Document Type : Research Paper
Authors
1 Associate Professor, Department of Business Management, Faculty of Economics, Management and Administrative Sciences, Semnan University, Semnan, Iran
2 Ph.D. in Finance - Financial Engineering, Faculty of Economics, Management and Administrative Sciences, Semnan University, Semnan, Iran
Abstract
This research examined the asymmetric effects of domestic economic policy uncertainty (DEPU) and global economic policy uncertainty (GEPU) on stock market index returns in Iran. The study focused on simultaneous analysis of economic policy uncertainty originating from both domestic and global sources within a nonlinear framework, as well as the stock market’s asymmetric responses to these uncertainties. It used the nonlinear autoregressive distributed lag (NARDL) model, as it enables dynamic analysis and distinguishes the market’s reactions to positive and negative shocks across different time horizons. The dataset consisted of quarterly observations from 1997 to 2024. In addition to the uncertainty indices, the model incorporated several control variables, including the exchange rate, global oil prices, the consumer price index, money supply, real non-oil GDP, and stock market liquidity. Before estimating the model, the statistical properties of the data—such as nonlinearity, stationarity, the presence of long-term relationships, and response symmetry—were examined to ensure the suitability of the NARDL approach and the validity of the results. The results indicated that positive and negative shocks to DEPU have significant positive and negative effects, respectively, on stock market index returns in both the short and long run. Furthermore, GEPU shocks exert significant short-term effects with a time lag: positive shocks increase, while negative shocks decrease stock market index returns. In the long term, however, only positive GEPU shocks have a significant positive impact. The control variables also exhibited significant effects on stock market index returns.
Introduction
Economic policy uncertainty (EPU) is widely recognized as a critical factor influencing financial markets, including stock market returns. In today’s interconnected global economy, both domestic and global sources of policy uncertainty play a pivotal role in shaping investor behavior, economic decision-making, and overall market stability. Given Iran’s repeated exposure to policy shifts, economic sanctions, and geopolitical tensions, the country presents a unique setting for analyzing the impacts of policy uncertainty. Uncertainties—whether originating domestically or globally—can affect the stock market in diverse ways, varying in timing, direction, and intensity. This study aimed to investigate the asymmetric effects of domestic economic policy uncertainty (DEPU) and global economic policy uncertainty (GEPU) on stock market returns in Iran over the period 1997–2024. The primary objective was to examine how different forms of policy uncertainty influence the behavior of the Iranian stock market, while accounting for the non-linear and dynamic nature of these relationships. Market responses are not only asymmetric but are also shaped by the specific nature of each uncertainty source and the market’s sensitivity to these factors. Therefore, a nuanced analytical approach is required to capture the interactions between policy uncertainty and stock market performance. To this end, the present study employed the nonlinear autoregressive distributed lag (NARDL) model, an advanced econometric technique designed to capture asymmetric responses to positive and negative shocks in EPU. The findings can provide valuable insights into the role of EPU in shaping stock market returns in an emerging market such as Iran.
Materials and Methods
The present study used the nonlinear autoregressive distributed lag (NARDL) model, an appropriate method for analyzing nonlinear relationships in economic time series data. The NARDL model allows for the differentiation between positive and negative shocks, offering a more nuanced understanding of how various forms of uncertainty impact market behavior. Unlike traditional linear models, which assume symmetric effects of shocks, the NARDL approach enables the examination of distinct effects arising from positive and negative policy uncertainty shocks on stock market returns. This asymmetry is central to the study, as it reveals how market responses vary depending on the intensity and direction of uncertainty—an essential aspect for comprehensively assessing the effects of EPU on stock market performance. The analysis used quarterly time series data spanning the period 1997 to 2024. Key variables included DEPU and GEPU, Iran’s stock market returns, the exchange rate, global oil prices, the consumer price index (CPI), money supply, real non-oil GDP, and stock market liquidity. The DEPU and GEPU indices were constructed using content analysis of news reports, a widely accepted method for measuring EPU. Based on the theoretical framework and following the model proposed by Shin et al. (2014), the nonlinear long-term specification for Iran’s stock market index returns is presented in Equation (1):
(1)
The analysis aimed to simultaneously analyze the asymmetric short-term and long-term effects of the variables, so Equation (1) was reformulated as a NARDL model in the form of an error correction model (ECM), as presented in Equation (2):
+
(2)
Before estimating the NARDL model, several preliminary tests were conducted to ensure its statistical validity and the suitability of applying this nonlinear model. These tests included the BDS test for nonlinear dependence, the ADF and PP tests for stationarity, the Zivot–Andrews test for structural breaks, the cointegration test of Pesaran et al. (2001) for long-run relationships, and the Wald test for asymmetry. The results confirmed that the data satisfied the necessary assumptions for valid estimation and that the NARDL model was appropriate for the analysis. Following the estimation of the NARDL model, several diagnostic tests were performed to assess the reliability of the results. These included the ARCH and Breusch–Godfrey tests to detect heteroscedasticity and autocorrelation in the residuals, as well as the CUSUM and CUSUMQ to check for parameter stability. The outcomes of these diagnostic tests indicated that the model was correctly specified and that the findings were robust and reliable.
Results and Discussion
The results from the dynamic NARDL model (Table 1) showed that, in the short term, the lagged value of the stock market index has a significant positive effect on its returns. This indicates that past market growth can stimulate future growth. In the short term, DEPU displayed asymmetric effects: positive shocks have a significant positive impact, while negative shocks have a significant negative effect. GEPU also showed delayed and asymmetric effects. Positive GEPU shocks are statistically insignificant, but their first and second lags exert significant positive impacts on stock market returns. Negative GEPU shocks are also statistically insignificant, although their first lag has a significant negative effect. Regarding the control variables, the exchange rate had a positive and significant effect in the current period, while its lag was statistically insignificant. Moreover, global oil prices can exert positive effects in the second and third lags, but not in the current period or the first lag.
Table 1. Results of Dynamic NARDL Model Estimation
p-value
t-Statistic
Standard error
Coefficient
Variable
0.00
3.12
0.09
0.30
LSP(-1)
0.02
2.32
0.10
0.23
LDEPU_POS
0.72
0.34
0.93
0.32
LDEPU_POS)-1)
0.01
-2.51
0.08
-0.22
LDEPU_NEG
0.14
1.45
0.70
1.02
LGEPU_POS
0.03
2.15
0.52
1.12
LGEPU_POS(-1)
0.02
2.30
0.56
1.29
LGEPU_POS(-2)
0.10
1.64
0.39
0.64
LGEPU_NEG
0.00
-2.66
0.54
-1.45
LGEPU_NEG)-1)
0.04
2.07
0.44
0.93
LEX
0.14
-1.45
0.51
-0.75
LEX(-1)
0.25
-1.15
0.30
-0.35
LOIL
0.23
1.18
0.44
0.53
LOIL (-1)
0.03
2.17
0.45
0.98
LOIL (-2)
0.00
4.82
0.29
1.42
LOIL (-3)
0.00
2.70
0.47
1.27
LCPI
0.72
-0.34
2.11
-0.73
LMS
0.60
-0.52
2.46
-1.29
LMS)-1)
0.02
2.49
0.41
1.03
LMS)-2)
0.69
-0.40
4.55
-1.82
LMS)-3)
0.54
0.60
0.17
0.10
LRNOGDP
0.72
0.35
0.10
0.03
LLIQ
0.04
2.00
0.10
0.20
LLIQ (-1)
=0.88 F-statistic=7.6836 Prob (F-statistic) =0.0000
Source: Research Results
The consumer price index was found to have a significant positive impact on stock market returns. Money supply has a significant positive effect only in the second lag. Real non-oil GDP is not statistically significant in the short term, whereas market liquidity is significant only in its first lag. The overall model significance was confirmed by the F-statistic.
The long-term results from the NARDL model (Table 2) indicated that DEPU exhibits significant asymmetric effects: positive shocks lead to a positive long-term impact on stock market returns, whereas negative shocks produce a negative long-term effect. GEPU also showed asymmetric long-term behavior, with only positive shocks having a statistically significant influence. Among the control variables, the exchange rate and real non-oil GDP exert positive and significant long-term effects on stock market returns, while global oil prices and the consumer price index have negative long-term effects. The ECM coefficient confirmed that, following any shock, the market gradually adjusts back to its long-term equilibrium.
Table 2. Long-Term Relationships and the Error Correction Model (ECM) Results From NARDL
p-value
t-Statistic
Standard error
Coefficient
Variable
0.01
2.41
0.14
0.34
LDEPU_POS
0.00
2.92
0.09
-0.28
LDEPU_NEG
0.00
2.70
0.37
1.02
LGEPU_POS
0.11
-1.60
0.55
-0.89
LGEPU_NEG
0.00
2.87
0.24
0.69
LX
0.04
-2.04
0.45
-0.92
LOIL
0.01
-2.39
0.80
-1.93
LCPI
0.40
-0.83
0.75
-0.63
LMS
0.01
2.47
0.56
1.38
LRNOGDP
0.33
0.97
0.22
0.21
LLIQ
0.02
-2.2
0.11
-0.25
ECM
Source: Research Results
Conclusion
According to the findings, both DEPU and GEPU exert significant and asymmetric effects on stock market returns in Iran. The results highlighted the importance of recognizing non-linear relationships when examining the impact of EPU, particularly in emerging markets such as Iran. In light of the findings, it is recommended that economic policymakers proceed with greater caution when announcing policies and avoid issuing contradictory or ambiguous signals that could trigger short-term market overreactions and instability. Moreover, achieving long-term economic stability requires careful attention to fundamental factors such as exchange rates, oil prices, and economic growth. The results also indicated that, in the short term, speculative and emotional behaviors play a substantial role in market fluctuations. These behaviors can be curbed by improving regulatory frameworks, for example by facilitating short-selling, futures contracts, and options trading. In the long term, enhancing investors’ financial literacy and encouraging long-term investment strategies can help reduce volatility driven by short-term speculation.
Keywords
Main Subjects
)