The paper focuses on the cyclical dynamics of business, credit, and investment in India, particularly in the context of the 2008 financial crisis which highlighted the importance of financial cycles in macroeconomic stability. The authors aim to understand the interactions between these cycles and to determine the role of domestic versus global financial factors. Existing research on business cycles is extensive; however, the analysis of financial cycles and their interplay with real economic activity is relatively less explored. This research is crucial for developing effective economic policy, particularly in managing economic fluctuations. The paper notes the Minsky's financial instability hypothesis and its three stages of instability: hedge, speculative, and Ponzi. These phases represent the evolution of debt levels and asset prices leading to booms and busts. The study also considers Borio's (2012) perspective of self-reinforcing interactions between risk perceptions, attitudes, and financing constraints. The authors note India's significant economic changes since 1991, including substantial banking sector digitization. Therefore, they incorporate a Digital Transformation Index (DTI), proxied by mobile subscriptions, to analyze its influence on the credit creation process. The primary research question is how the business, credit, and investment cycles interact in India, and what role domestic and global factors play in shaping these dynamics. This understanding is crucial for developing appropriate policy responses to economic fluctuations.

The literature review covers various aspects of the finance-growth nexus, examining both the role of financial development in economic growth (Schumpeter, 1911; McKinnon, 1973; Goldsmith, 1969; Gregorio, 1999; Levine, 1997; Arteta et al., 2001; Edison et al., 2002) and the cyclical fluctuations in macroeconomic variables. While extensive research exists on business cycles, the study of financial cycles is more recent and the research focuses on developed countries (Borio, 2012; Claessens et al., 2011a, 2011b; Adrian and Shin, 2010). The review highlights the strong relationship between business and financial cycles, as well as the synchronization of output cycles with credit and house prices (Claessens et al., 2011a, 2011b). Studies comparing the durations of business and financial cycles suggest that financial cycles tend to be longer (Rünstler, 2016). The impact of financial cycle shocks on macroeconomic variables and the usefulness of business cycle information in forecasting financial cycles are also discussed (Yong and Zhang, 2016; Jawadi et al., 2022). Methodological approaches to studying financial cycles, such as turning point analysis (Arthur and Mitchell, 1946; Bry and Boschan, 1971; Harding and Pagan, 2002), unobserved component models (Aikaman et al., 2010), and the Hodrick-Prescott and band-pass filters, are reviewed. The review also explores prior research on financial cycles in India (Behra and Sharma, 2019; Aravalath, 2020; Kumar et al., 2020; Paramanik et al., 2021; Saini et al., 2021), the role of global financial cycles (Cerutti et al., 2017; Silvia and Hélène, 2021), and the theories of Hawtrey, Hayek, Samuelson, and Hicks on business cycle causes.

The study uses annual data from 1980 to 2021 (2000-2021 for mobile subscribers) for various macroeconomic variables: Non-food Credit (NFGBC), Domestic credit to the private sector, Gross fixed capital formation (GFCF), Broad Money (M3), GDP growth rate, BSE Sensex volatility, Brent crude oil prices, CPI, and mobile subscriptions per 100 people. The Hodrick-Prescott (HP) filter with λ = 100 is employed to extract the cyclical components from each time series. Correlation and regression analyses are conducted to investigate the relationships between the cyclical components of GDP and other variables. A Granger causality test examines the causal relationships among the cycles. Finally, a structural vector autoregression (SVAR) model, with Blanchard and Quah's (1989) long-run restrictions identification scheme and two lags (determined by AIC, FPE, HQ, LR, and SC criteria), is used to analyze the dynamic interactions between the business (GDP), credit (NFGBC), and investment (GFCF) cycles. The SVAR analysis includes impulse response functions and variance decompositions to understand the long-run effects of shocks to each cycle on the others. Robustness checks are performed on the SVAR results to assess autocorrelation, normality, and heteroskedasticity of the residuals.

The HP filter reveals cyclical patterns in all variables. GDP shows 8 peaks and 9 troughs, with the longest cycles spanning 9 years. NFGBC has 6 peaks and 7 troughs, with a 12-year cycle. Domestic private credit shows 11 peaks and 8 troughs, with long cycles of 9 and 11 years. GFCF exhibits 10 peaks and 9 troughs, with an 8-year cycle. M3 has 6 peaks and 8 troughs, with a 13-year cycle. BSE shows 9 peaks and 10 troughs, with cycles around 3.5 years. Brent crude oil has 9 peaks and 8 troughs, with cycles around 5.7 years. Mobile subscriptions have one peak and one trough since 2000. Correlation analysis shows a moderate positive relationship between GDP and NFGBC (0.50), and a weak negative relationship between GDP and private credit (-0.20). A very weak positive relationship is observed between GDP and GFCF (0.10) and between GDP and BSE (0.05). A moderate positive correlation exists between GDP and M3 (0.34) and between GDP and mobile subscriptions (0.33). A positive correlation is also found between NFGBC and mobile subscriptions (0.50). Granger causality tests reveal a bidirectional relationship between GDP and NFGBC cycles, while crude oil prices and mobile subscriptions show a unidirectional causality with GDP and NFGBC, respectively. The SVAR analysis indicates a long-run positive relationship between GDP and both GFCF and NFGBC cycles. Variance decomposition from the SVAR shows that NFGBC explains a large portion of GDP cycle variance, while GFCF mainly explains its own variance. Robustness checks show the SVAR model is stable, except for a normality violation in GDP due to the COVID-19 shock.

The strong positive relationship between business and credit cycles supports the financial accelerator theory, implying that credit availability significantly impacts economic activity. The weak relationship between investment and other variables suggests that investment behavior is complex and influenced by factors beyond the scope of this analysis. The significant correlation between mobile subscriptions and credit cycles highlights the role of technological advancements in shaping credit markets. The findings imply that credit market regulations and policies should be carefully designed to mitigate business cycle fluctuations, and that attention to technological trends is important for financial stability policy. The lack of a significant relationship between GDP and global financial cycles (proxied by oil prices) suggests that India's domestic factors are more dominant in driving its business cycle.

The study demonstrates the interconnectedness of business, credit, and investment cycles in India. Credit cycles have a significant role in influencing business cycle dynamics. The impact of global financial cycles is less pronounced than that of domestic factors. Mobile subscriber cycles have a positive effect on credit cycles. Further research should consider the use of more sophisticated measures of DTI and explore the implications of various policy responses on cyclical interactions.

The study uses annual data, which limits the precision in identifying cyclical patterns. The use of mobile subscriptions as a proxy for DTI is a simplification and might not fully capture the complexity of digital transformation's effects. The analysis focuses on aggregate data, and regional variations in cyclical dynamics might be overlooked. The COVID-19 shock significantly affected various time series, which may have influenced the results.