Assessing the Effects of Macroprudential Policy on the Indian Macroeconomy

ABSTRACT

This study examines the impact of macroprudential policy (MaPP) on macroeconomic conditions. Using data from 2004–2005 to 2019–2020, this study finds that increased use of MaPP is associated with higher consumption and economic growth. However, this study does not find a significant effect on inflation. From a threshold perspective, the study reveals that MaPP has a positive impact on consumption and economic growth up to a threshold level of 34.00; however, above this level, MaPP has a negative impact on both variables. Similarly, MaPP has a positive impact on inflation rates below the threshold level but has a negative impact above it. From a tail risk perspective, MaPP has a positive impact on the bottom quantile of economic growth.

KEYWORDS:

• Consumption
• economic growth
• macroprudential policy
• QARDL
• threshold-regression

JEL:

• C22
• E44
• G10
• G28

Acknowledgments

This article is drawn from a chapter of the Ph.D. thesis of the first author Sanjiv Kumar, IIT, Hyderabad, India, and Swinburne University of Technology, Australia. The earlier version of this article was presented at the 16^th Bulletin of Monetary Economics and Banking International Conference, Bali, Indonesia (Webinar).

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1. The economic forecasts usually provide point estimates for the conditional mean of GDP growth and other targeted variables. However, such point forecasts ignore risks around the central forecast and, as such, may point to an overly optimistic picture of the state of the economy (Adrian, Boyarchenko, and Giannone 2019). Besides, in the case of the presence of a fat tail implies that standard approaches to measuring economic volatility do not fully characterize the tail of the GDP growth distribution. Therefore, standard measures are less useful for assessing the frequency and depth of large downturns. Adrian, Boyarchenko, and Giannone (2019) argued that conventional methods that focus on conditional mean underestimate downside GDP tail-risk. In this context, the econometric methods which focus on the risk associated with the tail event are more equipped to gauge the impact in the event of an extreme crisis.

2. Literature has identified these variables as the smallest and most robust set of variables that can assist in predicting the financial crisis (Akinci and Olmstead-Rumsey 2018; Cerutti et al. 2018; Padhan and Prabheesh 2019, 2022).

3. In order to check the threshold level, we construct the cumulative MaPP index which also includes the intensity of MaPP includes. For robustness check of index, we utilize the IMF MaPP provided by Alam et al. (2019).

4. In tail risk perspective literature has explored how MaPP measures influence tail-risk. Adrian et al. (2018), using the growth-at-risk methodology, identify the heterogeneous effects of the macro-financial variable on GDP growth distributions. Thus, the growth-at-risk (tail risk) methodology offers a useful framework to assess the impact of MaPP due to the importance of linkages between the financial sector and the real economy. Second, it also mitigates the negative effects of financial imbalances on economic growth, which is the primary objective of MaPP (European Systemic Risk Board ESRB 2015). However, so far, issues related to the impact of MaPP on the tail-risk perspective remain to be studied both in the short run and long run in the emerging economies context.

5. The starting period of the sample period is restricted by the implementation of MaPP. The end period is restricted due to COVID-19 as during the COVID period policymakers has extensively used the MaPP to limit the risk associated with COVID-19 which may influence the overall results. Therefore, we keep our sample period till 2020Q1. In future, studies can be done which analyze impact of MaPP during the COVID-19 crisis and post COVID-19.

6. As Indian economy has implemented the MaPP long before the financial crises (Nagaraj 2013). As these policies has been implemented to contain the systemic-risk and maintain the financial stability, but it is also important to understand apart from containing the risk related to systemic risk, how these policies affect the macroeconomic variables. So far there is no study has been conducted which analyze the macroeconomic impact of MaPP in case of Indian economy. Thus, through this study we try to attempt how these policies affect the selected macroeconomic variables.

7. For more information see https://www.stata.com/features/overview/threshold-regression/

8. We use the starting period from 2004-05Q1 to 2005-06Q4, as the Indian economy follows a financial year starting from April to March. We restricted the data period to 2019Q1-2020Q4. The ending period of the dataset is restricted due to the COVID-19-induced economic crisis, during which time the economy went through significant structural changes. There is a higher possibility of outliers influencing the overall results. Second, while estimating the model, EVIEWS automatically removes outlier samples from the estimation. Third, during the COVID-19 pandemic, policymakers aggressively used MaPP along with conventional policy, which may unduly affect the results. Therefore, we kept our estimation period to 2020Q1. Further studies could explore how MaPP behaved during COVID-19.

9. Money supply is proxied by M3 in the economy.

10. Reserve requirements has been used in countercyclical manner to contain credit, systemic risk, and ease liquidity (Tovar Mora, Garcia-Escribano, and Vera Martin 2012)

11. The cumulative index is used instead of individual date of announcement, as it is very challenging to exactly identify when MaPP measures were bindings on the borrowers and lenders (Akinci and Olmstead-Rumsey 2018; Verma 2018).

12. For construction of MaPP, which tools are selected and how index adjust the intensity of MaPP measures please follow Kumar, Prabheesh, and Bashar (2022).

13. The quantile ARDL model is more suitable when the data period is large. However, given that we have small datasets, results may vary depending on the availability of longer datasets. Therefore, these results should be interpreted with consideration of that the dataset is small.

14. For checking the unit root properties of the selected variable initially, we applied conventional ADF and PP unit root test, as these test do not incorporate the structural breaks thus there are chances of false rejection of null hypothesis. Thus, in order to incorporate structural break, we apply NP unit root test with structural break, and findings for ADF and PP unit root test are not reported in article but available from the authors.

15. Table 4 presents the long run results only, for short run and other diagnostic check results are reported in Appendix section Tables A1 to A5.

16. For analyzing the effects of MaPP on tail-risk of perspective we restrict our estimation to economic growth only. The tail-risk, concept is borrowed from the financial literature Value-at-Risk (VaR) and introduced in the economic literature in form of growth-at-risk (GaR) by Adrian, Boyarchenko, and Giannone (2019). As MaPP mitigate systemic risk and risk related to lower quantile in the economy, thus we selected economic growth as the dependent variable rather than inflation and output. Literature also support the argument of taking the effects of MaPP on GDP growth such as Sánchez and Röhn (2016), Duprey and Ueberfeldt (2018).

17. Not reported but available upon request.

18. This threshold level is created based on the cumulative macroprudential policy index. While creating the index we have also incorporated the intensity of the macroprudential tools similar to Kumar, Prabheesh, and Bashar (2022). For validation of the index we estimated again with IMF macroprudential policy database.

19. JB is Jarque Bera notmality test, LM is Breusch-Godfrey Lagrange multiplier test for autocorrelation and BPG Breusch-Pagan-Godfrey heteroscedasticity test.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.