- Email: [email protected]
Arising borders and the value of logistic companies: Evidence from the Brexit referendum in Great Britain
Finance Research Letters 20 (2017) 22–28
Contents lists available at ScienceDirect
Finance Research Letters journal homepage: www.elsevier.com/locate/frl
Arising borders and the value of logistic companies: Evidence from the Brexit referendum in Great Britain Artur Tielmann, Dirk Schiereck∗ Department of Business Administration, Economics and Law, Technische Universität Darmstadt, 64289 Darmstadt, Germany
a r t i c l e
i n f o
Article history: Received 21 July 2016 Revised 8 August 2016 Accepted 10 August 2016 Available online 13 August 2016 JEL classification: G10 G14 G15
a b s t r a c t The Brexit referendum may result in new border controls and a separation of Great Britain from the EU and Continental Europe. These consequences will impede the import and export of goods and can therefore have a strong effect on the valuation of logistic companies. We employ event study methodology and regression analysis, examining 107 logistic companies from continental EU countries and Great Britain. While the results indicate an overall negative value effect, UK based companies have a significantly poorer performance than logistic companies from Continental Europe. Companies that focus on the road transport as well as diversified firms are less affected. © 2016 Elsevier Inc. All rights reserved.
Keywords: Brexit Political risk Logistics industry Event study
1. Introduction On 23 June 2016 the British citizens (surprisingly) decided to instruct their government to detach their country from the European Union. This process can have a major impact on the logistics sector when border controls are installed and impede the import and export of goods from and to Great Britain. British companies have to expect additional VAT (value added taxes) and duties. Both factors negatively affect the overall economy in Great Britain and Continental Europe but specifically the logistics sector. Logistics industry is of course not the only industry affected by the Brexit but it is perhaps the most obvious one. According to the UK government, 44 per cent of all British exports are transported to EU countries. It can be expected that the Brexit will create borders over time which will have an impact on the logistics efficiency because of slower movement of goods. Moving goods across borders within the EU is easy and cheap at present. The only documentation needed for transporting goods from one country in the EU to another is a copy of the packing list or commercial invoice and the travel document (waybill, bill of landing or CMR note). At the moment, there is no customs clearance process and no duties applied. VAT doesn’t have to be handed over before the goods can be moved from the receiving port or airport. After the Brexit additional administrative burdens will apply and if the UK is treated like other countries from outside the EU, it will also be necessary to submit customs declarations to the authorities for goods leaving and entering the UK. As transportation follows production, logistics industry can be considered as an indicator and a concentrated aggregate of the overall consequences ∗
Corresponding author. E-mail addresses: [email protected] (A. Tielmann), [email protected], [email protected] (D. Schiereck).
http://dx.doi.org/10.1016/j.frl.2016.08.006 1544-6123/© 2016 Elsevier Inc. All rights reserved.
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
23
the Brexit has on manufacturing. Therefore, we expect to find sustainable negative stock price reactions for logistic firms in the aftermath of the Brexit referendum. As documented by Pastor and Veronesi (2012) changes in government policy affect stock prices and the magnitude of negative returns is large if uncertainty in government policy is high. We choose the Brexit event mainly for three reasons: (i) the outcome was unpredictable until the final day as both sides competed head-to-head (ii) the impact of the referendum result has to be significant as never before an EU member wanted to leave the European Union1 and the conclusions are not assessable yet (iii) a vote for “leave” implies an increase of uncertainty especially in the logistics sector as it is not clear what changes will come up regarding border controls, duties and taxes. There is already considerable empirical evidence indicating that political changes, such as elections, affect overall stock markets. Santa-Clara and Valkanov (2003), for example, examine stock returns in the U.S. presidencies elections. Further, Nippani and Medlin (2002) observe negative stock market reactions on the delay in the declaration of the U.S. presidential election winner of 20 0 0 on the domestic stock market. Indeed, most of the empirical evidence deals with politics in the U.S, but there are also studies with broader perspective or a different regional focus. Białkowski et al. (2008) investigate the stock market volatility surrounding national elections in 27 OECD countries and find that the country-specific component of index return variance can increase significantly, if several factors such as the political orientation of the government change are controlled for. Pantzalis et al. (20 0 0) find positive abnormal returns in the two weeks before elections across 33 countries. In particular, when the incumbent government loses the election, they find positive reactions in the market. And Döpke and Pierdzioch (2006) investigate the impact of political changes in Germany and find only weak evidence for an impact on the stock market. Our analysis tries to derive a more differentiated picture of political change by analyzing single stocks of a heavily affected industry instead of general stock market indices. 2. Sample construction and methodology We focus our analysis on the day after the EU referendum in the United Kingdom, when the national declaration of the result took place at breakfast time the following day. We set the event date accordingly to Friday, 24 June 2016. Since we focus on stock price reactions of logistic companies, we first collect all exchange-listed logistics stocks in the European Union. Then, we drop all firms and observations with illiquid trading patterns and check for confounding events (M&A; ad hoc announcements). This leaves us with a final sample of 107 observations - 21 for British enterprises and 86 companies headquartered in the rest of the 27 member countries of the EU. Table 1 summarizes the sample firms by country (Panel A) and by SIC Code (Panel B). We employ the methodology of the standard market model event study, as introduced by Dodd and Warner (1983) and Brown and Warner (1985). The cumulative abnormal return (CAR) for stock i during the event window [τ 1; τ 2] surrounding the event day t = 0 is calculated as follows: τ2
CARi,[τ 1,τ 2] =
t=τ1
(Ri,t − αˆ i − βˆi RM,t )
(1)
where CARi, [τ 1,τ 2] is the CAR i during the event window, Ri,t is the actually realized return of company i on day t, Rm,t is the return of the benchmark index of company i on day t, and αˆ i and βˆ i are the regression coefficients from an ordinary least squares (OLS) regression using a 252 trading day estimation period. As benchmark indices we use the Datastream’s value-weighted total return national stock market index of stock i’s country. Finally, the average CAR (ACAR) for a sample of N firms is calculated as follows:
ACAR[τ 1,τ 2] =
N 1 CARi, [τ 1,τ 2] N
(2)
i=1
ACARs are calculated for the interval [τ 1; τ 2] ∈ [−4; +10]. In order to test the significance of the ACARs, we apply the parametric test statistic following Boehmer et al. (1991), the BMP-test, and the nonparametric rank test following Corrado (1989), which was later refined by Corrado and Zivney (1992), the CZ-test. Finally, we perform a cross-sectional regression analysis to identify the logistic specific drivers of the CARs. The multivariate ordinary least squares (OLS) regression in its full specification takes the following form:
CARi,[τ 1;τ 2] =
β0 + β1 SIZE + β2 MT BR + β3 PERF ORMANCE + β4 DIV ERSIF ICAT ION + β5 ROAD + β6WAT ER + β7 AIR +β8U K + β9U K ∗ AIR + (3)
where CARi,[τ 1;τ 2] is the dependent variable, SIZE is the logarithm of the market value of a company in EUR in 2015, the year prior the event, MTBR is the market-to-book-ratio in 2015, PERFORMANCE is the past stock performance of the year before the event. DIVERSIFICATION is the amount of 4-digit SIC codes a firm has, ROAD is set to 1, if the company’s primary SIC
1
In fact, Greenland left the predecessor of the EU, the EWG, in 1985.
24
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28 Table 1 Descriptive statistics of the analyzed sample. Panel A: Country
Number of companies
Market value [in Mio. EUR]
Austria Belgium Croatia Cyprus Denmark Estonia Finland France Germany Greece Ireland Italy Luxembourg Malta Netherlands Poland Slovenia Spain Sweden UK
2 4 1 2 5 1 1 14 7 18 1 7 2 1 5 5 1 4 5 21
4341 7119 56 1344 27,639 455 318 44,279 45,832 7710 19,803 19,579 365 327 16,787 1345 322 46,402 768 40,116
Panel B: SIC Code
Number of companies
Description of SIC Code
4011 4212 4213 4215 4225 4226 4231 4412 4491 4499 4512 4513 4581 4731 4789
4 8 6 7 4 1 6 33 6 2 13 3 9 2 3
Railroads, line-haul operating Local trucking without storage Trucking, except local Courier services, except by air General warehousing and storage Special warehousing and storage, not elsewhere classified Terminal and joint terminal maintenance facilities for motor freight transportation Deep sea foreign transportation of freight Marine cargo handling Water transportation services, not elsewhere classified Air transportation, scheduled Air courier services Airports, flying fields, and airport terminal services Arrangement of transportation of freight and cargo Transportation services, not elsewhere classified
This table shows the number of companies and the cumulated market value per country in Panel A and the number of companies per SIC Group in Panel B.
is related to road transport,2 WATER is defined as 1, if the firms primary SIC is related to water transport 3 AIR is defined as 1, if the firms primary SIC is related to air transport, 4 UK is defined as 1, if the company’s headquarter is located in the UK and UK∗ AIR is defined as 1, if the company is focused on air transportation and its headquarter is located in the UK, 0 otherwise. The stock market data and balance sheet data are obtained from Datastream. Table 2 reports Pearson pairwise correlations for all explanatory variables. With exception of the combination UK∗ AIR – AIR the correlation coefficients are low, suggesting that multicollinearity is unlikely to be a problem. 3. Results Table 3 shows the results of the event study for the total sample (Panel A) and the two subsamples, UK based companies (Panel B) and companies from the EU mainland (Panel C). Overall, we document highly significant negative market reactions on the announcement day of the referendum decision to both subsamples and during the [0;+1] event window. However, the British subsample is significantly more affected than the Continental European companies which seems to be plausible as the Continental European logistic firms have more trouble entering one market while British firms are confronted with a more complicated access to all continental European national markets. In the week before the Brexit referendum we do not 2
The SIC Codes 4212,4213 and 4215 are defined as Local Trucking Without Storage, Trucking, Except Local and Courier Services, Except by Air. The SIC Codes 4412,4424, 4491 and 4499 are defined as Deep Sea Foreign Transportation of Freight, Deep Sea Domestic Transportation of Freight, Marine Cargo Handling and Water Transportation Services, Not Elsewhere Classified. 4 The SIC Codes 4512 and 4513 are defined as Air Transportation, Scheduled and Air Courier Services. 3
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
25
Table 2 Correlation matrix.
SIZE MTBR PERFORMANCE DIVERSIFICATION ROAD WATER AIR UK UK∗ AIR
SIZE
MTBR
PERFORMANCE
DIVERSIFICATION
ROAD
WATER
AIR
UK
UK∗ AIR
1 0.030 0.421∗ ∗ ∗ 0.328∗ ∗ ∗ −0.002 −0.275∗ ∗ ∗ 0.133 −0.024 −0.003
1 −0.036 −0.079 0.269∗ ∗ ∗ −0.125 −0.081 0.301∗ ∗ ∗ −0.005
1 0.281∗ ∗ ∗ −0.004 −0.408∗ ∗ ∗ 0.242∗ ∗ 0.096 0.141
1 0.064 −0.218∗ ∗ 0.0314 −0.055 −0.075
1 −0.390∗ ∗ ∗ −0.207∗ ∗ −0.007 −0.109
1 −0.331∗ ∗ ∗ −0.099 −0.175∗
1 0.123 0.528∗ ∗ ∗
1 0.448∗ ∗ ∗
1
This table presents Pearson pairwise correlations of the independent variables. All variables are defined in Section 2. significance at the 1%, 5% and 10% level, respectively.
∗∗∗
,
∗∗
and
∗
indicate statistical
Table 3 Event study results. Event window
Cumulative abnormal return
t-Test
Boehmer test
Corrado-test
Mean (%)
Median (%)
t-value
z-score
z-score
Nobs
−0.75 0.03 −0.89 −2.20 −1.22
−2.33∗ ∗ 0.33 −2.65∗ ∗ ∗ −5.67∗ ∗ ∗ −1.69∗
−2.68∗ ∗ ∗ 0.71 −3.72∗ ∗ ∗ −5.57∗ ∗ ∗ −1.69∗
−0.94 −0.45 −2.09∗ ∗ −3.56∗ ∗ ∗ −0.21
107 107 107 107 107
−8.07 1.27 −3.05 −7.99 −5.62
−4.23∗ ∗ ∗ −0.13 −3.29∗ ∗ ∗ −3.92∗ ∗ ∗ −3.22∗ ∗ ∗
−4.06∗ ∗ ∗ 0.46 −3.27∗ ∗ ∗ −3.92∗ ∗ ∗ −3.21∗ ∗ ∗
−2.03∗ ∗ 0.33 −2.26∗ ∗ −3.57∗ ∗ ∗ −1.99∗ ∗
21 21 21 21 21
0.13 −0.46 −0.62 −1.81 −0.43
−0.68 0.37 −1.47 −4.46∗ ∗ ∗ −0.34
−0.58 0.58 −2.37∗ ∗ −4.49∗ ∗ ∗ 0.21
−0.36 −0.65 −1.63 −2.88∗ ∗ ∗ 0.48
86 86 86 86 86
Cumulative abnormal return
two sample t-test
Wilcoxon test
࢞ Mean (%)
t-value
z-score
−3.41∗ ∗ ∗ −0.24 −2.09∗ ∗ −3.46∗ ∗ ∗ −2.76∗ ∗ ∗
−4.18∗ ∗ ∗ 0.24 −1.71∗ −2.30∗ ∗ −3.27∗ ∗ ∗
Panel A: UK and Mainland [−4;+10] −2.88 [−4;−1] 0.23 [0;0] −1.33 [0;+1] −4.20 [+1;+10] −1.78 Panel B: UK [−4;+10] −11.01 [−4;−1] −0.11 [0;0] −3.41 [0;+1] −9.15 [+1;+10] −7.50 Panel C: Mainland [−4;+10] −0.90 [−4;−1] 0.31 [0;0] −0.82 [0;+1] −3.00 [+1;+10] −0.39 Event window
࢞ Median (%)
Panel D: Difference UK and Mainland [−4;+10] −10.12 −8.20 [−4;−1] −0.42 1.74 [0;0] −2.59 −2.43 [0;+1] −6.15 −6.19 [+1;+10] −7.11 −5.19
This table shows the stock market reaction of the logistics stocks to the declaration of the result from the EU referendum on 24 June 2016. The ACARs are estimated for Panel A and the subsamples Panel B and Panel C in the same 5 event windows. The daily abnormal returns are calculated using the methodology of the market model with a 252 days estimation period. As the benchmark indices we use the Datastream’s value-weighted total return national stock market index of stock i’s country. All ACARs are tested for statistical significance by using the parametric BMP-test and the nonparametric CZ-rank-test. The stock market reaction to UK and mainland companies are presented in Panel B and Panel C. We also test for statistically significant differences between those two subsamples by using the parametric two-sample t-test and the nonparametric Wilcoxon rank sum test. ∗ ∗ ∗ , ∗ ∗ and ∗ indicate statistical significance at the 1%, 5% and 10% level, respectively.
find any significant abnormal reactions. This matches the newspapers’ reports in the week before the poll, when supporters and opponents competed in a close contest, and supports the hypothesis that capital markets did not anticipate the result. Remarkably, we find a second stock market decline for the UK subsample since 30 June on when Boris Johnson withdrew from his candidature as Great Britain’s next prime minister at a press conference on Thursday, 30 June. He left the race by saying: “I must tell you, […] that having consulted colleagues and in view of the circumstances in Parliament, that person cannot be me.” The market reacted immediately to the loss of the most popular face of the Brexit supporters. Fig. 1 illustrates the stock market reactions and points up the significant difference in market reactions between the British and the Continental European subsample.
07/04/2016
07/03/2016
07/02/2016
07/01/2016
06/30/2016
06/29/2016
06/28/2016
06/27/2016
06/26/2016
06/25/2016
06/24/2016
06/23/2016
06/22/2016
06/21/2016
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
06/20/2016
26
2.00% 0.00% -2.00%
ACAR
-4.00% -6.00% -8.00% -10.00% Mainland and UK
UK
Mainland
-12.00% Fig. 1. ACAR performance. This figure shows the ACAR performance of the sample firms from Monday, 20 June (t = −4) to Monday, 04 July (t = 10).
Table 4 Regression results. Dependent variable CARi,[0,1]
Dependent variable CARi,[ − 4,10]
Model 1 SIZE MTBR PERFORMANCE DIVERSIFICATION ROAD WATER
Model 2 ∗∗
−0.016 (−2.00) −0.001∗ ∗ ∗ (−3.48) 0.038 (1.55) 0.004 (1.36) 0.004∗ ∗ ∗ (2.8) 0.017 (1.04)
AIR UK UK∗ AIR CONSTANT Mean VIF Adjusted R2 F-value
−0.173∗ ∗ ∗ (−4.57) −0.007 (−0.29) 1.30 0.289 8.56∗ ∗ ∗
0.004∗ (1.83) 0.031∗ ∗ (2.27) 0.009 (0.55) −0.054∗ ∗ (−1.99) −0.053∗ ∗ ∗ (−2.65)
−0.048∗ ∗ ∗ (−3.2) 1.280 0.187 4.18∗ ∗ ∗
Model 3
Model 4 −0.009 (−0.38) −0.001∗ ∗ (−2.30) 0.020 (0.36) 0.001 (0.19) 0.041∗ (1.92) 0.013 (0.59)
−0.198∗ ∗ ∗ (−4.25) −0.003 (−0.04) 1.30 0.084 4.46∗ ∗ ∗
0.002 (0.41) 0.0210451 (1.11) −0.001 (−0.03) −0.074∗ ∗ ∗ (−3.18) −0.092∗ ∗ ∗ (−3.70)
−0.009 (−0.46) 1.28 0.11 6.27∗ ∗ ∗
This table shows the cross-sectional regression results for the entire sample of 107 observations. CARi,[τ 1;τ 2] is the dependent variable. For Models 1 and 2 we use the event window [0;+1], and for the Models 3 and 4 we use the event window [−4;+10]. The independent variable SIZE is the logarithm of the market value of a company in local currency in 2015, the year prior the event. MTBR is the market-to-book-ratio in 2015. PERFORMANCE is the past stock performance of the year before the event. DIVERSIFICATION is the number of 4-digit SIC codes a firm has. ROAD is defined as 1, if the company’s primary SIC is 4212,4213 or 4215, 0 otherwise. WATER is defined as 1, if the firms primary SIC is 4412, 4424, 4491 or 4499, 0 otherwise. AIR is defined as 1, if the firms primary SIC is 4512 or 4513. UK is defined as 1, if the company’s headquarter is located in the UK, 0 otherwise. UK∗ AIR is defined as 1, if the company is from the air transportation subsector and its headquarter is located in the UK, 0 otherwise. The stock market data and balance sheet data are obtained from Datastream. All regressions are checked for heteroscedasticity and the robust t-values are noted in parentheses below the coefficient values. ∗ ∗ ∗ , ∗ ∗ and ∗ indicate statistical significance at the 1%, 5% and 10% level, respectively.
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
27
Table 5 BHAR results. Event window
Buy-and-hold abnormal return
t-test
Skewness-adjusted Johnson test
Mean (%)
t-value
j-value
−2.11 −2.30
−2.61∗ ∗ −2.61∗ ∗
−2.63∗ ∗ ∗ −2.64∗ ∗ ∗
107 107
−12.37 −7.27
−5.50∗ ∗ ∗ −4.34∗ ∗ ∗
−5.52∗ ∗ ∗ −4.36∗ ∗ ∗
21 21
−0.30 −0.91
−0.30 −0.91
86 86
Buy-and-Hold Abnormal Return
two sample t-test
Wilcoxon test
࢞ Mean (%)
t-value
z-score
−5.25∗ ∗ ∗ −3.52∗ ∗ ∗
−5.05∗ ∗ ∗ −4.21∗ ∗ ∗
Median (%)
Panel A: UK and Mainland [−4;+10] −3.05 [+1;+10] −2.81 Panel B: UK [−4;+10] −14.15 [+1;+10] −10.12 Panel C: Mainland [−4;+10] −0.34 [+1;+10] −1.03 Event window
0.02 −1.25
࢞ Median (%)
Panel D: Difference UK and Mainland [−4;+10] −13.81 −12.38 [+1;+10] −9.08 −6.03
Nobs
This table shows a robustness check using the methodology of calculation buy-and-hold abnormal returns. As the matched portfolio we use the Datastream’s value-weighted total return national stock market index of stock i’s country. All BHARs are tested for statistical significance by using the t-test and the skewness-adjusted Johnson test. The stock market reaction to UK and mainland companies are presented in Panel B and Panel C. We also test for statistically significant differences between those two subsamples by using the parametric two-sample t-test and the nonparametric Wilcoxon rank sum test. ∗ ∗ ∗ , ∗ ∗ and ∗ indicate statistical significance at the 1%, 5% and 10% level, respectively.
To identify the logistic specific drivers of the negative value effects we use regression analyses. Table 4 summarizes the results. All models support the findings from the event study. UK based firms are experiencing significantly more negative effects than their peers on the mainland, as shown by the negative UK coefficient. As a second main driver for negative stock reactions we identify companies that are related to the air transportation sector. This may be due to the facts that the Brexit leads to the exclusion of British aviation companies from the Open Skies Agreement which the EU has contracted with a number of third countries (particularly the US), and that UK airlines will lose the opportunity to offer air transportation services within an EU Member State (the so called cabotage). We also find that a higher market value and higher marketto-book-ratio are connected to statistically significant more negative effects, although the economic significance of these factors are only limited. Diversified companies and enterprises engaged in road transportation experience less severe stock price declines than their peers using different infrastructure, as indicated by the positive coefficients DIVERSIFICATION and ROAD. As road transporters often have a more or less domestic business model these firms are less affected by arising borders. Maritime business models are more globally oriented and therefore less Brexit bound. 4. Robustness check To check whether our event study results are robust, we perform a buy-and-hold abnormal return (BHAR) analysis, as suggested in Ritter (1991) and Barber and Lyon (1997). The BHAR for stock i during the event window is calculated as follows:
BHARi,[τ 1,τ 2] =
τ2
[1 + Ri,t ] −
t=τ 1
τ2
[1 + RM,t ]
(4)
t=τ 1
where Ri,t is the actually realized return of company i on day t and RM,t is the return of the benchmark index of company i on day t. The results of the BHAR analysis presented in Table 5 confirm our previous findings. 5. Conclusion Our capital market analysis of the European logistics industry surrounding the Brexit referendum in the United Kingdom on Thursday, 23 June and the final decision announced one day later, shows significant negative value effects. Analyzing 107 logistic companies, we find that the decision to leave the EU has very strong, negative effects on the overall logistics sector, but in particular in the UK. From a subsector perspective air transportation companies are most affected due to the loss of being part of several flight agreements. Furthermore, we find that size and market-to-book-ratio have a statistically significant influence, but diversification and an involvement in the road transport helps to be less affected than specialized competitors. Up to now it is not clear when the British government will submit the official request for leaving the European Union and what the next step of “leaving” exactly means. As a result, the strong revaluation in the market can be obtained, based on the high level of uncertainty in the logistics market.
28
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
References Barber, B.M., Lyon, J.D., 1997. Detecting long-run abnormal stock returns: The empirical power and specification of test statistics. J. Financ. Econ. 43 (3), 341–372. Białkowski, J., Gottschalk, K., Wisniewski, T.P., 2008. Stock market volatility around national elections. J. Bank. Finance 32 (9), 1941–1953. Boehmer, E., Musumeci, J., Poulsen, A.B., 1991. Event-study methodology under conditions of event-induced variance. J. Financ. Econ. 30 (2), 253–272. Brown, S.J., Warner, J.B., 1985. Using daily stock returns: the case of event studies. J. Financ. Econ. 14 (1), 3–31. Corrado, C.J., 1989. A nonparametric test for abnormal security-price performance in event studies. J. Financ. Econ. 23 (2), 385–395. Corrado, C.J., Zivney, T.L., 1992. The specification and power of the sign test in event study hypothesis tests using daily stock returns. J. Financ. Quant. Anal. 27 (03), 465–478. Dodd, P., Warner, J.B., 1983. On corporate governance: a study of proxy contests. J. Financ. Econ. 11 (1), 401–438. Döpke, J., Pierdzioch, C., 2006. Politics and the stock market: evidence from Germany. Eur. J. Polit. Econ. 22 (4), 925–943. Nippani, S., Medlin, W.B., 2002. The 2000 presidential election and the stock market. J. Econ. Finance 26 (2), 162–169. Pantzalis, C., Stangeland, D.A., Turtle, H.J., 20 0 0. Political elections and the resolution of uncertainty: the international evidence. J. Bank. Finance 24 (10), 1575–1604. Pastor, L., Veronesi, P., 2012. Uncertainty about government policy and stock prices. J. Finance 67 (4), 1219–1264. Ritter, J.R., 1991. The long-run performance of initial public offerings. J. Finance 46 (1), 3–27. Santa-Clara, P., Valkanov, R., 2003. The presidential puzzle: political cycles and the stock market. J. Finance 58 (5), 1841–1872.
Contents lists available at ScienceDirect
Finance Research Letters journal homepage: www.elsevier.com/locate/frl
Arising borders and the value of logistic companies: Evidence from the Brexit referendum in Great Britain Artur Tielmann, Dirk Schiereck∗ Department of Business Administration, Economics and Law, Technische Universität Darmstadt, 64289 Darmstadt, Germany
a r t i c l e
i n f o
Article history: Received 21 July 2016 Revised 8 August 2016 Accepted 10 August 2016 Available online 13 August 2016 JEL classification: G10 G14 G15
a b s t r a c t The Brexit referendum may result in new border controls and a separation of Great Britain from the EU and Continental Europe. These consequences will impede the import and export of goods and can therefore have a strong effect on the valuation of logistic companies. We employ event study methodology and regression analysis, examining 107 logistic companies from continental EU countries and Great Britain. While the results indicate an overall negative value effect, UK based companies have a significantly poorer performance than logistic companies from Continental Europe. Companies that focus on the road transport as well as diversified firms are less affected. © 2016 Elsevier Inc. All rights reserved.
Keywords: Brexit Political risk Logistics industry Event study
1. Introduction On 23 June 2016 the British citizens (surprisingly) decided to instruct their government to detach their country from the European Union. This process can have a major impact on the logistics sector when border controls are installed and impede the import and export of goods from and to Great Britain. British companies have to expect additional VAT (value added taxes) and duties. Both factors negatively affect the overall economy in Great Britain and Continental Europe but specifically the logistics sector. Logistics industry is of course not the only industry affected by the Brexit but it is perhaps the most obvious one. According to the UK government, 44 per cent of all British exports are transported to EU countries. It can be expected that the Brexit will create borders over time which will have an impact on the logistics efficiency because of slower movement of goods. Moving goods across borders within the EU is easy and cheap at present. The only documentation needed for transporting goods from one country in the EU to another is a copy of the packing list or commercial invoice and the travel document (waybill, bill of landing or CMR note). At the moment, there is no customs clearance process and no duties applied. VAT doesn’t have to be handed over before the goods can be moved from the receiving port or airport. After the Brexit additional administrative burdens will apply and if the UK is treated like other countries from outside the EU, it will also be necessary to submit customs declarations to the authorities for goods leaving and entering the UK. As transportation follows production, logistics industry can be considered as an indicator and a concentrated aggregate of the overall consequences ∗
Corresponding author. E-mail addresses: [email protected] (A. Tielmann), [email protected], [email protected] (D. Schiereck).
http://dx.doi.org/10.1016/j.frl.2016.08.006 1544-6123/© 2016 Elsevier Inc. All rights reserved.
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
23
the Brexit has on manufacturing. Therefore, we expect to find sustainable negative stock price reactions for logistic firms in the aftermath of the Brexit referendum. As documented by Pastor and Veronesi (2012) changes in government policy affect stock prices and the magnitude of negative returns is large if uncertainty in government policy is high. We choose the Brexit event mainly for three reasons: (i) the outcome was unpredictable until the final day as both sides competed head-to-head (ii) the impact of the referendum result has to be significant as never before an EU member wanted to leave the European Union1 and the conclusions are not assessable yet (iii) a vote for “leave” implies an increase of uncertainty especially in the logistics sector as it is not clear what changes will come up regarding border controls, duties and taxes. There is already considerable empirical evidence indicating that political changes, such as elections, affect overall stock markets. Santa-Clara and Valkanov (2003), for example, examine stock returns in the U.S. presidencies elections. Further, Nippani and Medlin (2002) observe negative stock market reactions on the delay in the declaration of the U.S. presidential election winner of 20 0 0 on the domestic stock market. Indeed, most of the empirical evidence deals with politics in the U.S, but there are also studies with broader perspective or a different regional focus. Białkowski et al. (2008) investigate the stock market volatility surrounding national elections in 27 OECD countries and find that the country-specific component of index return variance can increase significantly, if several factors such as the political orientation of the government change are controlled for. Pantzalis et al. (20 0 0) find positive abnormal returns in the two weeks before elections across 33 countries. In particular, when the incumbent government loses the election, they find positive reactions in the market. And Döpke and Pierdzioch (2006) investigate the impact of political changes in Germany and find only weak evidence for an impact on the stock market. Our analysis tries to derive a more differentiated picture of political change by analyzing single stocks of a heavily affected industry instead of general stock market indices. 2. Sample construction and methodology We focus our analysis on the day after the EU referendum in the United Kingdom, when the national declaration of the result took place at breakfast time the following day. We set the event date accordingly to Friday, 24 June 2016. Since we focus on stock price reactions of logistic companies, we first collect all exchange-listed logistics stocks in the European Union. Then, we drop all firms and observations with illiquid trading patterns and check for confounding events (M&A; ad hoc announcements). This leaves us with a final sample of 107 observations - 21 for British enterprises and 86 companies headquartered in the rest of the 27 member countries of the EU. Table 1 summarizes the sample firms by country (Panel A) and by SIC Code (Panel B). We employ the methodology of the standard market model event study, as introduced by Dodd and Warner (1983) and Brown and Warner (1985). The cumulative abnormal return (CAR) for stock i during the event window [τ 1; τ 2] surrounding the event day t = 0 is calculated as follows: τ2
CARi,[τ 1,τ 2] =
t=τ1
(Ri,t − αˆ i − βˆi RM,t )
(1)
where CARi, [τ 1,τ 2] is the CAR i during the event window, Ri,t is the actually realized return of company i on day t, Rm,t is the return of the benchmark index of company i on day t, and αˆ i and βˆ i are the regression coefficients from an ordinary least squares (OLS) regression using a 252 trading day estimation period. As benchmark indices we use the Datastream’s value-weighted total return national stock market index of stock i’s country. Finally, the average CAR (ACAR) for a sample of N firms is calculated as follows:
ACAR[τ 1,τ 2] =
N 1 CARi, [τ 1,τ 2] N
(2)
i=1
ACARs are calculated for the interval [τ 1; τ 2] ∈ [−4; +10]. In order to test the significance of the ACARs, we apply the parametric test statistic following Boehmer et al. (1991), the BMP-test, and the nonparametric rank test following Corrado (1989), which was later refined by Corrado and Zivney (1992), the CZ-test. Finally, we perform a cross-sectional regression analysis to identify the logistic specific drivers of the CARs. The multivariate ordinary least squares (OLS) regression in its full specification takes the following form:
CARi,[τ 1;τ 2] =
β0 + β1 SIZE + β2 MT BR + β3 PERF ORMANCE + β4 DIV ERSIF ICAT ION + β5 ROAD + β6WAT ER + β7 AIR +β8U K + β9U K ∗ AIR + (3)
where CARi,[τ 1;τ 2] is the dependent variable, SIZE is the logarithm of the market value of a company in EUR in 2015, the year prior the event, MTBR is the market-to-book-ratio in 2015, PERFORMANCE is the past stock performance of the year before the event. DIVERSIFICATION is the amount of 4-digit SIC codes a firm has, ROAD is set to 1, if the company’s primary SIC
1
In fact, Greenland left the predecessor of the EU, the EWG, in 1985.
24
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28 Table 1 Descriptive statistics of the analyzed sample. Panel A: Country
Number of companies
Market value [in Mio. EUR]
Austria Belgium Croatia Cyprus Denmark Estonia Finland France Germany Greece Ireland Italy Luxembourg Malta Netherlands Poland Slovenia Spain Sweden UK
2 4 1 2 5 1 1 14 7 18 1 7 2 1 5 5 1 4 5 21
4341 7119 56 1344 27,639 455 318 44,279 45,832 7710 19,803 19,579 365 327 16,787 1345 322 46,402 768 40,116
Panel B: SIC Code
Number of companies
Description of SIC Code
4011 4212 4213 4215 4225 4226 4231 4412 4491 4499 4512 4513 4581 4731 4789
4 8 6 7 4 1 6 33 6 2 13 3 9 2 3
Railroads, line-haul operating Local trucking without storage Trucking, except local Courier services, except by air General warehousing and storage Special warehousing and storage, not elsewhere classified Terminal and joint terminal maintenance facilities for motor freight transportation Deep sea foreign transportation of freight Marine cargo handling Water transportation services, not elsewhere classified Air transportation, scheduled Air courier services Airports, flying fields, and airport terminal services Arrangement of transportation of freight and cargo Transportation services, not elsewhere classified
This table shows the number of companies and the cumulated market value per country in Panel A and the number of companies per SIC Group in Panel B.
is related to road transport,2 WATER is defined as 1, if the firms primary SIC is related to water transport 3 AIR is defined as 1, if the firms primary SIC is related to air transport, 4 UK is defined as 1, if the company’s headquarter is located in the UK and UK∗ AIR is defined as 1, if the company is focused on air transportation and its headquarter is located in the UK, 0 otherwise. The stock market data and balance sheet data are obtained from Datastream. Table 2 reports Pearson pairwise correlations for all explanatory variables. With exception of the combination UK∗ AIR – AIR the correlation coefficients are low, suggesting that multicollinearity is unlikely to be a problem. 3. Results Table 3 shows the results of the event study for the total sample (Panel A) and the two subsamples, UK based companies (Panel B) and companies from the EU mainland (Panel C). Overall, we document highly significant negative market reactions on the announcement day of the referendum decision to both subsamples and during the [0;+1] event window. However, the British subsample is significantly more affected than the Continental European companies which seems to be plausible as the Continental European logistic firms have more trouble entering one market while British firms are confronted with a more complicated access to all continental European national markets. In the week before the Brexit referendum we do not 2
The SIC Codes 4212,4213 and 4215 are defined as Local Trucking Without Storage, Trucking, Except Local and Courier Services, Except by Air. The SIC Codes 4412,4424, 4491 and 4499 are defined as Deep Sea Foreign Transportation of Freight, Deep Sea Domestic Transportation of Freight, Marine Cargo Handling and Water Transportation Services, Not Elsewhere Classified. 4 The SIC Codes 4512 and 4513 are defined as Air Transportation, Scheduled and Air Courier Services. 3
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
25
Table 2 Correlation matrix.
SIZE MTBR PERFORMANCE DIVERSIFICATION ROAD WATER AIR UK UK∗ AIR
SIZE
MTBR
PERFORMANCE
DIVERSIFICATION
ROAD
WATER
AIR
UK
UK∗ AIR
1 0.030 0.421∗ ∗ ∗ 0.328∗ ∗ ∗ −0.002 −0.275∗ ∗ ∗ 0.133 −0.024 −0.003
1 −0.036 −0.079 0.269∗ ∗ ∗ −0.125 −0.081 0.301∗ ∗ ∗ −0.005
1 0.281∗ ∗ ∗ −0.004 −0.408∗ ∗ ∗ 0.242∗ ∗ 0.096 0.141
1 0.064 −0.218∗ ∗ 0.0314 −0.055 −0.075
1 −0.390∗ ∗ ∗ −0.207∗ ∗ −0.007 −0.109
1 −0.331∗ ∗ ∗ −0.099 −0.175∗
1 0.123 0.528∗ ∗ ∗
1 0.448∗ ∗ ∗
1
This table presents Pearson pairwise correlations of the independent variables. All variables are defined in Section 2. significance at the 1%, 5% and 10% level, respectively.
∗∗∗
,
∗∗
and
∗
indicate statistical
Table 3 Event study results. Event window
Cumulative abnormal return
t-Test
Boehmer test
Corrado-test
Mean (%)
Median (%)
t-value
z-score
z-score
Nobs
−0.75 0.03 −0.89 −2.20 −1.22
−2.33∗ ∗ 0.33 −2.65∗ ∗ ∗ −5.67∗ ∗ ∗ −1.69∗
−2.68∗ ∗ ∗ 0.71 −3.72∗ ∗ ∗ −5.57∗ ∗ ∗ −1.69∗
−0.94 −0.45 −2.09∗ ∗ −3.56∗ ∗ ∗ −0.21
107 107 107 107 107
−8.07 1.27 −3.05 −7.99 −5.62
−4.23∗ ∗ ∗ −0.13 −3.29∗ ∗ ∗ −3.92∗ ∗ ∗ −3.22∗ ∗ ∗
−4.06∗ ∗ ∗ 0.46 −3.27∗ ∗ ∗ −3.92∗ ∗ ∗ −3.21∗ ∗ ∗
−2.03∗ ∗ 0.33 −2.26∗ ∗ −3.57∗ ∗ ∗ −1.99∗ ∗
21 21 21 21 21
0.13 −0.46 −0.62 −1.81 −0.43
−0.68 0.37 −1.47 −4.46∗ ∗ ∗ −0.34
−0.58 0.58 −2.37∗ ∗ −4.49∗ ∗ ∗ 0.21
−0.36 −0.65 −1.63 −2.88∗ ∗ ∗ 0.48
86 86 86 86 86
Cumulative abnormal return
two sample t-test
Wilcoxon test
࢞ Mean (%)
t-value
z-score
−3.41∗ ∗ ∗ −0.24 −2.09∗ ∗ −3.46∗ ∗ ∗ −2.76∗ ∗ ∗
−4.18∗ ∗ ∗ 0.24 −1.71∗ −2.30∗ ∗ −3.27∗ ∗ ∗
Panel A: UK and Mainland [−4;+10] −2.88 [−4;−1] 0.23 [0;0] −1.33 [0;+1] −4.20 [+1;+10] −1.78 Panel B: UK [−4;+10] −11.01 [−4;−1] −0.11 [0;0] −3.41 [0;+1] −9.15 [+1;+10] −7.50 Panel C: Mainland [−4;+10] −0.90 [−4;−1] 0.31 [0;0] −0.82 [0;+1] −3.00 [+1;+10] −0.39 Event window
࢞ Median (%)
Panel D: Difference UK and Mainland [−4;+10] −10.12 −8.20 [−4;−1] −0.42 1.74 [0;0] −2.59 −2.43 [0;+1] −6.15 −6.19 [+1;+10] −7.11 −5.19
This table shows the stock market reaction of the logistics stocks to the declaration of the result from the EU referendum on 24 June 2016. The ACARs are estimated for Panel A and the subsamples Panel B and Panel C in the same 5 event windows. The daily abnormal returns are calculated using the methodology of the market model with a 252 days estimation period. As the benchmark indices we use the Datastream’s value-weighted total return national stock market index of stock i’s country. All ACARs are tested for statistical significance by using the parametric BMP-test and the nonparametric CZ-rank-test. The stock market reaction to UK and mainland companies are presented in Panel B and Panel C. We also test for statistically significant differences between those two subsamples by using the parametric two-sample t-test and the nonparametric Wilcoxon rank sum test. ∗ ∗ ∗ , ∗ ∗ and ∗ indicate statistical significance at the 1%, 5% and 10% level, respectively.
find any significant abnormal reactions. This matches the newspapers’ reports in the week before the poll, when supporters and opponents competed in a close contest, and supports the hypothesis that capital markets did not anticipate the result. Remarkably, we find a second stock market decline for the UK subsample since 30 June on when Boris Johnson withdrew from his candidature as Great Britain’s next prime minister at a press conference on Thursday, 30 June. He left the race by saying: “I must tell you, […] that having consulted colleagues and in view of the circumstances in Parliament, that person cannot be me.” The market reacted immediately to the loss of the most popular face of the Brexit supporters. Fig. 1 illustrates the stock market reactions and points up the significant difference in market reactions between the British and the Continental European subsample.
07/04/2016
07/03/2016
07/02/2016
07/01/2016
06/30/2016
06/29/2016
06/28/2016
06/27/2016
06/26/2016
06/25/2016
06/24/2016
06/23/2016
06/22/2016
06/21/2016
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
06/20/2016
26
2.00% 0.00% -2.00%
ACAR
-4.00% -6.00% -8.00% -10.00% Mainland and UK
UK
Mainland
-12.00% Fig. 1. ACAR performance. This figure shows the ACAR performance of the sample firms from Monday, 20 June (t = −4) to Monday, 04 July (t = 10).
Table 4 Regression results. Dependent variable CARi,[0,1]
Dependent variable CARi,[ − 4,10]
Model 1 SIZE MTBR PERFORMANCE DIVERSIFICATION ROAD WATER
Model 2 ∗∗
−0.016 (−2.00) −0.001∗ ∗ ∗ (−3.48) 0.038 (1.55) 0.004 (1.36) 0.004∗ ∗ ∗ (2.8) 0.017 (1.04)
AIR UK UK∗ AIR CONSTANT Mean VIF Adjusted R2 F-value
−0.173∗ ∗ ∗ (−4.57) −0.007 (−0.29) 1.30 0.289 8.56∗ ∗ ∗
0.004∗ (1.83) 0.031∗ ∗ (2.27) 0.009 (0.55) −0.054∗ ∗ (−1.99) −0.053∗ ∗ ∗ (−2.65)
−0.048∗ ∗ ∗ (−3.2) 1.280 0.187 4.18∗ ∗ ∗
Model 3
Model 4 −0.009 (−0.38) −0.001∗ ∗ (−2.30) 0.020 (0.36) 0.001 (0.19) 0.041∗ (1.92) 0.013 (0.59)
−0.198∗ ∗ ∗ (−4.25) −0.003 (−0.04) 1.30 0.084 4.46∗ ∗ ∗
0.002 (0.41) 0.0210451 (1.11) −0.001 (−0.03) −0.074∗ ∗ ∗ (−3.18) −0.092∗ ∗ ∗ (−3.70)
−0.009 (−0.46) 1.28 0.11 6.27∗ ∗ ∗
This table shows the cross-sectional regression results for the entire sample of 107 observations. CARi,[τ 1;τ 2] is the dependent variable. For Models 1 and 2 we use the event window [0;+1], and for the Models 3 and 4 we use the event window [−4;+10]. The independent variable SIZE is the logarithm of the market value of a company in local currency in 2015, the year prior the event. MTBR is the market-to-book-ratio in 2015. PERFORMANCE is the past stock performance of the year before the event. DIVERSIFICATION is the number of 4-digit SIC codes a firm has. ROAD is defined as 1, if the company’s primary SIC is 4212,4213 or 4215, 0 otherwise. WATER is defined as 1, if the firms primary SIC is 4412, 4424, 4491 or 4499, 0 otherwise. AIR is defined as 1, if the firms primary SIC is 4512 or 4513. UK is defined as 1, if the company’s headquarter is located in the UK, 0 otherwise. UK∗ AIR is defined as 1, if the company is from the air transportation subsector and its headquarter is located in the UK, 0 otherwise. The stock market data and balance sheet data are obtained from Datastream. All regressions are checked for heteroscedasticity and the robust t-values are noted in parentheses below the coefficient values. ∗ ∗ ∗ , ∗ ∗ and ∗ indicate statistical significance at the 1%, 5% and 10% level, respectively.
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
27
Table 5 BHAR results. Event window
Buy-and-hold abnormal return
t-test
Skewness-adjusted Johnson test
Mean (%)
t-value
j-value
−2.11 −2.30
−2.61∗ ∗ −2.61∗ ∗
−2.63∗ ∗ ∗ −2.64∗ ∗ ∗
107 107
−12.37 −7.27
−5.50∗ ∗ ∗ −4.34∗ ∗ ∗
−5.52∗ ∗ ∗ −4.36∗ ∗ ∗
21 21
−0.30 −0.91
−0.30 −0.91
86 86
Buy-and-Hold Abnormal Return
two sample t-test
Wilcoxon test
࢞ Mean (%)
t-value
z-score
−5.25∗ ∗ ∗ −3.52∗ ∗ ∗
−5.05∗ ∗ ∗ −4.21∗ ∗ ∗
Median (%)
Panel A: UK and Mainland [−4;+10] −3.05 [+1;+10] −2.81 Panel B: UK [−4;+10] −14.15 [+1;+10] −10.12 Panel C: Mainland [−4;+10] −0.34 [+1;+10] −1.03 Event window
0.02 −1.25
࢞ Median (%)
Panel D: Difference UK and Mainland [−4;+10] −13.81 −12.38 [+1;+10] −9.08 −6.03
Nobs
This table shows a robustness check using the methodology of calculation buy-and-hold abnormal returns. As the matched portfolio we use the Datastream’s value-weighted total return national stock market index of stock i’s country. All BHARs are tested for statistical significance by using the t-test and the skewness-adjusted Johnson test. The stock market reaction to UK and mainland companies are presented in Panel B and Panel C. We also test for statistically significant differences between those two subsamples by using the parametric two-sample t-test and the nonparametric Wilcoxon rank sum test. ∗ ∗ ∗ , ∗ ∗ and ∗ indicate statistical significance at the 1%, 5% and 10% level, respectively.
To identify the logistic specific drivers of the negative value effects we use regression analyses. Table 4 summarizes the results. All models support the findings from the event study. UK based firms are experiencing significantly more negative effects than their peers on the mainland, as shown by the negative UK coefficient. As a second main driver for negative stock reactions we identify companies that are related to the air transportation sector. This may be due to the facts that the Brexit leads to the exclusion of British aviation companies from the Open Skies Agreement which the EU has contracted with a number of third countries (particularly the US), and that UK airlines will lose the opportunity to offer air transportation services within an EU Member State (the so called cabotage). We also find that a higher market value and higher marketto-book-ratio are connected to statistically significant more negative effects, although the economic significance of these factors are only limited. Diversified companies and enterprises engaged in road transportation experience less severe stock price declines than their peers using different infrastructure, as indicated by the positive coefficients DIVERSIFICATION and ROAD. As road transporters often have a more or less domestic business model these firms are less affected by arising borders. Maritime business models are more globally oriented and therefore less Brexit bound. 4. Robustness check To check whether our event study results are robust, we perform a buy-and-hold abnormal return (BHAR) analysis, as suggested in Ritter (1991) and Barber and Lyon (1997). The BHAR for stock i during the event window is calculated as follows:
BHARi,[τ 1,τ 2] =
τ2
[1 + Ri,t ] −
t=τ 1
τ2
[1 + RM,t ]
(4)
t=τ 1
where Ri,t is the actually realized return of company i on day t and RM,t is the return of the benchmark index of company i on day t. The results of the BHAR analysis presented in Table 5 confirm our previous findings. 5. Conclusion Our capital market analysis of the European logistics industry surrounding the Brexit referendum in the United Kingdom on Thursday, 23 June and the final decision announced one day later, shows significant negative value effects. Analyzing 107 logistic companies, we find that the decision to leave the EU has very strong, negative effects on the overall logistics sector, but in particular in the UK. From a subsector perspective air transportation companies are most affected due to the loss of being part of several flight agreements. Furthermore, we find that size and market-to-book-ratio have a statistically significant influence, but diversification and an involvement in the road transport helps to be less affected than specialized competitors. Up to now it is not clear when the British government will submit the official request for leaving the European Union and what the next step of “leaving” exactly means. As a result, the strong revaluation in the market can be obtained, based on the high level of uncertainty in the logistics market.
28
A. Tielmann, D. Schiereck / Finance Research Letters 20 (2017) 22–28
References Barber, B.M., Lyon, J.D., 1997. Detecting long-run abnormal stock returns: The empirical power and specification of test statistics. J. Financ. Econ. 43 (3), 341–372. Białkowski, J., Gottschalk, K., Wisniewski, T.P., 2008. Stock market volatility around national elections. J. Bank. Finance 32 (9), 1941–1953. Boehmer, E., Musumeci, J., Poulsen, A.B., 1991. Event-study methodology under conditions of event-induced variance. J. Financ. Econ. 30 (2), 253–272. Brown, S.J., Warner, J.B., 1985. Using daily stock returns: the case of event studies. J. Financ. Econ. 14 (1), 3–31. Corrado, C.J., 1989. A nonparametric test for abnormal security-price performance in event studies. J. Financ. Econ. 23 (2), 385–395. Corrado, C.J., Zivney, T.L., 1992. The specification and power of the sign test in event study hypothesis tests using daily stock returns. J. Financ. Quant. Anal. 27 (03), 465–478. Dodd, P., Warner, J.B., 1983. On corporate governance: a study of proxy contests. J. Financ. Econ. 11 (1), 401–438. Döpke, J., Pierdzioch, C., 2006. Politics and the stock market: evidence from Germany. Eur. J. Polit. Econ. 22 (4), 925–943. Nippani, S., Medlin, W.B., 2002. The 2000 presidential election and the stock market. J. Econ. Finance 26 (2), 162–169. Pantzalis, C., Stangeland, D.A., Turtle, H.J., 20 0 0. Political elections and the resolution of uncertainty: the international evidence. J. Bank. Finance 24 (10), 1575–1604. Pastor, L., Veronesi, P., 2012. Uncertainty about government policy and stock prices. J. Finance 67 (4), 1219–1264. Ritter, J.R., 1991. The long-run performance of initial public offerings. J. Finance 46 (1), 3–27. Santa-Clara, P., Valkanov, R., 2003. The presidential puzzle: political cycles and the stock market. J. Finance 58 (5), 1841–1872.