The Evolving Role of Central Banks

23 Kenya’s Transition from Direct to Indirect Instruments of Monetary Policy

Patrick Downes, and Reza Vaez-Zadeh
Published Date:
June 1991
  • ShareShare
Show Summary Details

Kenya has been engaged in far-reaching financial sector reforms aimed at improving monetary control and at improving the efficiency of financial intermediation. Exclusive reliance on credit ceilings, without an adequate mechanism for implementing such ceilings, combined with controls on interest rates had resulted in the progressive weakening of monetary control and had produced the inefficiencies in financial intermediation that are characteristic of the use of direct monetary policy instruments. This paper discusses the implementation and the sequencing of reform measures that led to the use of market-based techniques of monetary control.

The approach adopted by Kenya in moving from direct to indirect instruments of monetary control is instructive for three reasons. First, it was recognized that the liberalization of interest rates was a necessary prerequisite for the successful use of market operations to control money and credit aggregates. Hence, the authorities undertook a gradual relaxation of interest rate controls, focusing the removal of controls on treasury bill yields. Second, since government securities markets were to a significant degree segmented from other financial markets, the immediate and exclusive reliance on open market operations in government securities for purposes of monetary control was not desirable. Instead, it was necessary to broaden and deepen the market for government securities to remove the segmentation, and thus improve the transmission of policy effects from the monetary to the real sector. Third, Kenya introduced an innovative transitional arrangement in the form of a market-based enforcement mechanism for credit ceilings to improve control of money and credit, until open market operations could be used in sufficient volume.

The Issues

In 1989 and early 1990, the Kenyan monetary authorities faced major difficulties in implementing monetary policy.1 Credit ceilings were, in practice, the main instrument at their disposal to manage liquidity. These ceilings were not fully effective, however, since enforcement was weak; therefore, compliance was poor, and the ceilings could not in any event be used to fine-tune short-term liquidity movements. Furthermore, the monetary authorities were not in a position to sell sufficient government securities in the open market to control liquidity, because of an administered interest rate structure. As a result, any unprogrammed increase in reserve money—frequently due to government borrowing from the Central Bank of Kenya—could not be fully sterilized through the sale of government securities and resulted in a multiple expansion in the money and credit aggregates.

The development of a viable market in government bills and bonds was held back by two major impediments. The first was the growing gap between yields on such securities—particularly treasury bills—and effective bank lending rates. Through the application of fees and charges, banks had been able to increase effective rates above administered rates for a substantial portion of their lending to the private sector. For example, although the maximum rate for short-term lending was 15.5 percent in March 1990, the effective rate was estimated to be in the range of from 20 to 22 percent. At the same time, the maximum yield on treasury bills was about 14 percent and on treasury bonds about 19 percent, that is, not sufficiently high to generate voluntary demand for securities from the financial sector. Similarly, with treasury bill yields below the rate on time deposits of 15-16 percent, the nonfinancial sector also had no incentive to acquire bills on a voluntary basis. Under these circumstances, the sale of government securities was mostly to captive lenders, and actual quantities sold were well below desired levels.2

On the other hand, although, the spread between the rate on savings deposits earned at banks, and the yield on treasury bonds was about 6–7 percent in favor of treasury bonds, the authorities were not able to sell a sufficient volume at the prevailing bond rate to absorb excess liquidity. Thus, pricing was not the only factor holding back sales of bonds. The second impediment to increasing their sale, therefore, was identified as an insufficiently developed institutional structure in the government securities markets. In particular, the absence of an effective marketing and distribution system, as well as the absence of liquidity caused by the lack of a market maker was hampering the sale of bonds to the nonflnancial sector.

The reluctance to raise yields on government securities was based on the view that demand for such securities could be increased without further increases in yields and that, indeed, further increases in yields would not have much effect on demand. The fact that a yield on treasury bonds of 19 percent exceeded the rate on savings deposits by 6.5 percent, and that real interest rates were in excess of 5 percent, tended to support the hypothesis that informational and institutional shortcomings had effectively segmented the treasury bond market from other markets, thus severely hampering the transmission of monetary policy to the real sector.

Hence, rather than allowing yields to find their market-clearing levels in a market that was perceived to be segmented and narrowly based, the authorities undertook an ambitious program to broaden and deepen the market for government securities. Among these were the introduction of bearer bonds to preserve the anonymity of the investor; the expansion of the role of the Central Bank as market maker through outright purchases and sales designed to increase liquidity in securities markets; and finally, the publicizing of the pricing and availability of government securities, particularly to financial cooperatives, public sector employees, and the Asian community. These measures were then reinforced by a gradual increase in yields on treasury securities, particularly those on treasury bills. In late 1990, the authorities implemented the final step in the liberalization process, namely, the freeing of treasury bill rates. Simultaneously, the Central Bank established its operational capabilities in the money market and set up the data gathering mechanism needed to sterilize effectively any disturbances to the monetary base.

It was expected that these efforts to increase the sales of government securities would take some time to become fully effective. In order to ensure some degree of monetary control in the interim, the authorities implemented a transitional credit control measure to achieve a higher degree of compliance with the ceilings on bank credit to the private sector. In particular, banks were required to deposit 20 percent of any credit in excess of individual bank credit ceilings into an interest-free deposit with the Central Bank. The size of the deposit in relation to the excess credit was uniform across all banks and was calculated from the credit multiplier so as to reduce the growth in reserve money to levels consistent with credit ceilings. The effect of this penalty deposit scheme was to reduce the income from loans extended beyond the credit ceiling to approximately what could be earned by investing in treasury bonds.

Monetary Policy Problems and Arrangements

A review of monetary developments in Kenya in early 1990 identified two major problems: a permanently high variability in the stock of reserve money and, since the end of 1988, an unprogrammed expansion in monetary and credit aggregates. Both problems raised questions about the efficiency and effectiveness of the monetary policy arrangements. Table 1 illustrates the variability in reserve money and shows that government overdraft at the Central Bank has been one of the main causes of this increase. The monetary data in Table 2 show that growth in liquidity throughout 1989 was increasing rapidly.

Table 1.Sources of Variation in Reserve Money, 1987–89(In millions of Kenyan shillings; end of quarter)
Reserve money+8+1,769-943-602+627+ 1,137+ 771-707+ 702+1,371
Government overdraft+ 1,544+1,024-1,664+ 425+ 396-1,799-170-1,025-577-1,128
Net foreign assets-1,663+ 670-384-484-66-466+987-562-1,272+431
Advances and discounts to banks+5-5+51+ 166-216+ 152223+ 12-12+ 1,019
Treasury bills and bonds+ 207-162+ 425+ 63-500+ 1,789-27+398-13
Government stock-2,000
Uncleared effects-55+ 59-1,264-1,313-484+ 642+ 1,094-397-328+ 1,274
Other items (net)-177-186-48+ 177+934-490-704+ 168-683-212
Source: Central Bank of Kenya.
Source: Central Bank of Kenya.
Table 2.Monetary Data, 1987–89(Annual change in percent)
Reserve money18.91.817.317.316.717.3
Net domestic assets18.110.711.714.219.114.1
Total domestic credit20.
Money and quasi-money12.48.312.717.820.717.8
Sources: Central Bank of Kenya; and Fund staff estimates.
Sources: Central Bank of Kenya; and Fund staff estimates.

Present Policy Arrangements

The main instruments of monetary and credit policy in Kenya are quantitative ceilings on total domestic credit and net bank credit to the Government.3 The Central Bank controls bank credit to borrowers other than the Central Government through monthly credit ceilings on each bank. In practice, the Central Bank has attempted to enforce observance of these ceilings through moral suasion. The treasury tries to ensure compliance with the credit ceilings on outstanding lending by the Central Bank to the Government by controlling the level of the Government’s overdraft with the bank.

The main advantage of credit ceilings was that of a quick-acting, short-term monetary policy measure to restrict money and credit expansion. Kenya’s experience, however, particularly during 1990 and 1991, demonstrated how inefficient and ineffective credit ceilings are as the only instrument of monetary policy. The authorities were unable to reduce fluctuations in reserve money, nor could ceilings contain the significant growth in liquidity that had been building up in the system since the end of 1988. The ineffectiveness of this ceiling is illustrated by the fact that the four major banks in Kenya were in excess of their credit limits to the private sector by K Sh 3.0 billion at the end of January 1990, which is 14.3 percent on average above the ceilings (the banking sector as a whole was 8.5 percent above its aggregate limits).4

Credit controls also introduced rigidities and reduced competition in the financial system, and resulted in a misallocation of resources. In addition, borrowers and lenders in the financial markets have been successful in avoiding such controls by channelling lending to wholly owned nonbank financial institutions.

The Central Bank was not able to channel excess liquidity into the treasury bill and bond markets through sales of such securities. Current yields and institutional arrangements made it difficult to sell the necessary volume of government securities into the open market. Demand in these markets has been generated mainly by statutory liquidity requirements placed on the NSSF and insurance companies, and by the minimum liquidity requirements of the commercial banks and nonbank financial institutions. Efforts to achieve better monetary control have, therefore, concentrated on improving the ability of the Central Bank to sell government securities. This aim is being pursued along two avenues. First, yields on treasury bills and bonds are being allowed gradually to reflect market forces fully. Second, the institutional structure of the government securities market is being improved.

Simultaneously, the Central Bank is formulating on a continuing basis a reserve money program, entailing projections of the supply and demand for reserve money so as to determine the volume of government securities that must be sold in order to achieve the desired growth in base money. To this end, the bank is establishing a statistical framework for monitoring and forecasting influences on reserve money and has begun compiling on a daily basis the historical data required. This monetary programming framework will consist of a mechanism to (a) smooth short-term liquidity conditions and interest rates in the money markets, which should help the growth of the secondary market; and (b) ensure that monetary expansion occurs in line with its financial program. These statistical data are being supplied by the relevant departments of the Central Bank, the Ministry of Finance, and other government departments.

With a reliable statistical framework in place by mid-1990, the Central Bank started to conduct shadow, open market operations. These operations allowed the relevant central bank departments to gain expertise in monetary programming and in trading with the financial sector. This approach also allowed banks to become accustomed to this type of operation.

Reform of Interest Rate Policies

The Kenyan authorities recognized early that an environment of flexible interest rates is a vital prerequisite for the development of an effective monetary policy framework, operating through market-based instruments. This section reviews interest rate policies and the problems associated with these policies, as well as the reform efforts.

Present Interest Rate Policies

Kenyan authorities had been regulating interest rates of commercial banks since 1974 by imposing a maximum lending rate and a minimum savings deposit rate. In April 1989, the system was amended slightly: the maximum rate was split into a rate for short-term lending and one for lending with maturities of four years or more. This change brought the commercial banks’ long-term lending rate in line with that of the nonbank financial institutions. The maximum rate for long-term lending was applied to loans of more than three years in November 1989.

The effective interest rate structure, however, differed from this administered framework, mainly because of the existence of an active wholesale market and because of the application of fees and charges (Table 3). Effective interest rates on time deposits in June 1990 were 15–16 percent, and 7–9 percent on wholesale current accounts (current accounts are nominally non-interest-bearing). This implied that roughly 48–50 percent of the banks’ funding is at the minimum rate (savings accounts and non-interest-bearing current accounts) against 50–52 percent that is attracted at a higher cost.

Table 3.Interest Rate Structure, October 1989(In percent)
Effective bank lending rates22–24
Yield on treasury bonds (two years)24
Maximum long-term lending rate119
Maximum short-term lending rate117
Central Bank rediscount rate217.5
Interbank rate, overnight6–19
Discount rate on treasury bills114
Minimum savings deposits rate113.5
Wholesale current accounts7–9
Retail current accounts0
Certificates of deposit14–17

Administered rates.

Rediscount of government paper. The rediscount rate for private paper is 1 percentage point higher.

Administered rates.

Rediscount of government paper. The rediscount rate for private paper is 1 percentage point higher.

In general, the yields on government securities have been set more or less in line with administered rates. The discount rate on treasury bills was, in practice, predetermined by the treasury. Each week, the Central Bank offered for tender K Sh 2 billion in 90-day bills. Tenders were accepted or rejected on the basis of this discount rate. In practice, the accepted rate was only marginally different from the result of the previous week’s tender. The actual yield was slightly higher than the accepted discount rate.5 Bidders were, therefore, very cautious in their bids; and there is evidence that several market participants, who were in need of large amounts of treasury bills in order to comply with regulatory requirements, consulted with the authorities prior to the tender. Since the fourth quarter of 1988, there has been a tendency to increase the rate on treasury bills only gradually.6 As a result of the procedures, the amounts accepted are in most cases much lower than the amounts tendered, as Table 4 shows.

Table 4.Treasury Bill Tenders—Overview of the Amounts Allotted(In millions of Kenyan Shillings)
Amount Offered1Amount AcceptedAs Percentage of Amount Issued
Source: Central Bank of Kenya.

K Sh 2,000 million per tender.

Three tenders of K Sh 2,000 million and one, at the end of the month, of K Sh 5,000 million.

Source: Central Bank of Kenya.

K Sh 2,000 million per tender.

Three tenders of K Sh 2,000 million and one, at the end of the month, of K Sh 5,000 million.

Starting in September 1989, treasury bonds (with maturities of one, two, and five years) have been auctioned on a monthly basis. Previously, the Government only held auctions irregularly. The treasury fixes the amount of the issue, and the effective yield is determined by the price bids that are acceptable to the treasury.

Three important conclusions can be drawn from the preceding overview: (a) effective bank lending rates were significantly higher than the administratively prescribed rates; (b) the yields on government paper—and particularly on treasury bills—were set in line with administered rates; and (c) average savings deposit rates were significantly lower than the yields on government bonds (6-7 percent). Despite this latter yield gap, almost no nonbank savings are channeled into government paper. This fact was taken as indicative of the existence of severe institutional rigidities in the government paper market.

Problems Arising from Interest Rate Regulations

That yields on government paper are held closely to the administered bank lending rates creates a growing discrepancy with the effective bank lending rates. This is a major reason why the monetary authorities have been unable to sell sufficient quantities of government securities to control base money growth.

The fact that the yield on bills is 7–8 percent below the effective bank lending rates makes it difficult to attract genuine demand from the banking sector. Yet, in a financial market like Kenya’s, where financial intermediation is the predominant way of channeling savings from surplus to deficit units, this sector should be an important source for government borrowing. Second, the fact that the treasury bill rate is 2 percent below the wholesale time deposit rate makes this paper unattractive for other potential investors, such as large companies.

As regards treasury bonds, the fact that five-year bonds are yielding 2–3 percent less than short-term bank lending also means that the financial sector will not voluntarily hold bonds. Another economically relevant aspect of the yield structure, however, is the premium by which the bond yield exceeds the banks’ savings deposit rate.

The existing demand for government securities is largely generated by induced purchases by important captive lenders. It can be seen from Table 5 that for bills, bonds, and stock, the “parastatals and others” are by far the most important holders. At the end of the financial year 1988/89, parastatals and others held 52.3 percent of total domestic government debt. The NSSF and insurance companies (private and government owned) are the most important institutions in this category. They are subject to stringent holding requirements for bills and bonds. The parastatals are not subject to legal requirements, but they are frequently urged by the treasury to invest their surplus funds in government securities.

Table 5.Outstanding Government Domestic Debt(June, end of financial year)
In millions of Kenyan shillingsIn percent of totalIn millions of Kenyan shillingsIn percent of total
Direct Central Bank
advances to
Treasury bills15,59833.615,46631.7
Nonbank financial
Parastatals and
Treasury bonds9,16019.713,16927.0
Nonbank financial
Parastatals and
Government stocks13,92830.011,45123.4
Nonbank financial
Parastatals and
Total govern-
ment debt46,442100.048,833100.0
Source: Central Bank of Kenya.
Source: Central Bank of Kenya.

Only 29.8 percent of total domestic government debt is in the hands of banks and nonbank financial institutions. The data indicate that these securities are primarily held to comply with the liquid assets ratio or as a cushion for borrowing from the Central Bank.7 In June 1989, the commercial banks’ treasury bill portfolio amounted to K Sh 4.7 billion. The required amount according to the liquid assets ratio at that time was K Sh 4.2 billion. The residual, K Sh 0.5 billion is the “cushion.”

Another problem created by selling securities into captive markets is that—in addition to the lack of volume—the market in its present form is not liquid enough to permit the Central Bank to intervene through short-term purchases or sales to financial institutions. Most of the institutional investors will hold the securities until maturity.

The combination of below-market administered lending rates and credit ceilings has produced excess credit demand. This excess demand has reinforced banks’ preferences for safer, connected borrowers, thereby limiting the credit available to long-term projects and to new or more risky, entrepreneurial borrowers. This behavior has resulted in preferential treatment for larger companies and parastatals. A derived effect from this policy may be that the banks’ managerial skills are not as developed as they would be in an environment with more diversified investment opportunities.

Below-market administered interest rates have also tended to favor debt financing over equity financing, which, together with other institutional shortcomings, explains the decline of Kenya’s equity markets during the last decade. Recently, efforts have been undertaken to revitalize the capital markets, inter alia, through the establishment of a Capital Markets Authority.

Finally, below-market administered interest rates have prevented the authorities from considering the liberalization of exchange controls, because of the fear that this measure would induce large capital outflows.

Recent Policy Measures

The approach to interest rate liberalization has proceeded along two lines:

First is alignment of the administered bank interest rate structure with the market-clearing structure. Effective April 1, 1990, the ceilings on interest rates on deposits and on short- and long-term loans were raised by 1 percentage point, to 13.5 percent, 16.5 percent, and 19.0 percent, respectively. The ceiling on short-term loans was raised by a further 0.5 percentage point at the end of August 1990, to 17.0 percent. More important, the authorities removed the legal requirement stipulating that lending related fees and charges be inclusive of the loan interest rates subject to ceilings. This means that effective interest rates can significantly exceed the formal ceilings and that they are market determined.

Second is that yields on government securities are to be market determined. This was first and foremost required for treasury bills, as this market will be used for open market policy purposes. The same policy also holds for yields on government bonds, although institutional growth in this market is being emphasized equally. The rates on treasury bonds are fully liberalized, and the yields on treasury bills will be liberalized effective November 15, 1990, allowing price to be set by true and proper auction.

The move toward equilibrium interest rates was self-reinforcing: the gap between government securities’ yields and the effective bank rates narrowed, and the greater volume of government paper sold reduced the growth in reserve money caused by government borrowing from the Central Bank. Furthermore, the reduction in inflationary expectations lowered the effective bank lending rates and narrow the gap between them and government securities’ yields. The reforms of the administered interest rate structure will be completed by June 1991.

Institutional Reform of the Government Securities Markets

The previous section has pointed out that the relatively low yields are a major barrier to the smooth development of government securities markets. In addition, however, institutional and distributional rigidities have the effect of segmenting the government securities markets from other markets, most notably the banking markets. It was, therefore, necessary to undertake measures to broaden and deepen government securities markets before full interest rate liberalization could become effective.

In this regard, it is useful to draw a distinction between the treasury bill market and the bond market, inter alia, because the target group of potential investors differs. The treasury bill market will, at least in the initial stages, be mainly a market for professionals, that is, banks, other financial institutions, and large nonfinancial companies. The treasury bond market, on the other hand, should be adequately developed so as to offer investment opportunities for the nonfinancial sector as well. In this way, the small investor would be in a position to diversify his asset portfolio.

Reforms of the Treasury Bill Market

Although the treasury bill market suffered in the main from a yield that was too low, its development was also hindered by a number of institutional problems.


The main institutional problems in the treasury bill market concern the tender method. Efficient tender procedures are the heart of well-functioning government securities markets. They have a bearing on the marketability and the liquidity of the market, and for that reason also on monetary policy. A first requirement is that tenders must be “clean.” Currently, the treasury determines the discount rate on bills, and competition is undermined by informal consultations between large bidders and the authorities to determine which bids are acceptable. This has resulted in few bids in excess of the permitted yields, and, consequently, the sales volume has been less than the amounts on offer.

The Central Bank has now begun to preannounce auction dates and amounts on offer and to solicit sealed bids. It is planned to let the treasury bill tender amounts be determined by the reserve money program and not by the needs of the treasury. Bids will be accepted beginning with the highest, until the amount on offer is exhausted. During a transitional period, the Central Bank will seek to prevent excessive rate fluctuations in tenders by limiting the maximum change in the successive tender rates (e.g., 0.5 percent a year in any one direction). The market participants will be informed of this policy and of this range. As the tenders develop, the present practice of irregular “on tap” sales will also cease; a secondary market in treasury bills and bonds will remove the need for this type of operation.

A more competitive primary market should, in principle, provide a sounder foundation for monetary policy operations. Initially, the primary market will be used for monetary policy purposes. But since the use of the primary market for monetary policy purposes may create conflicts between monetary policy and fiscal policy priorities, the Central Bank will conduct monetary policy through the secondary markets.8

Primary Dealers

If the first two measures fail to generate turnover in treasury bills, then the Central Bank may ask a group of financial institutions to act as primary dealers. Participation in tenders would then be restricted to such dealers. In return, the dealers would be required to make a two-way market in government securities. The dealers might be established as subsidiaries of the commercial banks. They might serve also as market makers in other segments of the money market. Setting up a dealer network would require, of course, efficient supervisory guidelines, including in particular, minimum capital requirements.

Supplementary Measures

As the treasury bill market develops, the Central Bank will encourage other money market instruments as well. At present, the interbank market is exclusively an overnight market. The development of an efficient interbank market (with a variety of maturities) would allow banks to manage their liquidity more easily and would make the liquidity conditions more homogenous across the financial system. One way to encourage the interbank market is to make the lender-of-last-resort facilities penal. A liquid interbank market could provide also a benchmark rate for transactions in the fledgling secondary treasury bill market, and for that reason improve the efficiency of the latter.

Reforms of the Treasury Bond Market

The institutional rigidities and shortcomings that prevent the non-financial sector from investing in bonds mainly concern an inadequate marketing and distribution system; the absence of market makers to provide liquidity; the lack of transparency in tenders; and the unavailability of bearer instruments. As a result, the sale of securities to small investors has been minimal, despite the substantial interest differential in favor of government securities. The interaction of these measures with the recommendations regarding the yield should enable the Central Bank to exploit the market for government securities. A deepening of these markets will then allow the authorities to absorb part of the liquidity overhang that is currently in the system.

Special Issue of Treasury Bonds

A special issue of treasury bonds is meant to promote the sale of treasury bonds on a large scale. The target group is the nonbank sector, and especially the personal investor—public sector employees, farmers, and cooperatives. In order to acquaint this group with longer-term types of investment, one-year bonds are suggested for the first time. The Central Bank will fix the conditions of issue, and the bonds will be offered for subscription during a predetermined period.

Depending on the results of this issue, the authorities will decide whether to continue with a series of “special” issues, alongside the auctions for professionals, or to supplement the auction with non-competitive bids for small investors.

Purchase Facility for Small Volumes of Bonds

Holders of small amounts of bonds will want to be reassured on the liquidity of their investment. Professional investors are likely to be unwilling to purchase small quantities from personal investors; and the costs involved in locating a professional counterparty and completing the transaction are likely to be high. One way of enhancing the liquidity of small holdings of securities is for the Central Bank to establish a small-purchase facility. The Central Bank would stand ready (either directly or through the Post Office) to repurchase, on request, treasury bonds at current market rates (perhaps adjusted for a handling fee).

To avoid stifling the secondary market for major investors, and to prevent banks and others from gaining uncontrolled access to central bank funds, repurchases will be restricted, however, to personal investors completing the appropriate “application-to-sell” forms. A facility of this kind can increase the attractiveness of bonds at a low cost without harming the secondary market.

Bearer Securities

Bearer securities, as opposed to registered securities, provide many advantages, and the Central Bank is at present investigating actively their introduction—in particular, the security details of printing and the range of denominations to be offered. The issue of bearer securities should overcome the concern of investors that information regarding their holdings and interest income will be transmitted to the Commissioner of Inland Revenue. Bearer securities should encourage also the holding and trading of paper at places distant from Nairobi and Mombassa.

Effective Price Information

Offering coupon rates of interest that are closer to the effective yield could increase subscriptions. Many people may not realize that rates of return are dependent on the actual price paid for the securities, together with the coupon rate. They could be deterred by apparent returns of only 16 percent a year. The authorities will work toward improving the public’s appreciation of the actual yield they will receive on their investment, as part of the marketing campaign. If this does not appear to be successful, the authorities will consider higher coupon rates as a way of encouraging subscriptions.

Consolidating the Maturity Dates

For some time the Government has, on occasion, issued treasury bonds, but in 1989 it began a series of monthly tenders. Each tender included securities with one, two, and five years’ maturity. Rigid adherence to these maturities has produced a proliferation of stocks and maturity dates. There are 35 series currently on issue, each with a separate maturity date. Ten series will mature in 1990, eleven in 1991.

This multiplicity of stocks adversely affects their marketability. Reducing the number of series but increasing the volume on issue for each maturity date will widen the number of holders of each stock. The greater the number of holders and the larger the volume of stock, the more likely it is that secondary trading will occur.

This is to be achieved through two steps. First, a single maturity date will be chosen for each month—ideally the same date for each month of the year to simplify the trading process. Any date can be chosen; however, 10 of the 35 series currently mature on the fifteenth day of the month, so this might be a good initial choice. Second, the authorities will move away from the issue of strict one-, two-, and five-year maturities. By selecting a single maturity date for each month and issuing stocks that do not have an exact number of years to maturity, the authorities can reduce, over time, the number of maturity dates in each year.

An Asset Reserve Requirement

It was recognized that institutional reforms and monetary policy framework development were likely to be slow to take hold; hence, it was necessary to provide a backstop during the transitional period. Better compliance with the ceiling on commercial bank credit to the private sector could provide the necessary control of credit in the interim. Once the Central Bank is in a position to trade government securities in the open market, it will no longer be necessary to rely on credit ceilings to control monetary expansion.

Enforcement would take place by requiring noncomplying banks to place a non-interest-bearing deposit at the Central Bank equivalent to 20 percent of the amount by which the banks have exceeded their credit ceiling. The size of the deposit in relation to the excess credit would be uniform across all banks and would be such as to reduce the banks’ reserves at the Central Bank to the level that supports credit extended up to the ceiling only. It was expected that banks with excess credit positions will initially finance the required deposits through increased borrowing from the Central Bank. It was expected that access to such borrowed funds would be administered liberally during the transition phase, that is, prior to May 31, 1990.

A stricter enforcement of the credit ceiling on bank credit to the private sector would most likely also generate additional demand for government securities by financial institutions, in the event of unprogrammed growth in reserve money. Some of the excess liquidity created by government borrowing from the Central Bank would then flow into government securities.

Reserve Money Programming

The Central Bank has begun to develop the statistical base for a monetary programming exercise—to be able to forecast influences on reserve money. Changes in the supply of cash reserves result from transactions between the public (including the banks) and the Central Bank. These transactions can be divided broadly into nondiscretionary factors—those that are largely beyond the short-term control of the bank—and the Central Bank’s discretionary operations. Forecasting of influences on reserve money is concerned with the nondiscretionary transactions and the Central Bank’s policy action with discretionary operations. Changes in the demand for cash reserves depend on banks’ demand for reserve balances. This, in turn, depends on the growth of deposits in the banking system and cash ratio requirements. The main nondiscretionary transactions are listed below. Depending on their sign, the items will either increase (+) or decrease (−) the cash reserves of the banks:

(1) net purchase (+) or sale (−) of foreign exchange by the Central Bank;

(2) increase (−) in currency in circulation;

(3) net payments (+) by government (increase in overdraft or drawdown of deposits);

(4) net redemption (+) or issue (−) of treasury bills and bonds; and

(5) increase (+) in access to refinancing and advances windows at the Central Bank.

The main discretionary policy actions by the CBK that influence cash reserves are:

(6) sales (−) or purchases (+) of treasury bills;

(7) sales (−) or purchases (+) of treasury bonds; and

(8) central bank lending (+).

The sum of items (1) through (8) determines the overall increase in cash reserves. This increase has to be compared with the additional demand for cash reserves by the banking system to meet required cash ratios and the level of desired precautionary balances. By comparing the level of cash reserves that is likely to be supplied with the level that is forecast in line with the Fund-supported monetary program, the level of bank reserves that needs to be absorbed or injected can be determined. From these calculations, the volume of government securities to be sold or purchased in the Central Bank’s open market operations can be determined.

Once all the monetary data is available and in place, the Central Bank can start to train staff for “shadow,” open market operations, with a view to daily community operations in 1991. This strategy would, of course, need to be agreed to by the Monetary Policy Committee and other decision makers, and projections would need to be discussed on a regular basis in these forums. The Monetary Policy Committee should receive regular reports on monetary conditions to enable it to assess progress toward its longer-term monetary policy objectives and to assess the success of previous projections and decisions based thereon.

Finally, the Central Bank has undertaken efforts to ensure that its operation of monetary policy is well understood in the financial markets, including those parastatals who have invested a considerable amount of their portfolio in government securities. To this end, the objectives of monetary policy will be stated clearly, and any relevant information on monetary conditions publicized.


    Folkerts-LandauDavidMarcQuintyn andRichard J.Walton“Kenya: Toward Improved Monetary Control” (unpublishedWashington: International Monetary Fund1990).

    • Search Google Scholar
    • Export Citation

    Folkerts-LandauDavidPeter M.Garber andAftab S.Ahmed“Kenya: Development of Money and Capital Markets” (unpublishedWashington: International Monetary Fund1991).

    • Search Google Scholar
    • Export Citation
*The author is Deputy Division Chief of the Financial Studies Division of the International Monetary Fund. Peter Garber, Peter Heller, Maurice Kanga, Donald J. Mathieson, and V. Sundararajan provided helpful comments during the preparation of this paper.
1For a detailed review and description of the issues, see Folkerts-Landau and others (1990) and (1991).
2Banks and nonbank financial institutions demand treasury bills to satisfy a liquidity requirement; the National Social Security Fund (NSSF) and insurance companies are mandated to hold bills and bonds, and at times, parastatals are directed by the treasury to hold government bills and bonds, and at times, parastatals are directed by the treasury to hold government securities.
3These ceilings constitute performance criteria for the Fund-supported adjustment program. As a result of the problems mentioned in the previous section, the 1990 Article IV consultation mission recommended a change in the performance criterion for total domestic credit to a ceiling on net domestic assets, which will cover the “Other items net.” This, however, will not resolve other more general problems of compliance with the credit ceilings which are discussed below.
4The banking system as a whole could avoid breaching the credit ceilings only if it were willing to hold non-interest-bearing excess reserves at the Central Bank or to invest in government securities at below bank-lending rates for comparable maturities.
5The accepted discount rate for the March 5 tender was 14.0 percent, while the effective yield was 14.3 percent for 91 days.
6The treasury bill rate is also the benchmark rate for the Central Bank’s advances and rediscounts: the Central Bank sets its rate on bank borrowing when treasury bills are used as collateral and its rate for rediscounts of treasury bills at 1.5 percentage points above the maximum rate on treasury bills accepted by the Government in the most recent tender. The comparable rates for private paper are 2.5 percentage points above the maximum rate on treasury bills accepted by the Government in the most recent tender.
7Fifty percent of the commercial banks’ liquid assets under this ratio must be held in the form of treasury bills. Treasury bonds maturing within three months are considered equivalent to bills.
8No particular institutional or administrative impediment to the development of a secondary bill market exists. The reform measures, and especially the proposed increase in yields, would gradually generate secondary market transactions. In order to prepare for secondary market trading, Central Bank has to convert some or all of the current government overdraft into treasury bills. Proposals along this line have already been formulated. They should be implemented as soon as possible.

    Other Resources Citing This Publication