1. INTRODUCTION:

Late in 2019, the Chinese city of Wuhan published the novel COVID-19, which was incredibly well-read all around the world. The World Health Organization (WHO) was informed by the Chinese government about the novel corona virus on December 31, 2019, and since then, it has been disseminated worldwide. Furthermore, on January 30, 2020, the WHO declared the novel corona virus a global public health emergency and escalated it to a pandemic on March 11, 2020. Terrorist attacks, earthquakes, and other natural disasters of this nature are examples of past catastrophic occurrences that have affected stock markets and other significant economic operations. Along with the virus's growth, people began to fear it and label it as something of a societal stigma. In-depth research has been done in the literature on both the pandemic's beneficial and negative effects, including how they may affect the banking sector, climate change, and environmental issues.

It inflicted damage on virtually every economy, with stock markets and crude oil prices taking a significant hit. Crude oil, following gold, is recognized as a key hedging asset and has garnered the interest of financial investors and academics. Events of all kinds, whether they are social, economic, political, or cultural, have an impact on stock markets. Similar to COVID-19, this was a global pandemic and caused financial prices to fall everywhere. To prevent economic turbulence, governments all over the world implemented supportive policies for consumers, businesses, and the general public.

Due to the lockdown, global oil consumption has decreased to 3.9 million barrels per day. Because of the increased worldwide financial connectivity, the COVID-19 outbreak has shocked the financial systems and financial markets (Kazmi, 2021). In the period of uncertainty interdependence among macroeconomic variable increases and the opportunities for the diversification reduce and force the investors to search for the other investment opportunities to handle the risks (Xiaozhong et al. 2022). To diversify the investment usually investors switch between oil and gold. Oil is a very volatile product where as gold on the other side often seen as safe heaven asset (Reginer 2007; Baur and Luccey 2010). According to Bakas and Triantafyllou (2020), COVID 19 outbreak had a substantial impact on the financial sectors contributing to uncertainty. The crude oil value decreased to 37$ per barrel in April 2020 (Xiaozhong et al. 2022). Despite the fact that the gold is important for currency trading and hedging gold price volatility could have severe effect on financial markets.

In this context, this study aims to study an asymmetric relationship between oil price, gold price and stock prices before and after the covid 19 pandemic. There are main two contributions of this study to the existing literature; first, there are few studies which focus on the dynamic linkages of gold price, oil price and stock market prices. Moreover asymmetric relationship between the variables did not get much attention particularly with respect to the long run and short run relationship of the variables. Secondly, this study uses non linear Auto Regressive Distributed Lag (NARDL) Model which is better than linear ARDL as in non ARDL method; all independent variables are segregated into categories with positive and negative signs to evaluate the differential effects of macroeconomic variables on stock prices.

This paper is structured into five distinct sections. Section 1 provides an introduction to the topic, setting the stage for the subsequent analysis. Section 2 presents a comprehensive review of the existing literature, highlighting key findings and gaps in current research. Section 3 outlines the methodology employed in the study, detailing the research design and analytical approaches used. Section 4 encompasses the results and discussion, offering a thorough examination of the findings and their implications. The final section, Section 5, concludes the paper with a summary of key insights and offers policy implications based on the study’s outcomes.

2. REVIEW OF LITERATURE:

Awan, T. M. et al. (2021) investigated the volatility of crude oil and its relationship with the stock market during the COVID-19 pandemic, revealing that lockdowns imposed during the pandemic led to a significant decline in oil prices due to reduced demand. This study highlights the profound impact of COVID-19 on the oil market, emphasizing the need for further analysis of such exogenous shocks on commodity prices and their broader economic implications.

Mahajan, S. et al. (2021) examined the impact of the COVID-19 pandemic on both gold and the stock market from January 2020 to May 2020, using ARMA and GARCH models. Their primary focus was to assess the influence of the COVID-19 lockdown on the returns of gold and the Nifty index. The results indicated a significant adverse effect of gold on Nifty returns, suggesting that investors view gold as a safe haven during periods of heightened uncertainty. This study contributes to a deeper understanding of market asymmetry, investor behavior, and information processing in financial markets, particularly during crises.

Shaikh, I. (2021) analyzed the performance of the oil market during the COVID-19 pandemic, specifically examining its spillover connections with other asset classes. One prominent discovery was the heightened responsiveness of the crude oil market to misinformation related to the pandemic. The global pandemic panic index and pandemic sentiment index showed significant promise in explaining market behavior. An analysis of volatility spillover revealed a strong connection between the crude oil market and other markets, with the total connectedness index indicating an average contribution of 35% from spillover effects. During the initial stages of the pandemic, various macroeconomic and political events provided some support to the market, but the second phase saw the global crude oil market suffer negative consequences. The study concludes that infectious diseases amplify investor panic and anxiety, affecting market dynamics.

Albulescu, C. (2020) conducted a study to explore the impact of COVID-19, financial volatility, and U.S. economic policy uncertainty on crude oil prices using an ARDL model. The findings revealed that COVID-19 had a long-term negative impact on crude oil prices, underscoring the vulnerability of commodity markets to global health crises and policy uncertainties. This study adds to the growing body of literature on the interplay between macroeconomic factors and commodity prices during unprecedented events.

Syahri, A. et al. (2020) analyzed the relationship between gold, exchange rates, and the Composite Stock Price Index (CSPI) during the COVID-19 pandemic, focusing on data from January 2020 to June 2020. Their results showed that changes in gold prices had a substantial effect on the volatility of stock prices. The research uncovered a positive connection between CSPI and gold prices, while a negative correlation was observed between CSPI and exchange rates. This study highlights the intricate relationships between different asset classes during periods of economic uncertainty and the role of gold as a hedge against market volatility.

Despite numerous studies examining the relationships between oil prices, gold prices, and the stock market, there remains a notable gap in research specifically investigating the nexus between these variables. Most previous studies have focused on mean and volatility spillovers using models such as ARCH, GARCH, and E-GARCH (Bampinas and Panagiotidis, 2015; Chen and Qu, 2019; Gharib et al., 2021). These models have been instrumental in understanding market dynamics, but they often overlook the potential for structural breaks and asymmetries in the relationships between these variables.

Furthermore, many studies have analyzed the symmetric relationships between gold prices, oil prices, and stock market prices. However, the majority have not delved into the short-term and long-term dynamics of these relationships before and after the COVID-19 pandemic. This study aims to fill these gaps by analyzing the short-term and long-term relationships between gold prices, oil prices, and the BSE SENSEX before and after the pandemic.

The COVID-19 crisis has underscored the need for more nuanced analyses of financial markets. The pandemic has not only introduced unprecedented volatility but also highlighted the interconnectedness of global markets. Understanding how gold prices, oil prices, and stock markets interact, particularly in the face of such shocks, is crucial for policymakers, investors, and researchers.

This study makes a significant contribution to the existing literature by providing a comprehensive analysis of the short-term and long-term relationships between these key variables, considering the impact of structural breaks due to the pandemic. By employing advanced econometric models, including those that can account for structural breaks, this research aims to provide more robust insights into market dynamics during periods of crisis.

While previous studies have laid the groundwork for understanding the volatility and relationships between commodity prices and stock markets, there is a clear need for more targeted research that addresses the specific dynamics during the COVID-19 pandemic. This study seeks to bridge this gap by focusing on the unique context of the pandemic and its effects on the relationships between gold prices, oil prices, and the BSE SENSEX. Through this analysis, the study aims to enhance our understanding of market behavior during crises and contribute to more effective financial market strategies and policies.

3. METHODOLOGY:

3.1 Data: Two different time periods are assumed in order to capture the effect of Covid-19. For pre-COVID-19, the study uses the data from 2018 to March 2020 and for post COVID-19 this includes our observations during December 2020 till end of the year (assuming modern system stays). The study has been based on secondary analysis of the data that consist BSE database, London Bullion Market Association (LBMA) and West Texas Intermediate (WTI). This bifurcation allows for a direct investigation of changes in the correlations between different economic variables across these two distinct periods due to the pandemic. We chose the BSE database, along with those at London Bullion Market Association (LBMA) and West Texas Intermediate (WTI), for this study upon their appropriateness as well as overall pervasiveness in important economic factors.

3.2 Econometric Technique: In econometrics, the terms "symmetric" and "asymmetric" relationships refer to the nature of interactions between variables. In a symmetric relationship, the impact of a change in one variable on other remains consistent irrespective of whether the change is positive or negative. For instance, in a simple linear regression model examining the relationship between income and consumption, a symmetric relationship implies that an increase in income leads to a proportional increase in consumption, and a decrease in income leads to a proportional decrease in consumption. For example, if a 10% increase in income results in a 5% increase in consumption, a 10% decrease in income would similarly lead to a 5% decrease in consumption.

Conversely, an asymmetric relationship suggests that the effect of a change in one variable on another varies depending on the direction of the change. The relationship is not uniform for positive and negative changes. Consider the relationship between oil prices and airline stock prices: an increase in oil prices might significantly decrease stock prices due to higher operational costs, whereas a decrease in oil prices might not have as strong or might even have a positive effect on stock prices, as lower operational costs can lead to higher profitability.

The study examines the presence of asymmetric impacts between fluctuations in macroeconomic factors and stock prices using a Nonlinear Autoregressive Distributed Lag (NARDL) model. Introduced by Shin et al. in 2014, the NARDL model is particularly useful when the variables are integrated in a specific sequence. The optimal lag lengths are determined by considering the values of either the Akaike Information Criterion (AIC) or the Schwarz Information Criterion (SIC).

The study follows the interpretation provided by Bahmani-Oskoee and Saha (2015), who demonstrated the utility of this approach in capturing asymmetric relationships. In the NARDL framework, all independent variables are divided into positive and negative components to evaluate the differential effects of macroeconomic variables on stock prices.

The general form of the unrestricted error correction model for nonlinear ARDL can be represented as:

ΔLnstockprice =α₀ + Σα₁ c⁺ΔLngoldprice_t-1⁺ + Σα₂ c^-ΔLngoldprice_t-1^- + Σα₃ c⁺ΔLnoilprice_t-1⁺ + Σα₂ c^-ΔLnoilprice_t-1^- +ε_t

This model allows for the separation of positive and negative changes in independent variables, enabling a more nuanced analysis of their impacts on stock prices. Based on the review of the literature, the following hypotheses are constructed:

· H0: The relationship between stock prices, gold prices, and oil prices is symmetric.

· H1: The relationship between stock prices, gold prices, and oil prices is asymmetric.

This hypothesis framework is critical in determining whether the effects of changes in gold and oil prices on stock prices are consistent regardless of the direction of change. If the relationship is found to be asymmetric, it would suggest that market responses to increases and decreases in commodity prices are different, which has significant implications for investors and policymakers.

The study’s contribution lies in its comprehensive analysis of the short-term and long-term relationships between gold prices, oil prices, and the BSE SENSEX before and after the pandemic, considering the potential for structural breaks due to the COVID-19 crisis. By employing advanced econometric models such as NARDL, which can account for these structural breaks, the study aims to provide more robust insights into the market dynamics during periods of crisis. This approach is particularly relevant given the unprecedented nature of the COVID-19 pandemic and its far-reaching impacts on global financial markets.

While previous studies have laid the groundwork for understanding the volatility and relationships between commodity prices and stock markets, this study addresses a critical gap by focusing on the asymmetric relationships between these variables in the context of the COVID-19 pandemic. Through this detailed analysis, the study aims to enhance our understanding of market behavior during crises and contribute to more effective financial market strategies and policies.

4. RESULT DISCUSSION AND FINDINGS:

Analysis of Pre-COVID-19 Period (2018-2021)

The COVID-19 pandemic, which began in late 2019, had a dramatic impact on global economies and financial markets. The relationship between oil prices, gold prices, and stock markets was particularly affected. At the outset of the pandemic, oil prices experienced an unprecedented crash. This was driven by a sharp drop in global demand for oil due to widespread lockdowns and travel restrictions. The situation was exacerbated by a price war between major oil-producing nations, notably Saudi Arabia and Russia, leading to a supply glut.

The oil markets became extremely volatile, with prices fluctuating wildly in response to news about the pandemic, changes in supply dynamics, and government policies. For instance, in April 2020, West Texas Intermediate (WTI) crude oil futures fell below zero for the first time in history. Many oil companies, especially those heavily leveraged, faced severe financial distress, and some declared bankruptcy. The pandemic accelerated structural changes in the energy sector pushing companies toward cost-cutting measures and diversification strategies, including investments in renewable energy sources.

Gold, traditionally considered a safe-haven asset, saw a surge in demand as investors sought refuge from the economic uncertainty caused by the pandemic. This demand was further fueled by central bank stimulus measures and concerns about the stability of traditional currencies. Gold prices reached record highs during the pandemic, surpassing $2,000 per ounce in August 2020–levels not seen since the financial crisis of 2008.

Unlike stocks, which experienced significant volatility and declines in the early stages of the pandemic, gold prices remained relatively stable and even appreciated. This divergence highlighted gold's role as a hedge against economic uncertainty. Investors viewed gold as a reliable store of value amidst the economic turmoil, reinforcing its status as a crucial component of diversified investment portfolios.

Global stock markets experienced sharp declines in February and March 2020 as the severity of the pandemic became evident. Investors were deeply concerned about the economic impact of lockdowns, reduced consumer spending, and disruptions to global supply chains. Throughout 2020, stock markets exhibited heightened volatility, with frequent swings in response to news related to the pandemic, vaccine developments, and economic data releases.

However, some sectors, particularly technology and healthcare, outperformed others as the pandemic accelerated trends such as remote work and telehealth. Technology giants benefited from increased demand for their products and services, driving substantial gains in their stock prices. Conversely, sectors like travel, hospitality, and energy struggled due to the direct impact of the pandemic on their operations.

The pandemic underscored the importance of diversification in investment portfolios. Different asset classes responded variably to the global health crisis, highlighting the need for a balanced approach to risk management in financial investments.

To rigorously analyze the impacts of the COVID-19 pandemic on financial markets, we used the Augmented Dickey-Fuller (ADF) test to check for unit roots in the time series data. The null hypothesis of the ADF test posits that the sequence has a unit root (i.e., the data is nonstationary). The results of the ADF test, presented in Table 1, indicate that all the variables (oil prices, gold prices, and stock market indices) were nonstationary at their levels. However, after taking the first difference, the data became stationary, indicating that all the variables are integrated of order one, I(1). This finding is crucial for further econometric modeling, ensuring the validity of subsequent analyses involving these time series data.

To determine the appropriate lag length for our analysis, we employed the Vector Auto Regressive (VAR) Lag Length Criterion. The results are presented in Table 2, where the Schwarz Information Criterion (SIC) was used to identify the optimal lag length. According to these results, the minimum SIC value is observed at lag 2. Consequently, lag 2 has been selected for further analysis.

Table 3 presents the results of the asymmetric cointegration test, which examines the long-term relationships between the positive and negative fluctuations of gold prices, oil prices, and the SENSEX. To assess these relationships, the Wald test was employed. This test is particularly useful in determining whether there are cointegrating relationships among the variables under consideration.

The results of the Wald test are summarized in Table 3. The F-Statistic value of 2.078411 with a corresponding probability of 0.0550 suggests that the null hypothesis of no cointegration can be rejected at a significance level just above the conventional 5% threshold.

The asymmetric cointegration test thus confirms that fluctuations in gold and oil prices have a significant impact on the SENSEX in the long run. This implies that investors and policymakers should consider the interconnected nature of these markets when making decisions.

Table 1: Results of Augmented Dickey-Fuller Unit Root Test - Before

Variables	Level			Intercept			Decision
Variables	Intercept	Trend and Intercept	None	Intercept	Trend and Intercept	None
LNGold	-0.37(0.91)	-1.81(0.69)	1.59(0.97)	-77.47(0.00)*	-77.46(0.00)*	-77.43(0.00)*	I(1)
LNOil	-1.63(0.46)	-0.78(0.96)	-0.22(0.60)	-79.67(0.00)*	-79.70(0.00)*	-79.68(0.00)*	I(1)
LNSENSEX	-0.65(0.85)	-2.48(0.33)	1.97(0.98)	-69.67(0.00)*	-69.66(0.00)*	-69.62(0.00)*	I(1)

Note: * -1% level of significance

Table 2: Result of VAR lag length criteria

Lag	LogL	LR	FPE	AIC	SC	HQ
0	-9134.708	NA	0.005039	3.223177	3.226692	3.224402
1	45862.26	109916.3	1.90e-11	-16.17293	-16.15887	-16.16804
2	45491.23	157.7395	1.85e-11	-16.19761	-16.17301*	-16.18904*
3	45954.44	26.47203	1.85e-11	-16.19910	-16.16395	-16.18686
4	45968.69	28.43767*	1.85e-11*	-16.20095*	-16.15526	-16.18504

*indicates lag order selected by the criterion LR: Sequential modified LR test statistic (each test at 5%)

FPE: Final Prediction Error AIC: Akaike information criterion

SC: Schwarz information criterion HQ: Hannan-Quinn information criterion

Table 3: Wald Test reult

Test Statistic	Value	df	Probability
F-Statistic	2.078411	(5,5657)	0.0550
Chi-Square	10.39206	5	0.0549

Table 4 presents the estimation of long-run coefficients using the Nonlinear Autoregressive Distributed Lag (NARDL) model. This model is instrumental in capturing the asymmetric effects of the explanatory variables on the dependent variable, in this case, the SENSEX, both before and after the COVID-19 pandemic. The results in Table 4 reveal that, in the long run, only one lag of the SENSEX significantly impacts its current value before the COVID-19 pandemic. The coefficient for SENSEX (-1) is 0.10, with a P-value of 0.01, indicating that the lagged value of the SENSEX is positively correlated with its current value and is statistically significant at the 1% level. This finding implies that past performance of the SENSEX is a strong predictor of its future value, highlighting a momentum effect in the stock market before the pandemic. Investors and market analysts can use this information to make informed predictions about future market movements based on historical data.

However, the results also show that other variables, such as gold prices and oil prices, did not have a significant long-run impact on the SENSEX during this period. This insignificance suggests that, before the COVID-19 pandemic, the Indian stock market was relatively isolated from the long-term effects of fluctuations in these global commodities. This isolation could be due to various factors, including robust domestic economic policies, a stable macroeconomic environment, or a lack of immediate sensitivity to international commodity price movements.

The lack of significant long-term impact from gold and oil prices before COVID-19 contrasts with the heightened sensitivity observed during the pandemic. The pandemic-induced economic uncertainty and market volatility may have increased the interconnectedness between these commodities and the stock market, as investors sought to rebalance their portfolios and hedge against risks.

The significant coefficient for the lagged SENSEX highlights the importance of considering past market performance in long-term investment strategies. It suggests a degree of predictability and inertia in the stock market, where past trends and movements continue to influence future values. For policymakers understanding this momentum can aid in developing measures to stabilize the market during periods of volatility and ensure sustained growth.

Table 4: NARDL Long-run estimated coefficients

Variables	Coefficient	P value
SENSEX (-1)	0.10	0.01

Table 5 provides the short-run estimated coefficients derived from the Nonlinear Autoregressive Distributed Lag (NARDL) model. This model is pivotal for understanding the immediate impacts of fluctuations in independent variables, such as gold and oil prices, on the SENSEX. The short-run analysis reveals several key insights into the immediate impacts of changes in gold and oil prices on the SENSEX. The coefficient for a 1% decrease in oil prices is 0.03, with a P-value of 0.04, indicating a statistically significant positive impact on the SENSEX. This positive relationship can be attributed to several factors:

1. Cost Reduction for Industries: Lower oil prices reduce the operational costs for industries reliant on oil, such as transportation and manufacturing. This reduction can lead to higher profit margins, making these companies more attractive to investors.

2. Economic Outlook: A decrease in oil prices might signal improved economic conditions, such as increased global demand or better supply stability. This positive outlook can boost investor confidence, leading to more investment in equities.

3. Inflation Control: Lower oil prices can help control inflation, as oil is a critical input for various goods and services. Controlling inflation can lead to lower interest rates, making borrowing cheaper and stimulating economic growth, which positively impacts the stock market.

The coefficient for a 1% decrease in gold prices is 0.06, with a P-value of 0.06, also indicating a positive but marginally significant impact on the SENSEX. This relationship can be explained through:

1. Shift in Investor Sentiment: Gold is traditionally viewed as a safe-haven asset. A decrease in gold prices suggests that investors are shifting their preferences towards riskier assets, such as stocks, which typically offer higher returns.

2. Liquidity Movement: When gold prices fall, liquidity might move from gold to the stock market, as investors seek better returns. This increased liquidity in the equity market can drive up stock prices, thereby positively impacting the SENSEX.

3. Economic Confidence: Falling gold prices can also be interpreted as a sign of economic stability or optimism, reducing the need for safe-haven investments. This confidence can encourage more investment in the stock market.

The coefficient for the lagged value of the SENSEX (DSENSEXP (-1)) is 0.07, with a P-value of 0.001, indicating a strong and statistically significant impact. This highlights the momentum effect in the stock market, where past performance influences current market behavior. Investors often rely on historical trends to make decisions, and positive past performance can lead to continued investor confidence and investment, sustaining the upward momentum.

The short-run results underscore the sensitivity of the SENSEX to changes in global commodity prices and investor sentiment. The positive impact of decreases in both oil and gold prices on the SENSEX suggests that in the short run, market participants respond favorably to signals of economic stability and growth potential.

For policymakers and market analysts, understanding these short-term dynamics is crucial for developing strategies to stabilize the market during periods of volatility. It also highlights the importance of monitoring global commodity prices as they have immediate effects on investor behavior and market performance.

Table 5: NARDL short-run estimated coefficients

Variables	Coefficient	P value
DSENSEXP (-1)	0.07	0.001
Oil^-	0.03	0.04
Gold-	0.06	0.06

Diagnostic tests in econometrics are essential for verifying the reliability and validity of statistical models. They help in detecting potential issues that could undermine the accuracy of the model's results. Tables 6 and 7 provide the results of two critical diagnostic tests: the Breusch-Godfrey Serial Correlation LM Test and the Breusch-Pagan-Godfrey Heteroskedasticity Test.

Table 6: Results of Breusch-Godfrey Serial Correlation LM Test

F – statistic	1.569284	Prob. F (2,5655)	0.2083
*Obs R-squared**	3.144592	Prob. Chi-Square(2)	0.2076

Null hypothesis: No serial correlation at up to 2 lags

Table 7: Results of the Breusch- Pagan -Godfrey Heteroskedasticity Test

F Stat	1.23266	Prob. F (11,5657)	0.1100
*Obs R-squared**	12.6663	Prob. Chi-Square (11)	0.2400
Scaled explained SS	69.2744	Prob. Chi-Square (11)	0.2600

The Breusch-Godfrey Serial Correlation LM Test is designed to detect the presence of serial correlation, or autocorrelation, in the residuals of a regression model. The null hypothesis for this test posits that there is no serial correlation up to a specified number of lags. The test results show that the F-statistic is 1.5693, with a corresponding p-value of 0.2083. Additionally, the Observed R-squared is 3.1446, and its p-value is 0.2076. Since both p-values exceed the conventional significance level of 0.05, we fail to reject the null hypothesis. This indicates that the model does not suffer from significant serial correlation in the residuals.

The Breusch-Pagan-Godfrey Heteroskedasticity Test assesses whether the residuals of a regression model exhibit heteroskedasticity, which occurs when the variance of residuals is not constant across observations. The null hypothesis for this test is that there is no heteroskedasticity present. According to the test results, the F-statistic is 1.2327 with a p-value of 0.1100. The Observed R-squared is 12.6663, and its p-value is 0.2400. The Scaled Explained SS is 69.2744, with a p-value of 0.2600. All p-values are above the 0.05 significance level, suggesting that we fail to reject the null hypothesis of homoscedasticity. Consequently, these results indicate that the model does not exhibit significant heteroskedasticity, ensuring that the variance of residuals remains stable.

Analysis after Covid 19:

The analysis of the data from the post-COVID-19 period involved several key steps to ensure the robustness of the findings. The stationarity of the variables was assessed using the Augmented Dickey-Fuller (ADF) test, with the results detailed in Table 8. The test results indicate that all variables are non-stationary at their levels. However, after taking the first difference of each variable, they become stationary. This suggests that all variables are integrated of order one, denoted as I(1). This transformation is necessary for the reliability of subsequent econometric analysis.

To determine the optimal lag length for the Vector Auto-Regressive (VAR) model, the VAR lag length criteria were applied. The results of this selection process are shown in Table 9. According to the criteria, the Schwarz Information Criterion (SIC) reached its minimum at lag 1. Therefore, lag 1 has been chosen as the optimal lag length for further analysis. This choice ensures that the model adequately captures the dynamics of the variables while avoiding over fitting.

Table 8: Results of Augmented Dickey-Fuller Unit root test -During

Variables	Level			Intercept			Decision
Variables	Intercept	Trend and Intercept	None	Intercept	Trend and Intercept	None
LNGold	-1.72(0.41)	-2.54(0.30)	0.21(0.74)	-14.45(0.00)	-14.41(0.00)	-14.49(0.00)	I(1)
LNOil	-0.83(0.80)	-2.87(0.17)	0.93(0.97)	-11.22(0.00)	-11.21(0.00)	-11.15(0.00)	I(1)
LNSENSEX	-2.36(0.15)	-2.89(0.16)	0.08(0.70)	-13.45(0.00)	-13.58(0.00)	-13.49(0.00)	I(1)

Note: *** 1% level of significance

Table 9: VAR lag length selection

Lag	LogL	LR	FPE	AI C	SC	HQ
0	873.2210	NA	1.82e-08	-9.307176	-9.255340	-9.286172
1	1577.381	1378.195*	1.08e-11*	-16.74204*	-16.53470*	-16.65802*
2	1582.096	9.077800	1.13e-11	-16.69621	-16.33336	-16.54919
3	1588.372	11.879997	1.16e-11	-16.66708	-16.14872	-16.45704
4	1593.817	10.13270	1.20e-11	-16.62905	-15.95519	-16.35600

*indicates lag order selected by the criterion LR: Sequential modified LR test statistic (each test at 5%) FPE:

Final Prediction Error AIC: Akaike information criterion SC: Schwarz information criterion HQ: Hannan-Quinn information criterion

The post-COVID-19 analysis provides insights into the dynamics between gold prices, oil prices, and the SENSEX, focusing on both long-run and short-run effects. Table 10 presents the long-run estimated coefficients for the post-COVID-19 period. The results indicate that even after the pandemic, a single lag of the SENSEX continues to influence its current value. Specifically, the coefficient for SENSEX (-1) is -0.06 with a p-value of 0.01, suggesting a significant but negative long-term effect. This finding underscores the persistence of past values in shaping the current state of the stock market, reflecting the ongoing influence of historical market conditions on the present index levels.

However, the coefficients for gold and oil prices in the long run are statistically insignificant, indicating that these variables do not have a lasting impact on the SENSEX once short-term fluctuations are accounted for. The short-run results are detailed in Table 11. The coefficients reveal that an increase in gold prices has a substantial negative impact on the SENSEX, with a decrease of 6.5%. Conversely, a decrease in oil prices results in a 1.83% decline in the SENSEX. These findings highlight the immediate effects of changes in gold and oil prices on the stock market.

Gold's negative impact on the SENSEX can be attributed to its role as a safe-haven asset. During the COVID-19 pandemic, heightened economic uncertainty drove investors towards gold as a protective measure, leading to higher gold prices. The increased demand for gold often comes at the expense of investment in equities, thus contributing to a decline in the SENSEX.

In contrast, the relationship between oil prices and the SENSEX reflects the critical role of the energy sector in the stock market. Lower oil prices can depress the profitability of energy companies, which are significant components of the SENSEX. This decrease in profitability translates into lower stock prices for these companies, subsequently exerting downward pressure on the overall index.

Table 10: NARDL Long-run estimated coefficients

Variables	Coefficient	P value
SENSEX (-1)	-0.06	0.01

Table 11: NARDL short-run estimated coefficients

Variables	Coefficient	P value
DSENSEXP (-1)	0.07	0.001
Gold +	-6.5	0.04
Oil-	-1.83	0.06

Table 12 and Table 13 provide the results of diagnostic tests for autocorrelation and heteroskedasticity. The Breusch-Godfrey Serial Correlation LM Test results in Table 10 show that the p-value for the F-statistic and the Chi-Square statistic are both greater than 0.05. This indicates that there is no evidence of serial correlation in the residuals, confirming that the model is free from autocorrelation issues.

Similarly, the Breusch- Pagan- Godfrey Heteroskedasticity Test results in Table 13 indicate no significant heteroskedasticity problems. The p-values for the F-statistic, Chi-Square, and Scaled explained SS are all above the 0.05 threshold. This suggests that the variance of the residuals is constant across observations, further validating the robustness of the model.

Table 12: Results of Breusch-Godfrey Serial Correlation LM Test

F statistic	0.789576	Prob.F (2,308)	0.4550
Obs*R-squared	1.627208	Prob. Chi-Square(2)	0.4443

Table 13: Results of the Breusch- Pagan -Godfrey Heteroskedasticity Test

F Stat	14.88836	Prob. F (11,5657)	0.0100
Obs *R-squared	88.54458	Prob. Chi-Square (11)	0.0000
Scaled explained SS	319.3216	Prob. Chi-Square (11)	0.0260

Overall, the post-COVID-19 analysis reveals both persistent and evolving dynamics in the relationship between gold prices, oil prices, and the SENSEX. The findings highlight the ongoing influence of historical market conditions while also emphasizing the immediate effects of economic variables on the stock market in the short run. The diagnostic tests affirm the reliability of the model, providing confidence in the results and conclusions drawn from the analysis.

5. CONCLUSION:

The impact of COVID-19 on global financial markets, particularly the relationships between gold prices, oil prices, and stock markets, has been significant. The pandemic's onset led to unprecedented volatility in various markets. Despite the extensive research on the relationships between oil prices, gold prices, and stock markets, there remains a gap in studies specifically investigating these dynamics. Most research has focused on mean and volatility spillovers using models like ARCH, GARCH, and E-GARCH, often overlooking potential structural breaks and asymmetries in these relationships. Additionally, while many studies have analyzed symmetric relationships, few have explored the short-term and long-term dynamics of these relationships before and after the COVID-19 pandemic. This study aims to address these gaps by examining the short-term and long-term relationships between gold prices, oil prices, and the BSE SENSEX during these two distinct periods.

In this study, the analysis of the SENSEX’s response to global commodity prices before and after COVID-19 reveals significant differences compared to existing literature. Specifically, our findings indicate that the SENSEX’s sensitivity to decreases in oil and gold prices increased after the pandemic, with coefficients of 0.03 for oil and 0.06 for gold, highlighting a shift in market behavior. This contrasts with earlier studies such as Mahajan et al. (2021) and Shaikh (2021), which noted increased volatility in stock markets during COVID-19 but did not delve into the asymmetric effects or the detailed influence of specific commodity prices on the SENSEX. Furthermore, while Bampinas and Panagiotidis (2015) and Chen and Qu (2019) explored symmetric relationships and spillovers between commodities and stock markets, our study uncovers significant asymmetric effects, where the impact of price changes differs based on their direction. Additionally, this study incorporates of structural breaks due to COVID-19 reveals a heightened interconnectedness between commodities and the SENSEX, a nuanced insight not extensively addressed by other studies who examined pandemic impacts without focusing on such structural shifts. This study's emphasis on the post-pandemic shift in sensitivity and the identification of asymmetries provide new perspectives that extend beyond the more generalized findings in the literature.

While many studies converge on the broad impact of COVID-19 on financial markets, especially in terms of increased volatility and shifts in investor behavior, there are notable differences in the specific findings and methodologies. The convergence underscores a general agreement on the major trends induced by the pandemic, such as the safe-haven role of gold and the volatility in oil markets. However, the diversity in findings reflects the complexity of financial markets and the varied impacts of the pandemic across different asset classes and market dynamics. This variation highlights the need for continued research to fully understand the nuanced effects of global crises on financial markets.

REFERENCES:

1. Albulescu, C. COVID-19 and the United States financial markets’ volatility. Finance Research Letters. 2020: 38. https://doi.org/ 10.1016/j.frl.2020.101699

2. Awan, T. M., et al. Oil and stock markets volatility during pandemic times: A review of G7 countries. Green Finance. 2021: 15–27. DOI: 10.3934/GF.2021002

3. Baur, D. G., and Lucey, B. M. Is gold a hedge or a safe haven? An analysis of stocks, bonds, and gold. Financial Review. 2010; 45(2): 217–229. doi: 10.1111/j.1540-6288.2010.00244.x

4. Bakas, D. and Triantafyllou, A. Commodity price volatility and the economic uncertainty of pandemics. Economic Letters. 2020: 193. doi: 10.1016/j.econlet.2020.109283

5. Cui, X., Yen-Ku, K., Maneengam, A., Cong, P. T., Quynh, N. N., Ageli, M. M., and Wisetsri, W. Covid-19 and oil and gold price volatilities: Evidence from China market. Resources Policy. 2022; 79: 103024. https://doi.org/10.1016/j.resourpol.2022.103024

6. Meher, et al. The Impact of COVID-19 on Price Volatility of Crude Oil and Natural Gas Listed on Multi Commodity Exchange of India. International Journal of Energy Economics and Policy. 2020; 10(5): 422-431. DOI: 10.32479/ijeep.10047

7. Shaikh, I. Impact of COVID-19 pandemic disease outbreak on the global equity markets. Economic Research-Ekonomska Istraživanja. 2021; 34(1): 2317-2336. DOI: 10.1080/ 1331677X.2020.1863245

8. Mahajan, S., et al. Impact of COVID-19 on Stock Market and Gold Returns in India. Eurasian Journal of Business and Economics. 2021; 14(27): 29-46. DOI: https://doi.org/10.17015/ ejbe.2021.027.02

9. Regnier, E. Oil and energy price volatility. Energy Economics. 2005; 29(3): 405–427. doi: 10.1016/j.eneco.2005.11.003

Received on 27.08.2024 Revised on 14.02.2025

Accepted on 23.05.2025 Published on 20.08.2025

Available online from September 02, 2025

Res. J. of Humanities and Social Sciences. 2025;16(3):172-180.

DOI: 10.52711/2321-5828.2025.00029

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Creative Commons License.