Sensitivity of economic policy uncertainty to investor sentiment: Evidence from Asian, developed and European markets

Mobeen Ur Rehman (Informetrics Research Group, Ton Duc Thang University, Ho Chi Minh City, Vietnam and Faculty of Social Sciences and Humanities, Ton Duc Thang University, Ho Chi Minh City, Vietnam)
Nicholas Apergis (Department of Banking and Financial Management, University of Piraeus, Athens, Greece and School of Business, Law and Social Sciences, University of Derby, Derby, UK)

Studies in Economics and Finance

ISSN: 1086-7376

Article publication date: 19 June 2019

Issue publication date: 21 June 2019

Abstract

Purpose

This paper aims to explore the impact of investor sentiments on economic policy uncertainty (EPU). The analysis also considers the momentum effect, stock market returns volatility and equity pricing inefficiencies across markets, which, to the best of the authors’ knowledge, has not been addressed in the literature. The role of these control variables has collectively been considered to have important behavioral implications for international investors

Design/methodology/approach

Quantile regressions are used for estimation purpose, as it provides robust and more efficient estimates rather than those coming from the traditional regression model.

Findings

The momentum effect is negative and significant only at higher quantiles, while oil prices are positive and significant across all quantiles. The exchange rate exerts a negative and significant effect on EPU, whereas equity price volatility (i.e. investor sentiment) exerts a negative and significant impact on EPU in most of the quantiles.

Research limitations/implications

The results have important implications for international investors and policymakers, especially in terms of the breakdown of economic policy uncertainty across different sample markets. The breakdown of complete sample period into sub-samples acts as a robust analysis and documents the similarity of the results for the Asian and developed markets cases, but not in the case of the European markets.

Practical implications

The findings imply the importance of financial stability that impacts the accumulation of systemic risks and adds smoothness to the financial cycle in particular geographical areas.

Originality/value

The contribution of this paper is threefold. First, existing literature highlights and empirically tests the impact of economic policy uncertainty on different market, macro-economic and global control variables. The analysis, however, performs it in the reverse order, i.e. analyzing the impact of the momentum effect (investor sentiment variables), equity market inefficiencies and volatility (market variables) and exchange rates and Brent oil (control variables). Second, to check the sensitivity of economic policy uncertainty, the analysis analyzes a wide range of markets, segregated as emerging, developed and European regions over the sample period to generate region-wise implications. Finally, the analysis explores the relationship of aforementioned variables with economic policy uncertainty keeping in view the non-linear structure and prior evidence and investor sentiments and economic policy uncertainty in the regression model.

Keywords

Citation

Rehman, M.U. and Apergis, N. (2019), "Sensitivity of economic policy uncertainty to investor sentiment: Evidence from Asian, developed and European markets", Studies in Economics and Finance, Vol. 36 No. 2, pp. 114-129. https://doi.org/10.1108/SEF-01-2019-0040

Publisher

:

Emerald Publishing Limited

Copyright © 2019, Emerald Publishing Limited


1. Introduction

A series of various economic and financial crises, followed by steep declines and slow recoveries, raised significant concerns on the role of regulatory bodies regarding uncertainty in economic policies. Such uncertainties in economic policies are documented in the form of different proxies. According to Arouri et al. (2014), economic policy uncertainty defines limitations to certain parameters, thereby causing delays in economic activity and decision making, attributable to high uncertainty levels. However, reactions to any changes in economic uncertainty are weak if anticipated earlier. Increasing levels of uncertainty not only causes firms to delay their investments decisions, but also results in the reversion of preliminary spending patterns (Leduc and Liu, 2016; Pastor and Veronesi, 2012). Though a strand of literature provides evidence of the sensitivity of economic policy uncertainty to different macro-economic variables (Ali, 2001; Jones and Olson, 2013), economic policy uncertainty exhibit significant relationship with household savings (Giavazzi and McMahon, 2012), delays in firms’ entry (Handley and Limao, 2015), equity market volatility (Pastor and Veronesi, 2012), returns on assets (Brogaard and Detzel, 2015) and international oil prices (Arouri et al., 2014).

The effect of policy uncertainty on real economy has been an important topic of discussion over the past decade with a specific interest after the turbulence in global financial markets caused by the financial crisis of 2008-2009. Literature highlights different definitions and interpretations of economic policy uncertainty (EPU). For example, Bams et al. (2017) suggest that economic uncertainty arises from various sources, i.e. future equity returns, uncertain future stock prices and inflationary uncertainty. On the other hand, few studies measure and interpret economic policy uncertainty as an index comprising of different factors (Colombo, 2013; Antonakakis et al., 2013; Klobner and Sekkel, 2014; Karnizova and Li, 2014). A rather contemporary and comparatively new proxy proposed by Baker et al. (2016) measuring economic uncertainty has intensified over the past couple of decades. Existing literature documents the effect of policy uncertainty on macro-economic variables (Bachmann et al., 2013) however sensitivity of economic policy uncertainty to macro-economic and financial variables has rarely been explored. The link between the macro-economic variables and equity markets has an intuitive appeal. According to Chau et al. (2016), higher market volatility index (VIX) index causes fear and uncertainty, whereas lower values signal complacency and bullish market behavior. Gulen and Ion (2015) study the effect of EPU on both firms and industry level capital investments and find that it can affect investments. Liu and Zhang (2015) report EPU to be a decent predictor of stock market volatility. Investor sentiments induce large swings in business cycles; however, understanding its ability to induce changes in EPU helps policymakers in predicting and minimizing such fluctuations. During the past couple of decades, investor sentiments gained investor and policymakers attention in two aspects. First, is the advent of many proxies measuring investor sentiments, for example, the seminal work of Baker and Wurgler (2006) introduce investor sentiment index comprising of six proxies[1], following other proxies like AAII (American Association of Investor Sentiment Index) and consumer sentiment index. Second, a strand of recent literature examines the impact of such sentiment indices on different macro-economic and financial variables with evidences of significant relationship (Apergis et al., 2018). Such studies provide us an avenue of investigating the impact of investor sentiments on economic policy uncertainty. Furthermore, our proxies of investor sentiments include momentum effect of stock markets among others which comes under the herding behavior theory of behavioral finance.

Most of the existing studies use linear frameworks while analyzing the relationship between economic policy uncertainty and different macro-economic and financial variables given the fact that non-linear nature of both investor sentiments and economic policy uncertainty is also reported in literature. The residuals of these variables when modeled in a framework tend to exhibit non-linear behavior. To measure such non-linear relationship, quantile regression provides a clear way of understanding how the relationship among direct and conditioning variables or risk factors changes across distribution of conditional returns. Quantile regressions identify points in the conditional distribution where omitted variables have favorably and/or unfavorably influencing role. We may think of these omitted factors as representing idiosyncratic shocks, or as the receipt of bad news during the sample period by firms located in the lower quantiles. By using quantile regression, one can explore whether the underlying characteristics of firms’ market prices are consistent, given different degrees of good versus bad news. Such idiosyncratic shocks likely influence idiosyncratic volatility, which Campbell et al. (2001) and Goyal and Santa-Clara (2002) find to be a major component of the volatility of individual stock returns.

Our contribution in this paper is as follows. First, we investigate the impact of investor sentiments on economic policy uncertainty. This relationship has not been explored in the existing literature and therefore can curtail important implications for investors and policy makers. Investor sentiments have the ability to influence bearish and bullish market conditions and therefore can also induce uncertainty in economic policies, since the stock markets act as economic indicators. Though the effect of economic policy uncertainty on investor sentiments is documented in existing literature, the direction of relationship from investor sentiments to EPU presents itself as a new avenue. Secondly, we include a set of proxies to measure investor sentiments (i.e. momentum effect, equity market inefficiencies and stock market volatility) along with a combination of macro-economic variables (exchange rates and Brent oil prices) with strong theoretical justification in existing literature. The role of these control variables is collectively considered to have important behavioral implications for international investors. Third, to investigate the sensitivity of economic policy uncertainty, we analyze a wide range of markets, segregated as emerging, developed and European regions over the sample period to generate region-wise implications. Finally, our study takes a step forward by empirically analyzing (using quantile regression) the relationship between economic policy uncertainty and investor sentiments in the presence of macro-economic control variables keeping in view the non-linear structure in accordance with prior evidence of such non-linear relationship.

The results for quantile regressions provide robust and more efficient estimates rather than those coming from the traditional regression model. The momentum effect is negative and significant only at higher quantiles, while oil prices are positive and significant across all quantiles. The exchange rate exerts a negative and significant effect on EPU, whereas equity price volatility (i.e. investor sentiment) exerts a negative and significant impact on EPU in most of the quantiles.

Rest of the paper is structured as follows. Section 2 presents data and methodology. Section 3 provides data analysis followed by Section 4 concluding the results.

2. Data and methodology

The proposed model consists of economic policy uncertainty as the dependent variable, sensitive to the momentum effect, equity pricing volatility and pricing differences in the presence of international Brent oil prices and spot exchange rates. EPU represents the economic risk for a country, because of an uncertain path of government policy, leading towards an escalating risk premium and causing delays in individual and business spending until the uncertainty resolves. This EPU can interchangeably refer towards fiscal or monetary policy uncertainty, uncertain electoral outcomes, or tax regimes. Data on EPU are based on three main components: the newspaper coverage of economic uncertainty in relevance to policy issues, the provision set for the federal tax code for future years, and the disagreement across economic forecasters. China, India and Japan represent the major Asian economies whereas, for the case of developed economies, indices from the US, the UK and Europe (represented by a composite index) are used. Data for economic policy uncertainty are sourced from www.policyuncertainty.com/. The analysis uses the Market Integration index-MI (Connor and Korajczyk, 1989), which captures pricing differences across equity markets based on systematic risks across countries. They postulate that pricing errors, represented by an intercept term in the International Capital Asset Pricing Model, measure the extent of market segmentation; if all assets are priced according to their similar systematic risk, there is a perfect integration across stock markets and the value of the intercept equals zero. Pricing errors increase with higher official barriers, transaction costs and taxes to international asset trading. The MI index [equation (2) as an extension of equation (1)]; therefore, measures equity pricing differences across markets, represented by an absolute value of the intercept as:

(1) Ri,tRFi,t=αi,t+βi,tRWi,tRFi,t+εi,t
(2) Ri,t=αi,t+βi,tRWi,t+εi,t
where Ri,t is the returns on international indices, RFi,t is the risk free rate, β represents the systematic risk of the market proxied by S&P1200. The absolute value of α proxies pricing differences of equity returns. A zero value implies no mispricing. For equity market volatility, we source daily pricing data to take monthly averages and calculate as [logPt-logPt-1], where P denotes equity prices. The momentum effect is calculated by the stock markets’ trading volume. Data on equity prices for volatility and the stocks trading volume for the momentum effect are sourced from Datastream. Exchange rates data are sourced from the IMF and are relative to the US dollar. Global oil prices are closing spot prices (converted to a monthly basis by taking their daily average) measured as the West Texas Intermediate (WTI) crude oil prices. Data span the period 1995-2018 and are on a monthly basis.

Figure 1 represents the economic policy uncertainty trend over the sample period for three markets segments, i.e. Asian, Developed and European markets, respectively. Among Asian emerging markets, Figure 1 highlights more turbulence in the economic policy uncertainty of India as compared to China throughout the period. We witness that the economic policy uncertainty in India exhibit more sensitivity to the global financial crisis of 2008-2009. Japan however, among the developed economies exhibit similar sensitivity to the global financial crisis of 2008-2009. Economic policy uncertainty trend (among developed economies) between the US and the UK, with UK exhibiting a more inconsistent pattern, especially after 2000. Finally, all the European markets follow a comparatively inconsistent behavior. Overall picture suggests that the economic policy uncertainty is quite high and evident in most of our sampled countries. This uncertainty is not associated with the global financial crisis of 2008-2009 rather it maintains its turbulent behavior across the sample period. Few uncertainty declines are also visible for the case of Spain in 2003, which, however, follows a normal course ahead. Table I provides descriptive statistics for selected sub-sampled as well as complete markets. We witness that most of the series exhibit normality under complete sampling and sub-sampling periods. Furthermore, most of the series are positively skewed across different sample selection.

Table II presents the panel unit root properties of the model variables, with the results indicating various degrees of stationarity across the variables included in the modeling approach, as well as across the countries included in the sample analysis. Next, the analysis explains the suitability of the quantile regression model by first testing the presence of non-linearity in economic policy uncertainty using the Brock, Dechert, Scheinkman and LeBaron (BDS) test (Brock et al., 1996). The results are reported in Table III and provide strong evidence of non-linearity for different embedding dimensions of the BDS test. The results suggest that linear regression models might not be able to capture the sensitivity of economic policy uncertainty to included variables (i.e. the momentum effect, equity pricing volatility, pricing differences, Brent oil prices and spot FX rates) and, therefore, it could be a necessity to use the contemporary quantile regression testing methodology.

3. Data analysis

To analyze the sensitivity of economic policy uncertainty to the momentum effect, equity pricing volatility and pricing differences in the presence of international Brent oil prices and spot exchange rates, the empirical analysis focuses not only on the conditional mean, but also on the tails of the conditional distribution by estimating through a quantile regression framework. To this end, the analysis makes use of both OLS (mean results) and the quantile regression methodology, introduced by Koenker and Bassett (1978), as the relationship between economic policy uncertainty and the momentum effect, equity pricing volatility, pricing differences needs not be the same across the conditional distribution of oil returns (Du et al., 2015), especially in the presence of international Brent oil prices and spot exchange rates. The advantage of this methodological approach is that it is a semi-parametric method, which does not make any pre-suppositions about the parametric distribution of the error process. The τth conditional quantile is defined as the value Qτ(yt|yt-1, …, yt-q), such that the probability that economic policy uncertainty is conditional on its determinants will be less than Qτ(yt|yt-1, …, yt-q) is τ. By estimating at different quantiles, τ ε(0,1) we can get a set of estimates of the impact of determinants in different quantiles, running from 0.10 to 0.90.

Table IV reports both the mean and the quantile regression results for the full sample at different quantiles (0.10-0.90). Estimations based on the entire distribution focus on the mean and information about the tails of the distribution is lost. By contrast, quantile regressions provide robust and more efficient estimates. The momentum effect is negative and significant; pricing differences are positive and significant, only at higher quantiles. Among the control variables, the exchange rates remain insignificant across the majority of the quantiles. Stock price volatility (i.e. investor sentiment) exerts a negative and significant impact on EPU across all quantiles, except in the 0.90 quantile, with the overall distribution estimates highlighting the effect as statistically significant. WTI Crude oil prices are positive and significant across all quantiles, suggesting a major influence on economic policy uncertainty. These results suggest that the levels of economic policy uncertainty are sensitive to the raising global oil prices in the full sample; however, sub-sample findings are reported in the forthcoming discussion.

Table V presents the results of quantile regression for the case of Asian economies i.e. China and India. The momentum effect changes from negative in the full sample to positive in the Asian sub-sample; however, the significant coefficients are observed in the middle (0.40-0.50) and higher (0.80-0.90) quantiles. Pricing differences previously insignificant in the full country sample now become positive and significant in the case of Asian markets. The results in Table V highlight, therefore, the significance of the momentum effect, price volatility and pricing difference, but at varying quantile arrangements. Among the control variables, we can illustrate the significant coefficient values of exchange rates, along with Brent oil towards the economic policy uncertainty of Asian markets. This changing behavior of the exchange rates from the insignificant (in the full sample) towards significantly positive (the Asian market sub-sample) highlights its importance for the Asian countries that might be attributed to the higher sensitivity of their respective currencies compared with the US dollars, as well as to the increasing volume of foreign inwards remittances.

Table VI reports the results for the quantile regression framework for the case of the Developed markets sample. These results are somewhat different from the two previous country samples. They document that among all the variables, stock market volatility remains significant as before; however, the remaining variables remain insignificant. The momentum effect remains insignificant across all quantiles, while pricing differences document significant negative coefficients across the majority of the quantiles (specifically, from 0.50-0.90). Likewise, a disparity is also observed for the control variables where except in the higher quantiles, both the Brent oil and exchange rates remain insignificant. However, we can witness significant positive coefficients for the case of Brent oil and significant negative coefficients for the case of exchange rates in higher quantiles.

Finally, we report in Table VII the results for the European market sample, where different results are reported for equity pricing volatility. Unlike all other sample countries, pricing volatility exhibits an insignificant behavior across all quantiles. Similar results are also reported for the momentum effect (except for the last quantile). Equity pricing differences remain insignificant in lower quantiles; however, they exhibit a significant negative role in upper quantiles. In contrast, exchange rates render an insignificant behavior in the majority of the quantiles, except in the last two quantiles.

4. Conclusion

This paper investigated the impact of investor sentiment on EPU. We proxy investors’ sentiments through equity pricing volatility, while the momentum effect was captured through the trading volume and equity pricing inefficiencies. We also included global control factors, as WTI Crude oil and exchange rates. Keeping in view the non-linear structure of economic policy uncertainty and the current literature on behavioral finance, the analysis applied the quantile regression approach through which the evidence suggested a negative impact of investor sentiment on EPU. The analysis also divided the sampled markets into three different sub-samples, i.e. Asian, Developed and European markets. The full sample results suggested that equity price volatility remained as the most significant driver of economic policy uncertainty, along with Brent oil prices, implying that economic policy uncertainty was sensitive to international oil prices; however, exchange rate remained statistically insignificant. Pricing differences and exchange rates became significant for the case of Asian countries, along with pricing volatility. This may be attributed to the fact that the emerging markets of Asia receive huge remittance inflows from the developed countries, due to which their exchange rate variable per US dollar renders itself as a significant driver. The significance of equity pricing differences is due to the fact that the emerging status of Asian countries leads the equity pricing inefficiencies from their developed counterparts. The results from the developed markets resemble with the complete sample results, where economic policy uncertainty shows sensitivity only to equity price volatility in almost all quantiles, while both Brent oil prices and exchange rates only in higher quantiles. The results in the case of European countries sample differentiate from other markets, as equity pricing differences negatively influence economic policy uncertainty in these countries, with the remaining of the variables remaining statistically insignificant. However, Brent oil retains its position as an important determinant of economic policy uncertainty[2].

The results have important implications for international investors and policymakers, especially in terms of economic policy uncertainty across different sample markets. Our results highlight the presence of non-linear relationship between economic policy uncertainty and investor sentiments controlled by macro-economic conditions. The implications generated from increased stock market volatility and changes in investor sentiments can result in increased economic policy uncertainty thereby implying its effect on economic conditions of respective countries (Kang et al., 2017). Higher economic policy uncertainty leads to greater economic turbulence and financial riskiness that can affect not only investments but can also shake investors’ confidence in respective financial markets. Furthermore, breakdown of complete sample into further sub-samples acts as a robust analysis and documents similarity of results for the Asian and Developed markets, but not for European markets. The sensitivity of economic policy uncertainty is witnessed more in Asian economies, mainly attributable to their weak market efficiency where investor sentiments act as a more important indicator for investors decision making (Narayan and Rehman, 2017). These results are in accordance with Huerta et al. (2011) and Rehman and Shahzad (2016). Therefore, our findings imply importance of financial stability that impacts the accumulation of systemic risks and add smoothness to the financial cycle in particular geographical areas.

Figures

Economic policy uncertainty with structural break test

Figure 1.

Economic policy uncertainty with structural break test

Descriptive statistics

Statistic Economic policy uncertainty Momentum effect Price volatility Equity pricing inefficiencies Brent oil prices Exchange rates
Complete sample
Mean 1,776.39 11.526 0.543 375.675 51.385 78.432
SD 102.964 6.536 0.584 103.564 44.910 1.634
Skewness 434.013 0.234 3.854 4.829 13.029 2.605
Kurtosis 12.189 1.423 −4.584 −82.684 10.473 0.302
JB Stat. 67.431* 184.537 0.342* 79.493* 33.593* 35.674*
Emerging Asian market sample
Mean 97.012 11.993 0.563 58.941 53.695 40.454
SD 44.685 3.854 0.605 70.193 31.695 29.043
Skewness 1.602 0.059 −0.102 2.949 0.394 0.543
Kurtosis 4.604 5.584 1.104 14.594 1.594 1.305
JB Stat. 136.903* 147.830* 81.245* 36.594* 40.533* 54.534*
Developed market sample
Mean 113.534 18.434 0.053 907.42 53.695 53.534
SD 56.324 35.434 0.644 12.645 32.534 74.533
Skewness 1.324 −0.204 2.234 1.325 0.143 0.224
Kurtosis 7.109 1.204 20.645 3.705 1.324 1.435
JB Stat. 106.938* 132.534* 11.534* 22.535* 64.685* 131.534*
European market sample
Mean 343.53 4.204 0.534 137.432 52.645 120.534
SD 98.154 5.305 0.633 18.756 30.534 491.645
Skewness 2.653 1.604 1.645 1.324 0.105 4.745
Kurtosis 7.954 6.405 23.643 7.534 1.325 18.534
JB Stat. 26.406* 194.683 277.634* 12.103* 82.534* 14.645*
Notes:

Std. dev. presents standard deviation of the variables whereas JB Stat. represents Jarque Bera statistics for normality;

* presents significance level at 10 % or better

Unit root tests

Economic policy uncertainty Momentum effect Price volatility
Level 1st Differences Level 1st Differences Level 1st Differences
Countries D DT D DT D DT D DT D DT D DT
China 0.554 1.8534 2.5343 1.5343 47.5034 37.8485 0.5343 0.9434 0.6454 0.9534 34.0.65 65.9653
India 4.4534 7.9043 0.1564 0.5745 1.5343 8.9150 1.6343 4.6454 0.5656 1.3433 4.9653 4.3265
The USA 0.4242 2.2042 0.1534 0.1645 61.0522 12.9864 0.6454 0.654 37.7454 22.6454 0.8636 1.2356
The UK 1.5034 4.5043 0.6534 0.9024 8.9402 26.1975 0.2423 2.7454 43.7545 38.6353 76.2365 65.9653
Japan 0.5034 2.5343 0.7533 1.0943 8.4232 9.5893 0.1534 0.2534 24.9634 22.6454 59.9865 57.9532
Germany 0.4640 1.4953 0.4634 0.9254 2.5332 9.0581 0.5343 0.6454 31.6453 29.7432 80.9863 71.2356
Italy 18.8402 18.5342 0.6745 1.5342 1.0327 10.3226 0.1645 0.7454 3.9053 9.6453 22.3265 59.6532
France 5.6334 3.5003 0.7454 0.6442 1.8423 9.8355 0.63434 0.9434 93.7545 86.7454 17.9653 15.9865
Spain 0.3534 1.6034 0.2645 0.0532 2.2392 9.5670 0.6434 0.75354 24.9534 24.8565 70.2356 63.3265
Equity pricing
inefficiencies
Brent oil prices Exchange
rates
Level 1st Differences Level 1st Differences Level 1st Differences
D DT D DT D DT D DT D DT D DT
China 4.6454 3.7544 0.6343 1.5343 6.7454 6.4364 0.5364 1.1243 30.5343 12.7435 0.4365 1.9563
India 2.5343 6.7654 0.7454 0.6454 6.7454 6.4364 0.5364 1.1243 67.6454 6.8545 0.8032 0.8965
The USA 24.9856 5.7565 0.6454 1.5453 6.7454 6.4364 0.5364 1.1243 3.7565 12.6436 0.9653 1.3265
The UK 2.8565 4.6454 8.8543 31.6534 6.7454 6.4364 0.5364 1.1243 4.8232 16.8643 0.6598 0.2365
Japan 4.9756 3.8565 0.6453 0.8545 6.7454 6.4364 0.5364 1.1243 2.3744 9.0052 0.6532 0.8653
Germany 1.5343 3.9565 0.7454 0.8454 6.7454 6.4364 0.5364 1.1243 27.2756 12.8422 0.8956 0.2356
Italy 2.7654 4.9543 54.7655 203.5342 6.7454 6.4364 0.5364 1.1243 30.8533 16.5635 0.3265 1.5632
France 5.8656 1.7454 15.7454 57.5343 6.7454 6.4364 0.5364 1.1243 47.6356 25.6343 0.9565 0.5698
Spain 1.8664 6.7544 0.7645 0.9232 6.7454 6.4364 0.5364 1.1243 30.8565 16.7533 0.3652 0.2399
Note:

In the above table, D and DT represents unit root statistics without and with trend at level and first differenced values

BDS test statistics

Economic
policy
uncertainty
m
2 3 4 5 6
Countries
China 0.0447* (−0.0047) 0.1049* (−0.0092) 0.0969* (−0.0088) 0.0780* (−0.0074) 0.1054* (−0.0089)
India 0.0890* (−0.0042) 0.2420* (−0.0083) 0.2145* (−0.008) 0.1681* (−0.0067) 0.2592* (−0.0081)
The USA 0.1284* (−0.0049) 0.2973* (−0.0097) 0.2690* (−0.0093) 0.2152* (−0.0078) 0.3107* (−0.0094)
The UK 0.0586* (−0.0047) 0.1164* (−0.0093) 0.1128* (−0.0089) 0.0965* (−0.0075) 0.1125* (−0.009)
Europe 0.0805* (−0.0057) 0.1780* (−0.0112) 0.1662* (−0.0107) 0.1350* (−0.009) 0.1797* (−0.0108)
France 0.0433* (−0.0043) 0.0838* (−0.0084) 0.0815* (−0.0081) 0.0732* (−0.0068) 0.0770* (−0.008)
Germany 0.1029* (−0.0044) 0.2345* (−0.0087) 0.2147* (−0.0083) 0.1738* (−0.007) 0.2437* (−0.0084)
Italy 0.0670* (−0.0054) 0.1694* (−0.0107) 0.1517* (−0.0103) 0.1177* (−0.0086) 0.1762* (−0.0104)
Spain 0.0663* (−0.0052) 0.1286* (−0.0103) 0.1213* (−0.0099) 0.1061* (−0.0083) 0.1294* (−0.0099)
Japan 0.1172* (−0.0051) 0.3051* (−0.0101) 0.2698* (−0.0097) 0.2058* (−0.0081) 0.3304* (−0.0098)
Notes:

m denotes the parameter m in the embedding dimension and ε is the epsilon values. Standard errors values are reported in parenthesis;

* denotes significance level at 5% or better

Quantile regressions-full sample

Variables Mean results Q10 Q20 Q30 Q40 Q50 Q60 Q70 Q80 Q90
Intercept 89.5345* (212.5343) 40.553*** (3.745) 53.756*** (3.053) 66.325*** (3.524) 75.923*** (3.424) 81.634*** (3.634) 91.952*** (4.634) 101.745*** (4.663) 126.23*** (8.634) 167.52 (1890.534)
Momentum −88.635*** (9.854) 0.345 (0.745) 0.064 (0.189) −0.163 (0.174) −0.745* (0.426) −0.634 (0.305) −0.745 (0.363) −0.663*** (0.363) −1.834***(0.744) −793.53*** (85.634)
Price volatility −41.364 (−45.054) −11.745** (5.634) −10.593**(3.532) −12.643** (5.745) −12.952** (5.634) −16.745** (5.153) −22.732** (7.634) −28.363***(3.363) −36.634*** (5.734) −221.653* (197.523)
Pricing differences −0.006 (−0.034) 0.005** (0.012) 0.002 (0.003) 0.002 (0.001) 0.002 (0.004) −0.003 (0.005) −0.006* (0.004) −0.008** (0.004) −0.046***(0.005) 0.0335* (0.064)
Brent oil 3.536*** (0.833) 0.253*** (0.045) 0.363*** (0.045) 0.363*** (0.043) 0.634*** (0.036) 0.634*** (0.074) 0.734*** (0.074) 1.035*** (0.084) 1.745*** (0.346) 23.634*** (5.646)
Exchange rate −0.845 (0.075) 0.004 (0.053) 0.006 (0.001) 0.006 (0.004) 0.004 (0.002) −0.001 (0.064) 0.004 (0.002) 0.009 (0.006) 0.002 (0.003) −3.745*** (0.643)
Notes:

* p 0.10;

**p 0.05;

***p 0.01. Values in parenthesis represent standard errors

Quantile regressions-Asian markets

Variables Mean results Q10 Q20 Q30 Q40 Q50 Q60 Q70 Q80 Q90
Intercept 68.064*** (10.442) 37.745** (12.956) 52.534*** (8.543) 57.223*** (7.754) 58.745*** (7.965) 70.634*** (7.964) 88.534** (7.756) 86.534*** (10.634) 89.845*** (14.704) 76.845*** (16.745)
Momentum 1.745* (0.699) 1.534 (0.965) 0.745 (0.745) 0.854 (0.643) 0.986** (0.564) 0.886** (0.644) 0.855 (0.845) 0.845 (0.701) 1.564* (1.745) 4.674*** (1.464)
Price volatility −36.745 (38.053) −48.634** (17.534) −37.745** (14.856) −38.743** (13.935) −34.634** (15.856) −46.745*** (15.053) −57.804*** (18.804) −52.534** (21.745) −26.745 (24.053) 19.354 (28.545)
Pricing differences 0.064 (0.045) 0.089*** (0.031) 0.143*** (0.020) 0.061*** (0.021) 0.061*** (0.023) 0.037 (0.024) 0.023 (0.025) −0.007 (0.054) −0.045 (0.063) −0.043 (0.053)
Brent oil 0.320** (0.143) 0.075 (0.094) 0.147*** (0.085) 0.156*** (0.074) 0.213*** (0.076) 0.169** (0.078) 0.432* (0.063) 0.634** (0.053) 0.534** (0.043) 0.634** (0.435)
Exchange rate 0.644 (0.375) 0.674* (0.354) 0.391 (0.434) 0.497** (0.231) 0.441 (0.253) 0.629*** (0.254) 0.835*** (0.353) 0.634** (0.645) 0.745 (0.674) −0.435 (0.645)
Note:

Similar to those in Table IV

Quantile regressions-developed markets

Variables Mean results Q10 Q20 Q30 Q40 Q50 Q60 Q70 Q80 Q90
Intercept 85.434*** (25.645) 65.154** (24.654) 65.634** (21.243) 86.534*** (22.745) 91.644*** (23.345) 62.954*** (23.645) 55.834** (20.253) 90.534*** (22.645) 108.434*** (25.753) 142.534*** (49.423)
Momentum 0.634 (1.634) −0.954 (1.745) −0.365 (1.203) −1.053 (1.067) −0.845 (1.068) 1.534 (1.645) 1.985** (0.965) 1.245 (1.533) 1.364 (1.153) 2.534 (1.745)
Price volatility −24.634 (27.534) 9.845** (3.375) 3.754 (3.854) −1.745 (3.875) −8.956* (4.645) −16.845** (5.364) −27.956*** (6.745) −42.534*** (7.745) −56.374*** (8.653) −84.534*** (13.153)
Pricing differences −0.067 (0.005) 0.006* (0.003) 0.003 (0.003) −0.002 (0.003) −0.003 (0.004) −0.005* (0.004) −0.035** (0.006) −0.054* (0.005) −0.019** (0.005) −0.134* (0.005)
Brent oil 0.745 (0.153) 0.345 (0.091) 0.274* (0.083) 0.645 (0.073) 0.745 (0.085) 0.745 (0.091) 0.856 (0.154) 0.845*** (0.157) 0.975*** (0.253) 1.153 (0.235)
Exchange rate −0.363* (0.084) −0.036 (0.075) −0.046 (0.064) −0.098 (0.068) −0.374 (0.079) −0.043 (0.076) −0.064 (0.085) −0.245* (0.079) −0.253** (0.143) −0.535** (0.235)
Note:

Similar to those in Table IV

Quantile regressions-European markets

Variables Mean results Q10 Q20 Q30 Q40 Q50 Q60 Q70 Q80 Q90
Intercept 90.865 (120.236) 39.653*** (5.965) 50.865*** (6.465) 59.393*** (7.282) 65.563*** (7.865) 72.965*** (9.127) 83.658*** (10.266) 81.865*** (15.568) 215.865 (665.965) 144.326 (336.896)
Momentum −61.325*** (−16.563) 0.658 (0.698) 0.563 (0.862) 0.372 (0.433) 0.866 (0.965) 0.986 (0.536) 0.865 (0.865) 0.963 (0.865) −18.653 (29.865) −529.32*** (115.962)
Price volatility 41.265 (126.329) −7.568 (7.325) −2.239 (8.798) 0.529 (4.394) 1.632 (4.865) −3.689 (9.653) −10.956 (10.298) −11.556 (11.865) 391.236 (409.865) 15.356 (665.563)
Pricing differences −7.896*** (1.689) −0.135 (0.235) −0.039 (0.089) −0.048 (0.032) −0.095 (0.034) −0.1.6*** (0.235) −1.358** (0.235) −0.863*** (0.239) −23.765*** (1.639) −42.865*** (5.236)
Brent oil 40.265*** (5.865) 0.658* (0.123) 0.698*** (0.082) 0.703** (0.096) 0.898*** (0.168) 1.563*** (0.265) 1.865** (0.532) 3.829*** (1.563) 177.356*** (11.087) 173.865*** (35.026)
Exchange rate −0.865*** (0.236) 0.006 (0.007) 0.004 (0.006) 0.002 (0.004) −0.006 (0.005) 0.005 (0.008) 0.005 (0.008) −0.012 (0.016) −1.396** (0.203) −3.963*** (0.789)
Note:

Similar to those in Table IV

Notes

1.

For details see Baker and Wurgler (2006).

2.

These results are also in accordance with the findings of Rehman (2018).

References

Ali, A.M. (2001), “Political instability, policy uncertainty, and economic growth: an empirical investigation”, Atlantic Economic Journal, Vol. 29 No. 1, pp. 87-106.

Antonakakis, N., Chatziantoniou, I. and Filis, G. (2013), “Dynamic co-movements of stock market returns, implied volatility and policy uncertainty”, Economics Letters, Vol. 120 No. 1, pp. 87-92.

Apergis, N., Cooray, A. and Rehman, M.U. (2018), “Do energy prices affect US investor sentiment?”, Journal of Behavioral Finance, Vol. 19 No. 2, pp. 125-140.

Arouri, M., Rault, C. and Teulon, F. (2014), “Economic policy uncertainty, oil price shocks and GCC stock markets”, Economics Bulletin, Vol. 34 No. 3, pp. 1822-1834.

Bachmann, R., Elstner, S. and Sims, E.R. (2013), “Uncertainty and economic activity: evidence from business survey data”, American Economic Journal: Macroeconomics, Vol. 5, pp. 217-249.

Baker, M. and Wurgler, J. (2006), “Investor sentiment and the cross‐section of stock returns”, The Journal of Finance, Vol. 61 No. 4, pp. 1645-1680.

Baker, S.R., Bloom, N. and Davis, S.J. (2016), “Measuring economic policy uncertainty”, The Quarterly Journal of Economics, Vol. 131 No. 4, pp. 1593-1636.

Bams, D., Blanchard, G. and Lehnert, T. (2017), “Volatility measures and value-at-risk”, International Journal of Forecasting, Vol. 33 No. 4, pp. 848-863.

Brogaard, J. and Detzel, A. (2015), “The asset-pricing implications of government economic policy uncertainty”, Management Science, Vol. 61 No. 1, pp. 3-18.

Brock, W.A., Scheinkman, J.A., Dechert, W.D. and LeBaron, B. (1996), “A test for independence based on the correlation dimension”, Econometric Reviews, Vol. 15 No. 3, pp. 197-235.

Campbell, J.Y., Lettau, M., Malkiel, B.G. and Yexiao, X. (2001), “Have individual stocks become more volatile? An empirical exploration of idiosyncratic risk”, Journal of Finance, Vol. 56 No. 1, pp. 1-43.

Chau, F., Deesomsak, R. and Koutmos, D. (2016), “Does investor sentiment really matter?”, International Review of Financial Analysis, Vol. 48, pp. 221-232.

Colombo, V. (2013), “Economic policy uncertainty in the US: does it matter for the euro area?”, Economics Letters, Vol. 121 No. 1, pp. 39-42.

Connor, G. and Korajczyk, R.A. (1989), “An intertemporal equilibrium beta pricing model”, Review of Financial Studies, Vol. 2 No. 3, pp. 373-392.

Du, D., Gunderson, R. and Zhao, X. (2015), “Investor sentiment and oil prices”, Working Paper, No. 2, Northern Arizona University, W.A. Franke College of Business.

Giavazzi, F. and McMahon, M. (2012), “Policy uncertainty and household savings”, Review of Economics and Statistics, Vol. 94 No. 2, pp. 517-531.

Goyal, A. and Santa-Clara, P. (2002), “Idiosyncratic risk matters!”, Working Paper, Anderson Graduate School of Management, University of California at Los Angeles.

Gulen, H. and Ion, M. (2015), “Policy uncertainty and corporate investment”, The Review of Financial Studies, Vol. 29, pp. 523-564.

Handley, K. and Limao, N. (2015), “Trade and investment under policy uncertainty: theory and firm evidence”, American Economic Journal: Economic Policy, Vol. 7 No. 4, pp. 189-222.

Huerta, D., Perez-Liston, D. and Jackson, D. (2011), “The impact of TARP bailouts on stock market volatility and investor fear”, Banking and Finance Review, Vol. 3 No. 1, pp. 45-54.

Jones, P.M. and Olson, E. (2013), “The time-varying correlation between uncertainty, output, and inflation: evidence from a DCC-GARCH model”, Economics Letters, Vol. 118 No. 1, pp. 33-37.

Kang, W., de Gracia, F.P. and Ratti, R.A. (2017), “Oil price shocks, policy uncertainty, and stock returns of oil and gas corporations”, Journal of International Money and Finance, Vol. 70, pp. 344-359.

Karnizova, L. and Li, J.C. (2014), “Economic policy uncertainty, financial markets and probability of US recessions”, Economics Letters, Vol. 125 No. 2, pp. 261-265.

Klobner, S. and Sekkel, R. (2014), “International spillovers of policy uncertainty”, Economics Letters, Vol. 124, pp. 508-512.

Koenker, R. and Bassett, G. (1978), “Regression quantiles”, Econometrica, Vol. 46 No. 1, pp. 33-49.

Leduc, S. and Liu, Z. (2016), “Uncertainty shocks are aggregate demand shocks”, Journal of Monetary Economics, Vol. 82, pp. 20-35.

Liu, L. and Zhang, T. (2015), “Economic policy uncertainty and stock market volatility”, Finance Research Letters 159, Vol. 15, pp. 99-105.

Narayan, S. and Rehman, M.U. (2017), “Diversification opportunities between emerging and frontier Asian (EFA) and developed stock markets”, Finance Research Letters, Vol. 23, pp. 223-232.

Pastor, L. and Veronesi, P. (2012), “Uncertainty about government policy and stock prices”, The Journal of Finance, Vol. 67 No. 4, pp. 1219-1264.

Rehman, M.U. (2018), “Do oil shocks predict economic policy uncertainty?”, Physica A: Statistical Mechanics and Its Applications, Vol. 498, pp. 123-136.

Rehman, M.U. and Shahzad, S.J.H. (2016), “Investors’ sentiments and industry returns: wavelet analysis through squared coherency approach”, ACRN Oxford Journal of Finance and Risk Perspectives, Vol. 5, pp. 151-162.

Corresponding author

Mobeen Ur Rehman can be contacted at: Mobeen@tdtu.edu.vn