Journal of Energy Markets

Risk.net

The relationship between oil prices, global economic policy uncertainty and financial market stress

Sayyed Mahdi Ziaei

  • The NARDL technique confirms the long-run asymmetric relationship between global economic policy uncertainty, gold prices and oil prices.
  • Monetary policy influence on oil prices limited to negative impacts of it in the long-run
  • The results indicate long- and short-run relationships between negative OP and financial stress index
  • Oil prices inversely impact the financial stress index in the short and long runs

This paper introduces two models. The first model analyzes the impacts of global economic policy uncertainty (EPU), gold prices (GPs) and three-month US Treasury bill (TB) rates on oil prices (OPs) between 1997 and 2020. The second model examines the effects of OPs and US TB rates on the Püttmann Financial Stress Indicator (FSI) between 1979 and 2016. To capture the long- and short-run relationships between independent and dependent variables, an asymmetric nonlinear autoregressive distributed lag model is employed for both models. The results indicate that the effects of the independent variables are long term only. The results confirm the significant impact of global EPU and GPs on OPs in the long run. For the second model, OPs are the determining variable to explain financial stress in both the long run and the short run. Although the negative fluctuations in OPs have a positive effect on the FSI in the short run, their impacts on the FSI turn negative in the long run. The overall results underscore how OPs are affected by uncertainty and how they influence the FSI.

1 Introduction

Due to the global Covid-19 lockdown, crude oil prices (OPs) dropped into negative territory for the first time in history (April 2020), resulting in future OPs being cheaper than spot prices. The short-term glut in oil supply occurred due to the deep uncertainty over the future economic situation. OPs can be interpreted as one of the key indicators of global economic status. In general, economic downturns can be explained as the side-effects of recessions, while spikes in OPs can be attributed to economic booms. This paper investigates the impact of global economic uncertainty on OPs and how the latter contribute to financial market stress. Our study is, to the best of our knowledge, the first attempt in the literature to explore this subject on a global basis.

Most of the studies done recently focus on the effects of economic policy uncertainty (EPU) on financial markets (Nguyen et al 2020; Das et al 2019; Ersan et al 2019; Phan et al 2018; Li and Peng 2017; Arouri et al 2016; Li et al 2015; Wang et al 2014; Ziaei 2012, 2021). Their overall results imply the negative impacts of economic uncertainty on financial market volatility. Nguyen et al (2020) noted that both negative and positive changes in EPU have significant effects on bank credit growth. However, it seems that the impacts of EPU are greater in developing countries than developed countries. Kim et al (2020) believes that while emerging market economies react more to risk-aversion shocks, advanced economies are greatly sensitive to US policy uncertainty shocks.

Other researchers have investigated the impacts of EPU on aggregate demand and unemployment. Baker et al (2016) and Shoag and Veuger (2016) found that the pronounced impacts of economic uncertainty on unemployment, as explained by Jones and Olson (2013), led to declines in gross domestic product (GDP) during economic downturns. Similarly, some studies have focused on the relationship between economic uncertainty and OPs. Yang (2019) elaborated on the possibility of the coexistence between OP-shock uncertainty and economic policy outcomes. Yang (2019) noted that OPs act as the information receiver of EPU. The time factor plays a pivotal role in the strengthening of the relationship between EPU and OP. Having said this, Liu et al (2019) believe that many factors such as geopolitical crises, natural disasters, access to future supplies and current output are also determinants of OP volatility.

On the flip side, OP fluctuations have different effects on the economy, such as the supply-side aspect (Brown and Yucel 2002), which is the most well-known spillover effect. Kocaarslan et al (2020) elaborated on the five major effects of OPs in their study. They explained that OPs influence production costs, purchasing power, money demand, interest rates (money supply) and production structures through the supply-side, wealth transfer, real-balance, inflation and labor markets.

A substantial number of studies have evaluated the effects, comovements, spillover effects of returns and volatilities of both OPs and gold prices (GPs) (Tiwari et al 2020; Chen and Xu 2019; Teetranont et al 2018; Kanjilal and Ghosh 2017; Kumar 2017; Charlot and Marimoutou 2014; Ewing and Malik 2013). Lee et al (2012) and Zhang and Wei (2010) found a unidirectional causality between OPs and GPs, while Bildirici and Turkmen (2015) established a bidirectional relationship between them. Mensi et al (2015) employed the bivariate dynamic conditional-correlation fractionally integrated asymmetric power ARCH (DCC-FIAPARCH) model and noted the presence of a long-run relationship between OPs and GPs. Mokni et al (2020) applied regime-switching models and affirmed the link between oil shocks and GPs, as well as explaining how gold markets are strongly influenced by EPUs.

The effects of conventional and unconventional monetary policies on commodity prices including OP and GP have also been much investigated. Gospodinov and Jamali (2015) noted that uncertainty over the positive and negative impacts of monetary policies produces two different influences on commodity prices. Although a decrease in the target rate by more than expected (uncertainty effects) leads to a fall in the future prices of some energy and metal commodities, a rise in the target rate by more than the forecast rate leads to differing impacts on commodity prices and GDP. Kormilitsina (2011) employed the dynamic stochastic general equilibrium model to evaluate the claim that US monetary policy aggravated the slowdown created by OP shocks in the aftermath of World War II, and they found a positive relationship between the two factors. Rosa (2014) found that an unexpected asset-price-purchases policy led to a negative response to energy prices. Apergisa et al (2020) deduced that conventional and unconventional monetary policies had the same impact on GPs and OPs. However, they concluded that the effects of quantitative easing policies on OP volatility were more pronounced than conventional impacts.

The contribution of our paper is the following. First, this paper studies the asymmetric impacts of global EPU (GEPU), GPs and interest rates on OPs. While a growing body of research has analyzed the effects of economic policy on OPs, research on the asymmetric effects of GEPU on OPs is still limited. Second, we study the asymmetric impacts of OPs to determine how they could influence the financial stress index (FSI). Third, we employ a nonlinear autoregressive distributed lag (NARDL) in our research. NARDL is a particularly appropriate approach for our research as it evaluates the negative and positive impacts of each independent variable on the dependent variable. To the best of our knowledge, this is the first attempt to use NARDL to simultaneously investigate the asymmetric effects of GEPU on OPs and the asymmetric impacts of OPs on the FSI.

The main findings of our research are as follows. First, we find that GEPU significantly affects OPs in the long run. Second, while the impacts of positive and negative changes in GPs on OPs are widely observed in the long run, the monetary policy influence is limited to negative impacts (tightening monetary policy) in the long run. Further, the results indicate that negative fluctuations in OPs have negative impacts on the FSI in the long run.

The paper is structured as follows. Section 2 explains the data set we employ and descriptive statistics of variables, while Section 3 addresses the NARDL framework that is applied for asymmetric analysis. Section 4 summarizes the estimation results obtained from the empirical analysis and, finally, Section 5 discusses and presents the conclusions of our research.

2 Data

Monthly data for the first model is from January 1, 1997 to July 1, 2020, while the second model covers data from May 1, 1979 to December 1, 2016. Data for crude OPs is based on West Texas Intermediate. GPs and three-month Treasury bill (TB) market rates are obtained from Federal Reserve Economic Data, while information on GEPU and the FSI is derived from the EPU website.11 1 URL: http://www.policyuncertainty.com. Davis (2016) developed a monthly index of GEPU based on the EPU indexes of Baker et al (2016). The GEPU index is a GDP-weighted average of national EPU indexes for the 21 countries that account for 70% of global GDP. The FSI constructed by Püttmann (2018) is based on the reporting of selected indicators in five major US newspapers

3 Methodology

Owing to the mixed findings in the literature, we have used the NARDL model formulated by Shin et al (2014) to investigate the nonlinear association between global variables, ie, GPs, GEPU, and three-month TBs on OPs, and also the effects of OPs and interest rates on the FSI. The NARDL is an advanced form of the ARDL model that tests the nonlinear long-term equilibrium through asymmetric relationships between variables, and is an extension of the Pesaran et al (2001) and the Pesaran and Shin (1998) ARDL models.

According to Shin et al (2014) asymmetric long-run regression is written as

  yt =x0+α+xt++α-xt-+ut,   (3.1)
  Δxt =vt,   (3.2)

where yt is the dependent variable and xt is the independent variable. yt and xt are scalar I(1) variables. xt is defined as xt=x0+xt++xt-, where x0 is the initial value, and positive values (xt+) and negative values (xt-) are partial sum processes:

  xt+=i=1tΔxi+=i=1tmax(Δxi,0),xt-=i=1tΔxi-=i=1tmin(Δxi,0),   (3.3)

which show negative and positive fluctuations.

The error correction of the NARDL model and the long-run and short-run asymmetric relationship between dependent and independent variables are extended in the following, based on (3.1).

Model 1:

  ΔOPt =α+ΨOPt-1+β1+GEPUt-1++β1-GEPUt-1-+β2+GPt-1+  
      +β2-GPt-1-+β3+TBt-1++β3-TBt-1-  
      +i=1p-1μΔOPt-i+i=0q-1τ1+ΔGEPUt-i++i=0q-1τ1-ΔGEPUt-i-  
      +i=0q-1τ2+ΔGPt-i++i=0q-1τ2-ΔGPt-i-  
      +i=0q-1τ3+ΔTBt-i++i=0q-1τ3-ΔTBt-i-+εi.   (3.4)

Model 2:

  ΔFSIt =α+ΨFSIt-1+β1+OPt-1+  
      +β1-OPt-1-+β2+TBt-1++β2-TBt-1-  
      +i=1p-1μΔFSIt-i+i=0q-1τ1+ΔOPt-i++i=0q-1τ1-ΔOPt-i-  
      +i=0q-1τ2+ΔTBt-i++i=0q-1τ2-ΔTBt-i-+εi.   (3.5)

The dependent variable in (3.4) is the OP and the independent variables are the GEPU, the GP and the US three-month TB rate. Likewise, the dependent variable in (3.5) is the FSI and the dependent variables are the OP and the US three-month TB rate. The first difference between the variables is represented by Δ; ψ and βi denote the long-run relationship; and μ and τi show the short-run relationship between the variables. The + and - signs on the right-hand side of each equation refer to the partial sum of positive and negative changes.

For the regression (3.4),

  GEPUt+=i=1tΔGEPUi+=i=1tmax(ΔGEPUi,0),GEPUt-=i=1tΔGEPUi-=i=1tmin(ΔGEPUi,0),GPt+=i=1tΔGPi+=i=1tmax(ΔGPi,0),GPt-=i=1tΔGPi-=i=1tmin(ΔGPi,0),TBt+=i=1tΔTBi+=i=1tmax(ΔTBi,0),TBt-=i=1tΔTBi-=i=1tmin(ΔTBi,0).}   (3.6)

For the regression (3.5),

  OPt+=i=1tΔOPi+=i=1tmax(ΔOPi,0),OPt-=i=1tΔOPi-=i=1tmin(ΔOPi,0),TBt+=i=1tΔTBi+=i=1tmax(ΔTBi,0),TBt-=i=1tΔTBi-=i=1tmin(ΔTBi,0),}   (3.7)

based on Shin et al (2014). For the NARDL bound-test procedure, the following hypotheses should be tested.

For the regression (3.4),

  ψ=β1+=β1-=β2+=β2-=β3+=β3-=0.   (3.8)

For the regression (3.5),

  ψ=β1+=β1-=β2+=β2-=0.   (3.9)

To check hypotheses (3.8) and (3.9), the F-statistic of the Wald test is employed.

Moreover, to test the presence of long-run normalities, the following hypothesis (long-run symmetry) is tested:

  π+=π-,   (3.10)

where π+=-βj+/ψ and π-=-βj-/ψ.

4 Empirical results

In this section, we present the empirical findings from models 1 and 2. However, the results for nonstationary variables are checked based on the augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests (Dickey and Fuller 1979; Phillips and Perron 1988). The result of the ADF and PP tests (displayed in Tables 3 and 4) indicate that GEPU and FSI are stationary at the 1% significance level, while OP, GP and TB are stationary at the first difference in both models, which means that these variables are not integrated at the second-order difference.

Table 1: Descriptive statistics (model 1). [Obs., observations. SD, standard deviation.]
Variable Obs. Mean SD Max Min
OP 283 055.617 028.4070 0133.930 011.2800
GEPU 283 119.34 060.0482 0411.959 047.2237
GP 283 872.51 494.301 1840.80 256.197
TB 283 002.01 001.97 0006.17 000.01
Table 2: Descriptive statistics (model 2). [Obs., observations. SD, standard deviation.]
Variable Obs. Mean SD Max Min
FSI 452 101.09 00.76047 105.890 99.5425
OP 452 040.940 27.7364 133.930 11.2800
TB 452 004.549 03.676 016.30 00.01
Table 3: Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) test results for nonstationarity of variables (model 1). [***, ** and * denote statistical significance at the 1%, 5% and 10% levels, respectively.]
  ADF (test statistic) PP (test statistic)
     
    Intercept   Intercept
Variables Intercept trend Intercept trend
GEPU 0-0.745 0-3.715** 0-2.005 0-4.311***
ΔGPEU -13.286*** -13.286*** -29.031*** -32.448***
GP 00.180 0-1.686 00.131 0-1.754
ΔGP -13.449*** -13.491*** -13.528*** -13.558***
OP 0-2.508 0-2.565 0-2.102 0-2.046
ΔOP -11.204*** -11.210*** -10.735*** -10.636***
TB 0-1.937 0-2.266 0-1.766 0-2.103
ΔTB 0-6.055*** 0-6.058*** -11.644*** -11.630***
Table 4: Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) test results for nonstationarity of variables (model 1). [***, ** and * denote statistical significance at the 1%, 5% and 10% levels, respectively.]
  ADF (test statistic) PP (test statistic)
     
    Intercept   Intercept
Variables Intercept trend Intercept trend
FSI 0-4.912*** 0-4.934*** 0-7.607*** 0-7.619***
ΔFSI -11.400*** -11.401*** -61.918*** -66.763***
OP 0-2.346 0-3.119 0-1.865 0-2.340
ΔOP -13.922*** -13.906*** -13.558*** -13.540***
TB 0-2.369 0-5.401 0-1.695 0-3.442**
ΔTB 0-4.142*** 0-4.200*** -14.516*** -14.500***

To test the nonlinear co-integration based on the bound-test procedure, the Wald test (Table 5) is used to check the long-run relationship between the dependent and independent variables in both models. The result of the F-statistic confirms the rejection of the null hypothesis of no co-integration in models 1 and 2. Based on the presented finding, it is possible now to investigate the long-run and short-run asymmetric impacts of independent variables on the dependent variable in both models.

Table 5: Bound-test procedure result (Wald test). [*** denotes statistical significance at the 1% level.]
Model Co-integration hypothesis ?-statistic
1 ψ=β1+=β1-=β2+=β2-=β3+=β3-=0 5.229***
2 ψ=β1+=β1-=β2+=β2-=0 8.705***

We conduct diagnostic tests such as autocorrelation, heteroscedasticity and normality tests to check the standard regression assumption. Table 6 presents the diagnostic test results. The findings indicate that it is not possible to reject the null hypothesis for each test; therefore, we should accept the null hypothesis. This means that

  • there is no serial correlation between error terms,

  • error variance is homoscedastic and

  • the residual has a normal distribution (there is no violation of the standard regression assumption).

To find the optimal lag length in this study, the Schwartz information criterion was employed.

Table 6: Diagnostic tests. []
(a) Serial correlation (LM test)
  Model 1 Model 2
F-test 0.3241 0.823
p-value 0.7233 0.9914
(b) Heteroscedasticity (White test)
  Model 1 Model 2
F-test 0.1462 0.0891
p-value 0.5125 0.7130
(c) Normality
  First Second
  model model
χ2 0.464 0.291
p-value 0.9479 0.552

Table 7 reports the findings of (3.4). For model 1, the results imply that the OP is significantly affected by the GEPU and is influenced by economic uncertainty in the long run but not in the short run. We found that the positive and negative changes in GEPU lead to a significant effect on OPs in the long run. While positive changes in GEPU lead to a negative effect on OPs in the long run, negative changes in GEPU also lead to a positive effect on OPs in the long run. Further, while OPs appear to increase significantly in response to positive fluctuations in GPs in the long run, OPs appear to decrease significantly in response to negative fluctuations in GPs in the long run. This only means that a decrease in GPs will lead to a larger decrease in OPs, compared with a similar-sized increase in GPs leading to a larger increase in OPs. As for the impact of the three-month TB, OPs respond positively to positive TB changes in the long term. Although the Wald results in Table 8 confirm the asymmetric effects of positive and negative changes in both GP and three-month TB, there is a symmetrical impact of negative and positive fluctuations in GEPU. The Wald test for the short-run symmetry is not employed in this research as there are no statistically significant short-run estimators found in our results.

Table 7: NARDL estimated results (model 1). [***, ** and * denote statistical significance at the 1%, 5% and 10% levels, respectively. The independent variable is ΔOPt.]
Dependent   Standard  
variables Coefficient error ?-statistic
C 2.182788 0.975410 2.237817**
OPt-1 -0.143300 0.024307 5.895459***
GEPUt-1+ -0.023736 0.010479 2.264959**
GEPUt-1- -0.026219 0.013074 2.005433**
GPt-1+ 0.006958 0.001883 3.694009***
GPt-1- 0.011514 0.003076 3.743235***
TBt-1+ 0.796745 0.408931 1.948358*
TBt-1- 0.046564 0.215590 0.2155985
ΔOPt-i 0.391869 0.059254 6.613420
ΔGEPUt-i+ 0.024291 0.020948 1.159564
ΔGEPUt-i- -0.026202 0.023837 1.099186
ΔGPt-i+ -0.011810 0.013994 0.843915
ΔGPt-i- -0.016885 0.017426 0.968969
ΔTBt-i+ 4.264888 4.315765 0.988211
ΔTBt-i- -7.133425 2.266097 3.147891
Table 8: Wald test for presence of the long run symmetry (model 1). [** denotes statistical significance at the 5% level.]
Hypothesis ?-statistic
-β1+/ψ=-β1-/ψ 0.2821
-β2+/ψ=-β2-/ψ 4.800**
-β3+/ψ=-β3-/ψ 5.267**

Table 9 reports the findings of (3.5). For model 2, the results specify that negative OP changes lead to a decrease in financial stress in the long run. Conversely, negative OP fluctuations lead to an increase in financial stress in the short run. These findings indicate that the direction of the relationship is not identical within the long and short time horizons. Moreover, the positive and negative movements in TB did not influence the financial stress in the long run, whereas positive changes in monetary policy have significant asymmetric impacts on the FSI. The Wald results in Table 10 present the long-run symmetries in model 2.

Table 9: NARDL estimated results (model 2). [***, ** and * denote statistical significance at the 1%, 5% and 10% levels, respectively. The independent variable is ΔFSIt.]
Dependent   Standard  
variables Coefficient error t-statistic
C 22.77578 3.827158 5.951097***
FSIt-i -0.228965 0.038140 -6.003313***
OPt-i+ 0.001274 0.001117 1.140321
OPt-i- 0.002614 0.001321 1.978737**
TBt-i+ 0.019539 0.016720 1.168639
TBt-i- 0.003846 0.013911 0.276450
ΔFSIt-i -0.196925 0.046793 -4.208465
ΔOPt-i+ -0.010055 0.011072 -0.908134
ΔOPt-i- -0.024226 0.008604 -2.815706***
ΔTBt-i+ 0.178547 0.092614 1.927874**
ΔTBt-i- 0.057553 0.067189 0.856585
Table 10: Wald test for the presence of long-run symmetry (model 2). [*** denotes statistical significance at the 1% level.]
Hypothesis ?-statistic
-β1+/ψ=-β1-/ψ 9.45***
-β2+/ψ=-β2-/ψ 6.72***

5 Discussion and conclusion

The results of this study reinforce the relationship between GEPU and OPs. We found a significant relationship between GEPU and OPs in the long run. That said, the oil market also contributes significantly to the symmetrical relationship between the negative and positive effects. Economic uncertainty, as elaborated on by Baker et al (2016), can influence economic booms and busts; thus, it comes as no surprise that OPs, as a major indicator of economic conditions, are affected by GEPU. The results that are attributed to the impacts of GEPU – or, in other words, the effects of economic uncertainty among the top 21 economies in the world, which constitute 70% of global GDP – are most strongly felt in OPs. Yang (2019) believes that the correlation between GEPU and OPs occurs because the latter have uncertain outcomes. However, our results show that the impact of GEPU occurs only in the long run; thus, their correlation can be attenuated in the short run.

Our findings on the relationship between GPs and OPs are consistent with previous empirical studies (Tiwari et al 2020; Chen and Xu 2019; Charlot and Marimoutou 2014; Ewing and Malik 2013). While the positive fluctuations in GPs have a significant and symmetrical impact on OPs, the latter are affected more by negative GP changes. Nevertheless, the results corroborate the long-run relationship between the two variables. Mo et al (2018) and Das et al (2019) stated that, during times of economic and political uncertainty, gold, as a safe haven, can help diversify investment portfolios. Therefore, the nexus between the OP and GP changes, during periods of economic uncertainty, can be used as a proxy for measuring the financial crisis alarm index. For the relationship between the three-month TB rate (as a proxy of the interest rate or monetary policy) and OPs, findings indicate that the positive effects of TB as a sign of tightening monetary policy (either conventional or nonconventional) have a significant influence on OPs in the long run. It seems OPs act as an inflation hedge in times of OP volatility or during economic uncertainty. Although various studies have assessed monetary policy in situations of uncertainty (Ferrero et al 2019; Kimura and Kurozumi 2007), the question of what actually represents optimal monetary and fiscal policy in extreme situations, such as in times of pandemics, remains unanswered.

In the second model, a longer time frame of 31 years was employed, with the US FSI used as a dependent variable. Püttmann (2018) developed this index based on the frequency of reporting of 11 topics such as gold or silver, inflation, stocks, trade, etc, in five major US newspapers. Our findings on the OP effects on the FSI are fairly revealing. While declining OPs (negative changes) have a negative impact on the FSI in the long run, the impact is positive over the short time horizon. Thus, lower pressure on the supply side translates into a positive sign for the stimulation of aggregate demand in the long run. However, in the short run, the decline in OPs is interpreted as a sign of recession; as a result, the FSI responds positively to OP negative changes. What is more, the results explicitly show the positive impacts of the three-month TB rates on the FSI in the short run. A tightening of monetary policy has a positive impact on US FSIs. This seems to indicate that central banks adopting contractionary monetary policy is perceived by markets as a negative signal.

In conclusion, it is certainly true to say that there is a close relationship between OPs and GEPU. The results of this paper highlight the critical impacts of uncertainty, GPs and monetary policy on OP volatility while simultaneously showing the effects of OPs on the FSI.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

References

  • Apergisa, N., Chatziantonioub, I., and Cooray, A. (2020). Monetary policy and commodity markets: unconventional versus conventional impact and the role of economic uncertainty. International Review of Financial Analysis 71, Paper 101536, (https://doi.org/10.1016/j.irfa.2020.101536).
  • Arouri, M., Estay, C., Rault, C., and Roubaud, D. (2016). Economic policy uncertainty and stock markets: long-run evidence from the US. Finance Research Letters 18, 136–141 (https://doi.org/10.1016/j.frl.2016.04.011).
  • Baker, S. R., Bloom, N., and Davis, S. J. (2016). Measuring economic policy uncertainty. Quarterly Journal of Economics 131(4), 1593–1636 (https://doi.org/10.1093/qje/qjw024).
  • Bildirici, M. E., and Turkmen, C. (2015). Nonlinear causality between oil and precious metals. Resources Policy 46(2), 202–211 (https://doi.org/10.1016/j.resourpol.2015.09.002).
  • Brown, S., and Yucel, M. (2002). Energy prices and aggregate economic activity: an interpretative survey. Quarterly Review of Economics and Finance 42(2), 193–208 (https://doi.org/10.1016/S1062-9769(02)00138-2).
  • Charlot, P., and Marimoutou, V. (2014). On the relationship between the prices of oil and the precious metals: revisiting with a multivariate regime-switching decision tree. Energy Economics 44, 456–467 (https://doi.org/10.1016/j.eneco.2014.04.021).
  • Chen, R., and Xu, J. (2019). Forecasting volatility and correlation between oil and gold prices using a novel multivariate GAS model. Energy Economics 87, 379–391 (https://doi.org/10.1016/j.eneco.2018.11.011).
  • Das, D., Le Roux, C., Jana, R. K., and Dutta, A. (2019). Does Bitcoin hedge crude oil implied volatility and structural shocks? A comparison with gold, commodity and the US dollar. Finance Research Letters 36, Paper 101335 (https://doi.org/10.1016/j.frl.2019.101335).
  • Davis, S. J. (2016). An index of global economic policy uncertainty. Working Paper 22740, October, National Bureau of Economics Research (https://doi.org/10.3386/w22740).
  • Dickey, D. A., and Fuller, W. A. (1979). Distribution of the estimators for autoregressive time series with a unit root. Journal of the American Statistical Association 74(366), 427–431 (https://doi.org/10.2307/2286348).
  • Ersan, O., Akron, S., and Demir, E. (2019). The effect of European and global uncertainty on stock returns of travel and leisure companies. Tourism Economics 25(1), 51–66 (https://doi.org/10.1177/1354816618792398).
  • Ewing, B. T., and Malik, F. (2013). Volatility transmission between gold and oil futures under structural breaks. Internatinal Review of Economics and Finance 25, 113–121 (https://doi.org/10.1016/j.iref.2012.06.008).
  • Ferrero, G., Pietrunti, M., and Tiseno, A. (2019). Benefits of gradualism or costs of inaction? Monetary policy in times of uncertainty. Working Paper 1205, February 4, Bank of Italy, Social Science Research Network (https://doi.org/10.2139/ssrn.3431133).
  • Gospodinov, N., and Jamali, I. (2015). The response of stock market volatility to futures based measures of monetary policy shocks. International Review of Economics and Finance 37, 42–54 (https://doi.org/10.1016/j.iref.2014.11.001).
  • Jones, P. M., and Olson, E. (2013). The time-varying correlation between uncertainty, output, and inflation: evidence from a DCC-GARCH model. Economics Letters 118(1), 33–37 (https://doi.org/10.1016/j.econlet.2012.09.012).
  • Kanjilal, K., and Ghosh, S. (2017). Dynamics of crude oil and gold price post 2008 global financial crisis: new evidence from threshold vector error-correction model. Resources Policy 52, 358–365 (https://doi.org/10.1016/j.resourpol.2017.04.001).
  • Kim, Y., Lim, H., and Sohn, W. (2020). Which external shock matters in small open economies? Global risk aversion versus US economic policy uncertainty. Japan and the World Economy 54, Paper 101011 (https://doi.org/10.1016/j.japwor.2020.101011).
  • Kimura, T., and Kurozumi, T. (2007). Optimal monetary policy in a micro-founded model with parameter uncertainty. Journal of Economic Dynamics and Control 31(2), 399–431 (https://doi.org/10.1016/j.jedc.2005.10.003).
  • Kocaarslan, B., Soytas, M. L., and Soytas, U. (2020). The asymmetric impact of oil prices, interest rates and oil price uncertainty on unemployment in the US. Energy Economics 86, Paper 104625 (https://doi.org/10.1016/j.eneco.2019.104625).
  • Kormilitsina, A. (2011). Oil price shocks and the optimality of monetary policy. Review of Economic Dynamics 4(1), 199–223 (https://doi.org/10.1016/j.red.2010.11.001).
  • Kumar, S. (2017). On the nonlinear relation between crude oil and gold. Resources Policy 51, 219–224 (https://doi.org/10.1016/j.resourpol.2017.01.003).
  • Lee, Y., Huang, Y., and Yang, H. (2012). The asymmetric long-run relationship between crude oil and gold futures. Global Journal of Business Research 6(1), 9–15.
  • Li, X., and Peng, L. (2017). US economic policy uncertainty and co-movements between Chinese and US stock markets. Economic Modelling 61, 27–39 (https://doi.org/10.1016/j.econmod.2016.11.019).
  • Li, X. M., Zhang, B., and Gao, R. (2015). Economic policy uncertainty shocks and stock–bond correlations: evidence from the US market. Economics Letters 132, 91–96 (https://doi.org/10.1016/j.econlet.2015.04.013).
  • Liu, Z., Ding, Z., Lv, T., Wu, J. S., and Qiang, W. (2019). Financial factors affecting oil price change and oil–stock interactions: a review and future perspectives. Natural Hazards 95(1–2), 207–225 (https://doi.org/10.1007/s11069-018-3473-y).
  • Mensi, W., Hammoudeh, S., and Hoon, S. (2015). Precious metals, cereal, oil and stock market linkages and portfolio risk management: evidence from Saudi Arabia. Economic Modelling 51, 340–358 (https://doi.org/10.1016/j.econmod.2015.08.005).
  • Mo, B., Nie, H., and Jiang, Y. (2018). Dynamic linkages among the gold market, US dollar and crude oil market. Physica A 491, 984–994 (https://doi.org/10.1016/j.physa.2017.09.091).
  • Mokni, K., Hammoudeh, S., Ajmi, A. N., and Youssef, M. (2020). Does economic policy uncertainty drive the dynamic connectedness between oil price shocks and gold price? Resources Policy 69, Paper 101819 (https://doi.org/10.1016/j.resourpol.2020.101819).
  • Nguyen, C. P., Ha Le, T., and Su, T. D. (2020). Economic policy uncertainty and credit growth: evidence from a global sample. Research in International Business and Finance 51, Paper 101118 (https://doi.org/10.1016/j.ribaf.2019.101118).
  • Pesaran, H. H., and Shin, Y. (1998). Generalized impulse response analysis in linear multivariate models. Economic Letters 58, 17–29 (https://doi.org/10.1016/S0165-1765(97)00214-0).
  • Pesaran, M., Shin, Y., and Smith, R. J. (2001). Bounds testing approaches to the analysis of level relationships. Journal of Applied Econometrics 16, 289–326 (https://doi.org/10.1002/jae.616).
  • Phan, D. H. B., Sharma, S. S., and Tran, V. T. (2018). Can economic policy uncertainty predict stock returns? Global evidence. Journal of International Financial Markets, Institutions and Money 55, 134–150 (https://doi.org/10.1016/j.intfin.2018.04.004).
  • Phillips, P. C. B., and Perron, P. (1988). Testing for a unit root in time series regression. Biometrika 75(2), 335–346 (https://doi.org/10.1093/biomet/75.2.335).
  • Püttmann, L. (2018). Patterns of panic: financial crisis language in historical newspapers. Working Paper, April 27, Social Science Research Network (https://doi.org/10.2139/ssrn.3156287).
  • Rosa, C. (2014). The high-frequency response of energy prices to monetary policy: understanding the empirical evidence. Energy Economics 45, 295–303 (https://doi.org/10.1016/j.eneco.2014.06.011).
  • Shin, Y., Yu, B., and Greenwood-Nimmo, M. (2014). Modelling asymmetric co-integration and dynamic multipliers in an ARDL framework. In Festschrift in Honor of Peter Schmidt, Horrace, W. C., and Sickles, R. C. (eds), pp. 281–314. Springer (https://doi.org/10.1007/978-1-4899-8008-3_9).
  • Shoag, D., and Veuger, S. (2016). Uncertainty and the geography of the great recession. Journal of Monetary Economics 84, 84–93 (https://doi.org/10.1016/j.jmoneco.2016.11.002).
  • Teetranont, T., Chanaim, S., Yamaka, W., and Sriboonchitta, S. (2018). Investigating relationship between gold price and crude oil price using interval data with copula based GARCH. In International Conference of the Thailand Econometrics Society: Predicting Econometrics and Big Data, pp. 656–669. Studies in Computational Intelligence, Volume 753. Springer (https://doi.org/10.1007/978-3-319-70942-0_47).
  • Tiwari, A. K., Aye, G. C., Gupta, R., and Gkillas, K. (2020). Gold–oil dependence dynamics and the role of geopolitical risks: evidence from a Markov-switching time-varying copula model. Energy Economics 88, Paper 104748 (https://doi.org/10.1016/j.eneco.2020.104748).
  • Wang, Y., Chen, C. R., and Huang, Y. S. (2014). Economic policy uncertainty and corporate investment: evidence from China. Pacific-Basin Finance Journal 26, 227–243 (https://doi.org/10.1016/j.pacfin.2013.12.008).
  • Yang, L. (2019). Connectedness of economic policy uncertainty and oil price shocks in a time domain perspective. Energy Economics 80, 219–233 (https://doi.org/10.1016/j.eneco.2019.01.006).
  • Zhang, Y. J., and Wei, Y. M. (2010). The crude oil market and the gold market: evidence for co-integration, causality and price discovery. Resources Policy 35(3), 168–177 (https://doi.org/10.1016/j.resourpol.2010.05.003).
  • Ziaei, S. M. (2012). Effects of gold price on equity, bond and domestic credit: evidence from ASEAN + 3. Procedia: Social and Behavioral Sciences 40, 341–346 (https://doi.org/10.1016/j.sbspro.2012.03.197).
  • Ziaei, S. M. (2021). The impact of corporations and banking system leverage on renewable energy: evidence from selected OECD countries. Renewable Energy Focus 37, 68–77 (https://doi.org/10.1016/j.ref.2021.04.002).

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here