Abstract
1. Like many central banks in low-income countries, the Central Bank of Liberia (CBL) is modernizing its monetary policy framework to make its policy more effective in anchoring inflation and promoting macroeconomic stability. Much progress has already been made, but the modernization process is taking place under quite challenging circumstances. The country is facing a legacy of macroeconomic instability, a shortage of fit Liberian dollar banknotes is only now being addressed, financial markets are shallow, and the economy is highly dollarized.
MONETARY POLICY AND DE-DOLLARIZATION IN LIBERIA’S DUAL CURRENCY SYSTEM1
A. Introduction and Background
1. Like many central banks in low-income countries, the Central Bank of Liberia (CBL) is modernizing its monetary policy framework to make its policy more effective in anchoring inflation and promoting macroeconomic stability. Much progress has already been made, but the modernization process is taking place under quite challenging circumstances. The country is facing a legacy of macroeconomic instability, a shortage of fit Liberian dollar banknotes is only now being addressed, financial markets are shallow, and the economy is highly dollarized.
2. The authorities have already made significant progress in terms of strengthening the CBL’s institutional setup and improving the monetary policy framework and monetary operations. The passage of the amended and restated CBL Act of 2020 enhanced the institutional and operational independence of the CBL. It sets out price stability as the primary mandate of the CBL and generally disallows providing financing for the budget. The Act also provides for the establishment of the Monetary Policy Committee (MPC), with the main responsibility of formulating monetary and exchange rate policy in Liberia. In 2019, the CBL also initiated the process of modernizing its monetary policy framework – a move toward a more interest rate-based system, featuring a policy rate, an interest rate corridor, and standing credit and deposit facilities. The role of CBL bills was also reviewed, and they have become a key monetary policy instrument since 2019.
3. The conduct of monetary policy is complicated by Liberia’s history of macroeconomic instability, but progress has been made more recently. When inflation is rampant and the exchange rate depreciates quickly, monetary policy carries little credibility, and the general public is driven to using the U.S. dollar. In 2019, a decline in external assistance coupled with unsustainable fiscal policy, which relied on excessive central bank financing, and an accommodative monetary policy stance resulted in macroeconomic instabilities in Liberia. Both inflation and exchange rate depreciation climbed to 30 percent, while the economy slid into recession. Gross official foreign reserves dropped to 2.2 months of imports. However, the country started regaining macroeconomic stability in the context of the government’s reform program, which is supported by the IMF under the Extended Credit Facility (ECF) Arrangement signed in December 2019. Inflation declined sharply from about 30 percent at program inception to 5.5 percent in December 2021, thanks to ending monetary financing of the budget and prudent fiscal and monetary policies. This laid the foundation for better growth from 2020, but the pandemic delayed the takeoff. Liberia’s economy is estimated to have expanded by 5.0 percent in 2021 and is projected to grow by 3.7 percent and 4.7 percent in 2022 and 2023, respectively. Despite this welcome progress, it will take some time until macroeconomic stability becomes firmly entrenched in the eyes of the general public.
4. Monetary policy has also long been hamstrung by the lack of sufficient usable Liberian dollar banknotes, although this is now being addressed by the ongoing currency changeover. This exacerbated dollarization, thereby reducing the component of money that is under the direct control of the central bank and limiting the central bank’s ability to contain inflation by tightening domestic monetary conditions. To address this challenge, the authorities are in the process of rolling out the long-awaited currency changeover, which seeks to replace the entire existing stock of banknotes, which are highly worm out, and to meet the growing cash demand of the economy. An emergency order for LD 100 bills has already been received. All other denominations will be introduced in the second half of 2022. This will put an end to the periodic shortages of Liberian dollars, which threated financial and macroeconomic stability and spurred dollarization.
5. Low financial market depth is another obstacle for the efficient conduct of monetary policy. Liberia’s financial sector is small, financial intermediation is limited, and capital markets remain rudimentary (Figure 1). Government securities are issued only occasionally and there is no active secondary market for trading them. Shallow financial markets impair the transmission channels of monetary policy. Thin money markets, especially the interbank market, weaken the transmission of monetary impulse to the banking and financial sector. And the small size of the financial sector weakens the impact of banks’ lending conditions on a wider economy, including on aggregate demand, output, and inflation.
Liberia: Monetization, Financial Intermediation, and Dollarization Indicators
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Sources: IMF staff estimates and calculations.Notes: FCD: Foreign Currency Deposits. To isolate the impact of valuation changes, deposits and credits in foreign currency are evaluated at a constant exchange rate.Liberia: Monetization, Financial Intermediation, and Dollarization Indicators
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Sources: IMF staff estimates and calculations.Notes: FCD: Foreign Currency Deposits. To isolate the impact of valuation changes, deposits and credits in foreign currency are evaluated at a constant exchange rate.Liberia: Monetization, Financial Intermediation, and Dollarization Indicators
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Sources: IMF staff estimates and calculations.Notes: FCD: Foreign Currency Deposits. To isolate the impact of valuation changes, deposits and credits in foreign currency are evaluated at a constant exchange rate.6. Liberia’s high degree of dollarization places perhaps the most severe limitations on the effectiveness of monetary policy. Liberia operates a dual-currency system where both the Liberian dollar and the U.S. dollar are legal tenders. It is one of the most highly dollarized economies in the world. Over the period 2007-20, deposit and credit dollarization are estimated to have averaged 84 percent and 91 percent, respectively (Figure 1). Dollarization weakens the transmission channels of monetary policy since LD interest rates may not feed through to dollar interest rates or the real exchange rate. The literature suggests that dollarization renders the exchange rate highly volatile and increases exchange rate pass-through to domestic prices.2 Empirical evidence confirms that the degree of exchange rate pass-through to inflation is high in Liberia (IMF, 2014). Exchange rate movements in the wake of monetary policy decision could also give rise to destabilizing balance sheet effects, especially for banks with currency mismatches. Finally, dollarization constrains the central bank’s ability to act as lender of last resort and leads to a loss of seigniorage.
7. Now that the authorities have made some good progress with reestablishing macroeconomic stability and addressing the shortage of quality currency, the preconditions are falling into place to modernize the monetary policy framework and to initiate the process of de-dollarization. The objective of this paper is to offer some suggestions how to move forward on both fronts more concretely, drawing on the literature, cross-country experiences, and technical assistance already provided to Liberia.
8. The paper is structured as follows. Section B proposes next steps for enhancing Liberia’s monetary policy framework. Section C discusses how best to operationalize monetary policy under the enhanced framework. Section D offers some policy recommendations on promoting de-dollarization in Liberia. Section E concludes.
B. Enhancing Liberia’s Monetary Policy Framework
9. As Liberia’s monetary policy framework is transiting, some ambiguities have emerged in its current setup. While the CBL’s primary objective of price stability is clearly articulated, the CBL currently seems to follow both a reserve money and an interest rate operating target, which can create tensions when the targeted interest rate does not generate the targeted reserve money aggregate.3 Moreover, the framework seems not to explicitly articulate which variables could serve as intermediate targets or indicator variables to best guide the conduct on monetary policy. Monetary operations, especially the issuance of CBL bills, are not always well calibrated to existing liquidity conditions in the banking sector. CBL bills are issued in fixed amounts and at fixed interest rates, thereby not only limiting the interest rate transmission mechanism, but also price discovery and money market development, which are preconditions for an effective monetary policy transmission mechanism. To illustrate the point, despite a substantial 10-percentage points reduction in the policy rate from 30 percent in 2019 to 20 percent in November 2021, retail rates (deposits and lending rates) in the banking sector remained broadly unchanged. The standing facilities are also underutilized. Indeed, the standing deposit facility (SDF) was abandoned altogether in 2020.
10. These ambiguities could be resolved by adopting a flexible reserve money targeting framework.4 Thereby the CBL would clearly articulate reserve money as a medium-term indicative operating target while at the same time using interest-rate based instruments to conduct monetary policy in the short term and consistent with the monetary policy objective. Setting reserve money target over a medium-term horizon provides some flexibility for reserve money to deviate from the target path in the short term. This flexibility is important because significant seasonality could also complicate the accurate setting of the target, especially if viewed as end-period target. At the same time, allowing some flexibility for reserve money to deviate from the target path in the short term would in turn keep short-term interest rates reasonably stable (Laurens et al., 2015:18). Adopting a flexible reserve money targeting framework would require some adjustment to the conditionality under the current IMF-supported program that sets end-quarter point targets for net domestic assets and net foreign assets of the CBL, which are then internally further broken down into end-month targets.
11. The flexible reserve money operating target should be implemented within an interest rate corridor system. The authority could resuscitate the existing interest rate corridor system with its standing deposit and credit facilities. The corridor system would help limit volatility of short-term interest rates, which is critical for anchoring market expectations and reduce liquidity risk. The system allows the market to price liquidity within the interest rate corridor. Laurens et al., (2015) note that the corridor system renders commercial banks’ demand for central bank balances less sensitive to interest rates, and thus makes the market less sensitive to liquidity forecasting errors. To keep the market away from the hedges of the corridor (if short-term interest rates are persistently closer to the ceiling or floor) for an extended period, the corridor could be regularly realigned. The repositioning of the corridor would then signal the policy stance. This would serve the same signaling role as changing the policy rate under a system where short-term interest rate serve as an operational target and enhance monetary policy communication (Laurens et al., 2015; Maehle, 2020).
12. To make the interest rate corridor system work effectively, operational procedure for standing facilities should also be streamlined. In this regard, the CBL could reduce the administrative burden in executing these operations. One possible option being to delegate the approval process of the application for standing facilities to directors’ level at the CBL. The CBL should also ensure that these facilities are freely accessible and overnight with no limitation on the quantity, except for the standing credit facility (SCF), where it would be subject to the availability of collateral securities and haircuts. Accessibility of standing facilities will provide incentives for banks to hold limited excess reserves.
13. In conducting monetary operations, especially CBL bills auctions, the CBL could consider a variable rate tender for the current maturities: two-weeks, 1-month, and 3-months CBL bills. This would promote price discovery, facilitate correct pricing of liquidity, help build a domestic currency yield curve, and promote money market development. Maehle (2020) notes that the central bank could use flexible-rate fixed-quantity auctions to conduct monetary operations under a flexible monetary targeting framework. This would help the central bank offset the impact of autonomous factors on excess reserves and help keep total reserves consistent with the reserve money target.
14. Once the operating target has been clarified, intermediate variables or at least some indicator variables that could guide the conduct and monitor the implementation of monetary policy need to be chosen. In the context of Liberia, it seems appropriate to deemphasize rigid quantitative intermediate targets within the current framework and rather formulate and monitor monetary policy using multiple indicators that could provide useful information about the buildup of inflationary risks. This is not uncommon for central banks with a monetary targeting framework that is still evolving. Laurens et al., (2015) and Barth (2002) point out that central banks seeking to modernize their monetary policy sometimes find it useful to downgrade intermediate target on monetary aggregates to indicator or information variables. Instead of setting intermediate targets on monetary aggregates, central banks use monetary aggregates as indicators that supplement the assessment of short-term risks to price stability. Besides, key relationships upon which intermediate targets on monetary aggregates would be premised, such as predictability of money multipliers and velocity, have recently become relatively unstable in Liberia. There is also uncertainty about money demand, especially due to the ongoing currency changeover, which could influence the degree of dollarization and the general publics’ willingness to keep money in banks rather than holding cash.
15. The choice of indicator variables is guided by the degree of their correlation with inflation, their ability to predict inflation, and their timely availability. To this end, this paper evaluates the suitability of monetary aggregates, with and without foreign currency components, and the exchange rate to predict inflation in Liberia. In line with Berg and Borensztein (2000), the paper examines the relationships between inflation (CPI), the exchange rate (ER) and money in Liberia using a vector autoregression (VAR) analysis, estimated using monthly data for January 2007 to December 2020. The paper considers the following monetary aggregates: currency in circulation (CIC), the monetary base (MB), narrow money (M1) and broad money (M2). For MB, M1 and M2 the paper also considers the aggregate with and without foreign currency deposits.5
16. The empirical analysis of the relationship between inflation, exchange rate, and money suggests that in Liberia the exchange rate has the greatest predictive power for future inflation and that the domestic-currency component of narrow money and currency in circulation also contain useful information about future inflation.
The correlation analysis shows that both the exchange rate, domestic-currency component of narrow money, and currency in circulation are positively and significantly correlated with inflation (Table 1). The correlation is strongest for the exchange rate. The correlation between the other monetary aggregates and inflation is weak and statistically insignificant.
Liberia: Correlation Results
Liberia: Correlation Results
| ER | CIC | MB | M1 | M2 | |
| CPI | 0.6708*** | 0.1924*** | 0.0183 | 0.2222*** | 0.0984 |
Liberia: Correlation Results
| ER | CIC | MB | M1 | M2 | |
| CPI | 0.6708*** | 0.1924*** | 0.0183 | 0.2222*** | 0.0984 |
However, the results of VAR Granger causality test clarifies that only the exchange rate has a significant Granger effect on the inflation (Table 2).6 Neither currency in circulation nor monetary aggregates do. As a caveat in interpreting these results it is important to note that the results of the Granger causality test are sensitive to the choice of lag length used in estimating the VAR. In this paper, the Granger causality test was conducted using a VAR with a maximum lag length of 12 months based on Akaike information criterion (AIC).7
Liberia: VAR Granger Causality Test Results
Liberia: VAR Granger Causality Test Results
| VAR(CPI,CIC,ER) | VAR(CPI,MB,ER) | VAR(CPI,M1,ER) | VAR(CPI,M2,ER) | |
| Money | 0.4555 | 0.1133 | 0.2696 | 0.3675 |
| Exchange Rate (ER) | 0.0409** | 0.0026*** | 0.0161** | 0.0034*** |
Liberia: VAR Granger Causality Test Results
| VAR(CPI,CIC,ER) | VAR(CPI,MB,ER) | VAR(CPI,M1,ER) | VAR(CPI,M2,ER) | |
| Money | 0.4555 | 0.1133 | 0.2696 | 0.3675 |
| Exchange Rate (ER) | 0.0409** | 0.0026*** | 0.0161** | 0.0034*** |
The impulse response analysis indicates that shocks to the exchange rate, the domestic-currency component of narrow money, and currency in circulation have significant positive effects on inflation (Figure 2).8 The shock to the exchange rate has the largest and most persistent effect on inflation. Shocks to the domestic-currency component of narrow money and to currency in circulation have significant effects on inflation, though the effect of the shock on currency in circulation becomes significant only after 13 months. Shocks to other monetary aggregates also have a positive effect on inflation in Liberia, but the effect is not statistically significant (Figure 2).
Liberia: Impulse Responses of Inflation (CPI) to a One S.D. Innovations in the Exchange Rate (ER) and Money Aggregates Across Four Different VAR Models1
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Source: IMF staff estimates and calculations.Notes: CPI: consumer price index; CIC: currency in circulation; MB: monetary base; M1: narrow money; M2: broad money.The green dotted lines represent the impulse responses estimated with standard VAR while the blue solid lines represent impulse responses estimated with Jordá’s (2005) local projection method. The red dotted lines represent the lower and upper limit of the 90 percent significance confidence bands.1 For the impulse response functions, we report the results of monetary aggregates (MB, M1 and M2) that include only domestic-currency components for brevity. The results of monetary aggregates that include foreign-currency components suggest that monetary aggregates that include foreign-currency components have a positive but statistically insignificant effect on inflation.Liberia: Impulse Responses of Inflation (CPI) to a One S.D. Innovations in the Exchange Rate (ER) and Money Aggregates Across Four Different VAR Models1
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Source: IMF staff estimates and calculations.Notes: CPI: consumer price index; CIC: currency in circulation; MB: monetary base; M1: narrow money; M2: broad money.The green dotted lines represent the impulse responses estimated with standard VAR while the blue solid lines represent impulse responses estimated with Jordá’s (2005) local projection method. The red dotted lines represent the lower and upper limit of the 90 percent significance confidence bands.1 For the impulse response functions, we report the results of monetary aggregates (MB, M1 and M2) that include only domestic-currency components for brevity. The results of monetary aggregates that include foreign-currency components suggest that monetary aggregates that include foreign-currency components have a positive but statistically insignificant effect on inflation.Liberia: Impulse Responses of Inflation (CPI) to a One S.D. Innovations in the Exchange Rate (ER) and Money Aggregates Across Four Different VAR Models1
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Source: IMF staff estimates and calculations.Notes: CPI: consumer price index; CIC: currency in circulation; MB: monetary base; M1: narrow money; M2: broad money.The green dotted lines represent the impulse responses estimated with standard VAR while the blue solid lines represent impulse responses estimated with Jordá’s (2005) local projection method. The red dotted lines represent the lower and upper limit of the 90 percent significance confidence bands.1 For the impulse response functions, we report the results of monetary aggregates (MB, M1 and M2) that include only domestic-currency components for brevity. The results of monetary aggregates that include foreign-currency components suggest that monetary aggregates that include foreign-currency components have a positive but statistically insignificant effect on inflation.Liberia: Impulse Responses of Inflation (CPI) to a One S.D. Innovations in the Exchange Rate (ER) and Money Aggregates Across Four Different VAR Models1
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Source: IMF staff estimates and calculations.Notes: CPI: consumer price index; CIC: currency in circulation; MB: monetary base; M1: narrow money; M2: broad money.The green dotted lines represent the impulse responses estimated with standard VAR while the blue solid lines represent impulse responses estimated with Jordá’s (2005) local projection method. The red dotted lines represent the lower and upper limit of the 90 percent significance confidence bands.1 For the impulse response functions, we report the results of monetary aggregates (MB, M1 and M2) that include only domestic-currency components for brevity. The results of monetary aggregates that include foreign-currency components suggest that monetary aggregates that include foreign-currency components have a positive but statistically insignificant effect on inflation.Liberia: Impulse Responses of Inflation (CPI) to a One S.D. Innovations in the Exchange Rate (ER) and Money Aggregates Across Four Different VAR Models1
Citation: IMF Staff Country Reports 2022, 297; 10.5089/9798400219665.002.A002
Source: IMF staff estimates and calculations.Notes: CPI: consumer price index; CIC: currency in circulation; MB: monetary base; M1: narrow money; M2: broad money.The green dotted lines represent the impulse responses estimated with standard VAR while the blue solid lines represent impulse responses estimated with Jordá’s (2005) local projection method. The red dotted lines represent the lower and upper limit of the 90 percent significance confidence bands.1 For the impulse response functions, we report the results of monetary aggregates (MB, M1 and M2) that include only domestic-currency components for brevity. The results of monetary aggregates that include foreign-currency components suggest that monetary aggregates that include foreign-currency components have a positive but statistically insignificant effect on inflation.17. These results are consistent with the literature on highly dollarized economies in finding that the exchange rate channel is strong (Alvarez-Plata and García-Herrero, 2008; Baliño et al., 1999). However, in contrast to Dabla-Norris and Floerkemeier (2006) and Berg and Borensztein (2000), the paper establishes that broad money that includes foreign currency deposits has no material predictive power for inflation in Liberia.
18. This empirical analysis suggests that the exchange rate, domestic-currency component of narrow money and currency in circulation could serve as valuable indicator variables to guide the conduct of monetary policy in Liberia. Because the exchange rate has greater predictive power for inflation than the domestic-currency component of narrow money and currency in circulation, it should serve as the primary indicator variable while the domestic-currency component of narrow money and currency in circulation should be used as secondary indicator variables. Once flexible-rate CBL bill auctions will have been introduced, the short-term interest rates could also become valuable indicator variables. These variables should be viewed from a broader perspective of indicators that provide useful information in a timely manner to guide the conduct of monetary policy, including revisiting the target path for reserve money and the interest rate corridor as needed. They could serve as useful proxies of variables for which data is not readily available or of those variables that are observed with significant time lags (Masuch et al., 2003). They could contain some useful information about monetary conditions and the nature of shock hitting the economy. That said, they could play a key role in monetary policy decision-making process especially in countries with undeveloped financial markets (Laurens et al., 2015) like Liberia. They should not be seen as target variable to be rigorously pursued by the CBL or interpreted as intermediate targets in the strict sense of the word.9
19. The proposed flexible reserve money targeting framework would allow a seamless transition to a full-fledged interest-rate based approach. A period of flexible reserve money targeting will give the CBL sufficient time to prepare the ground and develop the necessary infrastructure for the next step. Money and interbank markets will develop, the interest rate channel of monetary policy will become stronger, money demand will grow more predictable, monetary policy will gain further credibility, a yield curve will start to emerge, and the risk that banks invest in CBL bills in full-allotment auctions rather than lend will be lower. As this happens, the CBL could deemphasize reserve money as the operating target and transition to an interest rate-based framework. Interest rates at which short-term CBL bills are auctioned on a full-allotment basis would serve as the monetary policy rate and be kept toward the mid-point of the interest rate corridor. The IMF-supported program would need to be redesigned accordingly and drop any explicit or implicit reserve money targets.
C. Implementing Monetary Policy Under Reserve Money Operating Targeting Framework
20. The implementation of monetary policy under flexible reserve money targeting is tantamount to effective management of liquidity in the banking sector. The objective is to ensure that there is adequate liquidity in the system at all times. Price stability is promoted directly through the impact of liquidity management operations on the domestic component of money supply and indirectly through the impact on the exchange rate.
21. The practical implementation of liquidity management operations should be guided by the assessment of prevailing liquidity conditions in the banking system.10 The approach is to derive a target path for total bank reserves that is consistent with medium-term target for reserve money and the forecasts for currency in circulation, as well as known cyclical patterns. This would be compared with the short-term forecast for bank reserves, based on evolution of central bank balance sheet items, to arrive at excess reserves in the banking sector. This would then guide the CBL’s short-term liquidity management or monetary policy operations, with CBL-bill auctions managing excess reserves so that reserve money returns to the target path over time.
22. It is important to note that effective liquidity management requires strong capacities for liquidity monitoring and forecasting. To this end, the CBL should regularly review and upgrade its liquidity monitoring and forecasting framework.11 Since liquidity management and forecasting require timely availability of high frequency data, the CBL would benefit from formalizing data sharing within the CBL and between the CBL and Ministry of Finance and Development Planning.
23. The CBL could also consider foreign exchange interventions as a last resort for liquidity management. Foreign exchange purchases should generally be reserved for gradually building up international reserves and as a means to steadily inject domestic currency in line with the evolving needs of the economy. On occasion the CBL could also use foreign exchange interventions to lean against excessive exchange rate volatility and to mitigate disorderly market conditions. Foreign exchange intervention for liquidity management purposes should be avoided to the extent possible.
24. For medium-term liquidity management purposes and to enhance monetary policy toolkit, the CBL could also consider strengthening the role of reserve requirements as a supplementary tool of monetary policy. A reserve requirement tool remains appropriate in countries where financial markets are underdeveloped and where the range of market-based tools is limited. But to make required reserves an effective tool, the CBL needs to ensure that all banks comply with the regulation and move away from the notion that reserve requirements are primarily a supervisory tool. The regulation on penalties for non-compliance should be reviewed to make sure that they are set at appropriate levels and then strictly enforced.
D. Promoting De-dollarization in Liberia
25. The authorities have made de-dollarization part of their economic reform package, with the time now ripe for a renewed implementation drive. The government committed to progressively raise the share of the wage bill paid in Liberian dollars from 20 percent until it reaches 50 percent and also increase the use of Liberian dollars to pay vendors (IMF, 2020). A similar commitment was made by the CBL for its spending. However, implementation remained largely stalled pending the establishment of macroeconomic stability and the availability of sufficient high-quality Liberian dollar currency. Indeed, after having make some good progress in this direction and increasing the share of the wage bill paid in Liberian dollars from 20 percent to 35 percent, the government was forced to backtrack due to insufficient sensitization of civil servants and severe shortages of Liberian dollars in late 2020. The de-dollarization trend also suffered a setback from late 2019 as substantial inflows of foreign exchange receipts could not be converted into Liberian dollars due to banknote shortages. As a result, the share of foreign currency loans and deposits started trending upward again from 2021 (Figure 1). Now that the Liberia has made good progress with achieving macroeconomic stability and with the rollout of the currency changeover, the time has come to re-initiate the de-dollarization process in Liberian.
26. The authorities should pursue a gradual and market-based de-dollarization strategy. Indeed, cross-country experience suggests that a package of market-oriented policies is critical for a successful de-dollarization, while non-market-based measures and forced de-dollarization usually fail and could, in any event, be unlawful in Liberia given the legal-tender status of the U.S. dollar (see box 1). De-dollarization takes a concerted effort and requires the support of both the government, the central bank, and the rest of the public, with the government taking the lead. Cross-country experience also suggests that de-dollarization takes time and therefore requires patience and persistence. For example, it took about two decades for Peru and Bolivia (between 2000 and 2019) and Israel (between 1984 and 2002) to make substantial progress on de-dollarization. Like these countries, Liberia needs time to entrench macroeconomic stability and policy credibility more firmly.
Liberia: Cross-country Experience with De-dollarization
The cross-country experience suggests that macroeconomic stability, especially price stability, is a precondition for successful de-dollarization. To achieve this, a number of dollarized economies, such as, Peru, Vietnam and Cambodia, pursued restrictive monetary policy based on monetary targeting frameworks within managed floating exchange rate regimes at the early stages of disinflation. The monetary aggregate targeting framework was instrumental in reducing and stabilizing inflation from the hyperinflation of the late 1980s and 1990s. The adoption of managed floating exchange rate regimes in countries like Peru between 1990 and 2000, Vietnam during the 1990s, and Cambodia was meant to mitigate risks associated with balance sheet effects and to anchor inflation expectations. Once inflation is anchored and the central bank’s credibility is established, there is a tendency to move from quantitative monetary targets to price target within the overall framework of inflation targeting (IT).1
Market-based approaches to de-dollarization have proven to be successful in promoting de-dollarization once the countries have made good progress on macroeconomic stability. A combination of prudential policies, which forces economic agents to internalize the risks of operating in foreign currency, and financial market development in domestic currency were instrumental in promoting de-dollarization in countries like Bolivia, Peru, Israel and Uruguay. Peru and Bolivia achieved substantial progress on de-dollarization between 2000 and 2019 through a combination of macroeconomic stability, development of financial markets in domestic currency, and prudential measures, such as higher regulatory requirements on foreign currency liabilities and higher provisions on foreign currency assets and limits on open foreign exchange positions (García-Escribano and Sosa, 2011 and Levy-Yeyati, 2021). Israel also achieved substantial progress on de-dollarization between 1984 and 2002 through a combination of macroeconomic stability, development of financial markets in domestic currency and active banking supervision to ensure that banks fully cover foreign-currency positions (Galindo and Leiderman, 2005).
Non-market-based approaches to de-dollarization such as forced conversion to domestic currency tend to be unsuccessful. Some examples include Pakistan in 1998, Argentina in 2001, Bolivia and Mexico in 1982, and Peru in 1985. In the case of Bolivia and Peru, forced conversion was followed by capital flight in the form of increases in offshore deposits, a sharp decline in financial intermediation, and a sharp increase in inflation (Reinhart et al., 2003; Galindo and Leiderman, 2005). In subsequent years, the policy was reserved because of these unintended consequences.
1 For example, Peru adopted IT in 2002 while Paraguay in 2011 following their success in reducing inflation rates.
27. Considering the international experience and Liberia’s idiosyncrasies, the following elements could make for a successful de-dollarization policy package:
Building a track-record of macroeconomic stability is the precondition for a successful de-dollarization. In this regard, the authorities should continue with reforms aimed at strengthening the role of monetary policy in promoting price stability. The implementation of anti-inflation policies and measures aimed at enhancing monetary policy credibility and central bank independence would go a long way. These should be complemented by measures aimed at instilling fiscal discipline and firmly staying away from monetary financing of the budget. Macroeconomic stability, especially price stability, reduces the hedging benefits of foreign currency and therefore boosts confidence in the local currency.
The currency changeover needs to be executed in an orderly fashion and implemented so as to guard against excessive currency injection. While the currency changeover should boost confidence in the Liberian dollar and hence foster de-dollarization, it also comes with operational and inflation risks. The CBL’s Currency Changeover Implementation Plan is designed to mitigate operational risks and should be closely adhered to. To avoid inadvertently injecting too much currency and reignite inflation, the CBL will also need to monitor monetary aggregates very closely through the planned new currency management system, the exchange rate, and other pertinent indicators as the currency changeover operation gathers speed later this year. It would be a big setback for de-dollarization if any of these risks materialized.
The CBL should continue to improve its monetary policy framework to enhance the effectiveness of monetary policy in anchoring inflation and promoting macroeconomic stability. As a prerequisite, the CBL should continue to build the capacity and necessary infrastructure to conduct sound monetary policy. These include technical capacity to analyze and monitor economic developments that may have implications for liquidity and monetary conditions and putting adequate policy tools to manage liquidity in the banking sector in place.
The authorities could consider prudential regulations that reduce banks’ incentives to focus on foreign currency deposits and foreign currency loans. One potential avenue in the context of Liberia is to review reserve requirements regulation. Currently, the regulation requires commercial banks to hold higher required reserve on Liberian dollar liabilities than on U.S. dollar liabilities, thereby promoting dollarization rather than discouraging it. In the shorter term, reserve requirement ratios on Liberian dollar and U.S. dollar liabilities should be harmonized in line with the CBL’s existing plans and commitments. In the medium to long term, the CBL could consider higher reserve requirements on U.S. dollar liabilities than on Liberian dollar liabilities. Other prudential measures could include higher liquidity requirements on U.S. dollar liabilities, higher capital requirements, provisions against foreign exchange exposure, and extra capital requirements on open foreign exchange positions. These measures would not only foster de-dollarization by increasing the cost of financial intermediation in foreign currency, but also enhance the resilience of the banking system against exchange rate, liquidity, and credit risks associated with financial dollarization.
Financial sector reforms, including strengthening the regulatory environment through risk-based supervision could also facilitate de-dollarization by forcing banks to internalize the risks and costs associated with financial intermediation in foreign currency.
Financial market development through the introduction of a variety of local currency-denominated financial instruments could further foster de-dollarization, as it did in Bolivia, Peru, and Israel (Levy-Yeyati, 2021; Galindo and Leiderman, 2005). To promote de-dollarization in Liberia, the government could review its debt management strategy and rebalance its debt from U.S. dollar to Liberian dollar denomination. Currently very little public debt is denominated in Liberian dollar currency because of its perceived high costs, but this could be addressed by selling government securities in variable-rate auctions to help ensure competitive pricing or considering price-indexed instruments.12 In the medium term, the government could consider issuing short- to medium-term debt securities on a regular basis and thereafter move to issue medium- to long-term government securities in local currency. This would provide an alternative to investing in foreign-currency assets and enhance the store of value function of the Liberian dollar. It would also help establish a domestic-currency yield curve, which would foster the development of private capital markets by providing a benchmark for pricing domestic-currency instruments and help banks fund and price long-term credit denominated in domestic currency.
Enhancing the efficiency of foreign exchange markets could also promote de-dollarization. It would improve ready market access to foreign exchange and thus reduce the need to hold foreign currency for precautionary purposes.
Government operations should be conducted in Liberian dollars to the extent possible. In the short term, the government should start progressively raising the share of its wage bill paid in Liberian dollars from currently 20 percent to 50 percent. Taxes and other government levies should also be increasingly collected in Liberian dollars. Some of these measures could be extended to other government institutions and the private sector could be encouraged to follow suit. But over time, the government should go further than just making and receiving payments in Liberian dollars and start stipulating contracts in Liberian dollars. This should be accompanied by careful sensitization.
Complementary steps could include administrative measures that have proven to be successful in contributing to de-dollarization in other countries. For example, Peru introduced regulation that required retailers and wholesalers to list prices in domestic currency in 2004 (García-Escribano, 2010). To discourage the use of foreign currency for payments, Bolivia imposed Tobin-type tax on foreign-currency financial transactions, especially on foreign currency-denominated bank accounts (Levy-Yeyati, 2021). However, for these measures to be successful, the government should sensitize the public adequately and well in time to avoid resistance from the public.
E. Conclusions
28. Liberia has made good progress with re-establishing macroeconomic stability and the unfolding currency changeover is addressing cash shortages and the poor banknote quality. With that, key preconditions are in place for Liberia to push ahead with modernizing its monetary policy framework and de-dollarization.
29. Regarding the modernization of the monetary policy framework, the paper recommends adopting a flexible reserve money operating targeting framework as a transitional arrangement. It involves setting a medium-term target path for reserve money consistent with price stability objectives, resuscitating the interest rate corridor system with standing deposit and credit facilities and revamped procedures, introducing variable rate tender auctions of CBL bills, and using CBL bills as the main instrument to manage banks’ excess reserves to ensure adequate liquidity at all times while steering reserve money toward its medium-term target path.
30. More generally, monetary policy should be guided by indicator variables rather than a rigid intermediate target. The analysis in this paper suggests that the exchange rate, as well as currency in circulation and the domestic currency component of narrow money, have the best predictive powers for inflation in Liberia and should therefore serve as the prime indicator variables. If the indicator variables signal an impending miss of price stability objectives, deviations from the targeted medium-term reserve money path should be allowed and, if the signals persist, the path should be revisited.
31. The flexible reserve money operating targeting regime will help prepare the ground and the infrastructure necessary for a successful transition to a modern monetary policy framework based on an interest-rate operating target. Money and interbank markets will develop, the interest rate channel of monetary policy will become stronger, money demand will grow more predictable, monetary policy will gain further credibility, and a yield curve will start to emerge. As this happens, the CBL could deemphasize reserve money as the operating target and transition to an interest rate-based framework. Interest rates at which short-term CBL bills are auctioned on a full-allotment basis would serve as the monetary policy rate and be kept toward the mid-point of the interest rate corridor.
32. Regarding de-dollarization, the paper lays out the various associated advantages and recommends that a concerted effort be made. De-dollarization would give monetary policy much stronger traction, strengthen the CBL’s lender-of-last resort function, earn Liberia seigniorage, and allow the exchange rate to serve as a shock absorber. Experiences from other countries suggest that it takes a package of market-based measures by the government and the central bank to bring de-dollarization about and that it takes time. In Liberia, such a package could include increasingly affecting government payments and revenue collection in Liberian dollars, stipulating government contracts progressively in Liberian dollars, gradually raising banks’ reserve and liquidity requirements for U.S. dollar liabilities above those for Liberian dollars, developing Liberian dollar debt markets, and supporting administrative measures. Macroeconomic stability and price stability would need to be preserved throughout to become more entrenched.
References
Acosta-Ormaechea, S., and D. Coble, 2011, “The Monetary Transmission in Dollarized and Non-Dollarized Economies: The Cases of Chile, New Zealand, Peru and Uruguay”, IMF Working Paper No. 11/87, (Washington: International Monetary Fund).
Alvarez-Plata, P., and A. García-Herrero, 2008, “To Dollarize or De-dollarize: Consequences for Monetary Policy,” DIW Berlin Discussion Paper 842, German Institute for Economic Research, Berlin, Germany.
Armas, A., A. Santos, and M. Tashu, 2015, “Monetary Policy in a Partially Dollarized Economy: Peru’s Experience with Inflation Targeting,” In: Santos, A., and A. Werner. Peru : Staying the Course of Economic Success, pages 191–206, (Washington: International Monetary Fund).
Baliño, T., A. Bennett, and E. Borensztein, 1999, “Monetary Policy in Dollarized Economies,” IMF Occasional Paper No. 171, (Washington: International Monetary Fund).
Barth, R., 2002, “The Framework of Monetary Policy,” In: Khan, M. S., S. M Nsouli and C-H. Wong, Macroeconomic management: programs and policies, pages 139–167, (Washington: International Monetary Fund).
Berg, A., and E. Borensztein, 2000, “The Choice of Exchange Rate Regime and Monetary Target in Highly Dollarized Economies,” Journal of Applied Economics, 3:2, 285-324.
Dabla-Norris, E. and H. Floerkemeier, 2006, “Transmission Mechanisms of Monetary Policy in Armenia: Evidence from VAR Analysis,” IMF Working Paper No. 06/248 (Washington: International Monetary Fund).
Erasmus, L., J. Leichter, and J. Menkulasi, 2009, "De-dollarization in Liberia-lessons from cross-country experience," IMF Working Paper No. 37 (Washington: International Monetary Fund).
Galindo, A. J., and L. Leiderman, 2005, "Living with Dollarization and the Route to Dedollarization," Inter-American Development Bank Working Paper.
García-Escribano, M., 2010, “Peru: Drivers of De-dollarization?”, IMF Working Paper No. 10/169 (Washington: International Monetary Fund).
García-Escribano, M. and S. Sosa, 2011, “What is Driving Financial De-dollarization in Latin America?”, IMF Working Paper No. 11/10 (Washington: International Monetary Fund).
Goujon, M., 2006, “Fighting inflation in a dollarized economy: The case of Vietnam,” Journal of Comparative Economics, Vol. 34: 564–581.
IMF, 2014, “Liberia: Third Review Under The Extended Credit Facility Arrangement And Request For Waiver Of Nonobservance Of Performance Criterion And Modification Of Performance Criteria—Staff Report; And Press Release,” IMF Country Report No. 14/197. (Washington: International Monetary Fund).
IMF, 2020, “Liberia—First and Second Reviews Under the Extended Credit Facility Arrangement, Request for Waivers of Nonobservance of Performance Criteria and Modification of Performance Criteria,” EBS/20/184. (Washington: International Monetary Fund).
Jordà, Ò., 2005, “Estimation and Inference of Impulse Responses by Local Projections,” American Economic Review, Vol. 95(1): 161–182.
Kokenyne, A., J. Ley, and R. Veyrune, 2010, “Dedollarization,” IMF Working Paper No. 10/188 (Washington: International Monetary Fund).
Laurens, B. J., K. Eckhold, D. King, N. Maehle, A. Naseer, and A. Durré, 2015, “The Journey to Inflation Targeting: Easier Said than Done The Case for Transitional Arrangements along the Road,” IMF Working Paper No. 15/136 (Washington: International Monetary Fund).
Levy-Yeyati, E., 2021, “Financial dollarization and de-dollarization in the new millennium,” Latin American Reserve Fund Working Paper.
Maehle, N., 2020, “Monetary Policy Implementation: Operational Issues for Countries with Evolving Monetary Policy Frameworks,” IMF Working Paper No. 20/26 (Washington: International Monetary Fund).
Masuch, K., S. Nicoletti-Altimari, and M. Rostagno, 2003, “The role of money in monetary policymaking,” BIS Papers, No. 19, pp-158-91.
Reinhart, C. M., K. S. Rogoff, and M. A. Savastano, 2003, “Addicted to dollars,” NBER Working Paper 10015 (Cambridge, MA; National Bureau of Economic Research).
Appendix I. Operational Guideline for Implementing Monetary Policy under Reserve Money Operating Targeting Framework
This appendix presents an ideal operational procedure for implementing monetary policy under reserve money operating targeting framework.
Step 1: Setting of reserve money target. In general, the target for reserve money is set based on the assumptions regarding inflation path (consistent with the policy objective), real growth and exchange rate path given money multiplier and velocity of money. In the context of IMF-supported program with Liberia, the target is guided by monetary conditionality under the ECF program, which consists of a floor on net international reserves (NIR) (which is the main component of net foreign assets (NFA)) and a ceiling on net domestic assets (NDA). From the quarterly targets, the central bank could derive some indicative monthly targets consistent with medium-term targets. However, to provide sufficient flexibility for short-term liquidity management in a manner that helps to stabilize short-term interest rates, reserve money targets should serve as medium-term targets or benchmarks that do not dictate, but only guide the medium-term evolution of the daily liquidity management operations.
Step 2: Deriving the target path for total bank reserves.1 The approach would be to derive a target path for total bank reserves (including required and excess reserves) that is consistent with the medium-term reserve money target (determined in step 1) and medium-term forecasts for currency in circulation.
Step 3: Preparing short-term liquidity forecasts. The approach is to produce short-term (weekly or monthly) liquidity forecasts based on the evolution of the main items of central bank balance sheet, with the variable of interest being the forecasts for total bank reserves. The forecasting horizon should cover at least reserve maintenance period and could also be extended to quarterly to ensure that overtime liquidity management is consistent with quarterly reserve money target. The forecasts should be regularly updated as new information becomes available. To this end, daily liquidity monitoring is key. Coordination with the Ministry of Finance and Development Planning with respect to government cashflow and domestic debt management is also essential. The accuracy of liquidity forecasts is also important as this could help reduce volatility in excess reserves, interest rates, and go a long way in enhancing the efficiency of the liquidity management.
Step 4: Determining excess reserves/liquidity to guide the size of monetary operations. Excess reserves/liquidity is calculated as the difference between the forecasts for total bank reserves (step 3) and the derived target path for total bank reserves (step 2). This would inform the size of monetary operations: how much CBL bills should be issued. That is, the calibration of monetary operations should be based on the estimates of excess reserves/liquidity, with CBL-bill auctions managing excess reserves so that reserve money returns to the target path over time.
Step 5: Conducting monetary policy operations.
The central bank announces the target amount of CBL bills for the auction based on the estimates of excess liquidity (step 4).
Then, let the market (banks) to determine the price (interest rate). For short-term (two-weeks) CBL bills, the pricing of liquidity could be guided by the establishment of interest rate corridor system, with interest rate on standing credit facility acting as a ceiling while the rate on standing deposit facility acts as a floor of the corridor.
To promote some competition, especially in Liberian context, banks could be allowed to submit multiple bids (maximum of 3 bids per bank).
Step 6: Reassessing the assumptions underlying reserve money target. To ensure that the reserve money target remains in line with evolving macroeconomic developments (fundamentals), the assumptions underlying the reserve money target (in step 1) should be regularly (e.g., quarterly, or semi-annually) reviewed and adjusted as needed (as new information become available). Periodic reviews of the key parameters that inform the monetary program should be done to quantify the deviations from the target and to assess the need to adjust the target. The periodic reassessment could draw on a broad range of information, including the information provided by the proposed indicator variables (exchange rate, currency in circulation and domestic component of narrow money), and economic assessment of the state of the economy. Once the underlying assumptions are updated, the circle starts again with step 1.
Prepared by Thabang Molise.
See for example Baliño et al., (1999), Alvarez-Plata and García-Herrero (2008) and Acosta-Ormaechea and Coble (2011).
In an interest rate operating target, the central bank accommodates banks request for liquidity without any limitation on the quantity. In a reserve money operating target, the central bank decides on the quantity in reference to its reserve money target and let the market decides on the interest rate. See also September 2021 MCM TA Report on Improving Monetary Operations in Liberia.
Flexible reserve money targeting is still relevant in countries with underdeveloped financial markets, with low level of financial intermediation. It provides a tool and a framework to protect the central bank from fiscal dominance or political pressure to meet fiscal objectives (monetary financing) that could lead to a loss of monetary control (Laurens et al., 2015:12).
All data are expressed in natural logs and are seasonally adjusted. The lag order of the VAR model is selected based on the Akaike information criterion (AIC).
The intuition behind this test is to establish whether the past values (information) of one variable can help predict the future values of another variable.
To address the shortcoming of the Granger causality test, the paper also performed some additional experiments with longer lag length (e.g., higher than 12 months) and establishes that domestic component of narrow money and currency in circulation also have some predictive power for future inflation.
For the VAR, the impulse response of variables to a shock are identified using the Cholesky decomposition. The ordering used is from money, to exchange rate and lastly inflation. Given high exchange rate pass-through in Liberia, it seems plausible to assume that inflation responds to exchange rate and money contemporaneous. The results with different ordering scheme are qualitatively similar to those reported in Figure 2. For robustness purpose, the paper also estimates the impulse responses of variables to a shock using local projections (Jordà, 2005).
The CBL should continue to allow exchange rate flexibility and only intervene in the foreign exchange market to address disorderly market conditions and as a means to steadily build up international reserves and inject domestic currency in line with the needs of the economy.
See Appendix I for more detail.
The IMF recently provided technical assistance to the CBL on improving liquidity monitoring and forecasting framework (See November 2021 TA Report on Improving Liquidity Monitoring and Forecasting Framework in Liberia).
In the immediate term to medium term, promotion of price indexation especially for longer maturities could also be explored. This would facilitate a switch from foreign currency-denominated instruments to domestic currency-denominated instruments, especially when monetary policy credibility is low. It would be more convincing to promote price-indexed instruments than domestic currency instruments when monetary policy credibility is low. Once monetary policy credibility has improved, price indexation could be phased out.
Using the implied target path for total bank reserves, and not reserve money, as a short-term operational target insulates the interbank money market from short-term fluctuations in demand for currency in circulations (which can be large and outside the control of the central bank – autonomous factor of liquidity) and therefore help reduce volatility of short-term interest rates (Laurens et al., 2015:17).
