e.g., through creation of sovereign wealth funds and other extra-budgetary funds); (iii) increase in size and complexity of financial operations; and (iv) need to have comprehensive and timely financial reports, the integrated approach tends to respond quicker to the needs of the government and to changes in the market. It also provides for a more economical use of scarce administrative resources.

The transition from separated to integrated structures, however, tends to be longer and more complicated than initially envisaged by the ministries of finance. As well as the need for legislation and new internal arrangements, both functions have to continue to be fully operational during the transition; usually there are no fully integrated information systems in place; and internal capacity needs to be upgraded substantially during the process.

Another major constraint is the capacity of the new units to retain and adequately remunerate the staff. Many DMUs will be unable to recruit experienced staff directly, and will rely on internal training programs. They will, however, often then be unable to pay staff in line with private sector analogues,17 and as the new DMU becomes more sophisticated, private financial institutions will be interested in recruiting people who have acquired enhanced expertise. In most countries, the DMU will be the largest treasury in the country—certainly when measured by market impact—and the expertise accumulated by these officials will be very valued by the market.

As noted above, in some countries the access to more flexible pay arrangements was one of the considerations in adopting an agency structure when setting up a DMU—even while it was still in the MoF.18 There are examples in Latin America where positive structural conditions have been created to retain qualified professionals in the DMU (see Box 3.5).

Brazil: Structural Conditions to Retain Qualified Professionals in the Debt Management Office

In Brazil, the creation of a specific career in 1986 to attract high-level professionals to the debt management office was a turning point. Selection of finance analysts is done through an open and competitive procedure. Salaries are similar to those paid to professionals working in other strategic areas, such as the Central Bank and the Tax Authority, and other benefits are provided, such as a pension fund and health insurance. Training is provided continuously to keep economists up to date and abreast of developments and to incentivize them to stay in the public sector. Rotation has reduced significantly since then.

A comprehensive approach is recommended in those countries that decide to move to an integrated structure. It would include (i) a detailed definition of the functions to be retained and added by the new entity and its relationship with key stakeholders (including respective responsibilities with others in the MoF and with the central bank); (ii) a detailed study of the best governance structures and organizational arrangement to facilitate the synergies of the new entity; and (iii) the support of reliable and comprehensive accounting and information systems. The work should also cover the management and staffing of units that might not be the focus of the integration process. CDM functions are very different from the payment and accounting functions of the traditional treasury. The latter may require more staff, but the skill requirements and processes are very different; hence also the managerial styles. This may be reflected in the internal structures of an integrated unit that covers all these functions, with the senior director general or chief executive delegating treasury process management to a deputy with relevant senior management experience.

A program of capacity building will be important, including a structured training program linked to business objectives and a realistic and attractive remuneration policy, as well as other measures to encourage job satisfaction that could retain and attract competent staff. Some offices have found a formal business plan useful, as a way not only to clarify objectives and identify capability gaps, but also to build a common culture and enhance individual commitment. This reform may take some years to be implemented; it is a project and, like other projects, there needs to be a process to identify priorities and sequencing, as well as tasks and dependencies.

References

  • Bangura, Y., D. Kitabire, and R. Powell. 2000. “External Debt Management in Low-Income Countries.IMF Working Paper WP/00/196, Washington, DC: International Monetary Fund. Available at http://lnweb90.world-bank.org/CAW/Cawdoclib.nsf/vewCrossCountryStudies/015988B62981189E85256CEE007AD1AB/$file/wp00196.pdf.

    • Search Google Scholar
    • Export Citation
  • IIF (Institute of International Finance). 2014. Sovereign Investment Relations: 2014 Evaluation on Investor Relations and Dissemination Practices by Key Emerging Market Borrowing Countries. Washington, DC: Institute of International Finance. Available at http://webcache.googleusercontent.com/search?q=cache:K9_IOAanTEIJ: https://www.iif.com/file/6572/download%3Ftoken%3DH34sui-y+&cd=2&hl=en&ct=clnk&gl=au.

    • Search Google Scholar
    • Export Citation
  • IMF (International Monetary Fund) and World Bank. 2003. Guidelines for Public Debt Management—Amended. Washington, DC: International Monetary Fund and World Bank. Available at http://www.imf.org/external/np/mfd/pdebt/2003/eng/am.

    • Search Google Scholar
    • Export Citation
  • IMF (International Monetary Fund) and World Bank. 2009. Developing a Medium-Term Debt Management Strategy (MTDS)—Guidance Note for Country Authorities Washington, DC: International Monetary Fund and World Bank. Available at http://go.worldbank.org/T7SB6VFEL0.

    • Search Google Scholar
    • Export Citation
  • Lienert, I. 2009. “Modernizing Cash Management.IMF Technical Notes and Manuals. Washington, DC: International Monetary Fund. Available at www.imf.org/external/pubs/ft/tnm/2009/tnm0903.pdf.

    • Search Google Scholar
    • Export Citation
  • Pessoa, M. and Williams, M. 2012. “Government Cash Management: Relationship between the Treasury and the Central Bank.IMF Technical Notes and Manuals. Washington, DC: International Monetary Fund. Available at www.imf.org/external/pubs/ft/tnm/2012/tnm1202.pdf.

    • Search Google Scholar
    • Export Citation
  • Shah, A. 2007. “Budgeting and Budgetary Institutions.World Bank Public Sector, Governance, and Accountability Series. Washington, DC: World Bank. Available at https://openknowledge.worldbank.org/bitstream/handle/10986/6667/399960PAPER0Bu10082136939301PUBLIC1.pdf?sequence=1.

    • Search Google Scholar
    • Export Citation
  • Storkey, I. 2009. “Operational Risk Management and Business Continuity Planning for Modern State Treasuries.IMF Technical Notes and Manuals. Washington, DC: International Monetary Fund. www.imf.org/external/pubs/ft/tnm/2011/tnm1105.pdf

    • Search Google Scholar
    • Export Citation
  • Williams, M. 2010. “Government Cash Management: Its Interaction with Other Financial Policies.IMF Technical Notes and Manuals. Washington, DC: International Monetary Fund. www.imf.org/external/pubs/ft/tnm/2010/tnm1013.pdf.

    • Search Google Scholar
    • Export Citation
2

These categories broadly equate to the familiar front-, middle-, and back-office functions that are characteristic of many DMUs. The relevant organizational issues are discussed in more detail below.

3

The development of a DMS and related debt management policies are not further discussed in this publication. A DMS is a potentially complex undertaking, at least for those countries sufficiently developed to have access to a range of financing options. For guidance, see IMF & World Bank (2009). There is substantial supporting literature, some of which has been brought together by the World Bank, at http://treasury.worldbank.org/bdm/htm/resource_publications.html (in particular, the “Cost-Risk Modeling” and “Debt Management Strategy” tabs).

4

This function is common to debt and cash managers. It tends, however, to play a greater role among debt management activities; in practice, the debt management database and related processes may serve both functions, even where they are institutionally separate.

5

Many of the arguments in this subsection are spelled out at greater length in Williams (2010).

6

Note that Tbills are part of government debt, whether they are issued for cash management or debt management purposes and whoever is responsible for their issuance. As stressed below, it is strongly advisable to have one source of issuance and one point of contact with the market; in any event, within the ministry, there should be an integrated oversight of the debt stock as a whole.

7

For active cash managers in the countries of northern Europe, it is quite usual for there to be 40–60 cash management transactions every day, while the number of debt transactions is much less.

8

In some countries, the DMO has a persistent problem keeping the database up to date (e.g., following successive loan disbursements). It is strongly preferable for all interactions with the creditor on disbursements, as well as on loan servicing, to go through the DMO rather than from the project management unit directly.

9

The central bank would normally be a member of the Public Debt Committee. Firstly, it is able to contribute its knowledge of domestic financial markets when discussing the appropriateness of different issuance strategies. Secondly, in most countries the foreign currency reserves are held on the central bank’s balance sheet, and the size and composition of those assets are relevant when considering the composition of government liabilities and the extent to which it is possible to hedge different parts of the balance sheet (sometimes referred to as sub-portfolio matching). There is a strong argument that the DMS should be developed taking account not only of cash assets, but of the full range of assets and liabilities on the balance sheet of the government (defined broadly to include the central bank). See, also, previous references in discussion of debt management functions above.

10

Sweden in 1789 and the Netherlands in 1841, for example.

11

Countries included in the survey are Argentina (ARG), Bolivia (BOL), Brazil (BRA), Chile (CHI), Colombia (COL), Costa Rica (CRI), Dominican Republic (DR), Ecuador (ECU), El Salvador (SLV), Guatemala (GUA), Honduras (HON), Mexico (MEX), Nicaragua (NIC), Panama (PAN), Paraguay (PRY), Peru (PER), and Uruguay (URU).

12

This chapter adopts as a general nomenclature “treasury” for units usually dealing with cash management and management of the TSA, and DMU or “debt unit” for units dealing with bond issuance and debt management functions. Should both functions be performed by an integrated unit, the term “integrated unit” is used.

13

The Central Bank of Brazil uses repo operations for market liquidity management. It uses Tbills and Tbonds issued by the Treasury as collateral. The Central Bank is forbidden by law to issue its own securities.

14

The Brazilian Treasury also issues medium-term Tbills of up to four years.

15

In general, it requires (i) authorization of the Board of the Central Bank, which sets the general financial conditions for issuance; (ii) a favorable opinion of the Interparliamentary Committee on Public Credit to external debt; (iii) the annual law that establishes an overall amount; and (iv) annual law which determines the needs of credit resources to finance the corresponding budgetary appropriations.

16

Although many countries outside Latin America issue Tbills with more flexibility, even when they operate a strict calendar for Tbonds.

17

In one of the countries where there is an integrated unit, the need to retain some highly skilled staff was the key reason to maintain two divisions (and two division chiefs) that should have otherwise been merged.

18

For example, in the United Kingdom, the DMO was able to establish its own pay structure; it still had to be approved by the civil service authorities, and annual increments must be consistent with government-wide policy, but it gave the DMO some flexibility to recruit and retain key skills.

The Key to Efficiency and Transparency