The paper estimates McCallum and Taylor monetary policy reaction functions, hybrids mixing instruments and targets from the two frameworks, and nominal feedback mechanisms for 20 emerging market economies. The choice of framework employed in analysing each country is informed by the corresponding institutional setting. McCallum–Taylor specifications with an interest rate instrument and a nominal income gap target perform better than benchmark Taylor reaction functions in describing monetary policy in inflation targeting economies. Estimating reaction functions for economies operating monetary and exchange rate targeting regimes produces mixed results, while the nominal feedback rules reveal a lean-with-the-wind behaviour in two out of three economies.
The policy strategy central banks in emerging economies follow can at times be difficult to follow (e.g., Calvo and Mishkin, 2003). In emerging economies fragile institutions, lack of central bank independence, and a scarcity of monetary instruments can render monetary control difficult. But following the crises in the 1990s many emerging market economies have been making considerable efforts to improve policy making institutions and empirical evidence could help in determining how central banks react to fundamental economic developments. Empirically estimated reaction functions can provide valuable information for discussing monetary policy and therefore may ultimately help in safeguarding macroeconomic stability (e.g., Billi, 2009).
The paper contributes by investigating monetary policy behaviour in 20 emerging market economies. The investigation seeks to answer the following questions: Are Taylor-type, interest rate, monetary policy reaction functions useful for understanding monetary policy performance in inflation targeting economies? Can McCallum-type, monetary base, reaction functions help in characterizing central bank policy in economies operating monetary targeting and exchange rate targeting policy strategies? For economies with a dearth of statistical information on the real economy and relatively low levels of institutional development, can a nominal monetary policy feedback rule embodying an inflation targeting mechanism approximate historical monetary policy conduct?
Understanding monetary policy behaviour in relation to the prescriptions of key analytical frameworks is relevant (e.g., Taylor and Williams, 2010). John B. Taylor argues that the 2008 financial crisis is at least partly to be blamed on a loose United States Federal Reserve monetary policy stance in comparison to the prescriptions of the Taylor rule (Wall Street Journal, 9 February 2009, page A19). Bernanke (2010) discusses the validity of that view.
The paper seeks to produce evidence about past monetary policy behaviour by estimating a family of reaction functions deriving from McCallum, 1988 and McCallum, 1999 and Taylor, 1993 and Taylor, 1999 contributions. The relevant empirical literature on emerging economies often adopts a framework without exercising judgment about which reaction function is adequate on the basis of the declared monetary policy regime and institutional idiosyncrasies. The investigation tackles that shortcoming by assembling a record of monetary policy institutions in 20 emerging markets which informs the subsequent empirical modelling.
The literature investigating monetary policy reaction functions uses much sophistication (e.g., Dolado et al., 2005 and Mavroeidis, 2010), but Kozicki (1999) argues that to be useful an empirical rule should be robust to minor variations. The strategy of the paper involves estimating benchmark, comparable, specifications using conventional econometric estimators to gain a better understanding of monetary policy in emerging market economies. In the spirit of McCallum (2000), the analysis runs a range of policy feedback rules in determining if the estimated central bank reactions are different depending on the type of instruments and targets considered for each economy.
The rest of the paper proceeds as follows. Section 2 gives an account of developments in monetary policy institutions in emerging market economies. Section 3 explains the monetary policy frameworks used in the empirical modelling. Section 4 describes the time series data and the family of monetary policy reaction functions. Section 5 runs the battery of reaction functions deriving from McCallum's and Taylor's contributions, and discusses the results. Section 6 concludes.
The paper investigates monetary policy behaviour in emerging market economies. The analysis empirically estimates McCallum and Taylor monetary policy reaction functions, hybrids mixing instruments and targets from the two frameworks, and nominal feedback mechanisms for 20 economies implementing diverse monetary policy strategies. The modelling finds that the behaviour of the inflation targeting economies is better captured with a hybrid McCallum–Taylor rule incorporating a nominal income target than with a benchmark Taylor-type rule. Countries pursuing a mix of monetary and exchange rate targets portray differences in the reaction of policy to domestic targets – output gap, inflation gap, or a nominal income target – and the exchange rate.
Future research on the topic could proceed by exploring several avenues for emerging market economies. Research along the lines of Orphanides's (2001) investigation for the US using Taylor's rule and real-time economic data could lead to better understanding central bank behaviour. Additionally, studies focusing on advanced economies incorporate a data-rich environment in modelling monetary policy, e.g., Bernanke and Boivin (2003). In that way the empirical analysis of central bank performance is closer to real decision-making process. Finally, some studies explore additional ways of incorporating rich data sets in understanding monetary policy, and this could be meaningfully applied for emerging economies as well. For example, Giannone et al. (2008) develop a method for extracting information from large time series data sets released at different times and with diverse lags that can be used in nowcasting, e.g. in assessing current quarter GDP growth to inform monetary policy decisions.
