Inflation volatility, economic growth and monetary policy in Bangladesh

Abstract

This article provides an overview of the trends and movements of CPI-inflation in Bangladesh since the early 1950s and examines the key issues in rule-based monetary policy for price stability, implying low and stable inflation, in this country. Under a fixed exchange rate system, inflation in Bangladesh was moderately high and volatile during the 1950s and 1960s. Since the country’s independence from Pakistan in 1971, inflation in Bangladesh has remained moderately high on average and highly volatile and persistent under a fixed-pegged exchange rate system or under a managed floating system since 2003. Using data from the early 1970s or earlier depending on data availability, the article undertakes both Granger-causality and the structural vector autoregression (SVAR) analysis with two models. The first model is comprised of such variables as inflation, the real interest rate, the real exchange rate and output growth, and the second model is comprised of the volatilities of money growth, real output growth and inflation. Then, based on the empirical findings, the article concludes that a rule-based monetary policy, namely monetary targeting or inflation targeting, remains appropriate for Bangladesh provided that it adopts a more flexible, if not freely floating, exchange rate system. The article suggests that the use of monetary policy to achieve multiple objectives under a fixed-pegged exchange rate system creates a time-inconsistency problem, reduces monetary policy credibility and makes it (monetary policy) ineffective in lowering inflation and its volatility. Low credibility of monetary policy in particular raises inflation persistence. Within the present monetary-policy framework in Bangladesh, the article illustrates how the fixed-pegged exchange rate system has generated money growth volatility in the presence of large-scale inflows of overseas workers’ remittances and readymade garments export earnings. This does not seem to be a concern of the central bank of Bangladesh (Bangladesh Bank); rather, it (Bangladesh Bank) pursues monetary-base targeting to keep inflation low and stable after considering economic growth. The consequent diminishing credibility of monetary policy has kept inflation volatile and persistent, which has adversely affected economic growth.

Keywords:

• inflation
• inflation volatility
• economic growth
• Granger-causality
• SVAR analysis
• monetary policy
• Bangladesh

JEL Classification:

• C22
• E52

Acknowledgement

The author would like to thank two anonymous referees and the editor of this Journal for their comments and suggestions for improving this article.

Notes

¹ The Latin American structuralist model of inflation considers structural rigidities as the key determinant of inflation in developing countries. Money growth is then seen as the effect rather than an independent source of inflation (Corbo, 1974).

² The open-economy trilemma states that the three desirable goals of an open economy, namely capital mobility (financial integration), independent monetary policy (ability to set the interest rate or control the money stock) and exchange rate stability (fixed or pegged exchange rate), cannot be maintained simultaneously; at least one of them must be sacrificed. Control over capital flows gives monetary policy some independence under a fixed-pegged exchange rate system. Although intermediate regimes are feasible, they involve trade-offs and create ambiguity and nontransparency in economic policy (Krugman et al., 2015). Monetary policy then becomes ineffective in achieving price stability.

³ At least three situations lead to a boom-bust cycle under a fixed-pegged exchange rate system. Firstly, if the pegged currency country adopts expansionary fiscal and monetary policies, the outcome is overvaluation of the real exchange rate. Real exchange rate misalignment often leads to a balance-of-payments crisis. Secondly, expansionary fiscal and monetary policies in the anchor-currency country translate into a boom-bust cycle in the pegged currency country. Thirdly, large-scale capital inflows can cause a boom-bust cycle under a fixed-pegged exchange rate system. Bentour and Razzak (2009, p. 2) have illustrated how a boom-bust cycle emerged in Hong Kong in the 1990s due to expansionary monetary policy in the anchor-currency country. Under a pegged system (ignoring country risk), the domestic nominal interest rate i equals the foreign interest rate i*, that is, i ≅ i*. If the domestic inflation rate π exceeds the foreign inflation rate π*, then the domestic real interest rate r falls below the foreign real interest rate r*. Under these circumstances, if the anchor-currency country lowers the real interest rate to a low or zero level, the real interest rate in the pegged currency country becomes negative. Hong Kong experienced this phenomenon from 1990 to 1999 when inflation in Hong Kong, which operates under a currency board system, was 7% while inflation in the United States was 3%.

⁴ Calvo and Reinhart (2002) have explained the apparent fixity of the exchange rates of currencies in developing countries as fear of floating.

⁵ To demonstrate the interactions among these variables, the author has developed a small structural model. The model is not reported but would be available upon request.

⁶ The data deployed for estimation purposes are drawn from Bangladesh Bank various issues of Economic Trends, Bangladesh Bureau of Statistics’ Statistical Yearbook of Bangladesh (various years) and Twenty Years of National Accounting (2003), IMF’s International Financial Statistics (various years) and the Government of Bangladesh’s Bangladesh Economic Review (2012).

⁷ Some formal econometric tests, namely the ADF test, the KPSS test and the Zivot–Andrews test, reject the null hypothesis that inflation in Bangladesh has a unit root. As inflation has high persistence, it is difficult to lower inflation following inflationary shocks without enhancing the credibility of monetary policy by adopting a rule-based monetary policy. On this issue, see Fuhrer (2011).

⁸ Additional empirical results (not reported) suggest that a positive inflation shock has a greater impact on inflation volatility than a negative shock does. This article does not investigate the issue further.

⁹ Section V empirically investigates the interrelations among money growth volatility, output growth volatility and inflation volatility.

¹⁰ The model does not include the source of inflation. However, Section IV shows that inflation in Bangladesh has monetary roots.

¹¹ Because a grossly appreciated real exchange rate induces capital flight, the real interest rate is required to rise as an inducement to slow down or stop capital flight. This follows the open-economy condition of real interest rate parity.

¹² Formally, the Bangladesh Bank adopted monetary targeting in 2005 (Bangladesh Bank, 2005).

¹³ An anonymous referee has expressed concern about the role of cost factors in inflation. The question raised is whether it is justified to rely on the Quantity Theory of Money that underwrites a rule-based monetary policy such as monetary targeting or inflation targeting in some respects when cost factors remain important in inflation. This article takes the view that although money growth remains the key, if not the sole, source of inflation over the long run, the monetary approach does not deny the impact of supply shocks or cost factors or both on inflation and its volatility in the short run. This point has been widely discussed in the monetary literature, including in four volumes of Handbook of Monetary Economics edited by Friedman and Hahn (1990; volumes 1 and 2) and by Friedman and Woodford (2011; volumes 3A and 3 B). Mishkin (2013) has reviewed this issue in detail. Supply shocks or cost factors or both affect the relative prices of products that are affected by these shocks. The shocks are transitory in nature and they do not affect inflation in the sense of causing a sustained increase in the price level unless the monetary authorities adjust the money growth to avoid the impact of shocks on output. This article does not examine the role of cost factors in inflation and its volatility.

¹⁴ Hossain (2006, 2010) has found that real broad money balances, real income, the deposit rate of interest, the US Treasury bill rate and the nominal effective exchange rate of the taka have a stable, cointegral relationship. The empirical results suggest some parameter instability in the broad money demand function during the early 1990s, which was caused by financial deregulation and reforms. The model has remained stable since the early 2000s.

¹⁵ For example, the share of broad money (M2) in GDP increased from 22.6% in 1991 to 52.4% in 2011.

¹⁶ The Bangladesh Bank’s website lists the objectives of monetary policy as follows: (1) price stability, (2) sustainable growth and development, (3) high employment, (4) economic growth and efficient use of resources and (5) stability of financial and payment system. See http://www.bb.org.bd/fnansys/regulator.php.

¹⁷ That is, FR = FR^$ NER where NER is the fixed exchange rate of domestic currency per unit of foreign currency ($).

¹⁸ For simplicity, the concepts foreign assets and foreign reserves are used synonymously.

¹⁹ In general, where CAB is the current account balance, is the change in NFA of nonbank entities or nonmonetary financial flows and is the change in NFA of the banking system or monetary financial flows. Current account surplus represents an increase in net official or private claims on nonresidents or the acquisition of reserve assets by the monetary authorities.

²⁰ For discussion on the specifications of these models, see Rajan and Gopalan (2015).

²¹ As money is neither neutral nor super-neutral in the short run, the monetary authorities attempt to stimulate economic activity when it perceives the need through ad hoc generation of inflation surprises. This creates monetary policy and inflation uncertainties. Furthermore, when there is political instability, the greater the ambiguity of monetary policy, the easier it is for the monetary authority to generate inflation surprises. This raises money growth and inflation on average (Cukierman and Meltzer, 1986). To escape accountability to the public, the monetary authorities who are responsible for achieving multiple objectives introduce monetary control procedures, which remain opaque and nontransparent.

²² An anonymous referee has inquired of the key requirements of credible monetary policy, which are suggested as follows. Firstly, to make monetary policy an independent policy instrument, the exchange rate has to be determined by market forces. This could be associated with market-based capital controls, if needed, until the financial sector is developed to absorb high volatility of exchange rates. The aim of this transitional arrangement is to stabilize the exchange rate or prevent it from overshooting or causing financial instability. Secondly, the objective of monetary policy has to be narrowed to price stability, measuring low and stable inflation. Thirdly, fiscal discipline has to be maintained to reduce pressure on monetary policy. Finally, the central bank should be given autonomy to conduct monetary policy and make it accountable to the parliament for its assigned task of keeping inflation low and stable.