Monetary policy credibility and macrodynamics: evidence from Ghana

ABSTRACT

We compute a time-varying metric of monetary policy credibility based on Ghana’s experience, using both symmetric and asymmetric approaches. We then follow-up with some empirical evidence to address the linkages between macroeconomic developments and central bank credibility. The empirical results reveal high and low credibility cycles with an average duration of 2 years over the study period. Particularly, higher levels of credibility were associated with stable domestic currency and lower nominal interest rates. This reinforces the notion that efficient monetary policy delivers higher central bank credibility with better outcomes for macroeconomic variables. In contrast, the level of credibility tends to worsen in the wake of weakening macro fundamentals which are not adequately countered by monetary policy decisions. There is therefore the need for efficient monetary policy formulation to achieve a stable macroeconomic environment in Ghana. This will in the long-run build policy credibility towards attaining the central bank’s medium-term inflation target.

KEYWORDS:

• Credibility
• inflation targeting
• macrodynamics
• Bayesian VAR
• expectation
• EGARCH
• threshold
• Ghana

JEL CLASSIFICATION:

• E52
• E58
• E6
• F31

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ According to CK2002 index, the central bank is considered fully credibility (CRED • =1), if expected annual inflation is lower than or equal to the inflation target. In contrast, it is non-credible (CRED • = 0) if expected inflation is equal to or higher than 20%.

² Their approach considers any deviation of inflation expectation from the inflation target (either positive or negative) as a loss of central bank credibility. Based on the same approach, de Mendonça and de Guimarães E Souza (2009) and Neuenkirch and Tillman (2014) propose a backward-looking measure of central bank credibility based on the past deviation of inflation from the target. Such backward-looking indicators are particularly germane for developing countries for which inflation expectation data are often unavailable.

³ DM2007 index was developed based on notion that negative deviation of expectation from target can also imply loss of credibility coupled with the fact that in practice the target is not a single value but a range. Thus, full credibility is obtained when the expected inflation is exactly equal to the midpoint of the inflation target, and the index decreases symmetrically and linearly when expectations deviate from target point.

⁴ DMGS2009 index was developed with the view that full credibility is not only achieved when inflation expectations are exactly equal to the midpoint target but it is also the case when inflation expectations are within the target range. Therefore, the focus on the midpoint by the preceding approaches is too restrictive and can lead to false conclusion.

⁵ The inverse LINEX credibility approach considers only positive deviation of inflation expectation from target as a loss of central bank credibility. Thus, the impact of negative deviations of central banks’ credibility are perceived to be less serious (if any) than that of positive deviations.

⁶ A similar issue has been addressed by de Mendonça and de Guimarães E Souza (2009) in the case of Brazil using OLS approach while LLR (2015) explored 18 emerging IT economies (excluding Ghana) using General Auto-Regressive Conditional Heteroscedastic (GARCH) approach. Both studies focused mainly on how short-term interest rate responds to variation in policy credibility.

⁷ The mean equation of our GARCH model is a second order autoregressive process augmented with the inflation rate, with a constant that is supposed to represent both the inflation target and the long-run equilibrium exchange rate. The time-varying conditional variance of exchange rate is supposed to follow EGARCH (1, 1) process, augmented with credibility indicator as additional determinant.