Indian Growth is Not Overestimated: Mr. Subramanian You Got it Wrong

Ashima Goyal and Abhishek Kumar

Subramanian (2019) argues that indicators like growth in export, import and private credit can be used to predict economic growth across countries. He finds these indicators were able to predict Indian growth before 2011 but fail to do so after the GDP estimation methodology was changed in 2011. This implies Indian growth was overestimated by 2.5% per annum post 2011. But on removing various flaws in his data and procedures used, we find these indicators underestimate growth before 2011 too. The most reasonable specification suggests that the underestimation before 2011 was higher than underestimation post 2011. Moreover, growth in a large number of countries is found to be either overestimated or underestimated, based on these indicators. His empirical design is therefore flawed. These regressions cannot be used for predicting growth or for concluding Indian growth is overestimated or for pointing to problems in the estimation methodology. Working Paper Mentioned in Popular Press Mentioned in Popular Press

Aggregate Demand Management, Policy Errors and Optimal Monetary Policy in India

Barendra Kumar Bhoi, Abhishek Kumar and Prashant Mehul Parab

This paper evaluates the rule-based interest rate policy for India since 2000 Q1, which has become more relevant in the flexible inflation targeting (FIT) regime. Based on results of the reduced form Taylor-rule, we observed two episodes of possible policy errors since 2001. First, in the aftermath of the global financial crisis, RBI brought down repo rate much below the level warranted by the Taylor rule that fueled inflation. Moreover, monetary policy tightening, followed thereafter, during March 2009 to June 2011 was insufficient in controlling prices. Second, despite favorable supply shocks, repo rate was above the rule-based policy rate during June 2013 to March 2016, leading to very high real interest rate. In the post-crisis period, we observed significant increase in the interest rate persistence, which could be attributed to RBI’s reluctance to cut policy rate despite softening of inflation, leading to growth slowdown. Our results suggest that the optimal policy rate ranges from 4% to 5% for the last quarter of 2018. Actual repo rate at 6.5% in December 2018 was 150 to 250 basis points above the optimal rate. RBI has reduced repo rate by a cumulative 110 basis points since December 2018. As the negative output gap has widened and inflation remains subdued, there is scope to cut the repo rate further. Working Paper

Money and Business Cycle: Evidence from India

Ashima Goyal and Abhishek Kumar

In this paper we take a New Keynesian model with non-separable money in utility to Indian data using maximum likelihood. The identification problem in isolating the effect of money on output and inflation is solved by adjusting real balances for shifts in money demand. Estimates with an extended model with relevant features like partial indexation in prices, markup shock and time varying inflation target, show that real balances do affect output and inflation even after correcting for money demand unlike results for the United States and Euro-zone. A regression estimate and multivariate structural vector auto-regression give similar results. Types of money matter. Reserve money has the largest impact, pointing to the importance of the informal sector. The estimated income elasticity of narrow money is more than twice that of broad money, pointing to the dependence of firms on banks. Interest semi elasticity of money demand is close to one. Responsiveness of output to real interest rate is high. We find that interest rate setting is quite persistent. Coefficient of lagged interest rate varies from 0.71 to 0.95. We conclude that there is a significant asymmetry in the role of money in India (an emerging economy) in comparison to United States and Euro-zone (advanced economies). Working Paper Published Version

Over-reaction in Indian monetary policy

Ashima Goyal and Abhishek Kumar

This paper explores challenges in implementing newly instituted inflation targeting in India. The evidence presented implies the aggregate demand channel is not weak in India. Supply side factors, however, remain important for inflation in contrast with advanced economies where in general inflation results from a demand shock. The rise and fall of inflation in the aftermath of the financial crisis closely followed supply shocks and thus can be attributed to bad and good luck respectively. Interest rate were increased to counter inflation originating from supply side that had significant output cost. An aggregate demand channel, which is stronger than expected, and inflation originating from supply shocks poses a significant challenge for the newly instituted inflation targeting regime. Published Version

A DSGE Model-Based Evaluation of the Indian Slowdown

Ashima Goyal and Abhishek Kumar

In this paper we take a New Keynesian Dynamic Stochastic General Equilibrium (DSGE) model to the Indian data using the Kalman filter based maximum likelihood estimation. Our model based output gap tracks the statistical Hodrick-Prescott filter based output gap well. Comparison of parameters of the model, impulse responses and forecast error variance decomposition between India and United States points to interesting difference in the structure of the two economies and of their inflationary process. Our estimates suggest higher value of habit persistence, more volatile markup and interest rate shocks in India. Markup shock plays much larger role in determination of inflation in India and this suggests the important role played by supply side factors. Impulse responses suggest higher impact of interest rate shock on output and inflation in comparison to US. Technology shock has less effect on output in comparison to US and this again suggest the presence of supply side bottlenecks. We use smoothed states obtained from the Kalman filter to create counterfactual paths of output and Inflation (during 2009 Q4 to 2013 Q2) in presence of a given shock. In the post 2011 slowdown, monetary shock imposed significant output cost and for a brief period of time made a negative contribution to the output gap. Working Paper

News, Noise and Indian Business Cycle

Ashima Goyal and Abhishek Kumar

We take a New Keynesian Dynamic Stochastic General Equilibrium (DSGE) model with various specification of technology, markup and interest rate shocks to the Indian data using Kalman filter based pure and Bayesian likelihood estimation. There are evidence that preference and interest rate shocks are important for output determination and markup and interest rate shock plays important role in determining inflation. We have evidence that anticipated interest rate shocks as in Grohé and Uribe (2012) are important for business cycles in India. Out of total variance explained by interest rate shock, one third is due to the anticipated shock. There is also evidence that significant portion of interest rate movement is anticipated. Anticipated technology and markup shock doesn’t play significant role. We also find that in terms of model fit the best anticipation horizon is one period for interest rate shocks. Anticipated correlated interest rate and markup shocks as in Walker and Leeper (2011) significantly improves the model fit and makes the monetary transmission stronger. Estimates suggest that markup shock has very low persistence. Correlated markup shock leads to anticipation of interest rate movement. There is evidence that permanent component of technology is not perfectly observed and it plays significant role in the business cycle. Working Paper

Active Monetary Policy and the Slowdown: Evidence from DSGE Based Indian Aggregate Demand and Supply

Ashima Goyal and Abhishek Kumar

We explore reasons for the strong Indian monetary transmission and response to other shocks estimated in a standard New Keynesian Dynamic Stochastic General Equilibrium (DSGE) model. While other responses are moderated, counterfactual analysis suggests that large cost shocks remain a primitive cause of inflation, and the strong transmission comes from large interest rate changes. Reducing the variance of the interest rate shock can significantly moderate the large output cost. Sensitivity analysis shows estimated excess EM volatility due to preference and technology shocks is reduced on introducing regime switching between multiple steady-states. The estimated model including multiple regimes is therefore used to obtain aggregate demand and aggregate supply schedules incorporating the policy response and identify their shifts during the Indian slowdown. A negative correlation estimated between factors shifting aggregate demand and supply aggravates shocks. The post 2011 slowdown is explained by severe demand contraction in response to adverse supply shocks. Habit persistence in consumption changes the slope of both aggregate demand and supply curves significantly. Working Paper Published Version

Usage of Formal Financial Services in India: Demand Barriers or Supply Constraints?

Abhishek Kumar, Rama Pal and Rupayan Pal

Estimating the role of both demand-side and supply-side factors in financial inclusion and its distribution is important for policy making. However, existing literature has primarily focused on supply-side factors. In this context, this paper estimates relative importance of removing demand-side barriers and eliminating supply constraints to enhance financial inclusion in India. It also measures the extent of concentration of usage of formal financial services among richer households. Results suggest that, while availability of banking services has a significant positive effect on usage of formal financial services, its contribution in inducing households to use formal financial services is very small compared to the contribution of factors, such as education, income, employment status, gender and social norms, that influence the demand for formal financial services. It highlights the importance of placing greater emphasis on addressing demand-side barriers, rather than on improving physical availability of banking services, to promote financial inclusion in India. Published Version

The effect of oil shocks and cyclicality in hiding Indian twin deficits

Ashima Goyal and Abhishek Kumar

The paper estimates the relationship between the current account and fiscal deficit, and the real exchange rate for India after controlling for output growth and oil shocks in an SVAR, for the managed float period 1996Q2 to 2015 Q4. It also examines the cyclicality of the current account, the size of each shock, and assesses whether aggregate demand, forward-looking behaviour, or supply shocks dominate outcomes. An FD shock raises the CAD, but high impact growth shocks and large variance oil shocks lead to overall divergence of the deficits. The CAD is found to be countercyclical. There is some support for the aggregate demand channel, but it is moderated by supply shocks and sectoral effects. Sticky consumption rather than forward-looking smoothing is supported. Working Paper Published Version

Determinants of Firm-Level Investment in India: Does Size Matter?

Abhishek Kumar and Parul Bhardwaj

The study estimates the dynamic panel version of augmented neoclassical investment model using ARDL specification. There are evidences in support of interest rate and credit channels of monetary transmission, both in the short as well as in the long run. Our evidence of interest rate channel is robust and is not driven by outliers on the basis of size, investment to capital and cash flow to capital ratio. We also correct for the presence of financially distressed and constrained firms. The heterogeneous impact of cash flow to capital stock ratio on investment spending of small and large firms provides further evidence in favour of working of credit channel. Working Paper

Growth of Finance, Real Estate and Business Services: Explorations in an Inter-Sectoral Framework

Sukanya Bose and Abhishek Kumar

The Indian growth experience over the past several decades has been service led. More recently, within services, Finance, Insurance, Real Estate and Business Services (FINREBS) has been the fastest growing sector, with its share in GDP rapidly rising to around 22 percent in a relatively short time-frame. What relation does the growth of FINREBS have with the rest of the sectors of the economy? Empirical exploration using input-output tables and econometric methods shows that FINREBS ranks low in backward and forward linkages compared to most other sectors of the economy. It is difficult to imagine FINREBS as a leading sector in the Hirschman sense. Rolling co-integration to study the evolution of long-term relationships shows an increasing co-movement in output of FINREBS and agriculture and allied activities. However, for most other sectors the association with FINREBS is insignificant or weak. Variance decomposition of forecast error corroborates that a large percentage of variation in the growth of FINREBS cannot be explained by other sectors of the economy, which gives FINREBS an autonomous character. The probable reasons for the ‘autonomous’ nature of growth in FINREBS are explored briefly in the paper. Working Paper Published Version

Macroeconomic Policy Simulations for the 14th Finance Commission

N R Bhanumurthy, Sukanya Bose, Parma Devi Adhikari, and Abhishek Kumar

This study attempts to construct a consistent macroeconomic framework for India to review the macro-fiscal linkages over the 14th Finance Commission period of 2015-19. The existing NIPFP model has been reworked to add a full-fledged real sector block comprising of agriculture, industry, services and infrastructure, with the overall economy comprising of real sector block, external block, monetary block, fiscal block and macroeconomic block. The estimated model was used for policy simulations for the 14th Finance Commission period. The various scenarios include (a) shock due to 7th Pay Commission award, (b) targeting deficit and debt (c) targeting higher growth and (d) the external shocks. The results suggest that while Pay Commission award would result in slightly higher growth compared to the base case, this also results in higher inflation, fiscal-revenue deficits, current account deficit as well as higher government liability. Further simulation results suggest that expenditure switching policy, which is the core of expansionary fiscal consolidation mechanism, by increasing higher government capital expenditure and reducing the government transfers could result in higher growth with a manageable fiscal deficit of 6 per cent that also brings down the government liability to its present level of 65 per cent. Working Paper

Trend and Cycles in Indian GDP

Ashima Goyal and Abhishek Kumar

Unobserved component models with several specifications are used to estimate Indian output gap and time varying growth rate. We consider models with deterministic trend, deterministic trend with one and two breaks, stochastic trend with and without breaks where we take both correlated and uncorrelated trend and cycle shock. We also estimate restricted cycle in trend and restricted trend in cycle versions of the stochastic trend model. Stochastic trend growth model with correlated and uncorrelated growth and cycle shock are also estimated. Our stochastic growth model nests HP filter and allows us to estimate augmented versions of HP filter. We obtain integrated likelihood of all these models in closed form and compare them using marginal likelihood and deviance information criteria. The best model is stochastic trend model with two structural breaks based on marginal likelihood criteria. Deterministic trend model with two structural breaks comes very close on the basis of likelihood and outperforms on the basis of deviance information criteria. Both marginal likelihood and deviance information criteria unequivocally reject HP filter. There is evidence of I) λ =1600 (commonly used smoothing parameter of HP filter) being at odds with Indian quarterly GDP data, II) negative correlation between trend and cycle shock, III) pure trend shock explaining at-least 40% of the variation in shock in our stochastic trend model with correlated shocks and two breaks IV) variation in cyclic process explaining 72% of the variation in growth rate in our stochastic trend model with correlated shocks and two breaks, V) cyclic (nominal) shock is also important for long run level of output, VI) two structural breaks in trend growth, one in 2002Q4 and another one in 2011Q1, VII) data favors a constant trend growth model with breaks rather than stochastic trend growth model, VIII) slowdown in the trend growth rate after 2011 based on both criteria where annual trend growth rate declined by 2%.