Ebook Monetary Policy Conduct Based on Nonlinear Taylor Rule: Evidence from South Africa
Many studies have tested the Taylor rule for monetary policy conduct internationally. However, there have been few studies carried out for emerging markets. Studies undertaken by Petersen (2007), Castro (2008) and Cukierman (2004) mainly focus on nonlinear models in developed economies such as the US and UK. Notably, there is a gap within emerging markets in particular South Africa that presents an opportunity for theutilization of nonlinear models to characterize the behaviour of the Reserve Bank using interest rate functions. Interest rate reaction functions have normally been formulated using the linear Taylor rule. This could be attributed to the notion that linear models on several cases are perceived to render reasonable approximations to the exact nonlinear interactions.
The Taylor rule spells out that the interest rate adjusts in accordance to the deviation of inflation from its target and real output from potential output. It also assumes in the US for instance, the federal funds rate is raised by 1.5 percentage points for each 1 percentage point increase in inflation (Taylor 1993). Further, an increase in the interest rate of that magnitude would raise real interest rates and help cool off the economy, thus reducing inflationary pressures. According to Taylor (1993), the rule also assumes that interest rates are reduced by 0.5 percentage point for each percentage point decline in real GDP below its potential. Such a reduction in the interest rate helps to mitigate a (growth cycle) recession and maintain price stability.
Recent findings by Castro (2008) point that there has been an increase in the usage of nonlinear model as central banks tend to have asymmetric preferences in their loss functions implying that weights assigned to negative and positive output gap and inflation could be different. With, the current dominance of financial instability (i.e. financial crises) the central bank tends to behave differently in the manner it adjusts its reaction functions to respond to economic booms and slumps. Furthermore, Castro (2008) shows that the failure by the US and UK to incorporate financial conditions in their monetary policy rule could have exposed them to the current financial crises.
The behavior of interest rate has been characterized by the use of smooth transition regression models, for example logistic smooth transition regression. This theoretical approach has been used extensively by Terasvirta and Anderson (1992). In their study, they argue that the Smooth Transition Regression model could be regarded as a regime switching model, whereby the transition from one regime to another occurs smoothly. For instance, from a low to high inflation regime (see e.g. Terasvirta (2006), and Castro (2008)) and Petersen (2007) argues that STR model is capable of justifying why and when the central bank adjusts its policy rule. The model requires the identification of a transition variable. This variable will indicate a point where a change from a low regime to a high regime takes place. This point of inflection is referred to as the threshold level. In this paper, we have used the grid search method to identify threshold levels as well as the speed of adjustment of the transition variable.
This paper contributes to current monetary debates through identifying how quickly interest rates move from a low to a high interest rate regime, estimating in the context of emerging markets. In addition, it identifies as well as shows the existence of threshold level of the transition variable for decision making. Further, we also evaluate the performance of linear and nonlinear models in providing accurate forecasts. To undertake this evaluation, we use the Diebold- Mariano (DM) and the Sign test to determine the forecasting performance of the linear and nonlinear models. The DM test allows for the evaluation of the performance of two models in terms of their ability to accurately predict. We find that linear models perform better over long horizons compared to nonlinear models. This shows the importance of nonlinear models over the short run period in describing how differently the central bank responds to positive as opposed to negative inflation or output gap to drive them towards the required targets.
The rest of the paper is organized as follows: Section 2 presents the literature. Section 3 describes the methodology applied. Section 4 outlines the data. Section 5 presents the results and discussion. Section 6 conducts forecasting and Section 7 concludes.
Download
PDF Ebook Monetary Policy Conduct Based on Nonlinear Taylor Rule: Evidence from South Africa
Posted in :
