Incomplete reform or opportunity: the role of the banking sector in the credit transmission mechanism in India

Abstract

The recent financial crisis in developed economies is attributed to the credit crunch and features of a free market economy. One main concern is the spreading of this crisis to emerging economies. This paper tests the importance of the banking sector as a credit transmission channel in India. The empirical analysis discovers a structurally stable long‐run relationship (immune to exogenous shocks) between bank credit and interest rate spread. This suggests that the reform process has not yet reached an extent where capital markets are fully competitive and banks' role in credit formation remains significant, suggesting India's reduced exposure to the current financial crisis.

Keywords:

• economic reforms
• credit channel
• financial crisis
• regime change
• cointegration

JEL classification:

• E51
• E60
• C32

Acknowledgements

I am thankful to the editor for encouragement and Raghbendra Jha for useful discussions. The research is part of my PhD dissertation funded by an ANU PhD Scholarship. The usual caveat applies.

Notes

1. Please note that the terms ‘bank lending channel’ and ‘credit channel’ are used interchangeably throughout the paper.

2. These include central government securities such as treasury bills with different periods of maturity.

3. As reported in RBI Weekly Supplement, 11 October 2003.

4. This is the only activity variable for which data are available on a monthly basis.

5. The index prior to 1997 could not be calculated because of inconsistency in the data.

6. This is because most of the external trade in India is dollar denominated. Moreover, the US has emerged as India's main trading partner in the last decade: the share of Indian exports to America increased by 9% to 25% during the period 1991–2000 as reported by the Federal Ministry of Commerce, India.

7. However the final form of the cointegrating vector is parsimonious, as the null hypothesis that the coefficient of exchange rate is zero cannot be rejected. The results remain qualitatively the same if the real effective exchange rate based on trade weights of 36 countries is used for the analysis.

8. The KPSS test, which does not have a bias in favor of accepting non‐stationarity, also indicates that each of these variables is I(1).

9. It should be noted the tests for endogenous structural breaks at unit root stage were applied and the null hypothesis of series being I(1) could not be rejected.

10. All the three tests assume unknown timing of the structural break.

11. The estimation is using the Gauss code provided by Bruce E. Hansen.

12. It should be noted that Hansen's analysis requires trimming of the data on edges. We have trimmed 15% of the data from both sides of the sample and hence the analysis is done using the data from November 1993 (month 20) to November 2001 (month 116).