Partial Integration of Indian Money, Forex, and Equity Markets Post-1991 Reforms: Cointegration Analysis and Vector Error Correction Modelling Using Monthly Time Series from 2015 to 2026

Overview

This study examines the degree of long-run integration across India's money, foreign exchange, and equity markets following the 1991 financial liberalization. Despite progressive opening to cross-segment arbitrage, full integration has not materialized due to structural policy constraints including the Reserve Bank of India's managed float regime, sterilized intervention practices, and capital-account restrictions. The research focuses on three representative monthly time series—the Weighted Average Call Money Rate, the logarithm of the USD/INR reference rate, and the logarithm of the Nifty 50 equity index—over a 134-observation period spanning January 2015 to February 2026. This sample period captures multiple structurally significant episodes: demonetization in 2016, Goods and Services Tax implementation in 2017, the Infrastructure Leasing & Financial Services crisis of 2018–2019, the COVID-19 shock in 2020, subsequent post-pandemic monetary tightening, and sustained rupee depreciation toward the 90–92 range. The investigation addresses whether these three market segments exhibit cointegrated behavior indicative of long-run equilibrium relationships and the extent to which policy constraints limit transmission mechanisms across segments.

Methods and approach

The empirical framework employs Johansen's maximum-likelihood cointegration test to identify the number of long-run equilibrium relationships among the three market variables. A Vector Error Correction Model is specified to capture both the long-run cointegrating relationships and short-run dynamic adjustments. The analysis includes estimation of adjustment coefficients to assess the speed and statistical significance of error correction in each market segment. Short-run dynamics are investigated through Granger causality tests to determine directional relationships among the variables, with particular attention to crisis episodes. Impulse response functions trace the dynamic responses of each variable to shocks in the others, while variance decomposition analysis quantifies the proportion of forecast-error variance in each series attributable to innovations in the remaining variables at different time horizons, extending to twelve months.

Key Findings

The cointegration analysis identifies a single long-run equilibrium relationship among the three markets, corresponding to a cointegration rank of one. This finding supports the characterization of partial rather than full integration across money, forex, and equity segments. Adjustment coefficients in the Vector Error Correction Model are statistically significant for both the foreign exchange and equity equations, indicating that these markets respond to deviations from long-run equilibrium. In contrast, the adjustment coefficient for the money-market rate is economically small and statistically insignificant, consistent with policy-constrained transmission mechanisms in the call money market. Short-run Granger causality tests reveal bidirectional relationships between USD/INR and Nifty 50, with intensification during crisis periods. Variance decomposition analysis demonstrates that rupee fluctuations explain up to 21 percent of equity market forecast-error variance at the twelve-month horizon, establishing a quantitatively meaningful linkage between foreign exchange movements and equity valuations over medium-term periods.

Implications

The documented partial integration has direct consequences for the effectiveness of monetary policy transmission in India. The economically small and statistically insignificant error-correction response in the money-market rate indicates that policy signals from the call money market do not fully propagate to foreign exchange and equity segments, limiting the scope of monetary policy influence across financial markets. The bidirectional short-run causality between forex and equity markets, particularly during crisis episodes, suggests that systemic stress can propagate across these segments independent of money-market conditions. The substantial contribution of rupee fluctuations to equity forecast-error variance—reaching 21 percent at twelve months—indicates that exchange rate volatility constitutes a material source of equity market uncertainty in India. These findings inform systemic-risk monitoring frameworks in emerging markets, where capital-account restrictions and managed exchange-rate regimes may create segmented transmission channels. The results also bear on portfolio allocation and hedging strategies in partially integrated financial systems, where cross-market linkages operate asymmetrically and are state-dependent.