Seasoned Equity Offerings Before and After BREXIT:
A Last Resort to Create Abnormal Returns or Dilution of
Shareholder Value?
University of AmsterdamCollege of Economics and Business BSc Economics & Business
Bachelor Specialisation Finance & Organisation
Author: B.E. Smelt
Student number: 10770917
Thesis supervisor: dhr. Dr. J.J.G. Lemmen Finish date: June 2018
PREFACE AND ACKNOWLEDGEMENTS
Before you lies the bachelor thesis ‘Seasoned Equity Offerings before and after BREXIT: A last resort to create abnormal returns or dilution of shareholder value?’. This thesis is the closing chapter of my life as a bachelor student. If I look back on writing my thesis it was a hard time starting, but everything became much more easier when I made a clear planning and I knew what I was doing.
The topic of my thesis was completely new to me. It caught my interest during the introduction lectures of the bachelor thesis course about finding a topic. After reading several papers I was certain to write my thesis about seasoned equity offerings. When I was reading these previous research papers I was thinking about in what way my research could contribute something new. And that is when I realised that it was particularly interesting to combine this topic to the recent BREXIT event. This field is not investigated much because there is not much data yet; also previous research about seasoned equity offerings is not much linked to big economic events in history. So, I think this research is pretty interesting and hopefully it will inspire others.
My special gratitude goes to my supervisor dhr. Dr. Jan Lemmen who helped me a lot with positive feedback and with good support in how to conduct my data research.
Enjoy reading, Bas Smelt
Statement of Originality
This document is written by student Bas Smelt who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document are original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
ABSTRACT
This study investigates the effect on stock prices of seasoned equity offering announcements before and after BREXIT. The results show that the announcement has got a significant negative effect on the stock performance. The volatility of the stock and the underpricing have got a significantly negative effect in the period after BREXIT, and this is said to be due to the more risk-averse investor. Furthermore, the two periods are shown to be significantly different from each other in the case that the stock returns after BREXIT are less than before BREXIT. This conclusion is in line with the stated hypothesis.
Keywords: Seasoned equity offerings, BREXIT, abnormal returns, information asymmetry, Chow-test.
TABLE OF CONTENTS
PREFACE AND ACKNOWLEDGEMENTS ... ii
ABSTRACT ... iv
TABLE OF CONTENTS ... v
LIST OF TABLES ... vi
LIST OF FIGURES ... vi
CHAPTER 1 Introduction ... 1
CHAPTER 2 Literature review ... 3
2.1 SEO Theory ... 3
2.2 SEO Empirical papers ... 4
CHAPTER 3 Methodology ... 7
CHAPTER 4 Data ... 11
CHAPTER 5 Results ... 16
5.1 Announcement day effect ... 16
5.2 Regression analysis ... 20
5.3 Chow-test ... 23
CHAPTER 6 Conclusion ... 25
REFERENCES ... 26
APPENDIX A Variable definitions ... 28
APPENDIX B Derivation Chow F statistic ... 29
LIST OF TABLES
Table 1 Explanatory variables and expected sign 10
Table 2 Number of seasoned equity offerings announcements 12
Table 3 Uses of the proceeds 14
Table 4 Summary statistics of the variables 15
Table 5 AARs and CARs 17
Table 6 AARs and CARs (small firms included) 19
Table 7 Three-day announcement abnormal returns 19
Table 8 Regression analysis 21
Table 9 Regression analysis (small firms included) 24
LIST OF FIGURES
CHAPTER 1 Introduction
This paper investigates the announcement-effect of seasoned equity offerings on stock prices before and after BREXIT, the exit of the UK from the EU. An on-going debate in financial literature is whether the issuance of equity is a favourable way of financing. According to Myers and Majluf (1984) a firm with ample financial slack is preferred to take all positive Net Present Value opportunities. These firms have a large amount of cash or marketable securities and have the ability to issue default-risk-free debt, but when it comes to financing by issuing debt it must be the case that management has got inside information. Management should act in the interest of the old stockholders if the information is so favourable that the cost to the old shareholder of issuing equity will outweigh the projects’ NPV. Lee (1997) was investigating whether firms that do issue equity are knowingly selling this equity overvalued. The results suggested that it is unfavourable to invest in firms were top executives sell their shares prior to the SEO. The top executives knowingly sold overvalued equity causing underperformance of the stock. This underperformance of the stock after an SEO can be due to negative information about the value of the firm. The issuance of equity gives a negative signal to investors about the firm value because the firm is not able to finance internally by using cash or debt. Ross (1977) found for example that higher debt ratios give a positive signal to investors. This higher debt ratio will have a constraint effect on the firm and thus signal a positive expectation of management about future cash flows.
Asquith and Mullins (1986) found that the announcement of common equity offerings reduce stock prices. In almost 80% of the sample the issue is associated with negative announcement excess returns. This conclusion is in line with the pecking order theory of Myers (1984) which states that internally generated funds are still preferred over issuing equity. More recent Allen and Soucik (2008) also found an underperformance of firms issuing seasoned equity in Australia, but they argued that the underperformance is dependent on the definition of the ‘long-run’. The firms conducting an SEO were underperforming in the first 5 years, but seem to turn their performance around in particularly in years 6 and 7.
Research about firms conducting an SEO has been done a number of times. However, there is little evidence provided about how the market responds during important economic events. More specifically, no research has been done about the recent BREXIT. The time period that can be examined is short, yet it is still of important value whether there is already a difference between stock returns before and after BREXIT. The announcement of the UK that it would be stepping out of the EU brought some questions on how the market would
respond. For firms it will be harder to trade because of the tariffs that the UK firms now have to pay. It is quite likely that firms are experiencing financial distress and are not able to finance by internally generated cash or debt. The information asymmetry and uncertainty around BREXIT will have investors doubting about positive expected cash flows by firms conducting an SEO. This leads to the question if BREXIT lead to firms issuing equity as a last resort because of financial distress or were they able to create abnormal returns?
In this paper we focus on testing whether the stock returns of UK firms before and after BREXIT differ from each other. It is expected that the BREXIT will have a negative effect on the stock returns because of the financial instability in this period. Investors will have a negative expectation about the firm value and the future cash flows due to the uncertainty. The null-hypothesis will therefore be: Stock returns after BREXIT are not different from before BREXIT. This hypothesis will be tested against H1: Stock returns after BREXIT are significantly less than before BREXIT. This hypothesis is in line with results found by Bayless and Chaplinsky (1996), who found that there exists a window of opportunity to issue equity at more favourable terms. In this research there does not exist such a window of opportunity because of BREXIT.
The sample consists of 87 seasoned equity offerings conducted by UK firms in the period 01/01/2015 - 31/03/2018. The referendum on BREXIT was held on June 23rd 2016, which falls almost in the middle of our sample period. In this paper we will calculate abnormal returns of stock returns by the market model around the announcement date of the seasoned equity offering.
The two-day abnormal returns calculated in this paper are more negative compared to previous research done about SEOs. A decline in abnormal returns is found of 7.4% before BREXIT and 9.3% after BREXIT. These more negative announcement abnormal returns are the results of the greater uncertainty and information asymmetry during BREXIT. During uncertain times like BREXIT investors particularly dislike a high volatility of the stock. They prefer the stock to be stable because investors become more risk-averse during uncertain times. Furthermore, the period before BREXIT and after BREXIT are tested against each other. The two subperiods are tested with a Chow-test and the result is that they are statistically different from each other.
This paper proceeds as follows: Chapter 2 will set out relevant literature about equity issuance. Chapter 3 describes the methodology. Chapter 4 describes the sample data was used. Chapter 5 includes the results of the analysis conducted. Chapter 6 will provide a clear conclusion and discussion about the results.
CHAPTER 2 Literature review
This section summarizes papers with theories about the issuance of equity by firms. The first part will cover papers written about the theory of SEOs. The second part will discuss relevant empirical papers about SEOs. The main focus in these papers is the announcement effect.
2.1 SEO Theory
Modigliani and Miller (1958) present a model about the cost of capital. This model gives answer to the question what the cost of capital is to a firm in a world in which yields are uncertain and capital can be acquired in several ways. Before this cost of capital used to be estimated by comparing alternative investment and financing decisions were ranked and compared with a subjective utility function. However, Modigliani and Miller stated that this utility approach had some serious drawbacks for normative and analytical purposes. Therefore, they came up with an alternative approach based on market value maximization. Under this approach any investment project and type of financing must raise market value of the firms’ shares. The market price reflects the tastes of the current owners as well as the tastes of the shareholders. The model captures the foundations of a theory of the valuation of firms and shares in a world of uncertainty. This model leads to an operational definition of the cost of capital as well as rational investment decision-making within a firm.
Myers and Majluf (1984) construct a model in which management has got information about the true market value of a firm that investors do not have. In the paper the authors address the problem, assuming rational investors and a rational firm, to figure out what the equilibrium share price is. Managers acting in the interests of the shareholder will only issue equity when the firms’ stock price is overvalued. The rational investors interpret the decision to issue equity as a negative signal about the true value of the firm. A firm does not have to issue equity when it has enough slack. The decision to issue equity will therefore always have a negative effect on the share price according this model.
However, according to Myers and Majluf (1984) there is one exception to this conclusion about negative signals around equity offerings. This is when the market knows the true firm value with certainty. In this case the issuance of equity only signals new investment. Since firms are not investing in negative-NPV projects, this will have a positive reaction on the stock price.
In line with investors knowing the true firm value, Parsons and Raviv (1985) provided a model that could explain why the chosen offering price is lower than the market price of the firms’ shares known by investors. This lower offering price is chosen by the underwriter to
attract investors with a high valuation of the firms’ new project. High valuation investors recognize that they will only be successful if the share is undersubscribed at the initial price. They are encouraged to buy at the initial price instead of the lowered price. Due to competitiveness of the market the price will eventually be driven to a level higher than the offering price.
The way in which the market prices firms’ equity should reflect the firms’ expected future profitability. Variables like net income, total assets, revenue, debt, etc. are being used in an effort to proxy future profitability. Koop and Li (2001) investigated the role of several of these explanatory variables. The authors used these variables to valuate IPO and SEO firms. The model included an error term that captures misvaluation defined as the difference between the maximum value of the firm and the market capitalisation at the time of the offering. The authors found that IPOs are underpriced compared to SEOs. A possible explanation for misvaluation of IPOs is that SEOs give a more fair valuation of the underlying assets of a firm.
Yet still the question rises why firms choose to do an SEO. DeAngelo et al. (2010) conclude that the probability that a firm conducts a seasoned equity offering is influenced by market-timing opportunities and corporate lifecycle stage. Firms time the market when they have an existing need for capital. The near term cash need is a primary motive for firms to conduct an SEO and not the chance to sell stock at a high price.
2.2 SEO Empirical papers
Most of the empirical papers about seasoned equity issues report a significant decline in stock returns on the announcement day of 2% - 3%. Masulis and Korwar (1986) document the stock price changes that occur on the announcement of seasoned equity offerings. The authors state that there are two major impacts on a firm regarding a stock offering. Firstly, the increase in equity capital lowers the firms’ leverage. Secondly, the proceeds are generally used to finance capital expenditure. These changes in leverage and revision of capital expenditure have got a similar change effect on the stock price. Masulis and Korwar (1986) found a statistically significant fall in the value of the common stock on the announcement of a seasoned equity offering. This fall in value of the common stock is supported by the fact that the announcement period stock returns are positively related to leverage change. In this same period Asquith and Mullins (1986) report a decline of 3.0% on the short-run stock returns at the announcement day. Both papers suggested that the change in leverage had an effect on the stock-offering announcement. Surprisingly both papers find different results, Masulis and
Korwar (1986) find a positive significant relation between change in leverage and stock returns whereas Asquith and Mullins (1986) find no significant relation.
The long-run stock returns following seasoned equity offerings are documented by Spiess and Affleck-Graves (1995). The authors wanted to determine whether managers have the ability to exploit overvaluation due to information asymmetry in the market. In this research they matched seasoned equity issuing firms to non-issuing firms of the same size in the same industry. The results stated that there was indeed a long-run underperformance of the stock following primary seasoned equity offerings. The underperformance is more severe for firms with a lower book-to-market ratio. The authors were not able to draw strong conclusions regarding the long-run underperformance. However, the magnitude and robustness leads the authors to conclude that the underperformance is due to management that can take advantage of firm-specific information and overvaluation. Lee (1997) was also investigating whether or not management knowingly sells overvalued equity. The author examines the relation between insider trading and long-run performance of the issuing firm. The results were in line with what is normally found as cumulative abnormal returns, namely a decline of 2.91%. However, the author did not find a close relation between top-executives’ trading and long-run performance of primary issues. Several possible factors that could explain this lack of a close relation are given. This can be due to cognitive bias in where inside traders pay no attention the possible worst-case scenario of a new project, underestimation of increased problems with free cash flows after the primary issue. The top-executive traders do not seem to know the degree of overvaluation. Finally, the author concludes that it is not desirable to invest in firms issuing primary seasoned equity because they underperform their benchmark. In this same period Rangan (1996) found a smaller decline of the stock returns after an SEO announcement. The decline in return was 1.32% and suggested that stock market temporarily overvalues issuing firms because management is able to manipulate their stock price by managing earnings.
This information asymmetry between management and shareholders is an important factor of the stock returns around SEOs. The research of Booth and Chang (2011) is an extension of the asymmetric information theory that states a positive relation: the larger the disagreement between management and shareholder, the larger the price drop is on the SEO announcement-day. The empirical research is partly based on the data from Spiess and Affleck-Graves (1995). The sample of SEOs is divided into two groups: dividend SEOs and non-dividend SEOs. Prior literature suggests that dividend payers have less information asymmetry than nonpayers. The findings of the author are in line with this prior literature.
Before the mid-80s the market did not differentiate between dividend-paying SEOs and non-dividend SEOs. Since then the market has responded less negatively to announcements of dividend-paying SEOs compared to non-dividend SEOs.
In order to shape this information asymmetry of firms, Autore (2011), Hull et al. (2012) and Karpoff et al. (2013) used the standard deviation of the daily stock returns as a proxy to measure for information asymmetry. In these three studies a negative significant relation was found between the volatility of the stock returns prior to the SEO and the future returns. The bid-ask spread is used by Hull et al. (2012), Corwin (2003) and Karpoff et al. (2013) to also measure for information asymmetry between firms and investors. A larger bid-ask spread compensates for uninformed investors that bear more risk of trading with informed investors.
To get a hold on the influencing factors of the issue on the future short-run returns the underpricing is necessary. The underpricing regarding SEOs is found by Corwin (2003) to be negative by 2.2%. Altinkilic and Hansen (2003) found an underpricing of roughly 3% during the 90s. Both papers stated that this underpricing is due to uncertainty about the financial situation of the firm and thus the market price of the stock. Corwin (2003), Autore (2011) and Hull et al. (2012) used the relative offer size as a proxy in their analysis to control for price pressure and found a positive significant relation. Furthermore, the trading volume and the percentage completion of the issue is used as a proxy to check the activity of speculators during SEOs by Kim et al. (2018). The effect is found to be positive for the SEO returns because a higher trading volume leads to a greater liquidity.
To capture the effect of the state of the firm Carlson et al. (2006) develop a rational theory of returns of firms conducting a seasoned equity offering. The model sets a scenario where the firm sells its output to the market with stochastic demand. The firm issues equity to finance investments and management has superior information about the firm value and growth opportunity. SEO firms are matched to non-SEO firms regarding their relative size and book-to-market ratio. Lee (1997), Teoh et al. (1998), Corwin (2003), Karpoff et al. (2013) all used the market value one day prior to the announcement as a variable to control for the state of the firm. Pilotte (1992) argued that there exists a difference in stock price responses regarding firms with growth opportunities and mature firms. The market-book ratios are used by Brav et al. (2000) and Brous et al. (2001) to control for the state of the firm.
In summary prior research found a decline in stock price returns on announcement day of about 2%-3%. Several variables are used to control for the information asymmetry, the issue characteristics and the state of the firm.
CHAPTER 3 Methodology
The stock markets’ reaction to the announcement of a seasoned equity offering is measured using daily abnormal returns. These daily abnormal returns are calculated with daily stock returns from Datastream. The daily abnormal returns for any stock are estimated by
𝐴𝑅𝑖𝑡 = 𝑅𝑖𝑡 − 𝐸(𝑅𝑖𝑡), where
t = day measured relative to the event, 𝐴𝑅𝑖𝑡 = abnormal return to firm i for day t, 𝑅𝑖𝑡 = return on firm i during day t,
𝐸(𝑅𝑖𝑡) = expected rate of return on firm i for day t.
𝐸(𝑅𝑖𝑡) is calculated by using a simple market model. This model is defined as 𝑅 = 𝛼 + 𝛽 ∗ 𝑅𝑚, where 𝑅𝑚 is the return on the market. The FTSE index of the London Stock Exchange is used as the benchmark for return on the market. This is because all the firms in the sample are listed on the FTSE. Datastream provides the daily returns of each firm 𝑅𝑖𝑡. The abnormal return is then calculated as the difference between the actual return to a firm and the return of the benchmark.
Average abnormal returns for each relative day are calculated by
𝐴𝐴𝑅𝑖𝑡 = 1
𝑁 ∑ 𝐴𝑅𝑖𝑡 𝑁
𝑖=1 ,
where N is the number of firms with abnormal returns during day t.
The cumulative abnormal returns for each firm i can then be calculated by summing the average abnormal returns as follows:
𝐶𝐴𝑅𝑖,𝐾,𝐿 = ∑𝐿𝑡=𝐾𝐴𝑅𝑖𝑡,
where the 𝐶𝐴𝑅𝐼,𝐾,𝐿 is for the period from t = day K until t = day L.
In order to perform this event study an estimation window is taken of 68 days before the announcement day until 21 days before the announcement day. The period chosen for the estimation window is in line with the estimation window chosen by Asquith and Mullins (1986) and Booth and Chang (2011). For the event window, a two-day average abnormal
return generated by each firm is examined. This two-day event window captures the effect of an announcement due to its timing relative to the markets’ trading hours. Day t = 0 is the day the news of an announcement is published in the Regulatory News of the London Stock Exchange. In most cases the news is announced on the previous day, t = − 1, and reported the next day. This is because the news can be announced after the market closes, causing the market to respond the next day and the reaction is on day t = 0. This is why a two-day announcement event window is chosen. The results of a three-day event window are also shown as a robustness check. The two-day return around the announcement for firm i is 𝐶𝐴𝑅𝑖,−1,0 where
𝐶𝐴𝑅𝑖,−1,0 = 𝐴𝑅𝑖,−1+ 𝐴𝑅𝑖,0,
and 𝐴𝑅𝑖,−1 is the abnormal return to firm i on the day prior to the announcement made in the Regulatory News of the London Stock Exchange. 𝐴𝑅𝑖,0 is the abnormal return on the day that the announcement is published in the Regulatory News of the London Stock Exchange. The Cumulative abnormal returns will be calculated for the whole sample period. Also cumulative abnormal returns will be calculated separate before and after the referendum of BREXIT. The referendum was held the 23rd of June 2016 and the outcome was announced in the evening, so the market reaction took place the next working day. So, 24th June 2016 is t = 0.
In order to support this event study the cumulative abnormal returns can be tested. These statistical tests are used to see whether the cumulative abnormal returns of firms conducting an SEO are significantly different from zero. So, the null hypothesis will be
𝐻𝑜: 𝐶𝐴𝑅𝑖,−1,0= 0
This hypothesis test will be tested three times: for the period as a whole, for the period ranging from 2015 until the BREXIT referendum and the period ranging from the BREXIT referendum until March 2018.
In order to explain the outcomes of the cumulative abnormal returns, several regression analyses will be performed. The main dependent variable is the two-day cumulative abnormal return. The explanatory variables are separated into three different groups. The first group of explanatory variables are linked to stock characteristics. Included in this group are the standard deviation of the stock returns within the estimation window, the trading volume of the stock and the bid ask spread one day before the SEO announcement.
The second group of explanatory variables are linked to the issue characteristics. Included in this group are the issue size, the percentage of completion of the issue, the issue discount, the underpricing and the amount of money raised. The third group of explanatory variables are linked to the firm characteristics. Included in this group are the years listed, the market value, the market-book ratio, the debt-equity ratio, and the amount of cash. The trading volume, the underpricing, the market value and the amount of cash are calculated as a logarithmic function in the regression model because of normality of these variables. Moreover, a dummy variable will be included whether the firm paid dividend to their shareholder in the month before the announcement of the SEO. See Appendix A for a complete derivation of each variable. Table 1 provides an overview of the variables used in the regression and their expected sign with the explanation of that sign.
There will be a regression conducted of the complete period from 2015 – March 2018, a regression of the period 2015 until the BREXIT referendum, and a regression of the period ranging from the BREXIT referendum until March 2018. See Appendix C for the VIF-test conducted to check for multicollinearity and which variables are dropped from the regression.
A Chow-test will be used to test if the coefficients are different for the period before BREXIT and the period after BREXIT. The Chow-test is used to determine whether the BREXIT referendum was a structural breaking point that splits the complete dataset in two separate datasets. The formula used to calculate the Chow F statistic is:
𝐶ℎ𝑜𝑤 𝐹 = (𝑅𝑆𝑆𝑝− (𝑅𝑆𝑆1+ 𝑅𝑆𝑆2)) / 𝑘 (𝑅𝑆𝑆1+ 𝑅𝑆𝑆2)/ (𝑛1+ 𝑛2− 2𝑘)
where
𝑅𝑆𝑆𝑝 = residual sum of squares complete regression, 𝑅𝑆𝑆1 = residual sum of squares before BREXIT, 𝑅𝑆𝑆2 = residual sum of squares after BREXIT, 𝑛1 = number of observations before BREXIT, 𝑛2 = number of observations after BREXIT, 𝑘 = number of variables used in the model.
The null hypothesis for the test is that there is no breaking point in the dataset. The null hypothesis will be rejected if the Chow F statistic falls in the critical region 𝐹(𝑘,𝑛−2𝑘).
Table 1. Explanatory variables and expected sign
Note: the variables shown in this table are only the variables that are used in the regression analysis.
Explanatory variables Expected sign Explanatory Description
Stock characteristics
STDEV
-
Higher standard deviation means more negative announcement returns
because of uncertainty Trading volume
+
Higher trading volume means more positive announcement returns because of greater liquidity Bid-ask spread
+
Higher bid-ask spread compensates for uninformed investors
Issue characteristics
Issue size
-
Higher issue size means more negative announcement returns because of dilution
Issue sold
+
Higher percentage of the issue sold means investors have more believe in the firm and has got a positive impact Underpricing
-
Higher underpricing means more negative announcement returns because of information asymmetry
Firm characteristics
Years listed
+
Older firms face less information asymmetry due to more data available
Market value
-
Higher market value means more negative returns because of
uncertainty about true value of firms Market-book ratio
+
Higher market-book ratio means more positive announcement returns
because of higher value in the market Debt-equity ratio
+
Higher leverage has a positive signal to the market about the stability of the firm
Cash
+
Higher amount of cash means more positive announcement returns
because of greater stability of the firm Dividend
+
Payment of dividend causes less information asymmetry about the firm
CHAPTER 4 Data
The initial sample of the data will be obtained from the London Stock Exchange. They provide an Excel-file named ‘Further Issues Summary’ which contains placings, open offers, further issues, exercise of options, etc. from the beginning of 1996 until present days. The data for placing & open offer for the period 2015 – March 2018 will be obtained. This gives an initial sample of 186 issues. To be included in the sample the firm must meet the following criteria:
1) The firm is listed on the AIM or UK Main Market. 2) The firm is located in the UK.
3) The firm is not an equity investment instrument or real estate investment trust. 4) The amount of money raised with the issue is more than zero.
5) The offering announcement was reported in the Regulatory News of the London Stock Exchange.
6) Required data for the regression analyses is available in the Datastream database.
The initial sample of the data is 102 seasoned equity offerings. The TIDM codes are manually checked on the London Stock Exchange to make sure the right TIDM code is used to retrieve the data from Datastream. This is due to several name changes of firms used in the sample size. The listing of each firm is also checked on the London Stock Exchange to make sure only well-established firms are used in the sample, because several firms were suspended from the London Stock Exchange due to irregularities. Also, small firms are excluded from the dataset. Reinganum (1981) and Roll (1981) discussed that these small firms are exposed to infrequent trading and this seems to be a powerful cause of bias in risk assessments with short-interval data. So, exclude for firms with a market value that is less than five million pounds, but the results of the CAR and regression with small firms are also shown to see the difference. At last the Regulatory News announcement of each issue is checked to make sure it was an open offer. This leads to a sample of 87 seasoned equity offerings. Table 2 reports the number of SEOs per quartile from 2015 until March 2018. Note the distinction in issues made in the fourth quartile of 2014. Overall the number of issues made is roughly equal.
The amount of money raised with the equity issue and the issue price are found in the ‘Further Issues Summary’ file and the number on shares issued and sold is manually checked in the Regulatory News database of the London Stock Exchange. The rest of the data is all
retrieved from the Datastream database. Table 4 reports the characteristics of the SEO issuing firm for the complete period, and the period before and after BREXIT. The first two lines report the variables regarding the stock characteristics, the next three lines report the variables regarding the issue characteristics, and the last four lines report the variables regarding the firm characteristics.
Table 2. Number of seasoned equity offerings announcements
Year Number of SEOs Percentage of sample
2015 - first quartile 5 5.75% 2015 - second quartile 7 8.05% 2015 - third quartile 5 5.75% 2015 - fourth quartile 6 6.90% 2016 - first quartile 7 8.05% 2016 - second quartile 4 4.60% BREXIT referendum 2016 - third quartile 6 6.90% 2016 - fourth quartile 9 10.34% 2017 - first quartile 6 6.90% 2017 - second quartile 9 10.34% 2017 - third quartile 8 9.19% 2017 - fourth quartile 13 14.94% 2018 - first quartile 2 2.29% Total 87 100.00%
Notes: Sample firms issuing SEOs were identified by accessing the London Stock Exchange ‘Further Issues
Summary’ file. Table 1 reports the number of SEOs for each quartile in the sample period. Note the breaking point in the middle of the sample period when the BREXIT referendum was held on 23 June 2016.
In table 4 can be seen that the standard deviation of the daily stock returns are larger for the period after the BREXIT. This increase can be due to the greater uncertainty after BREXIT. Trading volume also becomes larger after BREXIT and it is said to be that a higher trading volume increases the liquidity of the stock. However, according to Kim et al. (2018) higher trading volume indicates higher information asymmetry, which is the case in the period after BREXIT. The average of the bid-ask spread remains roughly the same indicating that there is not a greater dispersion in the bid-ask spread after BREXIT to compensate for uninformed investors of trading with informed investors. The issue size after BREXIT is larger indicating that firms are in a greater need of generating cash by issuing equity. This is not surprisingly because the expectation is that firms are in greater financial distress during
uncertain times. The underpricing is also larger for the period after BREXIT. According to Corwin (2003) and Altinkilic and Hansen (2003) underpricing of equity issues becomes larger with greater uncertainty. The uncertainty about the market price of the stock leads to the underpricing and eventually the stock price will get to a new equilibrium. The average amount of money raised is less in the period after BREXIT indicating that less investors are willing to invest large amounts of cash in SEO firms due to greater uncertainty. The leverage increases after the BREXIT but it can be the case that the firms in the period after BREXIT were already larger than the firm in the period before BREXIT. However, we can state that according to Masulis and Korwar (1986) that the issuance of equity lowers the firms’ leverage. There will be less proceeds of issuing equity after BREXIT due to uncertainty and that leads to a smaller fall in the firms leverage ratio.
Table 3 provides an overview of the uses of proceeds of the equity issuance. A distinction is made between takeovers, expansion in the sense of entering a new market segment, development of the products or facilities, performance in the sense of creating a better balance sheet and loan repayment. The uses of proceeds for better performance of the balance sheet is expected to be high because of firms that are in financial distress. Which is the primary reason to conduct a SEO according to Myers and Majluf (1984). Note the high number of uses for expansion in the third and fourth quartile of 2017.
In summary, the sample used in this paper contains 87 seasoned equity offerings in the UK for the period 2015 – March 2018. The period will be split into two separate periods to test whether there exists a breaking point in de variables due to the announcement of BREXIT.
Table 3. Uses of the proceeds
Takeover Expansion Development Performance
Loan Repayment 2015 - first quartile 1 3 0 1 0 2015 - second quartile 3 0 2 1 1 2015 - third quartile 1 2 1 1 0 2015 - fourth quartile 1 1 1 3 0 2016 - first quartile 0 3 0 3 1 2016 - second quartile 0 1 2 1 0 BREXIT referendum 2016 - third quartile 1 0 2 3 0 2016 - fourth quartile 2 1 5 1 0 2017 - first quartile 0 2 1 2 1 2017 - second quartile 2 2 1 3 1 2017 - third quartile 1 6 1 1 0 2017 - fourth quartile 2 6 1 2 2 2018 - first quartile 0 1 1 0 0 Total 14 27 18 22 6
Notes: the expansion stands for entering a new market segment or creating higher availability of the product.
Development stands for the development of the product of the firms or their facilities. Performance stands for creating a better and a healthier balance sheet.
Table 4. Summary statistics of the variables
Notes: STDEV reports the standard deviation of firm i in the estimation window period of 68 days before the announcement until 21 days before the announcement. Trading
volume reports the average turnover by volume within the estimation window. Issue size is the relative size of the issue measured as the amount of money raised divided by the market value of firm i one day before the announcement. Issue sold is the percentage shares that are sold divided by the shares available for the issue. The underpricing is calculated as the difference between the closing share price and the issue price divided by the issue price. The issue discount is calculated by taking the difference between the closing share price and the issue price divided by the closing share price. The bid-ask spread, market value, and the market-book ratio are all calculated one day prior to the announcement day. The debt-equity ratio is calculated by taking the amount of debt in the year prior to the SEO announcement divided by the market value one day prior to the announcement. The amount of cash is taken from the most recent available annual report prior to the announcement.
2015 - March 2018 2015 - BREXIT BREXIT - March 2018
Variable Mean Min Max Mean Min Max Mean Min Max
Firm characteristics
STDEV 0.073 0.002 3.279 0.039 0.005 0.147 0.096 0.002 3.279
Trading Volume (x million) 3.437 0.000 19.200 0.589 0.008 5.979 5.264 0.000 19.200
Bid-ask spread 0.047 0 0.229 0.048 0.000 0.187 0.046 0 0.229
Issue Size 0.273 0.002 1.451 0.245 0.002 0.649 0.304 0.009 1.451
Issue Sold 0.724 0.041 1 0.675 0.041 1 0.756 0.078 1
Underpricing 61.045 -0.193 4927 10.966 -0.193 139 93.172 -0.147 4927
Issue discount 0.189 -0.239 0.999 0.245 -0.239 0.993 0.154 -0.173 0.999
Money raised (x million) 28.7 0.072 448 35.7 0.178 448 24.2 0.177 370
Years listed 9.0 0.8 21.9 8.9 1.1 21.1 9.0 0.8 21.9
Market value (x million) 154 5 2.861 237 5 2.861 100 7 808
Market-book-ratio 3.26 -6.01 79.19 4.25 0.37 79.19 2.63 -6.01 21.42
Debt-equity-ratio 0.689 0.004 12.129 0.537 0.006 5.734 0.781 0.004 12
CHAPTER 5 Results
5.1 Announcement day effect
The announcement day effect for the total period and the two subperiods is tabulated in table 5. The table provides the average abnormal returns and the cumulative abnormal returns from twenty days prior to the announcement until twenty days after the announcement. In the table can be seen that the average abnormal return on the announcement day is more negative compared to the twenty days prior and after the announcement-day. Figure 1 reports the average abnormal returns in a graph. In the graph can clearly be seen that there is a negative spike at the announcement-day returns. The abnormal returns for the total period and the two subperiods are concentrated in a two-day announcement period, consistent with no information leakage. The average two-day announcement period abnormal returns for the total period is -8.6% with a t-statistic of 4.92. The average two-day period abnormal returns for the 2015 - BREXIT period is -7.4% with a t-statistic of 4.93. The average two-day period abnormal returns for the BREXIT - March 2018 is -9.3% with a t-statistic of 3.46. The results of a negative announcement effect are in line with what we expected. However the magnitude of the announcement effect is more than what is found in previous research. Asquith and Mullins (1986) found a two-day announcement effect return of -3.0%, Masulis and Korwar (1986) found for several industries a two-day announcement effect return between 2% and 3%. In more recent research Hull et al. (2012) found a three-day announcement effect return of -2.6%. The more negative announcement day returns found in this paper, can be due to the uncertainty during the time of the BREXIT. The announcement of the BREXIT referendum was on 27th of May 2015, from that moment it could be the case that investors became more risk-averse. The issuance of equity can be a negative signal to investors about the true value of the firm like Myers and Majluf (1986) stated, and in an uncertain period this announcement effect on abnormal returns can be more negative.
Table 6 reports the average abnormal returns and the cumulative abnormal returns from ten days prior to ten days after the announcement day for all the SEOs included with the firms that were expelled because of a market value less than five million pounds. The corresponding two-day announcement abnormal return for the total period becomes -1.4%, for the period before BREXIT the two-day abnormal return stays the same, and for the period after BREXIT the two-day abnormal return becomes 1.9%. Only the period before BREXIT
Table 5. Average abnormal returns and cumulative abnormal returns
Total period Before BREXIT After BREXIT
AAR CAR AAR CAR AAR CAR
Day (%) (%) (%) -20 -2.1 -2.1 -0.8 -0.8 -3.0 -3.0 -19 -0.4 -2.5 0.4 -0.4 -0.9 -3.9 -18 -0.7 -3.3 -1.4 -1.8 -0.3 -4.2 -17 -0.8 -4.1 -0.3 -2.1 -1.2 -5.4 -16 -0.1 -4.2 0.4 -1.8 -0.4 -5.8 -15 0.4 -3.8 -0.2 -1.9 0.8 -5.0 -14 0.5 -3.3 0.8 -1.1 0.2 -4.8 -13 -0.9 -4.2 -1.1 -2.2 -0.7 -5.5 -12 -0.2 -4.5 0.1 -2.2 -0.4 -6.0 -11 -1.2 -5.7 -0.3 -2.5 -1.8 -7.8 -10 -0.7 -6.4 -0.7 -3.1 -0.8 -8.5 -9 -1.9 -8.4 -1.3 -4.4 -2.4 -10.9 -8 -1.5 -9.8 -0.4 -4.8 -2.1 -13.1 -7 -0.5 -10.4 0.0 -4.8 -0.9 -14.0 -6 -0.9 -11.3 -0.1 -4.9 -1.4 -15.4 -5 -0.5 -11.8 -0.1 -5.0 -0.8 -16.2 -4 -0.3 -12.1 -0.6 -5.5 -0.1 -16.3 -3 -0.5 -12.6 0.0 -5.5 -0.8 -17.1 -2 -1.6 -14.2 -0.2 -5.7 -2.6 -19.6 -1 -1.3 -15.5 0.1 -5.6 -2.2 -21.8 AD -7.3 -22.8 -7.5 -13.1 -7.1 -29.0 1 1.1 -21.7 -0.2 -13.3 1.9 -27.0 2 -1.1 -22.8 0.0 -13.3 -1.8 -28.9 3 -0.2 -23.0 -0.3 -13.6 -0.1 -29.0 4 -0.3 -23.3 -0.8 -14.4 0.0 -29.0 5 -1.2 -24.5 -0.3 -14.7 -1.8 -30.8 6 -1.4 -25.9 -0.1 -14.8 -2.2 -33.0 7 -0.4 -26.3 0.0 -14.8 -0.6 -33.6 8 -0.6 -26.9 -1.3 -16.1 -0.2 -33.8 9 -0.3 -27.2 -0.2 -16.4 -0.4 -34.2 10 -0.8 -28.0 0.0 -16.4 -1.2 -35.4 11 0.3 -27.7 -0.3 -16.7 0.7 -34.7 12 -1.3 -29.0 -0.5 -17.2 -1.9 -36.6 13 -1.9 -31.0 -3.1 -20.4 -1.2 -37.8 14 -0.4 -31.4 0.2 -20.1 -0.9 -38.7 15 -0.7 -32.1 -0.6 -20.7 -0.7 -39.4 16 -1.1 -33.2 -1.2 -22.0 -1.1 -40.4 17 0.2 -33.0 -0.3 -22.3 0.6 -39.8 18 -0.6 -33.6 -0.9 -23.2 -0.4 -40.3 19 0.3 -33.3 0.4 -22.8 0.2 -40.0 20 -0.4 -33.6 0.5 -22.3 -0.9 -40.9 Two-day announcement return -8.6 -7.4 -9.3 t-statistic 4.92 4.93 3.46 N 87 34 53
is statistically different from zero, the two included small companies did not have a great impact on the abnormal returns. What is interesting is the two-day abnormal return for the period after BREXIT. With the inclusion of the thirteen smallest companies in this period the two-day abnormal return changed from -9.3% to 1.9%. Although the 1.9% is not statistically significant different from zero, yet still the inclusion of the thirteen small companies have got a severe positive impact on the two-day abnormal returns. Roll (1981) stated that this small firm effect is the cause of auto-correlation in the returns due to infrequent trading. This causes downward biased measures of portfolio risk and overestimates the risk-adjusted average returns.
Figure 1. Graph of AARs twenty days before and after announcement
In table 7 the three-day announcement abnormal returns are shown. The three-day announcement abnormal returns are less negative for the complete period compared and the period after BREXIT compared to the two-day announcement abnormal return. In figure 1 this can be seen with the positive spike at the first day after the announcement. The market restores itself one day after the announcement is made. The significance of all the three-day announcement abnormal returns is less than the two-day abnormal returns.
-8,0 -7,5 -7,0 -6,5 -6,0 -5,5 -5,0 -4,5 -4,0 -3,5 -3,0 -2,5 -2,0 -1,5 -1,0 -0,5 0,0 0,5 1,0 1,5 2,0 -20 -18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12 14 16 18 20 AAR Complete AAR Before AAR After
Table 6. Average abnormal returns and cumulative abnormal returns
Note: This table includes the small firms that were excluded in table 4.
Table 7. Three-day announcement abnormal returns Two-day announcement return -7.6 -7.5 -7.6 t-statistic 2.48 4.54 1.54 N 87 34 53
Total period Before BREXIT After BREXIT
AAR CAR AAR CAR AAR CAR
Day (%) (%) (%) -10 -0.6 -0.6 -0.8 -0.8 -0.6 -0.6 -9 -2.2 -2.8 -1.3 -2.1 -2.7 -3.3 -8 -1.3 -4.1 -0.6 -2.7 -1.7 -5.0 -7 -0.5 -4.6 -0.5 -3.2 -0.5 -5.5 -6 -0.7 -5.3 0.1 -3.1 -1.2 -6.7 -5 -0.3 -5.6 -0.4 -3.5 -0.2 -6.9 -4 -0.2 -5.8 -0.7 -4.2 0.1 -6.8 -3 -0.7 -6.5 0.0 -4.2 -1.1 -7.9 -2 -1.2 -7.7 -0.2 -4.4 -1.7 -8.6 -1 -1.2 -8.9 0.1 -4.3 -1.9 -10.5 AD -0.2 -9.1 -7.5 -11.8 3.8 -6.7 1 0.1 -9.0 -0.2 -12.0 0.3 -6.4 2 -1.0 -10.0 -0.1 -12.1 -1.5 -7.9 3 0.0 -10.0 -0.3 -12.4 0.2 -7.7 4 -0.2 -10.2 -0.9 -13.3 0.1 -7.6 5 -0.8 -11.0 -0.5 -13.8 -1.0 -8.6 6 -1.0 -12.0 -0.5 -14.3 -1.3 -9.9 7 -0.3 -12.3 -0.1 -14.4 -0.5 -10.4 8 -0.6 -12.9 -1.6 -16.0 0.0 -10.4 9 -0.4 -13.3 -0.3 -16.3 -0.5 -10.9 10 -0.9 -14.2 -0.3 -16.6 -1.2 -13.1 Two-day announcement return -1.4 -7.4 1.9 t-statistic 0.21 4.55 0.18 N 102 36 66
5.2 Regression analysis
Table 8 reports the results of the conducted regression analysis on the two-day announcement cumulative abnormal returns. Models (1), (2) and (3) show the results for respectively the total period, the period before BREXIT and the period after BREXIT with only the stock characteristic variables included. Models (4), (5) and (6) add the issue characteristic variables. Models (7), (8) and (9) contain all the explanatory variables.
Models (1), (2) and (3) are the base model for each period. These models include only the stock characteristic variables. Hull et al. (2012) used the daily standard deviation as an explanatory variable for abnormal returns and Autore (2011) used this as a proxy to measure for information asymmetry. For the total period and the period after BREXIT, models (1) and (3), the standard deviation of the daily stock returns is highly significant with a negative sign. The corresponding coefficients are -0.305 for respectively models (1) and (3), with t-statistics of 51.13 and 39.57. This implies if the stock returns prior to the SEO announcement have got a higher daily standard deviation, that this has a negative effect on the two-day announcement abnormal returns. The negative reaction of higher standard deviations is generally consistent with prior research. A reasonable explanation for this is that investors are more risk-averse after the BREXIT. When a stock has got a higher implied volatility it can be a sign to investors in the market that the firm is not performing well, the announcement of an SEO is than a sign to investors that the true value of the firm is different from what is known in the market according to Myers and Majluf (1986). This will have an impact in the way investors behave and have a negative impact on the two-day announcement abnormal returns.
The bid-ask spread in models (1) and (3) has a positive sign and a negative sign in model (2), although all three are not statistically significant. Adjusted R²s are respectively
0.43 for model (1), 0.00 for model (2) and 0.49 for model (3), which indicates that the standard deviation, the trading volume and the bid-ask spread do a poor job in explaining the cumulative abnormal returns in the period before BREXIT. Information asymmetry theory states that higher standard deviations have a negative effect and higher bid-ask spreads have a positive effect on future stock performance. For both variables this is the case. The coefficients of the trading volume are not statistically different from zero.
Models (4), (5) and (6) include the issue characteristics to the base model. Hull et al. (2012) stated that a greater relative issue size signals negative news to the market, because owners are taking greater advantage of an overvalued stock price. The variable underpricing states that a greater difference between the issue price and the share price one day prior to the
Table 8. Regression analysis 1 2 3 4 5 6 7 8 9 Intercept -0.078 -0.088 -0.070 -0.121 0.034 -0.198 -0.148 -0.197 -0.044 (0.98) (0.77) (0.66) (1.59) (0.23) (2.05)** (0.75) (0.57) (0.18) Stock characteristics: STDEV -0.305 0.019 -0.305 -0.305 0.857 -0.305 -0.3 -0.168 -0.298 (51.13)*** (0.03) (39.57)*** (68.13)*** (0.98) (55.37)*** (54.21)*** (0.12) (38.49)*** ln_Trading volume 0.000 0.001 -0.001 -0.004 -0.014 -0.002 -0.002 -0.009 0.005 (0.07) (0.17) (0.09) (0.78) (1.14) (0.30) (0.34) (0.57) (0.70) Bid-ask spread 0.204 -0.091 0.310 0.739 0.188 0.563 0.673 0.377 0.798 (0.73) (0.19) (0.74) (2.22)** (0.33) (1.59) (1.71)* (0.48) (1.97)** Issue characteristics: Issue size -0.073 -0.055 -0.049 -0.052 0.058 -0.088 (1.52) (0.52) (0.92) (1.02) (0.54) (1.48) Issue sold 0.037 -0.001 0.054 0.025 -0.036 0.032 (1.02) (0.03) (1.05) (0.63) (0.57) (0.55) ln_Underpricing -0.033 -0.018 -0.051 -0.033 -0.010 -0.049 (4.44)*** (1.31) (3.50)*** (4.11)*** (0.54) (3.29)*** Firm characteristics: Years listed -0.002 0.005 -0.007 (0.65) (1.75)* (2.33)** ln_Market value -0.003 -0.005 -0.008 (0.25) (0.28) (0.39) Market-book ratio 0.003 0.002 0.001 (1.89)* (0.88) (0.18) Debt-equity ratio 0.008 -0.037 0.012 (2.46)** (1.01) (2.78)*** ln_Cash 0.004 0.016 -0.002 (0.85) (3.23)*** (0.17) Dividend 0.029 0.095 0.027 (0.93) (1.47) (0.68) Adjusted R^2 0.43 0.00 0.49 0.65 0.18 0.77 0.68 0.48 0.80 N 87 34 53 85 34 51 85 34 51
Notes: Models (1), (4) and (7) are for the total period. Models (2), (5) and (8) are for the period before BREXIT. Models (3), (6) and (9) are for the period after
BREXIT. * Denotes the significance at the 10% levels, ** at the 5% levels, *** at the 1% levels. STDEV is the standard deviation. Issue size is the amount of money raised divided by market value. Issue sold is shares sold divided by shares available for the issue. Dividend is a dummy variable whether the firm issued dividend or
SEO announcement has got a negative effect on the future stock returns. Parsons and Raviv (1985) discussed that this higher underprcing is provided to attract high valuation investors and eventually drive the issue price upwards to the market equilibrium. But the underpricing suggests that the value of the firm in the market is not the true market value, yet it has a negative effect.
The coefficients of the standard deviation for models (4) and (6) follow the same reasoning given in models (1) and (3), they are still statistically significant at a 1% level. The bid-ask spread coefficient is positive in model (4), but only significant at a 5% level. This suggests that a greater bid-ask spread has got a positive effect on future abnormal returns for the total period. Overall, no statistical significance for the relative issue size in the three separate periods is being found. However, the sign of their coefficient is what we expected. The relative issue size has got a negative effect due to dilution, this is in line with theory of Asquith and Mullins (1986). The issue sold coefficients are positive in models (4) and (6) but negative in model (5) but are not statistically significant.
Regarding the coefficients of the underpricing they all have a negative sign for the three separate periods. The coefficients for models (4) and (6) are significant at a 1% level, stating that the greater underpricing has a negative impact on future abnormal returns after BREXIT due the greater uncertainty and information asymmetry. This is consistent with the reasoning of Parsons and Raviv (1985).
The coefficients of trading volume are negative in all three models, but not statistically significant. However, it is interesting why their coefficients have a negative sign. The expectation was that they would have a positive sign due to greater liquidity. However, Kim et al. (2018) also states that greater trading volume of the stock indicates a higher level of information asymmetry. During uncertain times investors become risk-averse and dislike high information asymmetries. The coefficient of trading volume becomes negative in that case.
Models (7), (8) and (9) include all the explanatory variables. The reasoning of the significance of the stock and issue characteristic variables remains the same as within the previous models. The inclusion of the firm characteristic variables has a significant effect for the years listed in models (8) and (9) respectively at the 10% and 5% level. The sign of the coefficient changes from positive before BREXIT to negative after BREXIT. It was expected that the coefficient would be positive because investors prefer older firms that are more stable during uncertain times. However, investors prefer smaller firms with maybe more potential growth after BREXIT. The debt-equity coefficients are positively significant in models (8) and (9) because this suggests greater stability of the firm. During uncertain times investors
prefer this greater stability. The positive significance of the amount of cash in model (8) can be reasoned by the fact that this was before BREXIT a reliable state of the firm. The firm has got a buffer for possible losses. After BREXIT this changed because the greater amount of cash gives the firm the ability to issue risk-free debt when financing, but a SEO announcement gives investors a different view about true firm value. Why are the firms conducting an SEO when they have more cash on hand? Although this coefficient is not statistically significant after BREXIT.
The adjusted R²s are 0.68, 0.48 and 0.80 for respectively model (7), (8) and (9). So the
model after BREXIT is best explained included with all the explanatory variables. Also a variance inflation factors test was conducted for the complete model to test for multicollinearity, and this resulted in no problem for the test because the values were below the generally accepted cut-off of 4.00.
Table 9 reports the regression results with the small firms included that were excluded in table 8. The standard deviations are still significant at a 1% level for the total period and the period after BREXIT. The interesting part is that the issue discount coefficients are not significant anymore for model (4), (6), (7) and (9). A possible explanation for this small-firm effect can be that the issue discount for small firms gives a signal to the market for potential growth. The R²s for the regressions made with the inclusion of small firms are except for
model (2), (5) and (8) less than in the regressions tabulated in table 5. These smaller R²s can
be the cause of infrequent trading of the smaller firms. 5.3 Chow-test
A Chow-test is needed in order to test whether the BREXIT is a structural breaking point that splits the data set. The separate regressions of the total period, the period before BREXIT and the period after BREXIT included the residual sum of squares to calculate the Chow F statistic. The Chow F statistic is 3.279 and with a critical 1% level the critical value is 2.496. We can conclude by this that the BREXIT is a statistically significant breaking point in my data set. Therefore the way the two-day announcement abnormal returns are explained is different from before BREXIT compared to after BREXIT.
This result is what was expected due to the uncertain times and greater information asymmetry after the BREXIT. See Appendix B for a complete derivation of the Chow F statistic.
Table 9. Regression analysis (small firms included) 1 2 3 4 5 6 7 8 9 Intercept -0.772 -0.135 -0.987 -0.967 -0.044 -1.258 -0.529 -0.191 -0.979 (1.22) (1.15) (1.20) (1.31) (0.31) (1.36) (0.67) (0.55) (0.61) Stock characteristics: STDEV -0.311 -1.219 -0.301 -0.309 -0.539 -0.299 -0.276 -0.591 -0.259 (21.92)*** (1.86)* (26.56)*** (12.28)*** (0.61) (13.15)*** (8.50)*** (1.06) (5.60)*** ln_Trading volume 0.049 0.007 0.062 0.059 -0.005 0.075 0.074 -0.006 0.096 (1.09) (0.69) (1.10) (1.09) (0.36) (1.18) (1.12) (0.46) (1.16) Bid-ask spread 3.377 0.587 4.729 3.507 0.875 5.363 3.424 0.537 5.059 (1.18) (1.40) (1.14) (1.27) (2.17)** (1.17) (1.24) (0.83) (1.05) Issue characteristics: Issue size 0.064 0.051 0.055 0.224 0.092 0.203 (0.95) (0.61) (1.01) (0.73) (1.26) (0.68) Issue sold 0.120 -0.059 0.278 0.131 -0.052 0.289 (0.93) (1.31) (1.02) (0.84) (1.09) (0.90) ln_Underpricing 0.034 -0.013 0.082 0.042 -0.008 0.079 (0.52) (1.02) (0.66) (0.60) (0.57) (0.64) Firm characteristics: Years listed -0.009 0.006 -0.007 (1.01) (2.12)** (0.63) ln_Market value -0.064 -0.007 -0.046 (0.84) (0.38) (0.44) Market-book ratio 0.000 0.002 0.008 (0.08) (1.22) (0.76) Debt-equity ratio -0.003 -0.047 -0.002 (0.72) (1.57) (0.64) ln_Cash 0.034 0.016 0.014 (1.04) (3.52)*** (0.36) Dividend 0.292 0.101 0.371 (1.16) (1.68)* (1.11) Adjusted R^2 0.09 0.12 0.11 0.11 0.31 0.17 0.15 0.63 0.19 N 102 36 66 93 33 60 93 33 60
Notes: Models (1), (4) and (7) are for the total period. Models (2), (5) and (8) are for the period before BREXIT. Models (3), (6) a nd (9) are for the period after
CHAPTER 6 Conclusion
The results of this study demonstrate the announcement of seasoned equity offerings reduce the stock performance. The findings of a negative announcement effect on stock returns are in line with previous research. However the magnitude of the decline in stock returns is much more than was found in previous research. The decline in stock returns on the two-day announcement return before BREXIT was -7.4% and after BREXIT it was -9.3%. These greater negative announcement returns are a possible consequence of the uncertain times during the period of the BREXIT. The information asymmetry between insiders and outsiders becomes larger, and outsiders become more risk-averse. These outside investors probably think that the insiders know the true value of the firm during this uncertain time, so the announcement of a seasoned equity offering has got a greater negative effect. However when the smallest firms that were excluded in the first place are added to the sample, than the two-day announcement effect after BREXIT becomes positive. A possible explanation is given due to infrequent trading, however this is not tested in this paper.
Regarding the regression analysis it is clear to see that the volatility of the stock has got a great impact on the stock performance. Due to greater risk-averse investors after the BREXIT this impact has become larger. The outside investors do not like uncertainty so they dislike larger standard deviation. The firm characteristic variables do a poor job in explaining the two-day announcement returns. Perhaps it was better to define the variables as averages instead of one-day values prior to the announcement.
Finally the two subperiods are tested against each other to see whether the two periods are different from each other. With the use of a chow test this was found to be the case. The stock returns after BREXIT are significantly lower than before BREXIT. The explanation given in this paper is the more risk-averse investor. However this is only assumed and could be tested in future research.
Limitations of this research are that the sample size is rather small. This is due to the relative small period that is tested. Also the firms are not compared on their relative size. For future research this would be relevant to implement in this research. So, expand the time period and match firms on their relative size to compare them with each other.
REFERENCES
Allen, D. E., & Soucik, V. (2008). Long Run Underperformance of Seasoned Equity Offerings: Fact or an Illusion? Mathematics and Computers in Simulation, 78(2), 146-154.
Altinkilic, O., & Hansen, R.S. (2003). Discounting and underpricing in seasoned equity offers.
Journal of Financial Economics, 69(2), 285-323.
Asquith, P., & Mullins, D. (1986). Equity issues and offering dilution. Journal of Financial
Economics, 15(1), 61-89.
Autore, D.M. (2011). Does Rule 10b-21 increase SEO discounting? Journal of Financial
Intermediation, 20(2), 231-247.
Bayless, M., & Chaplinsky, S. (1996). Is there a window of opportunity for seasoned equity issuance?
Journal of Finance, 51(1), 253-278.
Booth, L., & Chang, B. (2011). Information Asymmetry, Dividend Status, and SEO Announcement day Returns. The Journal of Financial Research, 34(1), 155-177.
Brav, A., Geczy, C., & Gomper, P.A. (2000). Is the abnormal return following equity issuances anomalous? Journal of Financial Economics, 56(2), 209-249.
Brous, P.A., Datar, V., & Kini, O. (2001). Is the Market Optimistic about the Future Earnings of Seasoned Equity Offering Firms? Journal of Financial and Quantitative Analysis, 36(2), 141-168.
Carlson, M., Fisher, A., & Giammarino, R. (2006). Corporate investment and asset-price dynamics: Implications for SEO event studies and long-run performance. Journal of Finance, 61(3), 1009-1034.
Corwin, S.A. (2003). The Determinants of Underpricing for Seasoned Equity Offers. Journal of
Finance, 58(5), 2249-2279.
DeAngelo, H., DeAngelo, L., & Stulz, R.M. (2010). Seasoned equity offerings, market timing, and the corporate lifecycle. Journal of Financial Economics, 95(3), 275-295.
Dierkens, N. (1991). Information asymmetry and equity issues. Journal of Financial and Quantitative
Analysis, 26(2), 181-199.
Dissanaike, G., Faasse, J., & Jayasekera, R. (2014). What do equity issuances signal? A study of equity issuances in the UK before and during the financial crisis. Journal of International
Money and Finance, 49, 358-385.
Hull, R.M., Kwak, S., & Walker, R.L. (2012). Explanation for market response to seasoned equity offerings. Journal of Economics and Finance, 36(3), 634-661.
Karpoff, J.M., Lee, G., & Masulis, R.W. (2013). Contracting under asymmetric information: Evidence from lockup agreements in seasoned equity offerings. Journal of Financial Economics,
Kim, D., Kim, J.S., & Seo, S.W. (2018). What options to trade and when: Evidence from seasoned equity offerings. Journal of Financial Markets, 37, 70-96.
Lee, I. (1997). Do firms knowingly sell overvalued equity? Journal of Finance, 52(4), 1439-1466. Loughran, T., & Ritter, J.R. (1995). The new issues puzzle. Journal of Finance, 50(1), 23-51. MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of Economic Literature,
35(1), 13-39.
Marsh, P. (1982). The choice between equity and debt: An empirical study. Journal of Finance, 37, 121-144.
Masulis, R., & Korwar A. (1986). Seasoned Equity Offerings: An empirical investigation. Journal of
Financial Economics, 15(1), 91-118.
Modigliani, F., Miller, M.H. (1958). The cost of capital, corporation finance and the theory of investment. The American Economic Review, 48(3), 261-297.
Myers, S.C. (1984). The Capital Structure Puzzle. Journal of Finance, 39(3), 575-592.
Myers, S.C., & Majluf, N.S. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13(2), 187-221. Parsons, J. E., & Raviv, A. (1985). Underpricing of seasoned issues. Journal of Financial Economics,
14(3), 377-397.
Pilotte, E. (1992). Growth opportunities and the stock price response to new financing. The Journal of
Business, 65(3), 371-394.
Rangan, S. (1998). Earnings management and the performance of seasoned equity offerings. Journal
of Financial Economics, 50(1), 101-122.
Reinganum, M.R. (1981). Misspecification of capital asset pricing: Emprical anomalies based on earnings' yields and market values. Journal of Financial Economics, 9(1), 19-46.
Roll, R. (1981). A possible explanation of the small firm effect. Journal of Finance, 36, 879-888. Ross, S.A. (1977). The Determination of Financial Structure: The Incentive-Signalling Approach. The
Bell Journal of Economics, 8(1), 23-40.
Schipper, K., & Smith, A. (1986). A comparison of equity carve-outs and seasoned equity offerings: share price effects and corporate restructuring. Journal of Financial Economics, 15(1), 153-186.
Spiess, D.K., Affleck-Graves, J. (1995). Underperformance in long-run stock returns following seasoned equity offerings. Journal of Financial Economics, 38(3), 243-267.
Teoh, S.H., Welch, I., & Wong, T.J. (1998). Earnings Management and the Long‐Run Market Performance of Initial Public Offerings. Journal of Finance, 53(6), 1935-1974.
APPENDIX A Variable definitions
This table describes the dependent variable and the explanatory variables in this study and how they are constructed. All the market and accounting data in this section comes from the London Stock Exchange and the Datastream database.
Dependent variable:
Cumulative abnormal return Normal returns are calculated with a market model, using [-68, -21] as the estimation window. The standard event window used is [-1, 0]. Stock characteristic:
Volatility of the stock The standard deviation of the stock returns within the estimation window [-68, -21].
Trading volume Average of the daily stock turnover within the estimation window [-68, -21].
Bid-ask spread The percentage difference in the bid-ask spread one day prior to the SEO announcement. Calculated as (Ask price – Bid price) / Ask price. Issue characteristic
Issue size Relative size of the issue calculated as the amount
of money raised, divided by the market value of firm i one day prior to the SEO announcement. Issue sold Ratio of the amount of shares that were sold by
the issue divided by the amount of shares made available by the firm to be sold.
Underpricing Percentage difference between the closing share
price on the day prior to the announcement and the issue price. Calculated as (Share Price – Issue Price) / Issue Price.
Issue discount Percentage difference between the closing share price on the day prior to the announcement and the issue price. Calculated as (Share price – Issue Price) / Share price.
Money raised Amount of money raised with the issuance of
