• No results found

The effect of Brexit on the European stock market : a Brexit event study

N/A
N/A
Protected

Academic year: 2021

Share "The effect of Brexit on the European stock market : a Brexit event study"

Copied!
38
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

1

“The effect of Brexit on the

European stock market”

, A Brexit event

study.

June 2017

University of Amsterdam, Amsterdam Business School

MSc Finance, Duisenberg honours programme in Corporate Finance and Banking Master Thesis

Jaap Stolp, 10688242

(2)

2 Abstract

This thesis investigates the initial impact of the Brexit on the stock market of six countries in the European Union, namely the United Kingdom, the Netherlands, Ireland, France, Germany and Ireland. This initial impact is shown via abnormal returns with the use of an event study for a sample of 1824 companies. The results show that the effect is the largest for Ireland, followed by France, Germany, the Netherlands and the United Kingdom. For Switzerland no significant effect is found. The most affected sector is the Oil & Gas sector, which holds for the average, as well as for the United Kingdom. Furthermore, the Consumer Services, Consumer Goods, Industrials and Financials sectors all experience significant negative cumulative average abnormal returns. The abnormal returns for the technology sector is only significant when the average is taken, while it is insignificant for the United Kingdom. For the sectors Basic Materials, Health Care, Telecommunications and Utilities no significant cumulative average abnormal returns are found. The results show that Ireland is affected the most. The sectors who depend on trade, are affected more than sectors who do not. In addition, sectors who depend on the free movement of labor, also experience significant negative returns.

Keywords Brexit – Event Study – European Union – Abnormal Returns – Sectors –

Uncertainty

(3)

3 Table of Contents Abstract 2 Table of Contents 3 1. Introduction 4 2. Literature Review 7 2.1 Uncertainty 7 2.2 Overall Effect 11 2.3 Sectoral Effects 13 2.4 Country Effects 14 2.5 Hypotheses 15 3. Data 16 3.1 Collection of Data 16 3.2 Descriptive Statistics 17 4. Methodology 20 4.1 Event Study 20 4.2 Robustness 22 5. Results 23 5.1 Countries 23 5.2 Sectors 25

5.3 United Kingdom Sectors 27

6. Robustness 33

7. Conclusion & Discussion 33

8. References 35

(4)

4

1. Introduction

It was 2015 when the re-elected David Cameron promised the United Kingdom to hold a binding referendum on the European membership. The 24th of June 2016 it was announced that a majority of 51.9% voted for “leave”, which was surprising, since the polls predicted that the majority would vote to remain. This means that the United Kingdom has to renegotiate its regulatory agreements with Europe. This unexpected outcome created uncertainty among investors and regulators, and will have an impact on varying costs and benefits for countries within Europe, and possibly even outside of Europe. In addition, the impact will also differ across sectors. In the past we have seen Algeria and Greenland leave the European Economic Community (EEC), but no country has ever left the European Union, which means there is no precedent to learn from. Research on the impact of the Brexit is scarce, since not much time has passed since the announcement. Yet, there are no proposals made for what happens after the Brexit, which makes it hard to estimate what the real impact will be. The question that should be asked here is: how big the initial impact of the Brexit is and whether European countries and/or sectors are affected differently by the Brexit. This thesis will try to answer this question by estimating the initial impact of the Brexit on the stock market, by looking at abnormal returns.

The five main issues regarding the Brexit will be explained to create a better understanding of how the Brexit influences Europe. The first issue that arises is what will happen to the businesses and the United Kingdom citizens who live in the Europe. International law states that the rights of the treaty remain, even when the United Kingdom leaves the European Union. This is, however, uncertain since new agreements can changes this law. Secondly, United Kingdom law will change in many aspects. This works two ways. In order to join the United Kingdom, a country has to adjust their laws and regulations to meet the requirements. Leaving the European Union would mean that the United Kingdom is free to change these laws and regulations. Besides laws and regulations implemented before joining the European Union, many new rules and policies, which became law for all members, were implemented in the past years. All these rules can be discarded when the United Kingdom is no longer part of the European Union. The United Kingdom will have to come to an agreement with the European Union regarding economic, social and political rights, and adjust their law according to these rights. Third, all funding policies will cease after the Brexit. This will have a great impact on the agricultural sector of the United Kingdom, since they receive significant funding from the European Union. The United Kingdom will have to decide whether to fund

(5)

5

this sector on their own, or to leave their agricultural sector with a large loss. Philippidis & Hubbard (2001) show that the overall effect will be positive if the United Kingdom leaves the Common Agricultural Policy, however this will increase income inequality. Fourth, is the issue of the transition period; Companies will need time to adjust to all the new policies and regulations, which will be costly. The United Kingdom will have to decide whether to slowly adjust their laws or to instantly adopt new laws. Last and foremost important, is whether there will be free trade and travel between the United Kingdom and the European Union. The removal of free trade and travel will induce great costs for all companies in the United Kingdom. It is utterly important for the United Kingdom to keep the costs of trade and travel low, since an increase in costs can give companies an incentive to move their businesses abroad. It will take time to evaluate every option the United Kingdom government has, regarding the old and new laws and regulations. These five issues will determine the magnitude of the impact of the Brexit on other countries and/or sectors that are connected to the UK and who depend on European law and regulations.

So far, only the issues regarding the Brexit are discussed, which could have a great impact, but there are outcomes were businesses and citizens are only affected mildly. The United Kingdom could decide to adopt the Norwegian or Swiss model, or they could rejoin the European Free Trade Organization (EFTA) or World Trade Organization (WTO). If the United Kingdom only wants to slightly decrease their involvement in the European Union, they could adopt the Norwegian model, meaning that they join the European Economic Area (EEA). This means that the United Kingdom will be part of the Single Market, and is able to freely move goods, services, people and capital. The United Kingdom can regulate their own external tariffs and are able to set up trade negotiations with non-European countries. This does, however, require a membership fee and European Union regulation concerning the Single Market. Another downside is that the United Kingdom will not be able to decide on the rules concerning the Single Market, the European Union sets the rules not the EEA. In short, joining the EEA brings economic advantages, but it has some political costs. In order to make a trade-off, we have to answer the question of what the United Kingdom wants to accomplish with the Brexit. If the United Kingdom wants less regulation but equal economic benefits, this might be a good option. But if the United Kingdom is aiming to become a single entity, this option should be discarded. A second option is to adopt the Swiss model, where the United Kingdom joins the European Free Trade Association (EFTA) and is free to choose which European Union initiatives they want to join via treaties. The United Kingdom won’t be part of the Single Market and will not be able to give their input when the European Union sets new regulations.

(6)

6

Although this approach is more flexible, it gives no certainty that the United Kingdom will be able to freely export their services which could lead to an increase in costs for United Kingdom companies. Baier et al. (2008) confirm that European Union members trade substantially more with other European Union members than they do with EFTA members. They estimate that United Kingdom trade will fall by 25% if they decide to join the EFTA. If the United Kingdom doesn’t want any European Union regulations, they could fall back to the WTO. The WTO is far less advanced concerning free trade and travel compared to the European Union or EEA, this means that the United Kingdom will have reduced access the European market. The European Union will set import tariffs for United Kingdom goods, which will hurt the export of the United Kingdom. Although the United Kingdom won’t have to meet European Union regulations, their goods will have to meet European Union requirements in order to be exported to the European Union. What we know so far is that the United Kingdom wants to leave the European Union, what we don’t know is to what extend they want to leave. It is important to know to what extend the United Kingdom is going to leave the European Union, since this will have an influence on the estimations of the total impact of the Brexit. The initial impact will be based on the expectations investors and regulators have. It could be that another reaction can be seen from the stock market if the future agreements are different from what investors and regulators expect.

The question this thesis tries to answer is whether the stock market in the United Kingdom is affected by the Brexit and if the stock markets in other countries i.e. Germany, the Netherlands, Ireland, France and Switzerland are affected differently. In addition, this thesis also looks at differences between sectors. The aim of this paper is to create a better understanding of the magnitude of the Brexit on stock returns in the United Kingdom, Germany, the Netherlands, Ireland, France and Switzerland, as well as the magnitude of the impact on several sectors. The magnitude of the announcement of leaving the European Union, will be tested by calculating cumulative abnormal returns using the market model. The abnormal returns will be calculated per company, where after the average will be calculated per country and sector, resulting in average abnormal returns. Summing up the average abnormal returns will provide cumulative average abnormal returns. These cumulative average abnormal returns will give an estimation of the initial impact of the Brexit. It is expected that the financial sector of the United Kingdom is negatively affected the most by the Brexit, since the financial sector of the United Kingdom has the largest proportion of foreign direct investment (FDI) (Tyler 2015). Furthermore, sectors of countries with high export to the United Kingdom are expected to have a larger reduction in stock prices, compared to countries with

(7)

7

less export.

This thesis contributes to the literature in the following ways. First of all, it adds to the literature by estimating the initial impact of the Brexit, where a significant impact on almost all countries is found. Although Ramiah, Pham & Moosa (2016) and Dhingra et al. (2016) already provide insight on the effect Brexit has on certain countries, this thesis uses a much larger sample. Secondly, this thesis estimates the initial impact on several sectors. This thesis uses a larger sample and a different classification of sectors compared to Oehler, Horn & Wendt (2017) to make estimation on the initial impact. Thirdly, this thesis adds to the research of Belke, Dubova & Osowski (2016) showing that effect of the Brexit spills over to other countries, by using less countries but more companies. This will result in a more narrow view on the researched countries, since not only the larger companies are used, but all also mid and small companies traded on the main stock exchange.

The remainder of this thesis is structured as follows. After the introduction previous studies will be presented and discussed in the literature review. Secondly, the methodology will be described, followed by a description of the used data. The found results will be presented, where after robustness is tested, and any shortcoming and possible improvements are discussed. Finally, the thesis will be concluded.

2. Literature review

The following section provides insight on how the Brexit affects countries and sectors. In the first section the uncertainty around the Brexit is discussed, including the implications for countries and sectors. The second section shows estimations of other researchers on the total impact of the Brexit. In the third section the previous research on the effect of Brexit on sectors is discussed. The last section provides an overview of which countries are most likely to be affected the most by the Brexit. To conclude, hypothesis will be formed based on the literature. 2.1 Uncertainty

As mentioned above, the Brexit is expected to have a negative influence on stock prices. Bloom (2014) contributes to the debate on how uncertainty affects economic and financial crises. His research tries to answer four questions: What are the stylized facts about uncertainty over time? Why does uncertainty vary? Do fluctuations in uncertainty matter? Has higher uncertainty worsened the Great Depression and recovery? He found that both macro and micro uncertainty rises in recessions. Micro uncertainty comes from firm-level indicators like future sales, future costs and future growth, while macro uncertainty comes

(8)

8 from GDP growth, climate change, volatility and exchange rates. All the uncertainty proxies raise during bad news events and recessions. Furthermore, he found that wages and income volatility appear to be countercyclical. The main cause for this phenomenon is that the lower end of the income distribution collapses during a recession. The last fact Bloom found, is that uncertainty is around a third higher in developing countries, which is caused by higher volatility in GDP growth rates, higher stock-market volatility and higher volatility in bonds. The variations in uncertainty are explained by the following concepts. First of all, bad news events which cause recessions, also increase uncertainty. This can be explained by the fact that predicting the future becomes more difficult when people have to deal with rare events. Another explanation is that recessions themselves increase uncertainty. This is because of slower economic growth during recessions and decreasing trading activity which in turn slows the spreading of information. It is clear that volatility and bad news events increase uncertainty. The question arises whether this matters. Bloom (2014) uses five concepts to explain this. The first concept are the Real-options, these are options companies have to make an investment or to hire a person. When uncertainty is high, the expected payoff of making an investment is low, making companies postpone costly projects. This also affects the costs of financing for projects. A loan will be costlier due higher risk-premium when the outcome of a project is uncertain. Also, a company will most likely increase their precautionary-savings during times of high uncertainty. In order to make future payments, companies save more cash or postpone projects. There are, however, some positive effects. Higher uncertainty may promote investing, since the upside can be higher.1 Besides higher upside, the Oi-Hartman-Abel effect states that companies who are able to expand during booms and contract during recessions, can reach a higher net output Bloom (2014). This does require the ability to expand or decrease production with relatively low costs. Taking all these effects into consideration, the increase in uncertainty did worsen the Great Depression and its recovery. Firstly, the housing collapse increased uncertainty, followed by the inability to estimate the total impact. Secondly, the investors and consumers were unable to estimate the impact of the unconventional monetary policy further raising uncertainty. Bloom concludes that the increase in uncertainty during the Great Recession accounted for one third of the total GD P loss.

Uncertainty is one of the main causes for instability in financial market s. As described above, the Real-options concept will be important for companies in Europe

(9)

9 during times of high uncertainty. The uncertainty caused by the Brexit will most likely spillover to other European countries. It is widely known that during crises and political events, volatility increases and spills over across markets which will affect other countries indirectly Belke et al. (2016). The main channels through which other countries are affected directly are by trade linkages, banking linkages and FDI. Based on trade it is expected that Ireland, the Netherlands and Belgium are affected the most. The spillover effect via the banking sector will be more pronounced due to high interconnectedness in Europe. France, Luxembourg, the Netherlands and Ireland will have more exposure because of high er FDI. Nevertheless, it is hard to estimate which country will be affected the most, since the Brexit influences the financial markets directly and indirectly in many ways. Belke et al. (2016) assess the increase in uncertainty caused by the Brexit on both the United Kingdom and Europe. Furthermore, they assess the impact of the Brexit on stock returns, sovereign credit default swaps, the pound & the euro and 10-year interest rates of 19 predominantly European countries. They use rolling estimation to test for spillover effects, where the total spillover can be attributed to the subprime-mortgage crisis, global financial crisis, sovereign debt crisis and the Brexit. Their results show that uncertainty spills over to other financial markets with magnitudes never seen before. The spillover index after the Brexit is around 30% larger than the previous maximum, which is caused by an increase in policy uncertainty and market volatilities. Moreover, the spillovers were not only of short-term. The effect of increased uncertainty is still noticeable, even after 3 months. In addition, they conclude that the perceived probability of the Brexit had large heterogeneous effects across countries. They believe that the expected decrease in economic activity could jeopardize the sustainability of some governments, where the main found “losers” were the GIIPS countries, namely Greece, Italy, Ireland, Portugal and Spain. These countries already have an ongoing sovereign debt crisis and foresee an even worse future when trading activities with the United Kingdom deteriorate even further.

What we know form the Brexit is that the United Kingdom will no longer be part of the European Union, but, we do not know to what extent. The uncertainty caused by the Brexit comes from the inability to estimate what agreements between the United Kingdom

and the European Union will be removed. Lannoo (2016) explains in his paper what agreements will be discarded when the United Kingdom leaves the European Union, which is the main cause for the increase in uncertainty. Before the financial crisis of 2008, the “Single market” covered basic rules for banking, financial and investment services. Other important elements such as rating agencies, derivative market and hedge funds were not

(10)

10 regulated on the European level. In the past few years these rules have become far more complex, and now cover most of the financial services in the European Union. The “equivalence” assessment is a tool used within the European Union, which recognizes a third country’s legal, supervisory and regulatory regime as equivalent to the framework of the European Union. This tool allows firms to provide services throughout the Europe with a single license, for the United Kingdom this is financial services, which accounts for about 8% of their GDP. The United Kingdom currently hosts 2250 companies using the free provision of services (FPS), so called “passporting” compared to 1000 for all other

European Union countries, which show the importance of “passporting” for the financial sectors in the United Kingdom. Lannoo (2016) concludes that there is an urgent need to shape a new deal, which will make it possible for companies to keep using their passports, instead of splitting their corporation. Besides regulatory consequences, the United Kingdom

will receive much less foreign direct investments (FDI). Today, only China and the United State receive more FDI than the United Kingdom. During the year 2015, the United Kingdom

was the top European destination for investments from emerging markets. Around 2200 inward investment projects were financed through FDI, creating more than 110.000 jobs.2 The uncertainty on the future position of the United Kingdom will have a negative effect on FDI, which will decrease the growth of jobs. Dhingra et al. (2016) explain this by providing three reasons why FDI might fall in the United Kingdom. First, the United Kingdom is an attractive export platform because they operate in the Single Market. This means that the tariff and non-tariff costs are fairly low when exporting to the rest of Europe. When the United Kingdom leaves Europe, these tariffs will most likely increase. Secondly, it would become more difficult for companies to manage their supply chains, because they will be subjected to different regulations and costs. On top of that, the migration of employees will become costlier due to migration controls. Third, the uncertainty on future trade agreements between the European Union and the United Kingdom will have a negative impact on FDI. Dhingra et al. (2016) use a statistical model to estimate why foreign investors invest in the United Kingdom, using data on 34 OECD countries from 1985 to 2013. Their data show that being a member of the European Union has a significant positive effect on inward FDI, which ranges between 14% and 38% depending on the statistical model used, with an average of 28%. They consider a

2 Retrieved from

(11)

11

scenario where the reduction in FDI is caused by the Brexit, which reverts to its current level in 2026, decreasing the total income between 1.8% and 4.3%.

2.2 Overall effect

Pain & Young (2004) contribute to the debate on the influence of the Brexit by assessing the consequences for the United Kingdom when they leave the European Union. They investigate this with the use of the NiDEM macroeconomic model to simulate the withdrawal of the United Kingdom under endogenous fiscal and monetary policies. They consider four broad “shocks”; the effects of a reduction in productive efficiency as a result of lower inward FDI, the impact of increased barriers to trade, the impact of an ex-ante fiscal windfall due to lower net transfers to the European Union and the impact of lower food prices. One important advantage of their approach is that they were able to allow for an endogenous monetary policy. It is important to account for this effect, because of how this policy affects the dynamics of adjustment to withdrawal and the impact on different sectors of the economy. Their paper highlights the following two points. Withdrawal of the United Kingdom from the European Union will cause disruptions of exports, increasing unemployment. On the contrary, the United Kingdom will be able to relax their monetary policy. In long-term, flexible wages and prices can be used to counter the increase in unemployment. While Pain & Young (2004) use the older NiDEM model, Ebell, Hurst & Warren (2016) use the newer NiGEM model to estimate the impact of leaving the European Union on the United Kingdom economy. They focus on four of the best understood economic implications of leaving the European Union; a decrease in trade with European Union members, a modest increase in tariffs, a decrease of FDI inflow and the repatriation of United Kingdom‘s contribution to the European Union. The most important implication is the decrease in trade, which can be seen as a decrease in demand for export products. This causes prices to decrease and will depreciate the Sterling. The depreciation in Sterling will make the export products cheaper for foreign countries, which will have a positive effect on exports in the long-run. The depreciation will in turn decrease imports, which will further deteriorate the terms of trade. This results in a decrease in consumption, wages and GDP. They conclude that the GDP of the United Kingdom will decrease by 2.7% in the long-run.

The impact of the Brexit on total welfare is estimated by Dhingra et al. (2016). They use a quantitative model of international trade to calculate the impact of leaving the European Union on the welfare of the United Kingdom population. The model uses a conventional static approach, where the gains of trade are mainly coming from comparative advantages. The model

(12)

12

does underestimate the welfare gains of staying member of the European Union, because it doesn’t take growth of productivity and other effects of trade into account. The authors do make some rough estimations of how much the model underestimates the welfare, where the actual effect could be double the estimations. In the past decades United Kingdom exports to the European Union have increased considerably, where in 1973 30% of United Kingdom exports went to the European Union which increased to 55% in 2008. HM Treasury (2013) found that the main benefit from leaving the European Union is that the United Kingdom no longer has to subsidize poorer countries which makes up 0.53% of GDP. The main cost of leaving the European Union is caused by a decrease in trade, which is the main focus of Ottaviano et al. (2014). They consider two views, the optimistic view and the pessimistic view; one where the United Kingdom will have the same access to the European Union as Norway and Switzerland have, and one where the costs of trade increase. The increase of costs can be explained by higher tariff barriers and higher non-tariff barriers, which result from new regulations & border control and the removal of the United Kingdom’s vote in further integration of the European Union. Their model estimates an overall welfare loss of -1.13% and -3.09% in the optimistic and pessimistic situation respectively. These estimates do not take the effect of trade on growth into account, thus the static model underestimates the effect. Sampson (2016) show that these static effects could be tripled if the dynamic effects of trade are taken into account. This means that leaving the European Union could lead to a welfare loss of almost 10%.

The research of the HM treasury (2016) adds to the paper of Dhingra et al. (2016), where they estimate the alternatives the United Kingdom has and the long term impact of these alternatives on the United Kingdom economy. HM Treasury considers three alternative models, the Norwegian model, the Swiss model or re-joining the WTO. They use the widely adopted gravity model to determine how the alternatives have an effect on trade and FDI, followed by an estimation of the consequences on productivity and GDP. They make some cautious, but realistic assumptions so that they are able to produce robust estimates. Their results show that United Kingdom GDP will decrease by -3.8% in the long-run if they adopt the Norwegian model, -6.2% for the Swiss model and -7.5% when the United Kingdom re-joins the WTO, which translates to a decrease of –£2.600, –£4.300, –£5.200 per household respectively. Furthermore, the decrease in tax receipts will outweigh any reduction in European Union contribution. They conclude that the overall benefit of European Union membership is significantly higher than any alternative.

(13)

13

2.3 Sectoral effects

The study of Ramiah et al. (2016) investigates the sectoral effects of Brexit on the British economy using the same method as Ramiah et al. (2013), by using event study methodology. They argue that besides costs of new tariffs and less access to the European market, there are some benefits to take into consideration. The United Kingdom can avoid the European regulations, save on contribution and are able to sign new deals with non-European countries. Although the long-term net effect is unclear, the short-term effect can be estimated. They test for price reactions in several sectors of the British stock market, where they allow for possible over- or underreaction by using cumulative abnormal returns for 10 trading days after the outcome. Their results confirm the proposition that Brexit has varying sectoral effects. The systematic risk increased for banking, construction and materials, household goods and home construction, real estate investment and services, real estate investment trusts, support services and travel and leisure sectors, while the systematic risk declined for the mining and tobacco industries. Interestingly, the other 31 sectors had no change in systematic risk. While it is too soon to predict long-term effects, the short-term reaction show significant differences between sectors. Milas et al. (2016) use the methodology of Knight (2006) to test for “winners” and “losers”, by calculating abnormal returns of stock listed on the FTSE, and test whether changes in Brexit sentiment have a statistically significant effect on these returns. They test for significant differences compared to the market, meaning that the results are only relative to the market instead of absolute. They use bookmakers’ odds as a measure of Brexit sentiment instead of polls. Bookmakers’ odds usually are more available, more active and in some cases are better predictors, for example in the case of the 2014 Scottish referendum. Their findings show that there are four times as many losers as winners, where a large share of the companies report Europe as an important business area. The finance sector had the most winners and losers, suggesting that the finance sector is affected the most by the Brexit. Construction, transportation and hospitality all had relatively more losers than winners. Although their methodology does not allow them to state the absolute difference, their results show that companies who depend on export are affected more than companies who do not.

Tielmann & Schiereck (2017) use event study methodology to examine the effect of Brexit on 107 logistics companies in the United Kingdom and Europe. They argue that new implemented border control will have a direct impact on the logistics industry, which can be seen as an indicator of the overall consequences of the Brexit, as around 44% of United Kingdom exports are transported to European countries. This industry is heavily dependent on

(14)

14

the terms of trade and when new border control is introduced, importing and exporting of products will be more expensive, worsening the terms of trade. They focus their study on the day after the referendum, by calculating the average abnormal returns of all the exchange-listed logistic companies in the European Union, using the market model. Their results show significant negative market reactions for all logistic industries, where the United Kingdom experiences the most negative effect. In addition to the market model abnormal returns, a buy-and-hold abnormal return analysis is performed as a robustness test, which confirm their previous findings. They conclude that the decrease in stock prices for this sector is driven by the increase in uncertainty.

2.4 Country effects

So far we have discussed the possible consequences of the Brexit on the whole economy of the United Kingdom, the possible consequences for ten sectors and the main causes for these consequences. Yet, there will also be consequences for the rest of Europe and possibly even outside of Europe. This thesis will focus on several European countries, namely the countries Netherlands, France, Germany, Ireland and Switzerland, where the bilateral trade with the United Kingdom is the largest. So far during 2017, 36% of all United Kingdom exports went to these countries. 3 With no doubt, the introduction of trade barriers will not only affect the United Kingdom, but other countries in Europe as well. Burke (2016) argues that the exposure of the Brexit on other countries will be severe, especially for Northern Ireland. As of yet, the future consequences for Northern Ireland have been ignored, while the Brexit can be seen as a threat for the Irish economy, the political relation and the cross-border cooperation. Today, the area of trade already is negatively affected by large margins and rapid fluctuations of currencies, which will deteriorate further when trade agreements come to an end. The Irish government estimated that 200.000 United Kingdom jobs depend on the trade with Ireland, and foresee a sharp decrease in jobs when the United Kingdom leaves the European Union. Officials also state that uncertainty will increase during the negotiations of new bilateral agreements, causing economic losses for both countries. Furthermore, the European support like program funds, will come to a stop resulting in more economic losses. Barret et al. (2015) support this view. They state that the bilateral trade between the United Kingdom and Ireland may diminish by roughly 20% and possibly more. The impact will differ across sectors. Export

3 Exports of UK march 2017 retrieved from

(15)

15

dependent sectors, like basic materials, food and beverages, suffer more losses. There are also indirect consequences for Ireland. As stated before, the FDI will diminish for the United Kingdom, slowing growth. This is likely to affect the growth of Ireland too. Overall, the effect of the Brexit on Ireland is expected to be severe.

There are different ways to estimate the effect of Brexit. Macroeconomic models usually are unable to include the effect of uncertainty. Bouoiyour & Selmi (2016) use uncertainty as a basis for their research, introducing the concept of internet concern. They use online information from Google Trends and Twitter, to predict the costs of the Brexit for France, Germany and the United Kingdom with the use of the quantile regression approach. Weekly data is used from January 2010 to July 2015, for a total of 268 observations, for the FTSE100, CAC40 and DAX30. First, their results show the benchmark effect of simple OLS regression, where the Brexit seems only significant for the United Kingdom and France. Second, their quantile regression estimations show a significant negative effect of the Brexit on the United Kingdom, France and Germany, with a more pronounced effect on Germany. Other explanatory variables, as gold price, VIX and WTI, are included to avoid omitted variable bias. The results show that the VIX has a significant negative impact on European stock returns, suggesting that an increase in uncertainty decrease stock prices. Both the methods of using Google Trends and Twitter data are fairly robust and show the same effect of Brexit on stock returns. Although this method is unable to estimate the full effect of the Brexit, the results are relatively intuitive, where the uncertainty exerts negative impact on Germany, France and the United Kingdom. Concluding, the market reactions on the increase uncertainty are sharply heterogeneous among the tested quantile levels and frequencies. Furthermore, the costs are not uniformly distributed among the investigated countries, where Germany is most likely to suffer more.

2.5 Hypotheses

Based on the discussed literature some hypotheses will be formed which will be tested to estimate the initial impact of the Brexit. Three tests will be performed to estimate this initial impact. Firstly, the impact on the countries will be estimated. Followed by an estimation of the initial impact on each sector. Concluded by an estimation of the initial impact on each sector within the United Kingdom. It is expected the Brexit influences all countries and sectors negatively.

(16)

16

As Burke (2016) and Barret et al. (2015) argue, the impact on Ireland is expected to be the largest.

𝐻0: 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝐼𝑟𝑒𝑙𝑎𝑛𝑑

< 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑖 𝐻1: 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝐼𝑟𝑒𝑙𝑎𝑛𝑑

≥ 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑖

As Hurst & Warren (2016), Barret et al. (2015), Tielmann & Schiereck (2017) and Belke et al. (2016) state, sectors who depend on European regulations and trade with the European Union are expected to suffer the most negative returns.

𝐻0: 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 < 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑖 𝐻1: 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠

≥ 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑖

The negative returns are expected to be larger for the Financials sector within the United Kingdom. 𝐻0: 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑈𝐾−𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 < 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑖 𝐻1: 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑈𝐾−𝐹𝑖𝑛𝑎𝑛𝑐𝑖𝑎𝑙𝑠 ≥ 𝐶𝑢𝑚𝑢𝑙𝑎𝑡𝑖𝑣𝑒 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝐴𝑏𝑛𝑜𝑟𝑚𝑎𝑙 𝑅𝑒𝑡𝑢𝑟𝑛𝑖 3. Data

This section provides information on the used data and the collection of data, followed by a description of the used data, where an overview is given for all sectors and countries.

3.1 Collection of data

Hurst & Warren (2016), Barret et al. (2015), Tielmann & Schiereck (2017) and Belke et al. (2016) all agree on the fact that countries and sectors who depend on trade will be affected the most by the Brexit. Thus it is most interesting to investigate countries with whom the United Kingdom trades the most. The countries which trade the most with the United Kingdom are;

(17)

17

Germany, the Netherlands, Ireland, France and Switzerland4. Daily stock data over the period January 2015 to December 2016 is downloaded from Datastream, including names, stock prices and Sic codes for the Euronext Amsterdam, Euronext Paris, Deutsche Boerse, London Stock Exchange, Irish Stock Exchange and Swiss Stock Exchange. All stocks traded on the exchanges are used so that not only the effect on the larger companies is estimated, but also the effect on smaller companies. The stock price represents the official closing price, the name is the stored name of the company in Datastream and the industry code used is the SIC 1 code. This SIC code is developed by the US government to provide a standard industry classification which covers all the economic activities of that company. This SIC code will be used to assign companies to certain sectors, which allows to group companies and make estimation on each sector. A company can have up to eight Sic codes, where Sic code 1 represents the business segment which earns the most revenue up to Sic code 8 which earns the least revenue. It is assumed that the business segment with the largest revenue is representative for the whole company. The STOXX 600 index represents 600 large, mid and small companies across 17 countries of the European region, which gives a good representation of the average return in Europe or the “market”. This index is used to calculate the market return, which is also retrieved from Datastream.

Stata is used to perform all data manipulations and calculations. The company returns are calculated using the differential in the log of the stock prices:

𝐶𝑜𝑚𝑝𝑎𝑛𝑦 𝑅𝑒𝑡𝑢𝑟𝑛𝑖𝑡 = log(𝑃𝑖,𝑡) − log(𝑃𝑖,𝑡+1)

Where after the company returns are winsorized within the estimation window, using the winsor2 command in Stata. This is done to remove some unexplainable price fluctuations, which could bias the results.

3.2 Descriptive Statistics

Table 1 provides summary statistics for the sectors within the Dutch, French, German, English, Irish and Swiss stock exchanges, for a total of 1848 companies. The London stock exchange is, with 632 listed companies, the largest exchange which makes up for 34.20% of the sample. Within the London stock exchange, the financial sector is the largest with 292 companies, representing 15.80% of the total sample and almost 50% of the exchange itself.

(18)

18

The Euronext Paris is the second largest exchange with 509 companies, where the largest sectors are financials and industrials with 5.68% and 5.09%, respectively. For the Deutsche Boerse the industrial sector is the largest, representing 3.84% of the total sample. A total of 227 companies are listed on the Swiss stock exchange where the financial and industrial sectors are the largest, consisting of 4.11% and 3.14%, respectively. There are no oil & gas companies listed on the Swiss stock exchange. The Euronext Amsterdam consists of 124 companies, not including any utilities companies. The largest sector is the financial sector consisting of 1.62%, followed by the industrial sector of 1.35% of the total sample. The Irish stock exchange is the smallest exchange with only a few listed companies. The largest sectors of the Irish stock exchange are the financial and consumer service sectors, consisting of 0.49% and 0.42%, respectively. The total sample consists of 1848 companies, where the largest sector is the financial sector, consisting of 30.09% of all the companies.

Sector # % # % # % # % # % # % # % Basic Materials 8 0.43% 21 1.14% 19 1.03% 28 1.52% 6 0.32% 11 0.60% 93 5.03% Consumer Goods 15 0.81% 71 3.84% 37 2.00% 38 2.06% 7 0.38% 20 1.08% 188 10.17% Consumer Services 14 0.76% 63 3.41% 39 2.11% 94 5.09% 8 0.43% 14 0.76% 232 12.55% Financials 30 1.62% 105 5.68% 44 2.38% 292 15.80% 9 0.49% 76 4.11% 556 30.09% Health Care 10 0.54% 56 3.03% 30 1.62% 21 1.14% 2 0.11% 27 1.46% 146 7.90% Industrials 25 1.35% 94 5.09% 71 3.84% 112 6.06% 5 0.27% 58 3.14% 365 19.75% Oil & Gas 5 0.27% 9 0.49% 12 0.65% 17 0.92% 3 0.16% 0 0.00% 46 2.49% Technology 15 0.81% 75 4.06% 48 2.60% 17 0.92% 2 0.11% 14 0.76% 171 9.25% Telecommunications 2 0.11% 3 0.16% 6 0.32% 6 0.32% 1 0.05% 2 0.11% 20 1.08% Utilities 0 0.00% 12 0.65% 7 0.38% 7 0.38% 0 0.00% 5 0.27% 31 1.68% Total 124 6.71% 509 27.54% 313 16.94% 632 34.20% 43 2.33% 227 12.28% 1848 100%

Sample summary statistics for the sample.

The table provides summary statistics for the 10 sectors for a sample of 1848 companies for the years 2015 and 2016. The sample selection is based on all the companies traded on the Euronext Amsterdam, Euronext Paris, Deutsche Boerse, London Stock Exchange, Irish Stock Exchange and Swiss Stock

Exchange. The data are taken from DataStream.

Table 1

Ireland Switzerland Total Stock Exchange

(19)

19

Table 2 provides the standard deviations of the company returns for all sectors and exchanges. It shows that the standard deviations of the company returns after the Brexit are higher

Country Population Estimation

window % Reduction Netherlands 0.023 > 0.020 -15% France 0.025 > 0.019 -22% Germany 0.024 > 0.020 -14% United Kingdom 0.018 > 0.016 -11% Ireland 0.037 > 0.022 -40% Switzerland 0.022 > 0.017 -23%

Sector Population Estimation

window Basic Materials 0.027 > 0.020 -25% Consumer Goods 0.022 > 0.018 -19% Consumer Services 0.022 > 0.018 -17% Financials 0.020 > 0.016 -19% Health Care 0.024 > 0.019 -22% Industrials 0.022 > 0.018 -17%

Oil & Gas 0.029 > 0.024 -19%

Technology 0.024 > 0.021 -14%

Telecommunications 0.019 > 0.017 -15%

Utilities 0.019 > 0.019 -2%

United Kingdom - Sectors Population Estimation window Basic Materials 0.024 > 0.023 -6% Consumer Goods 0.016 > 0.014 -10% Consumer Services 0.017 > 0.016 -10% Financials 0.017 > 0.014 -13% Health Care 0.020 > 0.018 -11% Industrials 0.018 > 0.016 -10%

Oil & Gas 0.023 > 0.021 -9%

Technology 0.023 > 0.019 -15%

Telecommunications 0.014 > 0.013 -4%

Utilities 0.016 > 0.016 -3%

Table 2

Showing standard deviations for the population and the estimation window of 250 days. Brexit increased the standard deviations across all Exchanges and Sectors.

(20)

20

compared to the standard deviations of the company returns of the whole population. As Belke et al. (2016) shows, this increase in volatility can cause spillover effects. It is expected that the spillover is the most pronounced for Ireland. The population standard deviation for the exchanges is the largest for Ireland with 0.037, and the lowest for the United Kingdom with 0.018. The Oil & Gas sector has the largest standard deviation of all sectors with 0.029. The lowest standard deviation found is 0.019 for both telecommunications and utilities. Within the London Stock Exchange, the largest standard deviation found was 0.24 for basic materials, close to Technology and Oil & Gas with 0.023.

As mentioned above, the increase in uncertainty positively affects the volatility, which in turn can cause spillover effects. So it is expected that the effect of the Brexit is most pronounced for the sectors and exchanges where volatility increased the most. Of the six exchanges, the volatility in Ireland changes the most, thus it is expected that Ireland will experience the largest negative abnormal returns. For the sectors, it is expected that basic materials, consumer goods, consumer services, financials, health care, industrials and oil & gas all experience more or less the same negative abnormal returns based on the increase in volatilities. For the United Kingdom the expectations are somewhat the same across sectors. Here, based on the increase in volatility, the Technology sector is expected to experience the largest negative abnormal returns, followed by Consumer Goods, Consumer Services, Financials, Health Care, Industrials and Oil & Gas.

4. Methodology

4.1 Event study

The effect of the Brexit on countries and sectors will be tested via an event study, where the standard market model is used to estimate the daily expected returns:

𝑅𝑖,𝑡 = 𝛼𝑖+ 𝛽𝑖𝑅𝑚,𝑡+ 𝜖𝑖,𝑡

Where 𝑅𝑖𝑡 and 𝑅𝑚𝑡 are the rates of return of company, 𝑖 and the European STOXX return at date 𝑡, respectively. The error term 𝜖𝑖𝑡 is assumed to be uncorrelated across firms, have a zero mean and is independent of 𝑅𝑚𝑡. An OLS regression is run to estimate the normal returns. In order to overcome autocorrelation and heteroscedasticity in the error terms, Newey-West HAC adjusted standard deviations are used. Abnormal returns are calculated as the difference between the realized return and the estimated return from the OLS regression:

(21)

21

𝐴𝑅𝑖,𝑡 = 𝑅𝑖,𝑡− 𝛼̂ − 𝛽𝑖 ̂ 𝑅𝑖 𝑚,𝑡

The parameters of the market model are estimated using an event window of 250 trading days, beginning one year before the event date on 24th of July 2015, where the event day is the 24th of July 2016. Average standardized abnormal returns are calculated to determine whether the average abnormal returns are significantly different from zero as where 𝜎̂ is the 𝑖𝑡 sample standard deviation of the abnormal returns within the estimation period.

The average abnormal returns per country and sector are calculated and standardized in the following way: 𝐴𝐴𝑅𝑖,𝑡 = 1 𝑁∑ 𝐴𝑅𝑖,𝑡 𝑁 𝑖=1 𝐴𝑆𝐴𝑅𝑖,𝑡 = 𝐴𝐴𝑅𝑖,𝑡 𝜎̂𝑖,𝑡 ~𝑡(𝑇 − 2)

A standard cross-sectional test will not give the right results, due to a sharp rise in volatility caused by the Brexit. This will result in a low power of the test. Table 2 clearly shows that the standard deviation of company returns after the Brexit are way higher, up to 40%, compared to the standard deviation of company returns in the estimation window.

The cumulative average abnormal returns are calculated for each sector and exchange to see what the total effect of the Brexit is for every individual exchange or sector. The cumulative average abnormal returns will also be calculated for every sector within the London stock exchange. This will allow a comparison between the effect on the whole sector and the effect on the sector within the United Kingdom. The same standard deviation as mentioned above is used to standardize the cumulative abnormal returns, where the cumulative abnormal returns over the event window are calculated as follows:

𝐶𝐴𝑅𝑖(𝑡1, 𝑡2) = ∑ 𝐴𝑅𝑖𝑡 𝑡2

𝑡=𝑡1

Taking the average of the cumulative abnormal returns result in the cumulative average abnormal returns for each exchange and sector:

(22)

22

𝐶𝐴𝐴𝑅𝑖,𝑡 = 1

𝑁∑ 𝐶𝐴𝑅𝒊,𝒕 𝑁

𝑖=1

The cumulative average abnormal returns will be standardized where after a T-test is performed to determine the significance of the average standardized cumulative abnormal returns.

𝐴𝑆𝐶𝐴𝑅𝑖,𝑡 =𝐶𝐴𝐴𝑅𝑖(𝑡1, 𝑡2)

𝜎̂ (𝑡𝑖 1, 𝑡2) ~𝑡(𝑇 − 2)

The hypothesis that the Brexit influences the stock market will be measured by testing if the average abnormal returns and the cumulative average abnormal returns are non-zero. The average abnormal returns will be tested for significance on the 24th of June, where the

cumulative average abnormal returns will be tested on the 24th of June as well as on the 8th of July. This will not only show the one-day effect of the Brexit, but also the 10-trading day cumulative effect. This will be done for the six stock exchanges, the ten sectors as well as the ten sectors within the London Stock Exchange.

𝐻0: 𝐸(𝐴𝐴𝑅𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒,𝑡) = 0 𝐻0: 𝐸(𝐴𝐴𝑅𝑆𝑒𝑐𝑡𝑜𝑟,𝑡) = 0

𝐻0: 𝐸(𝐶𝐴𝐴𝑅𝐸𝑥𝑐ℎ𝑎𝑛𝑔𝑒,𝑡) = 0 𝐻0: 𝐸(𝐶𝐴𝐴𝑅𝑆𝑒𝑐𝑡𝑜𝑟,𝑡) = 0

The outcome of the abnormal returns will give insight on where the initial impact of the Brexit is the largest, for countries, sectors as well as sectors within the United Kingdom.

4.2 Robustness

A non-parametric equality-of-medians test is performed to test for differences in median of company returns between the exchanges and the sectors. This non-parametric test is robust against outliers and extreme values, so this test will have a higher power compared to the non-parametric Kruskal-Wallis test. This will test the null-hypothesis that K samples were drawn from the population with the same median. The chi-squared test will test for significance. The table is presented in the appendix.

(23)

23

5. Results

5.1 Countries

Figure 1 provides a 20-trading day window for the cumulative average abnormal returns for the Euronext Amsterdam, Euronext Paris, Deutsche Boerse, London Stock Exchange, Irish Stock Exchange and Swiss Stock Exchange. The event day is the 24th of July 2016. The CAAR is the cumulative sum of the average abnormal returns per country.

Figure 1

First of all, in the ten trading days prior to the event day all exchanges seem to have negative abnormal returns, resulting in negative cumulative average abnormal returns for all exchanges. These results are only significant for Ireland. Secondly, the United Kingdom cumulative average abnormal return is the only cumulative return reaching above zero. The United Kingdom cumulative abnormal return sharply drops around the event day, and stabilizes slightly in the following days, ending around -5%. For France, the cumulative average abnormal return seems stable before the event day. Of the six exchanges, France experiences the second largest decrease on the event day. In the following days the cumulative average abnormal return of France stabilizes, ending with the second smallest negative cumulative average abnormal. The effect seems much more significant for the Irish Stock Exchange.

-. 1 5 -. 1 -. 0 5 0 .0 5 C u mu la ti ve a ve ra g e a b n o rma l re tu rn s in % -10 -5 0 5 10 Event Day

United Kingdom Netherlands

France Germany

Ireland Switzerland

(24)

24

Within three trading days the Irish Stock Exchange reaches a cumulative average abnormal return of less than -10%, decreasing further until trading day five. In the following days the Irish cumulative average abnormal return stabilizes only slightly. The Dutch cumulative average abnormal return doesn’t seem to be affected that much by the Brexit on the event day itself. However, in the following days the Dutch cumulative average abnormal return slowly decreases, ending below France. For Germany the Brexit seems to initially have a positive effect. Although the initial effect seems positive, the cumulative average abnormal return does decrease in the following days. Brexit has the smallest effect on the Swiss Stock Exchange. The Swiss cumulative average abnormal return decreases slightly after the event day, but the effect isn’t as significant as it is on the other stock exchanges. Overall, for a 20-trading day window the effect of the Brexit seems the largest for the Ireland, while the effect on Switzerland is the smallest. As stated in the hypotheses, the impact on Ireland was expected to be the largest. This is reflected by the results, confirming the hypotheses. Furthermore, Germany is the country which is the second largest “loser”, which also follows the expectations. This confirms the expectations of Burke (2016) and Barret et al. (2015). The effect on Switzerland is also as expected. Switzerland is the least dependent on European regulation, thus will not be affected as much by the Brexit.

(25)

25

5.2 Sectors

Figure 2 provides a 20-trading day window for the cumulative average abnormal returns for the sectors: Basic Materials, Consumer Goods, Consumer Services, Financials, Health Care, Industrials, Oil & Gas, Technology, Telecommunications and Utilities. The event day is the 24th of July 2016. The CAAR is the cumulative sum of the average abnormal returns per sector.

Figure 2

The only sector experiencing positive cumulative average abnormal returns is the Utilities sector. The small decrease on the event day is quickly countered by an increase in the following days. The effect on the Utilities sector seems ambiguous, but after ten trading days the cumulative average abnormal returns does fall below zero. The Telecommunications sector is the least affected by the Brexit. The effect of the Brexit on the event day is so small that it stabilizes in the following days. Consumer Services is the second most affected sector. The cumulative average abnormal return drops sharply on the event day, where after it further decreases, reaching below a cumulative average abnormal return of -5%. The Financials sector follows more or less the same trend as the Industrials sector. As expected, the cumulative average abnormal return decreases sharply for both sectors on the event day. Both sectors end

-. 1 -. 0 5 0 .0 5 C u mu la ti ve a ve ra g e a b n o rma l re tu rn s in % -10 -5 0 5 10 Event Day

Basic Materials Consumer Goods

Consumer Services Financials

Health Care Industrials

Oil & Gas Technology

Telecommunications Utilities

(26)

26

up at less than -5% ten trading days after the event day. Consumer Goods and Technology follow the same trend as Financials and Industrials, only the effect seems smaller. Both the Technology as the Consumer Goods sector reach a cumulative average abnormal return of slightly above -5% after ten trading days. The Health Care sector does not seem to be affected by the Brexit. On the event day the cumulative average abnormal return does not change for the Health Care sector. The sector does experience negative abnormal returns, resulting in a small but negative cumulative average abnormal return. The Basic Materials sector also is not affected on the event day itself, but does experience negative abnormal returns thereafter. At the end of the estimation window, the Basic materials cumulative average abnormal return reaches around -2.5%. The Oil & Gas sector seems to be affected the most by the Brexit. Ten trading days before the event day the sector already experiences the largest negative cumulative average abnormal returns. Although this does stabilize, the decrease on the event day is by far the largest. Even after the event day the Oil & Gas cumulative average abnormal returns decreases further, reaching around -7.5% ten trading days after the event window. Overall, the effect of the Brexit seems rather small on the event day itself. However, in the ten trading day-period after the 24th of June almost all sectors experience more negative abnormal returns. These results do not confirm the hypothesis that the Financials sector suffers the largest losses. Nevertheless, the results do confirm the expectations that sectors who depend on exports or the use of “passports”, suffer more losses than sectors who do not. This confirms the conclusion of Lannoo (2016) & Milas et al. (2016).

(27)

27

5.3 United Kingdom Sectors

Figure 3 provides a 20-trading day window for the cumulative average abnormal returns for the sectors: Basic Materials, Consumer Goods, Consumer Services, Financials, Health Care, Industrials, Oil & Gas, Technology, Telecommunications and Utilities for the London Stock Exchange. The event day is the 24th of July 2016. The CAAR is the cumulative sum of the average abnormal returns per sector in the United Kingdom.

Figure 3

Within the United Kingdom there are two sectors reaching above a cumulative average abnormal return of zero, namely, Utilities and Telecommunications. Nevertheless, these sectors do fall below zero ten trading days after the event day. The Utilities sectors does not seem to be influenced by the Brexit. It’s cumulative average abnormal return fluctuates around zero, whereas the cumulative average abnormal return of the Telecommunications sector decreases sharply after the event day. Health Care, Consumer Goods and Technology seem to follow the same trend. All three sectors experience a sharp decrease in their cumulative average abnormal returns, where after they revert slightly until the fifth trading day after the event day. Of these three sectors, Technology stabilizes the most, and reverts back to almost zero. The increase after trading day eight could be caused by the World Congress of Engineering, no other specific

-. 1 -. 0 5 0 .0 5 C u mu la ti ve a ve ra g e a b n o rma l re tu rn s in % -10 -5 0 5 10 Event Day

UK Basic Materials UK Consumer Goods UK Consumer Services UK Financials

UK Health Care UK Industrials

UK Oil & Gas UK Technology

UK Telecommunications UK Utilities

(28)

28

cause is found. After the fifth trading day, almost all sectors experience another drop. On the 28th of June the European Council got together do discuss the implications of the Brexit. Their conclusions are most likely the cause for this second drop. The sectors which are affected the most are Financials, Industrials, Oil & Gas and Consumer Services. These sectors experience large decreases on the event day. Again, these sectors see a small increase until the fifth trading day after the event day, where after they decrease again. The Oil & Gas sector is the sector with the largest negative cumulative average abnormal return ten trading days after the event day. The hypothesis that the Financials sector in the United Kingdom suffers the largest losses does not hold. Just as the Sectors across Europe, the sectors who depend on exports suffer more losses.

(29)

29

*p<.1, **p<.05, ***p<.01 AAR: Average abnormal return, CAAR: Cumulative average abnormal return.

Table 3 provides average abnormal returns per exchange, sector and sector within the London Exchange on the event day. Furthermore, it shows the cumulative average abnormal return per exchange, sector and sector within the London Exchange on the event day and ten trading days

Table 3

Average abnormal returns and cumulative average abnormal returns per exchange and sector.

Date

24/06/2016 24/06/2016 08/07/2016

Country AAR T-stat CAAR T-stat CAAR T-stat

Netherlands -0.49% -0.25 -1.04% -0.52 -4.68%** -2.37 France -1.61% -0.83 -3.49%* -1.81 -4.38%** -2.27 Germany 0.66% 0.33 -1.41% -0.70 -5.49%*** -2.71 United Kingdom -2.85%* -1.80 -2.40% -1.51 -5.30%*** -3.35 Ireland -3.60% -1.62 -4.79%** -2.16 -8.75%*** -3.94 Switzerland 0.58% 0.35 -1.04% -0.62 -2.39% -1.43

Sector AAR T-stat CAAR T-stat CAAR T-stat

Basic Materials -0.34% -0.17 -1.94% -0.95 -2.39% -1.17 Consumer Goods -0.99% -0.56 -1.65% -0.92 -4.14%** -2.32 Consumer Services -2.13% -1.18 -2.63% -1.46 -5.95%*** -3.29 Financials -1.56% -0.95 -2.31% -1.41 -5.14%*** -3.13 Health Care -0.11% -0.06 -2.45% -1.29 -2.95% -1.55 Industrials -1.72% -0.94 -2.83% -1.55 -5.25%*** -2.87

Oil & Gas -1.71% -0.73 -3.17% -1.35 -8.00%*** -3.40

Technology -0.84% -0.41 -1.83% -0.89 -4.36%** -2.12

Telecommunications -0.37% -0.22 -2.36% -1.43 -1.58% -0.96

Utilities -1.68% -0.90 -0.64% -0.34 -1.83% -0.98

United Kingdom - Sectors AAR T-stat CAAR T-stat CAAR T-stat

Basic Materials -1.00% -0.44 -2.30% -1.02 -1.38% -0.61 Consumer Goods -2.59%* -1.82 -2.82%** -1.98 -3.50%** -2.46 Consumer Services -3.72%** -2.37 -2.86%* -1.82 -6.26%*** -3.99 Financials -2.47%* -1.71 -2.22% -1.54 -5.52%*** -3.82 Health Care -2.14% -1.19 -0.05% -0.03 -1.36% -0.75 Industrials -3.90%** -2.43 -3.23%** -2.01 -7.03%*** -4.38 Oil & Gas -4.32%** -2.03 -2.69% -1.26 -7.72%*** -3.62

Technology -1.86% -0.96 -1.48% -0.76 -0.50% -0.26

Telecommunications -0.61% -0.46 0.12% 0.09 -1.06% -0.80

Utilities -1.44% -0.92 1.39% 0.89 -1.83% -1.17

(30)

30

after the event day.

Although most of the findings were expected, some findings were quite surprising. The average abnormal returns for Germany and Switzerland are positive on the event day. These initial results could suggest that the Brexit has a positive influence on Germany and Switzerland. A negative effect is found for the Netherlands, France, United Kingdom and Ireland on the event day, where the average abnormal returns are only significant for the United Kingdom. The United Kingdom is negatively affected by -2.85% on the event day. This would suggest that only the United Kingdom is affected on the event day. All the average abnormal returns for the sectors are negative, implying a negative effect of the Brexit on all sectors. However, no significant effects are found for the average abnormal returns. This implies that the initial impact of the Brexit did not have a significant effect on a single sector on the event day. This changes when the effect is estimated for the sectors within the London Stock Exchange. Again, all the sectors experience negative average abnormal returns. The sectors Consumer Goods, Consumer Services, Financials, Industrials and Oil & Gas all have significant negative average abnormal returns. Where the Oil & Gas had a negative abnormal return of 4.32%, the Industrials 3.90%, the Consumer Services 3.72%, the Financials -2.47%, and the Consumer Goods -2.59%. These results are in line with the discussed literature of Lannoo (2016) & Milas et al. (2016).The results for the sectors within the United Kingdom show that the initials effect is significant for some, but not all sectors. It was expected that sectors who depend on export or make use of the free movement of labor would be affected the most, since the Brexit could introduce tariffs and barriers. This is reflected by the results. Consumer Goods, Oil & Gas and Industrials heavily depend on exports for the United Kingdom5, while Consumer Services depend on the free movement of labor. As shown in the first column of Table 3, only certain sectors within the United Kingdom are negatively affected with significance. This does not fully reflect the impact of the Brexit, since a scope of one day is too short to draw conclusions.

The second column of Table 3 provides the cumulative average abnormal return for a window of ten days, ending after the event day. From the country section it becomes clear that Ireland suffers the largest losses with a cumulative average abnormal return of -4.79%, followed by France with -3.49%. Only these two countries experience significant negative effects, while the effects for the Netherlands, Germany, United Kingdom and Switzerland are negative, but not significant. This confirms the results found in Burke (2016) and Barret et al.

(31)

31

(2015), which states that Ireland suffers the largest losses. All the sectors have negative cumulative average abnormal returns up to the event day. Unfortunately, no significant effect is found for any of the sectors. Again, this changes when the estimations are made only for the sectors within the London Stock Exchange. Here we see significant negative effects for the sectors Consumer Goods, Consumer Services and Industrials, where Industrials is the largest loser with -3.23%. What is interesting to notice here, is that the average abnormal return for the Financials and the Oil & Gas sector is significant on the event day, while the cumulative average abnormal return is not. This means that these sectors had some positive abnormal returns before the event day, possibly expecting a different outcome of the Brexit.

The third column of Table 3 provides the cumulative average abnormal return for a window of twenty trading days, ending ten trading days after the event day. From the country section, it becomes clear that all countries but Switzerland have significant negative cumulative average abnormal returns. It was expected that the United Kingdom would be affected the most. Surprisingly, the Brexit seems to have a larger impact on Ireland than any other country, reaching a cumulative average abnormal return of -8.75%. Germany also seems to be affected more than the United Kingdom, where the cumulative average abnormal returns is -5.49%. The effect on the United Kingdom is with -5.30% the third most affected country. The effect on France and the Netherlands is slightly less significant, where the cumulative average abnormal returns are -4.38% and -4.68%, respectively. Overall the results make sense regarding the expected outcome of the Brexit, and confirm the discussed literature. Switzerland isn’t significantly affected, which could be due to the fact that Switzerland has more tools to deal with the change of regulations, since they are not a member of the European Union. The large effect on Ireland is most likely caused by the great dependence Ireland has on the United Kingdom. The United Kingdom accounts for almost 30% of Irelands exports6. The expectation of new import-tariffs, which will hurt Ireland’s economy, is reflected by the negative cumulative average abnormal return of the Irish Stock Exchange. As expected, France, Germany and the Netherlands also have negative cumulative average abnormal returns, who will most likely also suffer from the introduction of import-tariffs, where Germany will suffer the most since they account for more export and import than other countries.

The sector section provides a better view on which sectors expect to lose value due to changes in regulations, where it is expected that sector who rely on export and import are affected the most. For the sectors Basic Materials, Health Care, Telecommunications and

Referenties

GERELATEERDE DOCUMENTEN

tie hebben mogelijk gemaakt van belangrijke, ont- vreemde elementen van het decorum, zoals de beide kleine glas-in-loodramen aan de trap naar de tweede verdieping, de klapdeuren

The reading comprehension of English relative clauses by L1 Farsi speakers converge with their on-line relative clause processing results. There is a negative transfer from L1 Farsi

zaai in het najaar. Uitzaai in het voorjaar na stratificatie in de bak heeft ook lang niet altijd het ge­ wenste resultaat. Stratificatie zonder medium en met een ge­

vacuümtechnologie voor geconcentreerde inzameling van zwartwater geworden, die in Duitsland, en nu ook in Nederland steeds meer toegepast wordt (fig..

A visual representation of the main findings following from the integrated results in the form of a SWOT diagram; a schematic overview of the strengths, weaknesses, opportunities

In countries like Hungary and Poland where Fidesz and PiS are the only party in government, their political rhetoric on traditional populist issues seems to have deepened

In some countries local government structures make provision for dedicated education units as part of local governance, while in others national education

The legacy of Moulton’s directorate, as Tan shows, was tremendous, particularly evident in the division of the museum’s collection into a reference collection for scientific research,