Regulation and Supervision

49. There are a number of areas where RMA should develop its approach, both in substance and supervisory process. The RMA’s approach has a number of strengths, including intensive offsite supervision, based on extensive and frequent reporting by financial institutions; oversight of the banks’ preparation of their annual accounts, including meetings with auditors and bank management; and a readiness to hold financial institutions to its standards (i.e., not granting waivers and exemptions) and to say “No” when necessary. However, regulatory gaps should be filled over time, and the RMA should consider how it can move from a relatively prescriptive approach in some areas (for example, bank base rates, requirements to purchase reinsurance, restrictions on guaranteeing corporate bond issues) towards one based more on holding management responsible for how they run the business, the supervisory role being to assess whether they are meeting that responsibility and to respond where they are not. All of these changes will require RMA to rebuild supervisory capacity through recruitment and training.

50. Priorities should include:

• Ensuring that supervisors have a view of the full range of risks in financial institutions. This will require starting to use scenario and stress testing, for example requiring banks to quantify their key risks including credit, interest rate, liquidity and foreign exchange risks on the basis of certain assumptions or insurance companies to evaluate the impact of catastrophic events. In addition, the resumption of onsite work subsequent to the mission will be essential to monitor financial system soundness and to evaluate systems and controls, developments in credit standards and the adequacy of management and governance.

• Developing a policy framework and issuing regulations in key areas where there are gaps, including prudential standards applicable to insurance and reinsurance, and bank liquidity. The latter needs careful consideration given the lack of high quality liquid instruments that would normally be at the core of any requirement for stock liquidity. The alternative is to develop a mismatch limit approach as envisaged in Section 7 of the Prudential Regulations, and it is recommended that the RMA begin with such an approach.

• Developing a strategy for updating the bank capital adequacy framework in areas such as the definition of capital (introducing a minimum Core Tier 1 Equity concept from the Basel III framework), capturing operational risk (using the non-advanced approaches) and interest rate in the banking book (taking into account the results of stress tests that would help establish the extent of the risks); and developing a Pillar 2 framework to relate the capital adequacy of financial institutions to the supervisory assessment of risks in individual companies.

• Using the results of the results of the deeper risk assessment of individual financial institutions to develop supervisory strategies in relation to each company, which would include challenging on business strategy, controls and financial strength; in particular, any weak banks should be identified and challenged to develop remedial action plans. With a small number of financial institutions, RMA is well placed to take a more bank specific approach to its supervisory work, which would also address the variable financial conditions of the individual institutions.

• Developing a crisis management framework in case of a future bank failure. This should start with the authorities evaluating what additional information they may need in case of a crisis (for example, on immediately maturing liabilities) and ensuring that this information would be available at short notice, and making sure that practical preparations are in place, including updated contact details of key officials etc. In due course the work could extend to consideration of potential changes to the Bankruptcy Act to create a bank insolvency framework. In this context, it is important that planning for the introduction of deposit insurance resumes, as the availability of insurance can help support retail depositor confidence in case of crisis.

• Developing a policy framework on its approach to setting new standards and regulations, covering consultation with financial institutions and preparation of impact assessments. A number of regulatory changes have been introduced at short notice without consultation or reference to RMA’s objectives, triggering pushback from financial institutions which it has sometimes been hard for the RMA to resist in the absence of its own impact analysis.

• Strengthening its capacity in the framing of regulations and standards through the creation of a function within the Financial Regulation and Supervision Department dedicated to the preparation and promulgation of new standards.

51. The RMA should also build on its existing readiness to take macroprudential policy measures with:

• A regular review of sector-wide risks, extending the analysis in the Financial Sector Performance Review; and

• An internal policy statement on future choice and use of macroprudential tools.

The objectives should be to move away from the current reliance on one tool, the adjustment of risk weightings, and, in the current context, to ensure that supervisors seek to identify and address what may be the next source of stress rather than focusing only on present stresses related to current account pressures.²⁰ It must be emphasized that macroprudential tools cannot substitute for an appropriate monetary and fiscal stance in addressing overheating pressures.

52. Debt-to-income limits should in particular be considered, in conjunction with the development of monetary policy tools. Some macroprudential tools used in other countries appear to have limited relevance at present. Loan-to-value ratio limits would have limited impact given that high levels of collateralization are part of established lending practice. Limiting credit to deposit ratios could be introduced if and when money and capital markets develop, at which point banks would have the scope to manage balance sheets to meet such minimums. Making monetary policy implementation more effective should be the starting point and priority. However, the RMA could also consider setting debt-to-income ratio limits, which have been effective in other countries, and may also support increased focus on individual credit assessment. Limits on overall credit growth could be considered but would be hard to calibrate.

53. The authorities also need better to distinguish macroprudential measures from those designed to encourage lending to particular sectors and to take action preemptively. Recent changes to risk weighting have been driven mainly by analysis of risks in bank lending, but are also designed to encourage lending to targeted sectors such as agriculture—reflecting the wider government objective to stimulate sectors with lower demand for imports. The adequacy of overall capital requirements resulting from recent changes is hard to assess, and the RMA’s recent increases are consistent with a higher risk environment, but increasing the requirements at an earlier stage (for instance as proposed in early 2012) might have helped moderate the rapid build-up in credit.

Developing the Financial Sector

54. A number of obstacles have frustrated the development of Bhutan’s financial markets, notwithstanding the significant expansion of the banking sector. The challenges include:

• The absence of fully market-determined interest rates and the development of a yield curve to support pricing in financial markets generally.

• Inflexibility in credit provision—entrenched approaches to banking products and services, partly reflecting the absence of any alternatives to bank finance, limited openness to foreign participation and low penetration of technology.

• Limited availability of risk management instruments—in insurance (i.e. limited scope of products, especially for life insurance and low penetration generally) and no scope for hedging of risks through financial transactions.

• Still limited infrastructure and business services, with no ratings agency, other sources of independent evaluations of risk in financial assets or local accounting, auditing and actuarial services; and limited expertise in modern finance.

• Government ownership and control of large parts of the economy, limiting the scope for broader participation in ownership of financial assets. (The government is committed to retaining its most significant interests.)

• Limited understanding of finance amongst investing public.

• The difficulty of achieving scale in a country of only 700,000 people.

55. Equally, Bhutan has a number of advantages that could be exploited:

• Institutional and high net worth investors including the NPPF, life insurance, the funds of the new fund management company, and DHI which manages the government’s investments and investor interest from abroad.

• A number of large public and private sector companies with a need for longer term finance and a government with a deficit to finance.

• A legal framework that is broadly understanding of and ready to support the enforcement of financial contracts.

• High quality infrastructure in payments and settlement.

• Policy-makers ready to take a strategic approach as evidenced, for example, by the government project on accounting and auditing standards.

56. While a number of initiatives have been taken, and improvements in supervision and regulation will contribute, more is needed. The recent expansion of the financial sector has contributed to financial development by bringing in foreign capital and expertise and adding to listings on the RSEB. Developing regulation and supervision will have the same effect. Strong supervision that, for example, addresses weak banks and requires all banks to recognize full costs of their lending decisions (for example, by addressing the risks in maturity mismatches), will support the better allocation of resources through bank intermediation as well as promoting confidence and helping to safeguard financial stability. Good supervisory process enables institutions to plan for the development of their business. However, more is required.

Present Situation

61. Treasury bills were introduced in December 2009 to replace RMA bills with the objectives of improving government cash management, allowing the use of market-based instruments of monetary policy, and providing a benchmark for short term interest rates. These objectives have not been met because of erratic issuance, unpredictable and insufficient amounts, small market size, limited market participation, and the lack of other financial instruments (see Figure 11).²¹

View Full Size

Government Financing, July 2011-July 2013

Citation: IMF Staff Country Reports 2014, 178; 10.5089/9781498357708.002.A004

Sources: RMA and MoF

• Download Figure
• Download figure as PowerPoint slide

62. Treasury bills are also not being issued at market-determined interest rates. The issuance of Treasury bills started out as a conventional market auctions, but the MoF found it difficult to accept the interest rate outcome. Since February 2013, the interest rate has been fixed at 3 percent which is far below deposit interest rates and RMA’s policy rate. A so-call “tap-sale” mechanism has been used in which the MoF determines the interest rate and the amount. On redemption, Treasury bills are often rolled over and the same holders keep the security at unchanged interest rates. Maturities have varied from 14 to 90 days, with issuance primarily at 30 days.

63. Despite a substantial financing requirement in recent months, the government has continued to rely on bank financing rather than T-bill issuance. First, an overdraft facility (OD) is available at the RMA up to a limit determined as 10 percent of average revenue over the last two years. This limit was recently raised from Nu 1.9 billion to Nu 2.1 billion reflecting this guideline. The RMA charges an interest rate, currently 4 percent, at 1 percentage point above the rate for the most recent Treasury bill issuance. Second, the Ways and Means (W&M) account with the largest commercial bank (BOBL) can be overdrawn without a limit. The interest rate charged is currently 5.5 percent. As shown in Figure 11, the government will first utilize the cheaper overdraft in RMA before making use of Ways and Means. It is also noticeable that the issuance of Treasury bills has to a large extent been detached from the financing requirement. In the period under review, Treasury bills have not been issued for monetary policy purposes.

Discussion and Recommendations

64. Developing a government securities market is the most critical element for making monetary policy more effective and for enhancing financial markets in Bhutan. The availability of sufficient amounts of government securities at market-determined rates will provide a range of benefits. First, financial institutions will be given an investment alternative which at present is lacking for necessary portfolio adjustments and liquidity management, including meeting prudential liquidity requirement. Second, the central bank will directly and indirectly have a basis for effectively managing banking sector liquidity. Third, the government will have access to market financing necessary for efficient cash management. Finally, building a (risk-free) benchmark yield curve will have a number of positive externalities for financial sector development.

65. The current arrangements for financing the government hamper the development of a Treasury bill market and money markets more generally. The main sources for financing of government should be gradually shifted from bank to market financing as the government cash management improves and the financing requirement can be covered through an operational issuance program of Treasury bills. At that point, the Ways and Means account with a commercial bank should not be allowed to be overdrawn and the use of the overdraft facility in RMA should be made more restrictive. Even though drawings stay within the legal limits, this arrangement weakens incentives to strengthen government cash management.²² Consideration should be given to limiting direct access to financing from the central bank to emergency situations in which access to other funding sources is difficult. Moreover, overdraft drawings should have a limited maturity and not generally be rolled over. This would put additional pressure on the Department of Public Accounts to impose strict discipline on budgetary agencies to follow their cash plans and give incentives to an active management of public debt.

66. Although the main responsibility for a meaningful Treasury bill issuance program rests with the MoF, the central bank has also an obligation to persistently promote the development of a short-term securities market. Without such a market, the implementation of monetary policy will be less effective and the transmission of interest rates more limited. It must therefore be a joint responsibility of the MoF and RMA to establish a reliable issuance calendar for Treasury bill and to ensure that meaningful amounts are issued also when there are no immediate financing needs of the Government. Establishing an issuance calendar of Treasury bills (used for fiscal and monetary policy purposes) will require a proven capacity at the central bank to forecast liquidity and at the Ministry of Finance to forecast government cash flows.

67. The arrangement in place for using Treasury bills both for government financing and monetary policy purpose should be fully utilized.²³ This is a progressive and forward-looking decision by the authorities which could have important advantages for market development and the use of RMA bills should not be resumed. A combined issuance program using Treasury bills for both fiscal and monetary purposes would broaden the basis for establishing reference rates for the pricing of financial assets. On its part, the RMA should for liquidity management purposes preferably issue Treasury bills with shortest maturities, with maturities and amounts being determined by the expected liquidity management needs. Issuance of Treasury bills for both fiscal and monetary purposes will necessitate robust arrangements to prevent the government from utilizing the liquidity that has drained from the system for monetary policy purposes, as well as an appropriate communication strategy with market players to communicate the policy direction.

68. Market financing of government cash requirements and liquidity absorption at market rates will imply increased costs to the budget and to the RMA. These costs are outweighed by the benefits of having a more developed financial sector and better tools for achieving stability in the economy. It is also not obvious that the cost of Treasury bill issuance for monetary purposes should be born exclusively by the RMA. The present arrangement as specified in the November 2009 Memorandum of Understanding (MoU) clearly states that the cost of Treasury bills issued for monetary policy purposes should be borne by the RMA. In general, providing liquidity to the banking system generates interest earnings for the central bank while draining liquidity using market-based instruments carries costs. Too much focus by a central bank on the profit position may defer necessary market operations.

69. It is important to establish an acceptance that the cost of monetary policy ultimately—directly or indirectly—must be borne by the budget. This is because the finances of the government and the RMA must in the final analysis be considered on a consolidated basis. Taking sterilization costs directly into the budget—for example through the issuance of Treasury bills for monetary policy purpose—may be a better alternative than having to repeatedly recapitalize the central bank.²⁴ In order to provide certainty to the institutional arrangement safeguarding RMA’s financial position and its ability to conduct monetary policy, the RMA and the MoF could consider adjusting the existing MOU on this point. A binding cost-sharing arrangement should be established in which the government’s willingness to protect the central bank for taking on these costs as part of implementing monetary policy is clearly stated. The revised MOU should specify how the RMA could be reimbursed for the costs that cannot be absorbed on its own balance sheet in connection with issuing Treasury bills for sterilizing liquidity. This would support to the independence and autonomy of the RMA. Public knowledge of such an arrangement will strengthen confidence in the RMA, as it would not be constrained by profit considerations when it needs to tighten monetary policy. Although the Treasury bills issued for different purposes are identical for the investors, there should be full transparency as to the distribution of issued amounts between the MoF and RMA.

70. Apart from cost sharing, cooperation and coordination between the MoF and the RMA on debt and liquidity management should be strengthened. While the management of public debt is clearly the responsibility of the MoF, the RMA should be given the opportunity to comment on the debt strategy before final decisions are taken. The choice of instruments, maturities, and amounts may have implications for monetary policy implementation and should therefore be factored in, if possible. The existing Treasury Bills Management Committee should meet regularly (e.g. weekly)²⁵ and be fully used as a platform for sharing all relevant information in relation to the implementation of fiscal and monetary developments between the two institutions on an ongoing basis. Furthermore, in determining the short-term borrowing requirement of the government, there should be some flexibility to take into account factors which can influence RMA’s liquidity management.

71. The technical infrastructure supporting a Treasury bill program needs to be strengthened. At present, the ownership records are maintained in an Excel spreadsheet at the RMA. An automated depository system should be developed to facilitate transfer of ownership resulting from secondary market trades and allow for temporary transfers in connection with the use of securities as collateral and in repurchase transactions. Ideally, the depository should be integrated with the payment system to ensure delivery versus payment. As the payments system recently adopted by the RMA is based on the RBI system, the RBI could be approached for support in extending this system to include a depository.²⁶

72. Along with the technical infrastructure, a viable Treasury bill program should also be supported by market infrastructure. This should include regulations and market practices, and clear guidelines for trading and settlement. The RMA should assume the main responsibility for this task, in close collaboration with MoF.²⁷ It would be useful if the traders themselves established a Code of Conduct. For example, under the umbrella of the Financial Institutions’ Association of Bhutan (FIAB), a forum of bank treasurers should be set up. One of the first tasks of this group should be to discuss and adopt a Code of Conduct for trading in the domestic money market and, perhaps, foreign exchange market. This activity should be utilized as a vehicle for training and increasing trading skills among bank treasurers.

Present Situation

73. The RMA continues to rely heavily on direct instruments and administrative intervention to implement monetary policy. As such, the use of the reserve requirement (cash reserve ratio (CRR)) and liquidity regulations (SLR) have been central tools. Banks have separate CRR accounts, and there is not a provision for banks to meet the requirement on average during a maintenance period. Efforts have been made to move in the direction of more market based instruments, but the use of RMA bills and, more recently Treasury bills, has not achieved the intended objectives. In connection with the rupee stabilization measures introduced in March 2012, a number of new monetary policy instruments were announced with the intention to make monetary policy more effective. This included a “lender of last resort” facility (RMA Short-Term Liquidity Adjustment Window (RSTLAW)) which would provide funds up to 90 days. This facility was priced 1 percentage point above the Policy Rate which also was introduced. It is currently set at 6 percent. The facility has been used only once, mainly because banks lack collateral but also reflecting excess liquidity. The intention was also that the Policy Rate should set the price for central bank funds in ordinary market operations, but as such operations have not yet taken place the Policy Rate has so far had no practical purpose. Base rate regulation was also introduced (see above) with the aim of increasing transparency and improving the transmission of monetary policy. While there may have been some positive effects in meeting the first objective, it does not appear that transmission has been improved by the introduction of base rates.

74. A fairly recent administrative measure to influence banking system liquidity is to sweep balances of selected accounts in commercial banks to the RMA overnight. This is a normal practice in connection with the management of Treasury Single Accounts (TSA) in central banks where funds collected by commercial banks or any unused balances on settlement accounts are swept overnight to the central bank. In Bhutan, sweeping has been used to manage liquidity, and the practice has been extended to a number of large accounts (currently 8), including mostly, but not only, hydropower project accounts. The daily amounts involved vary between Nu 500 million - Nu 1.2 billion. The sweeping was discontinued at the end of May 2012 when liquidity became tight, but was resumed in the beginning of July 2013. This indicates that although the stated objective is to sterilize volatile hydropower related funds, the measure has also been used as a general liquidity management tools (see discussion below).

75. There is a limited interbank market in Bhutan, with banks and other financial institutions making unsecured placements with each other on a regular basis. These placements are almost never short-term, and normal maturities range from 90 days to one year. The interest rates are said not to fluctuate much (according to a “gentlemen’s agreement”), but follows the interest rates for corporate deposits. An indication of this activity is shown in the Figure 12 below. Both demand and time deposits with other banks qualify as Quick Assets in relation to the SLR, and bank placements make up about one-third of all Quick Assets.

View Full Size

Banks’ Placements with Other Domestic Banks, January 2012-June 2013 ^1/

Citation: IMF Staff Country Reports 2014, 178; 10.5089/9781498357708.002.A004

Source: RMA.^1/ Only quarterly data is available for 2013.

• Download Figure
• Download figure as PowerPoint slide

76. There are no foreign exchange transactions between banks, and banks are only allowed to deal in foreign exchange with their customers at rates fixed by the RMA. The RMA fixes buying and selling rates daily based on the US$/INR closing rate the day before. Banks are obliged to use these rates in their retail transactions and they are not allowed to make transactions within this margin. (Under a similar system in India, banks are allowed to trade within this margin). Banks can at any time sell convertible currencies to the RMA when they are approaching a fixed limit (US$10 million for larger banks and US$5 million for smaller banks). All sales and purchases of foreign exchange with the RMA take place at the mid-rate. This arrangement gives no incentives to interbank trading.

Discussion and Recommendations

77. An operational liquidity framework will be an important mechanism to develop the capacities for formulating monetary policy and analyzing financial stability. By monitoring the factors influencing liquidity developments and banks’ behaviors, the central bank can develop a better understanding of current and future trends affecting liquidity developments in the banking sector. The standard approach for most central banks is to make forecasts in order to observe banks’ reserves and calibrate open market operations to deal with disrupting fluctuations. Technical assistance has been given in the past to set up a liquidity monitoring framework, but availability of data to the RMA has made the updating very difficult. As the most difficult and important data input is government cash flows, the ongoing TA to the Department of Public Accounts on cash management and the introduction of a TSA should improve the basis for liquidity forecasts. It is important that the RMA becomes involved in this work so that the cash flow projections can be directly entered into the liquidity forecasts. Cash flow forecasts would also need to reflect information about large hydropower related flows. It is important to note that not all cash flows have immediate liquidity effects, such as foreign exchange transactions and other transactions between the government and RMA. Under a fixed exchange rate arrangement, the RMA should develop its capacity to offset domestic sources of volatility in bank reserves to avoid corresponding adverse fluctuations in interest rates.

78. The reserve requirement system (CRR) should be redesigned to enable reserve requirements to be more effective as a supplement to other monetary instruments. In particular, the practice that required reserves are kept in separate CRR accounts should be changed. Banks should have only one account with the RMA which should be used to meet the reserve requirements as well as serving as an ordinary settlement account. Having one current account in RMA would also make it possible to allow banks to meet the reserve requirement on average over the maintenance period. The averaging mechanism would provide more flexibility in banks’ liquidity management and reduce the demand for excess reserves. In an averaging system, individual banks may draw on their reserve account in case of unexpected liquidity needs, but they would then have to overfulfill later in order to meet the requirement. This gives the banks more responsibility and incentive to manage their own liquidity more actively, which in itself is a precondition for improving the monetary transmission mechanism. The current practice that the RMA makes the posting on the banks’ reserve account will become unnecessary. At the present relatively low level of the reserve ratio (5 per cent), this is now a good opportunity to make the change. Should the ratio increase in the future, the averaging provision could be combined with a minimum daily requirement in order to limit the potential for a large reserve drawdown on any given day. This floor could be lowered as experience with the new system increases. At the end of the maintenance period, non-fulfillment should be penalized.

79. Reserve requirements are an effective but a blunt instrument which should be used with care. The RMA has (understandably) actively used changes in the reserve ratio in response to tight liquidity resulting from the measures taken in March 2012 to deal with the rupee shortage.²⁸ The RMA should in the future avoid frequent and large changes in the cash reserve ratio, as this can be disruptive for financial intermediation. Changes in the reserve ratio have a direct impact on the money multiplier and reserve requirements generally serve as a tax on financial intermediation, which may result in wider spreads between retail deposit and lending rates. As the reliance on market-based instruments in liquidity management increases, large and frequent changes in the reserve requirement should not be necessary.

80. The application of sweeping is an administrative tool for direct regulation of liquidity and should not be used as a general liquidity instrument as alternative tools are developed. Future use of this arrangement should be analyzed and clarified in cooperation with commercial banks. Although the arrangement can serve a useful purpose, namely to sterilize huge, very short-term and potential destabilizing funds in connection with the hydro-power projects, the selection of accounts could be reviewed.

81. With a stronger focus on short-term liquidity management, the RMA should consider introducing a short-term standing lending facility. The RSTLAW is much too long term, and banks need to become more active and responsive in their short-term liquidity management. Many central banks have introduced an overnight standing facility, which allows banks to borrow funds overnight. The arrangement should normally not be extended beyond overnight, to give banks incentive to repay any outstanding balances in the morning when the interbank market opens. As it will take time to develop an interbank market in Bhutan, the short-term facility could be extended to, say, one week. To encourage banks to first try to find the funds needed from other banks before making use of the central bank facility, the short-term loan facility should be offered at a reasonably penal interest rate relative to the cost of borrowing in the interbank market. This interest rate on the facility should be determined by the RMA, and it will normally form the ceiling for the interbank market rate. The purpose of the facility is therefore to induce banks to take a reasonable risk in their liquidity management which will help to increase activity in the short-term money market and reduce interest rate volatility. Access to the RMA standing loan facility would be at the discretion of the banks that have appropriate collateral. It is important that occasional use of this facility should be considered as a normal part of a bank’s liquidity management, and no stigma should be attached to frequent drawings.

82. It should be emphasized that the standing lending facilities are meant only to provide very short-term liquidity support. Banks with longer-term liquidity problems or structural liquidity needs in the banking system should be addressed through other measures. The RMA should consider the need for introducing a properly designed Lender-of-Last-Resort (LoLR) facility for banks that are faced with longer-term liquidity problems. The already existing RSTLAW, which can extend credit up to 90 days in the form of an overdraft, could perhaps serve such purpose in the meantime should the need arise.²⁹ LoLR support should not be considered as part of monetary policy implementation, although the liquidity effects of longer-term support have to be taken into account.

83. Many central banks have also introduced a standing deposit facility to provide an interest rate floor in the interbank market. As a standing facility, banks can deposit any excess funds overnight at their discretion. Similar to the overnight lending facility, the interest rate on the deposit facility should have a penalty element in the form of a low rate and be lower than the deposit rates offered in the interbank market. This should discourage its use and provide incentive to first try to place available funds in the interbank market. The need for a standing deposit facility is not pressing in Bhutan, but such a facility would in due course assist monetary policy in case of a sudden capital inflow by preventing interest rates from falling significantly (and in this respect be an alternative to sweeping). The cost of these operations has to be kept in mind, as the RMA would need to pay interest on all excess reserves. A standing deposit facility will only become relevant when the RMA is able to manage the banking system liquidity in such a way that the use of the standing facilities would be for marginal funds only. The RMA should consider introduction of these tools in the context of a broader package aimed at fostering market development and effective monetary policy implementation.

84. As part of a more market-based monetary policy framework, some increased flexibility also in the foreign exchange market should be considered. Banks could—as in India—be allowed to determine the exchange rate for transactions with their retail customers within the margin prescribed by the RMA. Allowing banks to compete on price for attracting foreign currency transactions is necessary to create a more competitive environment. Banks should also be allowed to sell foreign exchange a bit cheaper as part of their overall banking relations with their customers. Also, the foreign exchange transactions between the RMA and the commercial banks should not be at mid-rate and available at all times. As a start, the RMA should provide a daily window (at a fixed time) with a buy and sell rate around a parity rate of the USD/Nu rate determined on the basis of the US$/INR market rate. This would give banks an incentive to obtain a better price from each other. At a later stage, the RMA should only open the window on a few predetermined days during the week. A functioning inter-bank market is internationally the main mechanism for facilitating an efficient distribution of foreign exchange from institutions with a surplus to institutions with a deficit. Giving banks incentives to become more responsive and active also in this area will support the orientation toward a more effective implementation of monetary policy. This established practice of stable interest rates appears to have contributed to the build-up of substantial maturity and interest rate mismatches on banks’ balance sheets.

Present Situation

85. The transmission of interest rates is currently weak in Bhutan. Despite large swings in banking system liquidity and macro-economic developments, banks’ retail deposit and lending rates have remain remarkable stable (see Figure 13). Although banks are in principle free to determine their interest rates, there appears to have been limited attempts to make adjustments in their portfolios by changing their interest rates.³⁰ As discussed earlier, there has also been very limited impact on bank funding costs coming from monetary policy instruments, such as Treasury bills (see Figure 14).

View Full Size

Banks’ Deposit and Lending Rates (mid-rate), 2002-Aug 2013

Citation: IMF Staff Country Reports 2014, 178; 10.5089/9781498357708.002.A004

Source: RMA.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Treasury Bill Activities, July 2011-July 2013

Citation: IMF Staff Country Reports 2014, 178; 10.5089/9781498357708.002.A004

Source: RMA.

• Download Figure
• Download figure as PowerPoint slide

Discussion and Recommendations

86. Liquidity management should aim at influencing the functioning of money and securities markets. This would strengthen the transmission channel through interest rates and pave the way for changes in the rates determined by the central bank (policy rates) having an effect on the economy through changes in bank retail interest rates and asset prices. A reasonably competitive and responsive banking sector is necessary in order for interest rates to respond to changing market conditions so that the allocation of funds from savers to debtors in the retail market is influenced in the desired direction. Usually, there are two important interest rate transmission mechanisms. The first goes from the policy rate to the interest rates in the wholesale interbank and Treasury bill markets. As these interest rate changes are influencing the banks’ cost of funds, the second channel will lead to adjustments in banks’ retail deposit and lending rates. By integrating the two channels, the transmission of monetary policy decisions would become significantly stronger. The challenge for most central banks trying to influence interest rate transmission through changes in the policy rate is to have a liquidity management framework that ensures short-term market interest rates are determined close to the policy rate. It is also important in a fixed exchange rate regime that changes in the policy rate can be used to correct market interest rates from diverting too much from the interest rate of the anchor currency.

87. Developing these transmission channels in Bhutan will take some time, particularly as banks’ current lending practices should be changed. Because banks mostly lend medium to long term at fixed interest rates, a tightening of monetary policy with corresponding increase in funding costs will have a first-round effect of declining profits and perhaps capital depletion. The effect on bank lending and on overall demand would be stronger if interest rates on existing loans could be adjusted, and not only the interest rate on new loans.

88. In practical terms, liquidity management should attempt to ensure stability in bank reserves in order to establish a link between quantities (bank reserves) and the price of reserves (market interest rates). Large volatility in banks’ reserves makes liquidity planning by banks difficult and without a consistent framework to absorb large swings in liquidity flows, it will be difficult to establish an interbank market. In addition to monitoring and forecasting the supply of reserves, the approach must also be to determine banks’ demand for reserves and “true” excess reserves (or surplus reserves).

89. For Bhutan, the demand for bank reserves is primarily linked to the banks’ need to meet the reserve requirement (CRR) and the liquidity requirements (SLR). The latter is unusual as banks normally will meet prudential liquidity requirements with other interest bearing assets rather than unremunerated current accounts at the central bank. A credible Treasury bill program, as discussed earlier, should provide more appropriate assets for meeting the SLR in the future. In any event, banks often may want to hold additional reserves over the legally required minimum (CRR or SLR) for clearing purposes and other contingencies. This portion of banks’ reserves is often referred to as banks’ “own demand” for reserves. Banks’ own demand cannot be directly observed, but has to be estimated or critically assessed in some manner. Surveys can be used as a starting point to capture banks’ expressed preferences, but they should be supplemented by frequent contacts with individual banks. In a liquidity framework with a focus on bank reserves, it will be necessary to manage “true” excess reserves (also referred to as surplus reserves), which is the residual portion above statutory requirement and banks’ own demand.

90. In order for the link between banks’ reserves and market interest rates to work efficiently, the market for bank reserves should be in approximate balance. Any “surplus reserves” held beyond the demand for reserves will then become particularly important for liquidity management. A rise (decline) in surplus reserves will normally induce banks to expand (reduce) their balance sheets and to lend more (less) on the interbank market, putting downward (upward) pressure on short-term interest rates. An analysis of the interrelationship between the level of surplus reserves and interest rates will become important in a monetary regime with interest rates as the operating variable.