Panel Cointegration

∆ is the first difference operator. Pedroni suggests some adjustments for each of all test statistics (both for panel unit root tests and panel cointegration tests) described above that produces standard normal distributions. In this study, we report the adjusted values so that in all cases the reported test values can be compared to the standard normal distribution. This is the case for both the cointegration and unit root tests. 3

Table 1 shows the results of panel cointegration tests. Under the null hypothesis of no cointegration, the test statistic is -2.564 for Panel PP test, -2.696 for Panel ADF test, -2.208 for Group PP test, -2.877 for Group ADF test, 74.450 for Fisher statistic from trace test, and 75.410 for Fisher statistic from maximum eigen-value test for case 1. The null hypothesis is rejected at the 5% significance level in every test. Table 1 also indicates that similar results are obtained for case 2 and case 3. Thus, it can be said that exports, relative prices, and world trade volume have a cointegrating relation in all cases.

Rising economic performance has enlarged energy demand, carbon emissions and global warming. Policy makers need to avoid global warming. Therefore, energy-growth nexus is important. This paper empirically investigates the relationship between energy consumption and economic growth for a panel of G20 countries over the period 1990-2016. For this purpose, the paper considers the panel cointegration and panel vector error correction model. Panel cointegration test set out a long-run equilibrium relationship. Long-run relationship is estimated using a Fully Modified OLS (FMOLS) and Dynamic OLS (DOLS). Panel Granger causality and Vector Error Correction Model results show that bidirectional relationship between energy consumption and GDP. It is indicates that “feedback hypothesis” is valid for G20 countries. Keywords: economic growth; energy consumption; panel unit-root; panel Granger causality; cointegration

Although the past econometric literature deemphasized the importance of the so called price variables, this may be explained partly by the lack of reliable and long- run time series data; partly by the econometric techniques which suffered from a number of problems that now we can cope with using panel unit root tests and panel cointegration; finally, the econometric specification usually was not the most appropriate to capture the price variables. In our econometric specification, we take the growth rate of investment as the appropriate index which essentially indicates the acceleration or deceleration of capital accumulation. As independent variables, we include the incremental rate of profit as an index of profitability, the real prime interest rate as an index of financial conditions, the growth rate of real GDP as our proxy for the demand conditions or acceleration effects and we also use lags of the dependent variable to capture the possible cumulative and diffusion of technology effects of past investment activity. The major advantage of such econometric specification is that in our measure of profitability, we dispense altogether with the capital stock and the difficulties, theoretical and empirical, associated with its measurement as produced (means of production) goods. As a consequence, the gross investment expenditures, used in our profitability variable (see below), have the advantage that their measurement is straightforward and common across countries and also over the years.

The unit root test is widely employed to investigate the presence of stationarity of the variables. It is well known that panel unit root tests are more powerful compared to the individual time series. There are several types of unit root tests for the panel data, namely the Breitung (2000), Hadri (2000), Levin, Lin and Chu (2002), Im, Pesaran and Shin (2003) and Fisher type tests. In this stuy, we use the panel unit root tests developed by Levin, Lin and Chu (LLC) (2002) and Im, Pesaran and Shin (IPS) (2003) to check the stationarity properties of the variables. LLC test supposes that there exists a common unit root process across the cross-sections while IPS test suggests that there exists individual unit root processes across the cross-sections. In these tests the null hypothesis is that there exists a unit root. In LLC test the alternative hypothesis is that there exists no unit root, alternative hypothesis is that some cross-sections do not have a unit root in IPS test (Pao and Tsai, 2011). The main characteristic of LLC test is that it suggets a homogeneous autoregressive root under the alternative hypothesis. Conversely, the main feature of IPS test is that it enables for a heterogeneous autoregressive coefficient under the alternative hypothesis (Hamit-Haggar, 2012). 4.2. Panel Cointegration Tests

Testing for cointegration was initiated by Engle and Granger (1987) via implementing residual based tests for cointegration. More robust tests where suggested by Phillips (1987) as well as Phillips and Ouliaris (1990). Multivariate tests for cointegration were developed by Johansen (1988, 1991), Johansen and Juselius (1990) as well as Stock and Watson (1988). The idea to conduct tests for unit roots within a panel system originates from Quah (1994). Tests for panel cointegration were suggested by Pedroni (1999, 2004), Kao (1999) and Westerlund (2007), among others.

Pedroni [5] extends his residual-based panel cointegration tests [5] for the models, where there are more than one independent variable. He maintains several residual-based null of no cointegration panel cointegration test statistics. Pedroni developed seven cointegration statistics to test for the null of no-cointegration among the variables. The four statistics – within-dimension panel cointegrationtests pool the autoregressive coefficients (ϕ i ) across different members for the unit root tests on the residuals. The next

The evidence suggesting the existence of a long-run link between the investment and savings to GDP ratios in many advanced economies (Feldstein-Horioka, FH, puzzle after Feldstein and Horioka, 1980) is considered one of the six major empirical puzzles of contemporary macroeco- nomics (Obstfeld and Rogo¤, 2000). In a world with no barriers, capitals should seek the most pro…table investment opportunities regardless of national borders. Thus, in any individual coun- try, capital formation (I ) should not be constrained by domestic savings (S), as the di¤erence between the two aggregates is the current account balance (B), which in such a world does not need to be zero. Hence, the evidence of a savings-investment relationship was indeed puzzling. After thirty years, during which the issue has been investigated in a huge literature (for a recent review, see Apergis and Tsoumas, 2009, who list nearly 200 references), the puzzle is still essen- tially open, as existing studies leave much to be desired. Tests of the long-run validity of the puzzle based on cointegration methods have as a null hypothesis the absence of a relationship, i.e., no puzzle. Hence, the prevailing conclusion that no puzzle exists (e.g., Kim, 2001) may be simply due to the notiorous lack of power of cointegration tests with the small or moderate samples typically used. Panel cointegration studies, such as Pelgrin and Schich (2008), do have higher power, but are based on tests valid only for independent units. If this assumption does not hold these tests are strongly biased against the null hypothesis of no relationship (Banerjee, Marcellino and Osbat, 2004), so that the reported rejections of the hypothesis of no relationship may be spurios. In fact, in Di Iorio and Fachin (2010) we showed that, using a bootstrap panel cointegration test valid for dependent units, evidence of a saving-investment long-run relation- ship cannot be found for a panel of 18 economies including the core of OECD 2 , but only some

The traditional model of demand for military expenditure developed in Smith (1980) presents military expenditure of a country as function of civilian output and of the ‘strategic environment’ which in turn is a function of military expenditures of other countries. In particular, the illustrative focus there was on the relationship between a superpower and other countries. Under the assumption that military expenditure of a superpower can be interpreted as a credible signal of threat, two behaviours could be envisioned: (i) free riding; (ii) leader/follower relationship. In the first case, the empirical association between military expenditures of a superpower and its allies turns to be negative because a country can ‘free-ride’ so reducing its contribution to the production of the public good of security. In the alternative case, a countr y is intended to ‘follow’ the leader so increasing the military expenditure. In fact, military expenditure of the superpower signals an increase in threat. The latter can be either allies or foes. Eventually, several papers confirmed the interdependence between the military spending of countries within an alliance at regional level [see among others Murdoch and Sandler (1984), Smith (1989), Sandler and Murdoch (1990)]. In this vein, this paper is intended first to verify whether the interdependence between US and European countries is confirmed in the period 1988-2013 when using a panel cointegration analysis. Then, by means of FMOLS and DOLS regressions, we provide long-run elasticities. In brief, we are able to verify whether European countries followed or free rode on US military spending. Secondly, we also consider interdependence among a panel of twenty European countries.

However, recent empirical and theoretical studies on the subject, for example those of Wirth (2014), Cifor et al. (2015), Chang and Wang (2017). Mertzanis (2018) Bhattacharya et al. (2017) highlighted the importance of having an adequate institutional infrastructure. Control of corruption, respect for democratic principles, respect for the laws and the legislative authority are all important elements for a policy aimed at promoting the beneficial use of renewable energies to succeed. Indeed, it is interesting to note that the results of various studies, examining the contribution of institutional quality in explaining the relationship between renewable energies and economic growth, are mostly mixed. This lack of consensus may be the consequence of the differences in the samples, the techniques used and the institutional variable selected. Our paper aims to contribute to this debate by considering a satisfactory set of institutional measures. As we have pointed out, in this paper we study the role of institutions in sustainable growth in the MENA region. To this aim, we make use of several robust panel cointegration methods in the presence of structural breaks to account for changes that may be identified in the studied relationships.

In this section we adopt FMOLS procedure from Christopoulos and Tsionas (2004). In order to obtain asymptotically efficient and consistent estimates in panel series, non-exogeneity and serial correlation problems are tackled by employing fully modified OLS (FMOLS) introduced by Pedroni (1996). Since the explanatory variables are cointegrated with a time trend, and thus a long-run equilibrium relationship exists among these variables through the panel unit root test and panel cointegration test, we proceed to estimate the Equation (2) by the method or fully modified OLS (FMOLS) for heterogeneous cointegrated panels. This methodology allows consistent and efficient estimation of cointegration vector and also addresses the problem of non-stationary regressors, as well as the problem of simultaneity biases. It is well known that OLS estimation yields biased results because the regressors are endogenously determined in the I(1) case. Following cointegrated system for panel data

We explore the long-run relationship between the unemployment rate and the labor force participation rates for U.S. over the period of 1976-2015. We use U.S. state level data and panel cointegration techniques that are robust to cross-sectional heterogeneity, cross-sectional dependency, omitted variable bias and endogeniety issues. We find evidence that on average these two variables are cointegrated and are inversely related. Similar to studies that employ U.S. country level data, this study further questions the empirical relevance of the unemployment invariant hypothesis for the case of the U.S. Keywords: Unemployment Rate, Unemployment Invariant Hypothesis, Panel Cointegration

After testing for stationarity of the variables, we then test for the existence of a long-run relationship among the variables. As discussed in BANERJEE, MARCELLINO, and OSBAT (2004), panel cointegration tests can be largely oversized in the presence of cross-unit long- run relationships. Not accounting for such relationships makes it more likely to obtain a finding in favor of cointegration, which may be false. An alternative panel cointegration test was proposed by WESTERLUND (2007). The tests are general enough to allow for a large degree of heterogeneity, both in the long-run cointegrating relationship and in the short-run dynamics, and dependence within as well as across the cross-sectional units. Also, WESTERLUND ’s (2007) tests have good small -sample properties with small size distortions and high power relative to other popular residual-based panel cointegration tests, such as PEDRONI (1999, 2004). WESTERLUND (2007) developed four error-correction-based panel cointegration tests. Two tests are designed to test the alternative hypothesis that the panel is cointegrated as a whole, while the other two tests the alternative that at least one unit is cointegrated. The author considers the following error correction model where all variables in level are assumed to be integrated of order 1:

Firms are established objectively to maximize their value and that of their shareholders; this can be achieved via payment of dividend and investment in profitable ventures. Studies conducted in both developed and developing economies could not solve the problem of dividend dynamism. It is against this background that this study is conducted to determine the effect of firms’ performance on dividend policy of 21 listed financial companies in Nigeria purposively selected for a period of 20 years (1997-2016). Secondary data was collected and used for analysis. Hausman and Wald test were conducted to choose between fixed effect and random effect, and fixed and pooled OLS respectively. The study conducted correlation matrix test, panel unit root and panel cointegration test while all study objectives were tested using multiple regression of the fixed effect analysis. The outcome from the regression reveals that all the independent variables significantly affect dividend payout ratio of the sampled companies. It is clear from the analysis that performance affects dividend decisions in both short and long runs. As such, managers of these companies should sustain effective utilization of their assets and should also strive to increase the value of their equity by investing larger portion of their earnings into profitable ventures. Keywords: Performance, Dividend Policy, Financial Sector, Accounting Based Measures, Market Based Measures, Tobin’s Q, Market Value Added JEL Classifications: G3, M41

alone, but rather based on the order of integration of both the money stock and the real variable, meaning that comprehensive neutrality tests can only be evalu- ated if money and real output are nonstationary and thus subject to permanent shocks. King – Watson (1997), who analysed the robustness of neutrality re- sults to alternative identifying assumptions, underlined the necessity of applying preferably as many tests as possible in the hope that this will make results more confident. Furthermore, the possible presence of cointegration is in itself suf- ficient for rejecting the money neutrality proposition. This is why Westerlund – Constantini (2009) believe that cointegration tests should be considered as an integral part of neutrality tests and not only as a diagnostic preliminary as is usually the case. Considering the need for a selective approach and the problem of cross-sectional dependency when analysing a sample of countries, Bullard (1999) stresses the importance of accurate unit root and cointegration tests, and suggests that a panel approach might be more appropriate in this respect (Wester- lund – Constantini 2009). This acknowledgement is in line with our research per- spective. Relatively few papers applied panel (especially cointegration-based) analysis when testing the long-run money neutrality. Westerlund – Constantini’s (2009) paper on panel cointegration for ten industrialised countries covering the period from 1870 to 1986 suggested that the null hypothesis of no cointegration between money and real output can be rejected and thus that the neutrality of money can also be rejected. Their two conclusions are of great importance for our study, namely, that (1) most series have missing observations, which not only makes the panel unbalanced, but also reduces the effective number of time-series observations, thus making the cross-sectional dimension a very important source of information, and (2) the cointegration test must be evaluated as an integral part of the neutrality tests with serious efforts pointed towards detecting time-series integration properties.

The main question of this section as well as the next one is to know the distinction between the long-term and short-term effects of fiscal decentralization on economic growth through panel cointegration. However, before any analysis it is advisable to start with the degree of integration of the variables. If they are all stationary, that is I (0), the distinction between long-term and short-term relationships would be superfluous. If, on the other hand, some or all of the relevant variables are not stationary, the distinction becomes meaningful. The non-stationarity of the variables in our model is analyzed using the panel stationarity test of Im et al. (2003).

This study uses panel data covering the period from 1984 to 2014, panel cointegration as well as Dynamic Ordinary Least Square method to investigate the effect of trade openness on manufacturing growth in Economic and Monetary Community of Central Africa (EMCCA) countries. The results reveal two effects. Firstly, there is a positive and significant effect of Foreign Direct Investment and investment on manufacturing growth. Secondly, there is an ambiguous effect of trade openness on manufacturing growth. Indeed, trade openness affects either negatively the manufacturing growth or has no effect on manufacturing growth in EMCCA countries. Whatever the case, EMCCA countries should develop their manufacturing sector before fully trading with foreign countries.

In this paper, we have examined the causal relationship between total energy consumption and output level for a panel of G7 countries. The novelty of the paper is that we propose a new panel co- integration test in a nonlinear smooth transition regression framework and estimate nonlinear panel vector error correction model. Although conventional linear panel cointegration tests suggest that energy consumption and output level are not co-integrated, we find a strong evidence of cointegration among these variables using newly proposed nonlinear cointegration tests. This result suggests that adjustment of these variables to the long-run equilibrium level is inherently nonlinear.

This study investigates the evidence of efficient market hypothesis for the firms listed in Karachi Stock Exchange (KSE), which signifies the nature of the stock market. According to this hypothesis, the empirical information available is not enough to predict the present movement of share prices. After using the Panel cointegration approach between current share price and past share price in the light unknown structural break on the daily data of 75 selected firms over the period June 2004 to March 2014, comprising 2370 observations per firm. The study found that the firms listed in Karachi Stock Exchange are inefficient firms, therefore, for the case of KSE – 100 offer information which can be used to make economic profits. Hence Karachi Stock Exchange provide enough predictable information using past trends so that investor can gain economic profit from it, so it will take time for the market to become mature and the creation of competition.

In our model, if our variables [q, y, p, gp, pop] contain a panel unit root, the issue arises whether there exists a long-run equilibrium relationship between the variables. We use the Pedroni (2004) test to examine for panel cointegration that allows both for heterogeneity in the intercepts and slopes of the cointegrating equation. In his test, Pedroni (2004) provides seven statistics to test the null hypothesis, which means no cointegration in heterogeneous panels. One part of seven statistics is entitled “within dimension” which takes into account common time factors and allows for heterogeneity across countries. The other part is termed “between dimension” that allows for heterogeneity of parameters across countries. The established seven statistics by Pedroni (2004) that we employ are as follows: