(1) Current Situation

4. There already exist a significant number of uncollateralized interbank (I/B) transactions in Myanmar and the CBM collects data on interbank (I/B) transactions daily, but the data are not effectively used. The CBM is making a monthly internal analysis on the data, although the analysis itself remains confined to the volume and direction of interbank transactions rather than on interest rate developments, the yield curve, or duration structure of the market. The CBM is publishing graphs on “interbank market rates,” but it is outdated (the most recent graph shows rates development during February–December 2017) and its presentation can be improved.

5. There are two types of uncollateralized I/B transactions for which the CBM collects data separately. One is the typical I/B borrowing/lending (I/B borrowing market) and the other is I/B deposit placing/receiving (I/B deposit market). For the former, trading parties conduct the transaction as borrowing and lending. As for the latter, deposit placing banks have the deposit accounts at deposit receiving banks and their deposit amount is increased by the placing amount. According to the interviewed banks, the CBM net is used for the settlement of funds for both types of transactions. Banks also explained in the interview that the choice of types of transaction depends on such reasons as accounting or which side initiates the transaction (one bank reported that when it needs funds, it uses the I/B borrowing market and when it wants to use excess funds, it uses the I/B deposit market).

6. Some data collected by the CBM show a discrepancy. The CBM collects data from both the borrowing/deposit receiving side and the lending/deposit placing side. As for the I/B borrowing market, figures for borrowing and lending typically match. However, reported figures for the I/B deposit market do not match and the data gap is rather considerable. For example, the total amount reported for I/B deposit receiving during February–December 2017 was 2.1 trillion MMK while that for I/B deposit placing amounted to 2.5 trillion MMK. According to the CBM staff, this discrepancy was caused by: (1) some banks not reporting their I/B deposit transactions at all; (2) the coverage of reporting deposit transactions differing among banks (some banks include branch level transactions but some banks do not); (3) some banks including interest payments as I/B deposit transactions; and (4) simple misreporting. The CBM staff are trying to correct the situation, but a discrepancy remains.

Graphs Published on the CBM Website

Citation: IMF Staff Country Reports 2019, 087; 10.5089/9781498305198.002.A001

Source: The Central Bank of Myanmar website.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Graphs Published on the CBM Website

Citation: IMF Staff Country Reports 2019, 087; 10.5089/9781498305198.002.A001

Source: The Central Bank of Myanmar website.

• Download Figure
• Download figure as PowerPoint slide

Graphs Published on the CBM Website

Citation: IMF Staff Country Reports 2019, 087; 10.5089/9781498305198.002.A001

Source: The Central Bank of Myanmar website.

• Download Figure
• Download figure as PowerPoint slide

7. Analysis shows that banks are conducting both types of uncollateralized transactions rather actively, while the total number and amount of the I/B deposit market are higher than those of the I/B borrowing market. As regards the number of transactions by durations, O/N is the most frequently traded duration, followed by 1W and 1M in the I/B borrowing market. As for the I/B deposit market, 1M is the most frequently traded duration, followed by O/N and 3M. The aggregated trade weighted average interest rates (TWAR) for various durations during February–December 2017 show the positive yield curve, ranging from about 6 percent for O/N and 10 percent for 1Y.

Table 2.

Summary of I/B Transactions by Type and Duration (Total, February–December 2017)

Source: Central Bank of Myanmar.Note: The data for the I/B deposit market use that of the receiving side.

Table 2.

Summary of I/B Transactions by Type and Duration (Total, February–December 2017)

	Number of Transactions		Transaction Volume (MMK bil.)		Trade Weighted Average Interest Rate (percent)
	Borrowing	Deposit	Borrowing	Deposit	Borrowing	Deposit
O/N	236	236	875	1,146	5.75	6.06
1W	178	38	309	143	5.91	6.49
2W	59	12	112	42	6.88	7.11
1M	143	323	211	517	8.18	8.87
3M	103	155	101	167	8.87	9.27
6M	10	61	16	38	9.07	9.31
1Y	20	13	12	18	9.68	9.83
Call	-	50	-	17	-	8.26
Other	31	20	74	62	-	-
Total	780	908	1,710	2,152	-	-

Source: Central Bank of Myanmar.Note: The data for the I/B deposit market use that of the receiving side.

View Table

Table 2.

Summary of I/B Transactions by Type and Duration (Total, February–December 2017)

	Number of Transactions		Transaction Volume (MMK bil.)		Trade Weighted Average Interest Rate (percent)
	Borrowing	Deposit	Borrowing	Deposit	Borrowing	Deposit
O/N	236	236	875	1,146	5.75	6.06
1W	178	38	309	143	5.91	6.49
2W	59	12	112	42	6.88	7.11
1M	143	323	211	517	8.18	8.87
3M	103	155	101	167	8.87	9.27
6M	10	61	16	38	9.07	9.31
1Y	20	13	12	18	9.68	9.83
Call	-	50	-	17	-	8.26
Other	31	20	74	62	-	-
Total	780	908	1,710	2,152	-	-

Source: Central Bank of Myanmar.Note: The data for the I/B deposit market use that of the receiving side.

8. As for transaction parties, foreign banks’ presence is larger on the borrowing side in the I/B borrowing market, while their presence is high on both sides in the I/B deposit market. The interviewed banks suggested that they conducted arbitraging activities across and within the types of I/B transactions when the opportunities arise.

Table 3.

I/B Transaction Parties and Shares of Foreign Banks (Total, February–December 2017)

Source: Central Bank of Myanmar.Note: The data for the I/B deposit market use that of the receiving side.

Table 3.

I/B Transaction Parties and Shares of Foreign Banks (Total, February–December 2017)

		Borrowing/Receiving Deposit		Lending/Placing Deposit
		Borrowing	Receiving	Lending	Placing
Number of banks		20	18	16	28
	Foreign banks	13	5	8	8
Foreign banks’ share of transaction number (percent)		82	3	38	30
Foreign banks’ share of transaction volume (percent)		53	3	42	56

Source: Central Bank of Myanmar.Note: The data for the I/B deposit market use that of the receiving side.

View Table

Table 3.

I/B Transaction Parties and Shares of Foreign Banks (Total, February–December 2017)

		Borrowing/Receiving Deposit		Lending/Placing Deposit
		Borrowing	Receiving	Lending	Placing
Number of banks		20	18	16	28
	Foreign banks	13	5	8	8
Foreign banks’ share of transaction number (percent)		82	3	38	30
Foreign banks’ share of transaction volume (percent)		53	3	42	56

Source: Central Bank of Myanmar.Note: The data for the I/B deposit market use that of the receiving side.

9. Weekly data suggest that traded amounts and interest rates for each duration fluctuate rather widely. The CBM collects data daily, but there are many days without any activities, depending on the duration. Accordingly, weekly analysis of data was conducted on interest rates and transaction amounts during February–December:

Weekly I/B Transaction Volume and TWAR (February–December 2017)

Citation: IMF Staff Country Reports 2019, 087; 10.5089/9781498305198.002.A001

Source: Central Bank of Myanmar.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Weekly I/B Transaction Volume and TWAR (February–December 2017)

Citation: IMF Staff Country Reports 2019, 087; 10.5089/9781498305198.002.A001

Source: Central Bank of Myanmar.

• Download Figure
• Download figure as PowerPoint slide

Weekly I/B Transaction Volume and TWAR (February–December 2017)

Citation: IMF Staff Country Reports 2019, 087; 10.5089/9781498305198.002.A001

Source: Central Bank of Myanmar.

• Download Figure
• Download figure as PowerPoint slide

10. What drives these fluctuations of transaction amounts and rates is not clear. This analysis shows that the upper range of the market rates seemed to be limited by the CBM policy rate (10 percent). And there are gaps in rates in different I/B markets, which suggests that there is still room for arbitrage between markets. At the same time, I/B markets are not developed enough yet, as there are still weeks without any transactions, as shown.

Table 4.

Number of Weeks with No or Small I/B Transactions (February–December 2017)

Source: Central Bank of Myanmar.Note: The data for I/B deposit market use that of the receiving side.

Table 4.

Number of Weeks with No or Small I/B Transactions (February–December 2017)

	Types	O/N	1W	1M	3M
Number of weeks without any transactions	Borrowing	11	5	3	10
	Deposit	8	29	3	5
Number of weeks with total transaction amount less than 1 bil.	Borrowing	2	1	2	7
	Deposit	0	1	1	3

Source: Central Bank of Myanmar.Note: The data for I/B deposit market use that of the receiving side.

View Table

Table 4.

Number of Weeks with No or Small I/B Transactions (February–December 2017)

	Types	O/N	1W	1M	3M
Number of weeks without any transactions	Borrowing	11	5	3	10
	Deposit	8	29	3	5
Number of weeks with total transaction amount less than 1 bil.	Borrowing	2	1	2	7
	Deposit	0	1	1	3

Source: Central Bank of Myanmar.Note: The data for I/B deposit market use that of the receiving side.

(2) Recommendations

11. The advisor recommends that the authorities find out in more depth what motivates banks to use different types of I/B transactions in the first place. It is not quite clear yet how banks choose one type of I/B transaction over another on any given occasion. Focused interviews with banks will increase the capacity to analyze the collected data.

12. It is recommended that efforts to correct data discrepancy should be continued and that this issue should be solved as soon as possible. It is necessary to find the data discrepancy in a timely manner, correct the data, and ask the reasons for misreporting by communicating with banks. However, because the volume of reporting mistakes is a challenge for the available human resources, the CBM may want to consider increasing the number of staff allocated for the task of data collection and correction.

13. The advisor also recommends better use of the data by publishing the TWAR and total transaction amounts for each type of transaction, possibly weekly, with due consideration of the operational feasibility. It may be ideal if the daily data are available for key durations, but that would not be feasible for some time. An alternative for the time being would be to publish weekly TWAR for key durations, where transactions are concentrated, and total transaction amounts for each type of transaction, with due consideration of the operational feasibility. Interviewed banks indicated their need for market reference interest rates data published by the CBM.

Table 5.

Example of a Publication Template (Summary of I/B Transaction during the Week Starting on X/X)

Note: The periods of O/N, 1W, 1M, and 3M are selected, as their trading frequency and volumes are relatively higher and they are typically the important durations for making analyses on short-term money markets. As for the transaction volume, it is proposed to publish only the total volume, as the transaction volume of each duration can be small and traded parties may be able to infer other transaction details. In order to avoid the noise information, it is proposed not to publish TWAR for the duration, where less than a 1 billion MMK transaction is made during the week.

Table 5.

Example of a Publication Template (Summary of I/B Transaction during the Week Starting on X/X)

								(%, MMK bil.)
I/B borrowing/lending					I/B deposit receiving/placing
0/N	1W	1M	3M	Total transaction amount	O/N	1W	1M	3M	Total transaction amount
o.oo	x.xx	y.yy	z.zz	WW.W	o.oo	x.xx	y.yy	z.zz	WW.W
This summary is based on data collected from banks. Total transaction amount covers all the transactions in each type, not limited to durations mentioned in the table. All transactions are uncollateralized. TWAR (Trade Weighted Average Interest Rate) is not shown if no or only small transactions are conducted for the duration of the week. O/N includes transactions whose durations are 1–3 days. 1W includes transactions whose durations are 4–10 days. 1M includes transactions whose durations are 25–45 days. 3M includes transactions whose durations are 75–104 days.

Note: The periods of O/N, 1W, 1M, and 3M are selected, as their trading frequency and volumes are relatively higher and they are typically the important durations for making analyses on short-term money markets. As for the transaction volume, it is proposed to publish only the total volume, as the transaction volume of each duration can be small and traded parties may be able to infer other transaction details. In order to avoid the noise information, it is proposed not to publish TWAR for the duration, where less than a 1 billion MMK transaction is made during the week.

View Table

Table 5.

Example of a Publication Template (Summary of I/B Transaction during the Week Starting on X/X)

								(%, MMK bil.)
I/B borrowing/lending					I/B deposit receiving/placing
0/N	1W	1M	3M	Total transaction amount	O/N	1W	1M	3M	Total transaction amount
o.oo	x.xx	y.yy	z.zz	WW.W	o.oo	x.xx	y.yy	z.zz	WW.W
This summary is based on data collected from banks. Total transaction amount covers all the transactions in each type, not limited to durations mentioned in the table. All transactions are uncollateralized. TWAR (Trade Weighted Average Interest Rate) is not shown if no or only small transactions are conducted for the duration of the week. O/N includes transactions whose durations are 1–3 days. 1W includes transactions whose durations are 4–10 days. 1M includes transactions whose durations are 25–45 days. 3M includes transactions whose durations are 75–104 days.

Note: The periods of O/N, 1W, 1M, and 3M are selected, as their trading frequency and volumes are relatively higher and they are typically the important durations for making analyses on short-term money markets. As for the transaction volume, it is proposed to publish only the total volume, as the transaction volume of each duration can be small and traded parties may be able to infer other transaction details. In order to avoid the noise information, it is proposed not to publish TWAR for the duration, where less than a 1 billion MMK transaction is made during the week.

14. JICA agreed to provide IT solutions to support the CBM’s effort on this front. Additional IT solutions were necessary to conduct such a weekly task. With the acknowledgment of the CBM, the advisor discussed this issue with resident advisors of JICA, and they showed their willingness to provide necessary IT solutions. The sample database was provided to the JICA advisors with the CBM’s permission. JICA is communicating with the CBM staff to get the information on the current system configuration at the CBM for the data collection so that the agency can propose the best solution to match the CBM’s needs and estimate how long it will take to provide this IT solution.

15. In addition, based on these weekly data, the CBM should make further analyses of the market interest rate developments and factors behind those developments through market intelligence activities. In the medium to long term, it is also recommended that the CBM publish periodically its analyses on I/B market developments as a part of the “Financial Market Report in Myanmar.”

(1) Current Situation

16. The CBM issued the instruction for market repo transaction in Myanmar in December 2017. It is stated in this instruction, “Repo Transactions shall be conducted under contract forms that are consistent with Global Master Repurchase Agreement (GMRA) and international standard.” The daily reporting requirement is also stipulated in the instruction.

17. There were several repo transactions actually conducted by market participants in Myanmar. One transaction was conducted between two domestic private banks, and several transactions were conducted between a private domestic bank and one of the State-Owned Banks (SOBs) in March 2018. The repo rates of transactions involving one of the SOBs for O/N transactions were about 1 percentage point lower than that of uncollateralized O/N transactions. As mentioned in the following section, most foreign banks do not own government securities in their portfolio, so they cannot be expected to play a major role in the repo market. But it can be expected that SOBs play a major role in developing this repo market, as repo can provide credit risk control measures, which may fit their investment strategy, and the use of repo can contribute to reducing their excess holding of liquidity. SOBs’ activities in I/B markets to provide market liquidity have been limited, but, considering the large amount of excess liquidity SOBs have, it is expected that SOBs will become the major investors in the I/B repo market.

18. To delve into the development of government securities markets, several interviews were conducted with market participants. It was indicated that one of the major reasons foreign banks cannot invest in Myanmar government securities comes from their funding structure. Because foreign banks cannot pay any interest rates on their customer deposits, they tend to rely on the very short-term funding, such as a call deposit and interbank borrowing. Under this circumstance, foreign banks find it difficult to invest in government securities whose shortest duration is three months, due to the liquidity risk. Another reason may be the credit risk concern, as no country rating is available for Myanmar yet.

19. At the same time, it was indicated that some of the major investors slipped out from investing in government securities as a result of the introduction of the auction system and paperless issuance. Uncertainties on issuance prices, resulting from auctions and the need to revise their internal rules, resulting from paperless issuance, prevented those traditional investors of government securities from continuing their investment, and banks are eager to pay higher interest rates than government securities to obtain deposits from them. These traditional investors include big religious organizations and even the government pension fund. But it is not clear whether this kind of information is shared by relevant authorities.

20. Meetings with market participants indicated that there exists no forum for communication between the practitioner level market participants and relevant officials, namely at the CBM and MoPF. They have been separately conducting bilateral meetings with each market participant, but information is not fully shared among them. The demerit of this lack of forum was most obviously revealed when MoPF intended to introduce the bond switch last autumn. Communication among MoPF, market participants, and the CBM officials were not enough and, as a result, this bond switch was not accepted by banks due to operational reasons.

(2) Recommendations

21. As for the repo market, it is recommended to discover through interviews with banks the obstacles that may be hindering further development of the market. Possible obstacles include regulatory constraints, the presence of excess liquidity, and the lack of a proper legal framework and operational capability. Interviews with market participants, especially with SOBs, will be beneficial in understanding what is necessary to promote the market development.

22. As for the government securities market, the advisor recommends establishing the triparty working group, consisting of an issuer/regulator (MoPF and possibly SECM for the secondary market development), market participants (major investing banks and securities companies), and the custodian (the CBM) to discuss issues to enhance primary and secondary market activities. The terms of reference of this working group should address such issues as: (1) major traditional investors’ concerns over shying away from auctions with paperless issuance; (2) regulatory constraints, which hinder potential investors from actually participating in the market; and (3) other issues that prevent this market from developing (for example, accounting). Currently there is no “government securities working group” periodically held among these members at the practitioner level, and communication between the authorities and market participants is not enough. Sharing information and tackling issues collectively are vitally important for further development of this market.

(1) Current Condition

23. In March 2018, the CBM issued the reserve requirement schedule, which maintained the almost three months’ lag between the reserve base period and the maintenance period, as before. This follows the recommendation made in January that, considering the lack of information of operational feasibility to shorten the lag, it is recommended to keep the existing schedule at least for the time being². However, it was also pointed out that the CBM would want to shorten this timing gap as the first step to increase its monetary policy effectiveness in its efforts to modernize its policy framework.

24. In response to this recommendation, the CBM expressed its intention to shorten the timing gap between the reserve base period and the maintenance period from the current three months to two months. At the same time, it turned out that all banks submitted their balance sheet data of the end of March 2018 before the end of May, while until last year, the belated submission of the balance sheet data from banks was the basic cause of this long timing gap of three months. This earlier submission of the balance sheet data by all banks indicated that it could be operationally feasible to shorten the timing gap.

(2) Recommendations

25. The advisor recommends starting to consider shortening this timing gap, as it was shown that it is possible for all the banks to submit their balance sheet data earlier than in the past. Further communication with some banks may be necessary to ensure their commitment to submit their balance sheets in a timely manner.

26. In addition, the macro- and micro-level impact of the shortening of this gap should be analyzed. And its result should be shared with the CBM officials responsible for the prudential areas, for example, the Financial Institutions Supervision Department, and their opinions should be sought. If shortening the gap is approved by the CBM, it should be communicated with banks in a timely manner in order to make time for the necessary preparation.

(1) Current Condition

27. The CBM has two credit facilities; one is an overnight facility and the other is “discount window” lending. The terms and conditions for the use of the overnight facility is clearly stipulated in “Current Account and Overdraft Guideline (Version 1.0),” issued by the CBM in December 2015. Overnight facility is provided if intraday overdraft is not repaid on the same day. Accordingly, the use of the overnight facility is essentially limited to fill the negative amount in a bank’s current account held at the CBM. The use of the overnight facility is “permitted for up to three consecutive days against the collateral held in the CBM” and “[i]nterest for overnight shall be charged at official central bank lending rate” (which is currently set at 10 percent), and “[b]anks having overnight facility beyond three consecutive days shall be subject to penalty set by the CBM” (which is currently set at 2 percent, so that the overall rate is 12 percent, after applying a penalty). As a result, this overnight facility cannot be used to fill the required reserves or other funding needs by banks, unless the bank has the negative account balance at the CBM.

28. There exists another reason the use of the overnight facility is limited. As mentioned earlier, the penalty rate is attached to this facility for the use of more than three consecutive days, while “discount window” lending does not have any penalty and can borrow for longer periods (92 days) without any penalty. Accordingly, it is considered that banks naturally prefer using a discount window facility when they face the need for short-term liquidity support from the CBM.

29. At the same time, however, terms and conditions for the use of the credit facility of discount window are not very clear. This facility is based on the instruction issued on November 14, 1995, under the old central bank law (Section 71), and, according to the CBM staff, the legal basis for this discount window has been inherited to Section 86 of the new central bank law, which reads:

• “In the circumstances of requesting for financial assistance to improve liquidity by the banks, the Central Bank may, on such terms and conditions as it may from time to time determine, grant financial assistance at appropriate interest rate for periods not exceeding 92 days based on the following:

• (a) The bank is solvent and can provide adequate collateral to support the loan;

• (b) The bank needs such financial assistance to preserve the stability of the financial system.”

However, as the instruction was issued in 1995 based on the old central bank law, terms and conditions for this discount window is not very clear for banks. For example, it is not clear whom to contact, what procedure is necessary, and what haircut is applied for the collateral³.

30. In addition, the current credit facility framework of the CBM essentially excludes its use by the foreign banks, as most of them do not own government securities. The only eligible collateral for the overnight facility and the discount window is government securities, but most of the foreign banks do not own government securities due to the reason mentioned earlier in this report. Under this situation, for example, any simple mistakes in the liquidity management of foreign banks can cause disturbances in the financial system.

31. And there exists the legal issue as well. The discount window facility is legally regarded as the “lender of last resort” (LLR) facility. The legal basis for this facility, Section 86 as described above, is stipulated under Chapter XIII: “Lender of Last Resort.” This legal basis may cause the situation where banks avoid using this facility due to the possible stigma attached to the facility, that is, the fear of being regarded as the bank in need of special help from the CBM.

(2) Recommendation

32. The advisor recommends that the CBM holistically review the framework of its various facilities to ensure smooth market functioning and maintain stability of the overall financial system. This review should include: (1) how to include foreign banks in the CBM’s framework of short-term liquidity support; (2) how to clarify the terms and conditions of the central bank’s discount window facility; (3) how to set the legal basis of the “discount facility” as a monetary operation tool rather than a lender of last resort; (4) how to transform the discount window facility into the standing credit facility with the overnight maturity; and (5) how to streamline the use of the overnight facility and the discount window facility. It would help if the CBM issued another instruction, clarifying the terms and conditions of the discount window facility, and send it to all banks, both foreign and domestic. The advisor suggests that doing so would reestablish the discount window facility, based on Chapter VII (“Maintaining Monetary Stability”), Section 45 of the new CBM Law, rather than the current LLR basis, to avoid possible stigma.

(1) Current Situation

33. The CBM needs to respond to the change in the fiscal year, starting from this year, from April–March to October–September. In the transition to the new fiscal year, the government has set the gap fiscal year from April–September 2018. The advisor supports the officials of MPAD of the CBM to establish weekly liquidity projection worksheets for this gap fiscal year, based on fiscal information provided by MoPF, which clarified its planned expenditure and income amounts. It was expected that the fiscal expenditure would spike at the end of the gap fiscal year (September 2018), as had been the case at the end of the fiscal years in the past (every March).

34. Another challenge is envisaged for the liquidity forecasting of the next fiscal year. It is necessary to prepare for the next fiscal year, October 2018–September 2019, for which no previous patterns of government expenditure and reserve money (both currency in circulation and current account balance) are available. Informed adjustment will be necessary in forecasting liquidity conditions.

(2) Recommendation

35. In order to ensure the smooth preparation for the next fiscal year, a recommendation was made to start preparing the liquidity forecasting for the next fiscal year as soon as possible. The response to the change of the fiscal year is the challenging task, and starting this task as soon as possible may be helpful in providing insights on necessary points to consider at an early stage.