Nigeria: Selected Issues
Author: International Monetary Fund. African Dept.
Publication Date:
08 Feb 2021
eISBN:
9781513568461
Language:
English
Keywords:
CBNS toolbox; policy reaction function; CBN lending; inflation expectation; framework effective; Open market operations; Monetary base; Inflation; Monetary policy frameworks; Price stabilization; Global
Download PDF (622.7 KB)

Selected Issues

Abstract

Selected Issues

Nigeria—Strategy for a Monetary Policy Reset1

14. This paper takes stock of Nigeria’s monetary policy framework and proposes a strategy for a reset. Despite reputable levels of autonomy and far-reaching powers, the Central Bank of Nigeria (CBN)’s monetary policy effectiveness has been weak as demonstrated by sustained high inflation above the target range of the CBN and weak transmission mechanism. More recently, the central bank’s financing of the fiscal deficit2 (Figure 1) has resulted in large buildup of Open Market Operations (OMO) bills (Figure 2)—including issuance to foreigners—which has been costly for the CBN and has increased Nigeria’s susceptibility to capital outflow risks.

Figure 1.
Figure 1.

CBN Lending, 2017–20

(Billions of naira)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Central Bank of Nigeria.
Figure 2.
Figure 2.

Stock of CBN Bills 2017–20

(Billions of naira)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Central Bank of Nigeria and staff calculations.

15. The CBN’s commitment to price stability has been compromised by its multiple objectives. Section 2 of the 2007 CBN Act postulates (1) monetary and price stability; (2) safeguarding international value of the legal tender currency; and (3) promotion of a sound financial system as the CBN’s principle objectives without prioritization. Historically, Nigeria’s stabilized exchange rate regime along with the CBN’s broad engagement in development financing3 has failed to anchor inflation expectations.

16. Nigeria’s inflation performance has lagged behind emerging market and regional peers (Figure 3). Combined with the CBN’s long-standing preference to keep stabilized exchange rates, positive inflation differential relative to Nigeria’s trading partners have resulted in persistent real appreciation during normal times and large step depreciations following crises (Figure 4). The COVID-19 crisis and its severe macroeconomic fallouts question the sustainability of the current exchange rate and monetary policy set up. Defusing the balance of payment pressures call for a move to a fully unified and market-determined exchange rate. And this cannot be done in a sustainable manner without an overhaul of the monetary policy framework—including to thwart the risk of a spiral between inflation and depreciation expectations.

Figure 3.
Figure 3.

Consumer Price Index

(YoY percentage change, period average)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Figure 4.
Figure 4.

Real and Nominal Effective Exchange Rate

(Index number)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

17. This paper looks at the strategy for monetary policy reset in Nigeria. Our analysis focuses on three questions: (1) What are the CBN’s de facto policy objectives? (2) What are the key impediments to CBN’s monetary policy operational framework?; and (3) What can be done to provide a credible anchor to inflation expectations? We find that the CBN’s monetary policy has been procyclical, with interest setting mainly driven by a tendency for exchange rate stabilization. We also find that multiple objectives of the monetary policy regime and heterodoxy have eroded effectiveness and clarity of the monetary policy operational framework. As a way forward, we propose a flexible monetary targeting (FMT) regime and an overhaul of CBN’s policy toolbox.

A. What Are the CBN’s De Facto Policy Objectives?

18. How have the CBN’s de jure parallel mandates shaped its de facto policy objectives? To examine this question, we estimated a Taylor rule-type policy reaction function following the classical approach proposed by Clarida, Gali, and Gertler (1998)4.

  • Behavioral Model: In line with Clarida, Gali, and Gertler (1998), we constructed the following forward-looking Taylor rule model, which has been augmented to include real exchange rate and oil price (given Nigeria’s oil dependence) as explanatory variables. In this class of models, the most critical parameter is β, which captures response of central bank’s target interest rate to changes in expected inflation: the reduced-form equation (1) implies that if β > 1 the central bank’s target real interest rate (i.e., real policy rate) adjusts itself to stabilize inflation (i.e., real interest rises when expected inflation increases); but if β < 1 it moves to accommodate inflation (i.e., real interest declines when expected inflation increases).5 Likewise if γ < 0, real interest rates behave pro-cyclically (i.e., it decreases in response to an increased output gap).
    it=ρit−1+(1−ρ)(i*+β(Etπ¯t+s−π*)+γyt−1+δREERt+μOILt+φDummyt)+ϵt(1)

    (where i, y, π¯, REER, OIL denote nominal interest rate, output gap, inflation rate, real effective exchange rate, and crude oil prices; i* and π*natural interest rate and inflation target; Et-1 an expectation operator conditional on information available at time t - 1, and ε a statistical disturbance)

  • Empirical model: Following Clarida, Gali, and Gertler (1998), equation (1) was rearranged by substituting expected inflation (Etπ¯t+s) with future inflation minus forecast error π¯t+s−Etπ¯t+s. This transforms ail right-hand side variables to observed macroeconomic variables. But it also renders the regression into ‘an-error-in variable’ model—for which an OLS estimator is known to be biased. The likely presence of autocorrelation also implies that it may not be efficient Thus, we use GMM as the preferred estimation methodology, which has become the norm in this strand of literature. Lagged (up to 4 lags) variables of output gap, inflation, interest rate, and oil price have been used as instrument variables.

  • Data: See Table 1 for data description

Table 1.

Nigeria: Data Summary 1/

article image

Quarterly (1995Q1 to 2020Q1)

19. Appendix I summarizes the outcome of the estimated regression equations. The regression equations were fitted using nine different behavioral policy horizons (reacting to inflation; from 4 quarter lag to 4 quarter lead) to test the robustness of the estimation outcome. In ail estimations, the results show that the CBN’s monetary policy reaction to macroeconomic developments was largely procyclical—and fell short of establishing a credible record of the primacy of price stability.

  • Procyclicality: Most striking feature is that the CBN’s policy rate has been reacting to economic cycles (i.e., output gap) in a procyclical manner (i.e., decreases during booms and increases during busts) as signified by the fact that the estimated value of γ stays negative and statistically significant for ail behavioral policy horizons (i.e. from s=-4 to 4). The result is consistent with Kaminsky, Reinhart, and Vegh (2004) and suggests that Nigeria’s economic cycles have likely coincided with oil-led capital flow cycles—thus output expansions (contractions) tend to be accompanied by exchange rate appreciation (depreciation) pressures; and the desire to maintain nominal exchange rate stability has subordinated policy rate to exchange rate defense— decreasing (increasing) during economic booms (busts) thus amplifying output volatility5.

  • Inflation blindness. Primacy of price stability requires the coefficient for expected (or actual if s<0) inflation β should be greater than one and statistically significant. However, Table A1 shows that apart from one quarter backward horizon (s=-1) these conditions fail to be met with the signs of the estimated β flip-flopping and lacking statistical significance. This implies de facto that the CBN’s monetary policy has been inflation blind. 7

B. Is the CBN’s Operational Framework Effective and Coherent?

Orientation

20. Nigeria’s monetary policy framework has an ambiguous orientation. The CBN follows, de jure, a monetary targeting (MT)8 but behaves differently from that typical of an MT central bank. MT central banks generally use broad money as the intermediate target—a nominal anchor to pin down inflation expectation. However, given that broad money is outside a central bank’s direct control, it relies on reserve money as the operating target given the relatively stable relationship between the two. Then the operating target becomes the guide post for an MT central bank’s day-to-day operation: a deviation of the actual level of reserve money from the target path (or target growth rate) triggers liquidity absorbing or injecting reactions (e.g., through open market or treasury operations) to bring back the actual to the target path. Thus, the single most important hallmark of an MT regime is a stable growth rate of reserve money, which is consistent with the broad money target. In contrast with, say, the U.S. during the period (1975–82) when the Federal Reserve officially adopted an MT, or Tanzania since 2001 (Figure 6), the historical behavior of Nigeria’s reserve money (Figure 5) is characterized by very large volatility9--too large to be deemed a de facto MT regime.

Figure 5.
Figure 5.

Nigeria Reserve Money

(Percent, Y-Y Growth)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Central Bank of Nigeria.
Figure 6
Figure 6

Tanzania Reserve Money

(Percent, Y-Y Growth)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Haver.

21. Nigeria’s monetary policy regime cannot be characterized as an inflation targeting (IT) either. An IT regime is defined by a central bank using its inflation forecast as the intermediate and short-term interest rates as the operating target. Its operation is characterized by day-to-day liquidity operation aimed at aligning interbank rates to policy rate and periodic adjustment of policy rate following a Taylor-type policy rule—in which nominal policy rate responds to changes in inflation forecast by more than 100 percent. The hallmarks for an IT regime thus are (1) stable and close relationship between short-term interest rates and the policy rate and (2) the strength of the policy rate’s response to expected inflation. As discussed, the CBN’s policy reaction function is characterized by inflation blindness and procyclicality. The comparison of short-term interest behavior between Nigeria and selected countries (Figures 7 and 8) furthermore illustrate that Nigeria’s interbank rate behavior is far from ones expected in an IT.

Figure 7.
Figure 7.

Nigeria Interest Rates

(Percent, Per Annum)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Central Bank of Nigeria.
Figure 8.
Figure 8.

Uganda Interest Rates

(Percent, Per Annum)

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Haver.

Tools

22. The CBNs toolbox does not suffer from a shortage of instruments but the tools are not ail used in an orthodox way. In fact, the CBN is equipped with most tools needed by modem central banks including ones to (1) generate stable demand for reserves (cash reserve requirement); (2) inject or absorb day-to-day liquidity (CBN bills, repo, T-bill auction); and (3) cap or floor interbank and money market rates (e.g., standing lending and deposit facilities). However, many of these tools are calibrated inadequately or modified in a heterodox fashion (Table 2). Nigeria’s interest corridor does not bind with access restrictions hindering conventional functions of standing lending facility (SLF) and standing deposit facility (SDF). And asymmetry of cash reserve requirement represses financial intermediation including by making effective CRR to keep increasing over time (Table 2). The CBN introduced a minimum loan to deposit ratio (LDR) in June 2019 in a bid to prop up credit growth. While LDR prompts banks to shift liquid assets to loans at individual bank levels, its effectiveness has been limited as seen by recent decline in sequential credit growth.

Table 2.

Nigeria: An Analysis of CBN’s Key Monetary Policy Tools

article image

23. The unorthodox use of tools has negatively affected the effectiveness of monetary policy, including through interest rate segmentation and elevated balance of payments risks.

  • Liquidity fragmentation. Since November 2019, the CBN significantly reshuffled its monetary policy toolkit by (1) prohibiting OMO purchase by nonbank residents; including to rollover maturing OMOs; and (2) using discretionary CRR to sterilize liquidity released from the ensuing reduction in the OMO stock. The combination of these resulted in a divergence in liquidity conditions between the interbank—which remained tight due to absorption though CRR—and liquidity in the domestic bond market1. The diversion resulted in commensurate segmentation of interest rates—with interest for T-bills (91 days) declining to average 2 percent in 2020Q3 but interbank rates staying on average at 9–10 percent. Since September 2020, interbank bank and OMO rates converged significantly toward T-bills—likely reflecting less stringent use of CRR and increasing share of banks in the OMO market, which is facilitating bank-led interest rate arbitrages. However, this has further weakened MPR’s relevance as the interest rate benchmark.

  • External risks. The current level of OMO rates (½-3 percent) present risks for capital outflows given the prevalent share of foreign portfolio investment (about $13 billion; about 40 percent of international reserves) in this market segment. For now, foreign investors remain in OMOs despite alternative destinations that offer higher interest rates (e.g., Egypt and Ghana, where domestic interest rates are in double digits) forced by a System of FX rationing2.

C. A Strategy for a Monetary Policy Reset

Establish a Nominal Anchor (Near Term)

24. The shift should start with resetting the monetary policy regime. One practical option is a flexible monetary targeting (FMT), which is a transitory monetary regime between MT and IT; before moving to an IT. An FMT would combine a flexible reserve money operating target (average over a prespecified period, e.g., bi-monthly) with an interest rate band target (i.e., an interest rate corridor).3 Under this framework, a deviation of the actual reserve money growth from the target would trigger corrective liquidity operation through fixed-quantity flexible-rate open market operations. For this, the CBN’s current interest corridor will need to be refreshed by removing access restrictions on both SLF and SDF.

Establish Primacy of Price Stability (Medium Term)

25. A shift to a rules-based operational framework cannot work in a sustainable manner if a central bank’s multiple objectives keep colliding with one another. The danger of colliding objectives is more prevalent in commodity-dependent economics like Nigeria where both business and capital flow cycles are led by commodities. For example, an upswing in business activities during a commodity boom lifts credit and deposit growth, which under a rules-based (and unaccommodating) MT, makes excess reserves to fall below usual levels. But the ensuing liquidity tightness brings upward pressures on interest rates. This may attract more capital flows, thus further exacerbating pressures on the exchange rate. In situations like this, a rules-based MT will need to establish a priority ordering to assure price stability. This calls for changes in the CBN Act. Despite its significant improvement, the 2007 CBN Act leaves important gaps as it misses (1) a priority ordering among multiple objectives and (2) explicit prohibition from taking instructions from a third party. The law should be amended to fix these gaps to ensure proper autonomy and effectiveness of the CBN.

26. The primacy of price stability also cannot be achieved without a shift to a more flexible exchange rate regime. According to Mundell-Fleming trilemma, central banks cannot use monetary policy for stabilizing inflation while also stabilizing exchange rates under an open capital account. More specifically, it suggests that under free capital flows, central banks lose control over interest rates—which are determined by covered interest parity. Thus, an IT regime, which ordinarily hinges on central banks’ control over interest rates, cannot be implemented in parallel with a stabilized exchange rate. The same principle also holds for an MT regime, wherein interest rates are more volatile (than in an IT) but their average levels are closely linked to reserve money growth (i.e., the higher the latter, the lower the former). Thus, a move to an FMT requires a shift to a more of payment pressures brought by the COVID-19 global crisis.

Address Fiscal Dominance (Near to Medium Term)

27. Eliminating fiscal dominance is also a key precondition to establishing de jure and de facto primacy of price stability. Even if the CBN’s primary legal mandate were price stability, the CBN is unlikely to be held accountable, particularly by the government, if this mandate collides with fiscal financing in an environment where the government keeps asking the CBN for it

  • Restore adherence to legal limits (near term). In recent years, legal limits on temporary advance to Federal Government—5 percent of previous year’s revenue—have been beached repeatedly; with no apparent consequences. For example, 2018 CBN Financial Statement reports 5.3 trillion naira as its overdraft and short-term advance balance—which is twice the size of the Federal Government revenue in 2017. Nonadherence to the legal limits should be stopped.

  • Fix the cause of the problem (near term). The main reason for the nonadherence was inadequate spacing of financing resources in the budget planning process, as well as persistently large fiscal deficits. Domestic borrowing requirement should be adequately budgeted and met by marketable debt instruments. Beyond the immediate term, a complete removal of CBN’s credit to government will require higher domestic revenue mobilization.

  • Wind down outstanding loan stock (medium term). This can be done by gradually switching the CBN’s legacy advance stock owed by Federal Government of Nigeria (FGN) by new FGN bonds, which is set to be absorbed (e.g., by integrating into conventional bond issuance program) mainly by domestic institutional and foreign portfolio investors. The reserve drainage caused by new FGB bond4 calls for liquidity injection under an FMT. This should be done through a matched winding down of OMO bill stock—if necessary through buybacks. A hypothetical example (Figure 9) where N5 trillion of CBN credit stocks are replaced by FGB bonds—70 percent of which are taken up by nonbanks—shows that banks will be able to supply N3.5 trillion of new loans using the space freed up from reduced OMOs (-N5 trillion ).

Figure 9.
Figure 9.

CBN Loan to FGN Bond Shift: Illustration

(Change in stock, trillion Naira) 1/

Citation: IMF Staff Country Reports 2021, 034; 10.5089/9781513568461.002.A001

Source: Staff estimates.1/ Assuming a hypothetical example of 5 trillion reduction of CBN advances financed by FGB bonds; nonbanks are assumed to absorb 70 percent of new FGN Bonds. * denotes additional changes caused by multiplier effects (which is caused by emergence of excess reserves).2/ Dotted red and black charts denote changes in ODC reserve deposits and FGN deposits at CBN respectively.3/ Change in reserves (zero) minus change in required reserves (=CRR ratio*-3.5); CRR ratio (assumed)=0.275.

28. Finally, the reset of monetary policy operational framework cannot be completed without cleaning up the CBN’s toolbox.

  • Normalize interest rates (near to medium term): Since September 2020, interbank and OMO interest rates converged significantly toward T-bills rate at 1–3 percent range. At the moment, the ultra-low interest rate is maintained through de facto administrative controls (e.g., effective shut down of FX market) and liquidity segmentation. As the crisis passes, and with the lifting of these controls, interest rates will need to be normalized; mainly through their realignment toward MPR. This can be achieved by either (1) nullifying the prohibition of OMO holding by nonbank residents including domestic institutional investors and reversing the OMO to CRR shift5 or (2) engineering a shift from CRR to T-bills5.

  • Normalize CRR (near term): The CRR should be reset by eliminating asymmetry. Symmetric CRR will help leveling the playing field and reduce financial repression. Going forward, there is a need to gradually decrease CRR, with a view to reducing effective tax on financial intermediation, for example by pairing it with plans to reduce structural liquidity injected for quasi-fiscal objectives (e.g., AMCON bond7).

  • Reconsider LDR (near term). The minimum LDR policy has initially spurred credit but is reaching its limits and carries with its financial stability risks. Therefore, consideration should be given to its discontinuation.

References

  • Bhattacharya, R. I. Patnaik and A. Shah, 2011, “Monetary Policy Transmission in an Emerging Market Setting”, IMF Working Paper (WP/11/5).

    • Search Google Scholar
    • Export Citation

  • Clarida R., J. Gali, and M. Gertler, 1997, “Monetary Policy Rules in Practice: Some International Evidence”, NBER Working 6254.

  • International Monetary Fund, 2015, “Evolving Monetary Policy Frameworks in Low Income and Other Developing Countries”, IMF Policy Paper.

    • Search Google Scholar
    • Export Citation

  • Maehle, N., 2020, “Monetary Policy Implementation: Operational Issues for Countries with Evolving Monetary Policy Framework”, IMF Working Paper (WP/20/26).

    • Search Google Scholar
    • Export Citation

  • Kaminsky, G., C. M. Reinhart and C. A. Vegh, 2004, “When It Rains, It Pours: Procyclical Capital Flows and Macroeconomic Policies, NBER Working 10780.

    • Search Google Scholar
    • Export Citation

  • Kuttner, K. and P. C. Mosser, 2002, “The Monetary Transmission Mechanisms: Some Answers and Further Questions”, FRBNY Economic Policy Review, May 2002.

    • Search Google Scholar
    • Export Citation

Appendix I. CBN’s Policy Reaction

Nigeria: GMM Estimation Outcome

article image
1/ **, *, # means statistical significance at 5, 10, and 15 percent.2/ s denotes lag (+) and lead (‐) for inflation variable.
1

Prepared by Jack Ree (AFR).

2

Our analysis suggests that the fiscal financing was largely sterilized rather than monetized and resulted mainly from weak PFM practices (e.g., insufficient budgeting of market financing needs and lack of flexibility in the PFM framework for its correction within a fiscal year). However, the large increase in CBN’s credit to government led to perception of fiscal dominance—thus negatively affecting inflation expectations.

3

The CBN has historically been granted generous powers to engage in financial or economic development, in which CBN typically contributes capital to interest-subsidized lending schemes operated by banks, mainly through on-lending modalities. The ensuing credit risks are either borne by participating banks or shared with the CBN.

4

By fitting Taylor-rule type of a regression model to post-1979 G3 data, Clarida, Gali, and Gertler (1998) find that the monetary policy regime since 1979 has shifted to a soft-hearted inflation targeting—whereby a central bank raises nominal rates sufficiently enough to push up real rates in response to a rise in expected inflation. The paper then attributes the successful disinflation in the post-1970s to inflation targeting regime—which puts primary focus on inflation-given its merits as a nominal anchor for monetary policy given its simplicity and ease of rebuilding credibility for central banks.

5

As illustrated by Clarida, Gali, and Gertler (1998), the inflation stabilizing condition (if β > 1 ) is driven by the fact that nominal interest rate can be decomposed to real interest rate and inflation expectation, which can be proxied by Etπ¯t+s. Substituting and rearranging equation (1) makes coefficient for Etπ¯t+s positive if and only if β -1 >0, thus implying the desired inflation stabilizing condition.

6

The procyclicality suggests that CBN may possibly be a nominal exchange rate targeting central bank. But directly testing such a proposition is not feasible under a stabilized exchange rate regime as deviation of nominal exchange rate from the target would be countered mainly by FX market intervention. Without controlling this, the interest (rate’s reaction to nominal exchange rate would not be distinguishable. And administrative controls would further obscure such reaction.

7

The insignificance of inflation coefficient—despite significance of policy rate reaction to output gap—suggests weak causal link between output gap and inflation, likely reflecting prevalence of supply-side factors in inflation dynamics in Nigeria.

8

CBN’s latest available Monetary, Credit, Foreign Trade and Exchange Policy Guidelines (2018/2019) states that monetary targeting framework will remain the monetary policy strategy and will be complemented by an appropriate

9

The standard deviation of Nigeria’s reserve money growth (y/y) is 5 times that of Tanzania during recent years (2017–2020) and 15 times larger than the U.S. during 1975–82.

1

During October 2019 to October 2020, CBN’s OMO stock decreased by N7 trillion—with the OMOs held nonbanks decreasing by N7.2 trillion. The reduced OMO holding by nonbanks increased nonbanks’ deposits at banks by the equivalent amount (i.e., N7.2 trillion) as the CBN repaid the maturing OMOs through the banking System.

2

Another mitigating factor is the large yield premium given at OMO auctions (about 500–600 bps higher relatively to secondary markets).

3

IMF (2015), “Evolving Monetary Policy Frameworks in Low Income and Other Developing Countries”, IMF Policy Paper. An FMT can be implemented when legal and economic preconditions for an IT is lacking and is conducive to creating these preconditions by stoking financial market development.

4

Non-bank and bank buyers of FGN bonds make their payments using bank deposits (non banks) or reserve deposits (banks). This is then remitted to government’s account at the CBN by CBN debiting relevant banks’ reserve

5

This can be done by bringing back OMO stocks toward October 2019 level while neutralizing reserve effects on banks by corresponding release of CRR. The release can be done without changing CRR ratio until discretionary CRR balance is depleted.

6

This can be done by issuing T-bills until T-bill rates realign with the policy rate while offsetting the reserve effects on bank by corresponding release of CRR. The proceed from the T-bills can be credited to government’s deposit account at the CBN for budget pre-financing—with the FGN and CBN sharing the burden of negative interest margin. The CRR release can be done without changing CRR ratio until discretionary CRR balance is depleted.

7

Winding down of AMCON bond can be done, for example, by floating AMCON Bond to the market at prevailing interest rate—if needed after granting government guarantee.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Nigeria: Selected Issues
Author: International Monetary Fund. African Dept.
Volume/Issue:
Volume 2021: Issue 034
Publisher:
International Monetary Fund
ISBN:
9781513568461
ISSN:
1934-7685
Pages:
65
DOI:
https://doi.org/10.5089/9781513568461.002

CBN Lending, 2017–20

(Billions of naira)

Stock of CBN Bills 2017–20

(Billions of naira)

Consumer Price Index

(YoY percentage change, period average)

Real and Nominal Effective Exchange Rate

(Index number)

Nigeria Reserve Money

(Percent, Y-Y Growth)

Tanzania Reserve Money

(Percent, Y-Y Growth)

Nigeria Interest Rates

(Percent, Per Annum)

Uganda Interest Rates

(Percent, Per Annum)

CBN Loan to FGN Bond Shift: Illustration

(Change in stock, trillion Naira) 1/