Changing transmission of monetary policy on disaggregate inflation in India

Highlights

• We investigate the effects of monetary policy on aggregate, sectoral, and disaggregate inflation in India.

• The response of aggregate inflation to monetary shocks has improved in India.

• Credit and asset price channels appear to be active in monetary transmission.

• Prices in manufacturing sectors respond more aggressively than in the agriculture sector.

• A cost channel is present at the disaggregate level, preventing the intended monetary transmission.

Abstract

This paper investigates the time-varying effects of monetary policy on aggregate, sectoral, and disaggregate inflation in India from 1997 to 2017 using a large dataset of 439 variables. We find that the effectiveness of a contractionary monetary policy in controlling aggregate inflation has improved over time. This improvement in the policy's effectiveness can be attributed to better transmission through credit and asset price channels. In investigating disaggregate inflation, we find that a contractionary monetary policy is more effective in reducing inflation in the manufacturing sector than in the agricultural sector. Further, the sacrifice ratios in all manufacturing sectors have improved over time. However, the commodities prices of some sectors respond positively after a monetary contraction, which demonstrates the presence of a cost channel in the Indian economy. Our findings suggest that the monetary authority in India should have an interest rate rule that incorporates sectoral inflation and reacts to each with different intensity.

Introduction

The effectiveness of monetary policy in achieving a stable rate of overall inflation largely depends on how it affects firm’s pricing decisions in different sectors. The effects of monetary policy on disaggregate inflation are expected to differ from those on aggregate inflation due to the different channels of monetary transmission mechanisms, degrees of price stickiness, etc. Altissimo et al. (2009); Bils and Klenow (2004); Baumeister et al. (2013); and Clark (2006) show that there are important differences in the dynamics of inflation at the aggregate and sectoral level. Thus, the measure of aggregate inflation may hide the dynamics of disaggregate inflation. Accordingly, Balke and Wynne (2007) argue that the aggregate price index alone may not be the most appropriate guide for formulating monetary policy. Instead, Aoki (2001) argues that stabilizing the relative prices of different commodities around their optimal value is one of the most effective goals for a central bank. Therefore, it is crucial that monetary authorities understand how the prices of various commodities in different sectors respond to monetary policy action.

However, empirical evidence on the effects of monetary policy on disaggregate inflation is inconclusive. For instance, Balke and Wynne (2007); Baumeister et al. (2013); and Lastrapes (2006) show that monetary policy has a differential effect on relative commodity prices at the disaggregate level. In contrast, Boivin et al. (2009) show that monetary policy plays only a slight role in explaining the variations in individual prices (see also De Graeve and Walentin, 2011). This paper re-examines the effects of monetary policy on aggregate and disaggregate prices in a large emerging economy, India, from 1997 to 2017.

Our study extends the literature on the effects of monetary policy on disaggregate price levels in three directions. First, we analyse the effects in a large emerging market economy (EME) such as India; in contrast, most previous studies have focused on advanced economies (AEs) (Bernanke et al., 2005; Ellis et al., 2014). However, it is well established that the conduct and effects of monetary policy in EMEs differ widely from those in AEs (Frankel, 2010; Lahiri and Patel, 2016). In general, EMEs suffer from less developed political and institutional systems (Fraga et al., 2003). Further, their financial markets are less developed, with many weaknesses that present a hostile environment for monetary authority. The pro cyclic nature of fiscal policy, with a large fiscal and budget deficit, also hinders monetary policy transmission (Talvi and Végh, 2005). In most cases, this leads to an uncompetitive banking system that can be used as a tool to drive financial repression (Denizer et al., 1998). Emerging economies suffer from frequent adverse supply shocks that can be attributed to the fact that primary products, such as agriculture, form a larger share of these economies. Capital flow in emerging economies also demonstrates a pro cyclic, boom-and-bust nature (Kaminsky et al., 2004). Large capital flow and the overvaluation of currency with sudden stops in inflow, which lead to recession, are common features of emerging economies (Mendoza, 2002). These salient features of EMEs generate a unique monetary transmission mechanism that requires a separate study to understand how various commodities of different sectors respond to monetary action.

Second, we analyse the effects of monetary policy on disaggregate price levels in India in a time-varying framework using a large dataset. Previous studies have estimated the effect in a constant parameter model (e.g., Misra and Suresh, 2014). However, India’s monetary policy framework has evolved continuously due to developments in the economy on various fronts. First, in 1998 the Reserve Bank of India (RBI) adopted a multiple indicator approach, rather than a monetary aggregate approach, in the conduct of monetary policy. The post-1991 reform period has seen increasing openness in the economy, deregulation of interest rates, and developments and innovations in financial markets. It has also seen volatility in the financial markets, which has rendered the money demand function unstable. Therefore, using a multiple indicator approach, multiple variables—such as interest rate, exchange rate, inflation rate, rates of return in different markets, bank credit, output, trade, capital flows, monetary aggregates, and fiscal indicators—have been used to make policy decisions. Notably, the RBI adopted a flexible inflation targeting (FIT) framework in 2016 because the stagflation that arose after the 2008 global financial crisis posed serious challenges to the monetary policy and brought into question the credibility of a multiple indicator approach. Consequently, on the recommendations of the Urjit Patel Committee report, a new monetary policy framework agreement—the FIT framework agreement—was signed by the Government of India in February 2015. Under this new framework, the RBI targets inflation measured in terms of consumer price index at 4% with a (±) 2% tolerance band. Therefore, the effectiveness of monetary policy in containing aggregate and disaggregate inflation and the relative importance of its transmission mechanism may have changed over time. Furthermore, previous studies have used a small-scale VAR model to analyse the effects of a monetary policy. However, Bernanke et al. (2005) argue that using a small-scale VAR model may generate omitted variable bias, as central banks may base a decision on information that may not be reflected in a small-scale VAR model. To address this issue, we estimate the effect of monetary policy using a large dataset of 439 variables. To the best of our knowledge, this is the first paper to study the effect of monetary policy at the individual commodity level in India using a large dataset with a time-varying framework.

Our third contribution is to estimate the sacrifice ratio in manufacturing and its different sub-sectors in response to monetary policy. Most previous studies have calculated the sacrifice ratio for the aggregate output in India (Mitra et al., 2015). However, different manufacturing sectors may react differently to monetary policy shocks: Some sectors might be more responsive to monetary policy shocks than others. Therefore, estimation of the aggregate sacrifice ratio may obscure the differential effects. To overcome this limitation, we calculate the sectoral-level sacrifice ratio for India’s manufacturing industry within a time-varying framework. We define the sacrifice ratio as the decrease in sectoral output in the manufacturing sector due to a one-percentage-point decrease in average inflation. This is calculated as the ratio of the cumulative response of sectoral output in the manufacturing sector to the cumulative response of average inflation up to 40 months due to a 100-basis-point increase in the interest rate. Sectoral calculation of the sacrifice ratio in each manufacturing sector will allow the monetary authority to identify sectors with low and high sacrifice ratios. It will also enable them to identify which manufacturing sectors of the economy do not respond to their actions.

Specifically, our study seeks to answer the following questions.

• What is the effect of monetary policy on inflation at the aggregate, sectoral, and individual commodity level in India?

• How has the effectiveness of monetary policy evolved at the aggregate, sectoral, and individual level over time?

• Has the dispersion of the responses of prices of different commodities changed over time?

• What is the sacrifice ratio, at both the aggregate and sectoral level, in the manufacturing industry in India, and how has it evolved over time?

• What is the relative importance of different channels for monetary transmission in India?

Our key findings are as follows. (i) A contractionary monetary policy reduces overall and sectoral inflation in India, and its effectiveness has improved over time. (ii) An increase in the interest rate affects disaggregate inflation more aggressively in manufacturing sectors than in sectors related to agriculture. (iii) Some individual commodities’ inflation rates respond positively; i.e., they show signs of a “price puzzle.” (iv) However, the price puzzle has weakened in the recent past. (v) Dispersion of the responses of commodities’ inflation rates has increased in recent years. (vi) The cost of controlling inflation—i.e., the sacrifice ratio—in manufacturing industries has declined in all sectors. (vii) Credit and asset price channels explain the time-varying response of aggregate inflation.

The rest of the paper is as follows. Section 2 reviews related literature on the relationship between the interest rate and inflation in India. Section 3 explains our time-varying parameter factor augmented vector autoregression (TVP-FAVAR) model. Our large dataset is described in Section 4. Section 5 analyses our results, and Section 6 performs a robustness check. Section 7 concludes.