5 Data, Empirical Results & Analysis
The chapter commences by presenting a brief summary of the characteristics of the final sample.
Subsequently, the results of the announcement effects, the long-run stock performance and the multivariate analysis are presented and analyzed. Lastly, we analyze the long-run stock performance based on the market reactions and discuss predictable patterns while trying to establish a potential trading strategy.
Figure 10: Number of SEOs by country and year
Another point to reflect upon is the distribution of the SEOs per year. The yearly average accumulates to 30 SEOs undertaken. It is evident that up to 1996 and during the financial crisis years of 2008 and 2009, the number of SEOs was relatively low. Moreover, the year 1999 & 2000 account for years in which SEO activity is below average. On the other hand, many events have been conducted a priori to the financial crisis of 2007. After the financial crisis, the number of SEOs returns to a level which is similar to the development between 1997 and 2007. This SEO distribution can potentially bias end results, particularly with regard to the calendar time portfolio approach. Given the low coverage in the described years, there may exist calendar months which are dominated by large companies. However, we will account for this potential bias as we will resort to a robustness check throughout the CTP analysis by comparing the CTP results considering all months with an approach that only considers month that displays at least 20 observations.
5.1.2 Market Capitalization
Many scholars explain the impact that an overrepresentation of small cap companies on the study results has as explained in the section 3.3 “Independent Variables”. Therefore, the following section will feature the market cap composition of the event companies.
Graph 2: Market value in € mil by year and largest company in Market value terms in the respective year
Graph 2 displays the overall market values by all event companies within the respective year and compares it to the highest market value observed by any company within the year. This shall enable a potential detection of an event company which is dominating a year in market capitalization terms. It becomes evident that, on the one hand, in most years the high number of events leads to a broad diversification and the biggest company by market capitalization only accounts for a comparably small fraction of the overall observed market values. On the other hand, the years between 1992 – 1995 and 2007 – 2009 include companies which account for more than one third of the overall market value.
This is owing to the small number of SEOs undertaken in the years 1992 – 1995, 2008 and also 2009.
However, the year 2007 amounts to 42 SEOs throughout the year. Therefore, the high contribution by the biggest company to the overall market capitalization occurs with respect to this company’s immense size. Within these years, a bias might therefore arise as the return results of the dominating companies might shape the accumulated results, and they, as a result, might not be entirely representative. However, the use of equally- and value-weighted averages should minimize this potential bias. Furthermore, the robustness check given the CTP approach should diminish the bias.
0 200,000 400,000 600,000 800,000
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Market Value Max MV by 1 company
Table 3: Market value distribution and summary
Table 3 displays that the companies included in this study demonstrate diverse market capitalization backgrounds. Nevertheless, skewness towards companies with a market cap of € 2 billion and less is recognizable, which in the final analysis might have an impact on the thesis results. Consequently, the application of both equally- and value-weighted approaches is selected as this might emphasize the influence these small cap stocks have on the end result and furthermore account for such. With regard to countrywide anomalies, it is recognizable that the composition within all countries exhibits sound similarities. In addition, the market value draws a picture which coincides with the number of equity issues. This consequently further supports the decision of regarding the four markets as homogeneous due to the fact that the events within each market seem to be partitioned in similar fashion.
5.1.3 Size of the Issue
Another characteristic to be presented is the size of the issue. Accordingly, the graph will give an overview over the total amounts issued per year throughout the entire sample whilst putting them into perspective by the comparison with the number of SEOs. Afterwards, the introduced table should serve to disclose a little more detail.
Market value (in € mil) Entire Sample France Germany Netherlands Belgium
< 100 37 15 16 4 2
100 - 500 102 29 32 31 10
500 - 1,000 68 19 31 13 5
1,000 - 2,500 101 26 44 20 11
2,500 - 5,000 80 39 27 13 1
5,000-7,500 55 31 14 8 2
7,500 - 10,000 35 15 15 5 0
10,000 - 15,000 33 17 7 8 1
15,000 - 20,000 31 13 12 5 1
20,000 - 30,000 33 16 9 6 2
30,000 - 50,000 30 10 13 6 1
> 50,000 27 15 11 1 0
∑ 632 245 231 120 36
Maximum 166,931 139,870 166,931 50,400 40,104
Minimum 3 3 16 18 25
Average 9,855 11,852 10,262 6,565 4,618
Median 2,766 4,731 2,177 1,438 1,156
Number of companies
Market Value (in € mil)
Graph 3: Issue proceeds by year compared to the number of SEOs
The development of the yearly accumulated issue size over the examination period resembles the number of SEOs undertaken per year to a vast majority. Nonetheless, three years reveal differences which have to be mentioned. The year 2005, in which the most SEOs were undertaken by a clear margin, is generating comparably little issue proceeds. While, the year still accounts for the second largest amount of proceeds, the proceed amount turn out smaller proportionally. On the other hand, both the year 2000 and 2002 feature considerable proportional ascents in issue proceeds compared to the respective SEO number. Previous research (e.g. Asquith & Mullins, 1986; Choe et al., 1993;
Masulis & Korwar, 1986) has documented a negative relationship between the issue size and the corresponding returns. Therefore, these two years yield a potential bias given the proportionally vast issue proceeds generated by the modest amount of SEOs conducted in the respective years.
Particularly, the potential bias can be introduced in the CTP approach as the returns of the correspondent event companies are pooled together in a portfolio, resulting in a possible heavy influence on the compound monthly portfolio return.
Table 4: Issue proceed summary yearly and per event
Table 4 displays the absolute issue as well as the relative issue size on a yearly and an event base. It becomes evident with regard to the yearly data that on average per year only 9 % of the market
0 10,000 20,000 30,000 40,000 50,000
0 20 40 60 80 100
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Issue Proceeds (€ mil)
Number of SEOs
Number of SEOs Issue Proceeds
Total Average Maximum Minimum Issue Proceeds (in € mil) 291,244 13,869 38,908 13
Relative Issue Size 9.17% 53.19% 1.20%
Issue Proceeds (in € mil) 460.83 16,495 0.3
Relative Issue Size 14.44% 359.77% 0.11%
Yearly per Event
capitalization of the issuing company is generated by the equity issues. The year with the highest relative size is 1994, which only features one company that issues equity. For the relative issue size per event, a significant gap between the maximum and minimum is recognizable. Several companies within the sample issued equity in excess of their actual market value (16). Nevertheless, the average equity issuance accounts for approximately 14.4 % of the companies market cap. This is achieved as the majority of events (424) issues equity which equals less than 10 % of their market cap.
5.1.4 Frequent Issuers
The last characteristic that will be illustrated briefly is the number of frequent issuers. Former research evaluated the impact that frequent equity issues have on the stock price development of the respective companies as described in section 3.3 ‘Independent Variables’.
Graph 4: Number of companies issuing equity frequently & Comparison of Number of SEOs and Frequent Issues per country
The two graphs display some details that are worth mentioning. Firstly, 277 issues within the sample (~44 %) arise from companies which issue equity for the second or a consecutive time within the examined period. As a result, the sample consists of 355 individual companies undergoing 632 SEOs.
Secondly, the number of frequent issues within each country represents the distribution of the overall occurring SEOs. The number of companies that do issue equity frequently in our sample accounts for approximately 36 % (129 out of 355) of the overall number of companies. The majority of frequent issuers undertake a second equity issuance within the examined period. However, splitting up the consecutive issues, it becomes evident that these companies issue equity in close proximity. In conjunction with the aforementioned former research, this can have an impact on the results as they
60
32 17
10
3 3 3 1
2 3 4 5 6 7 8 9
Number of issues during examined period
245 231
120
36
116 103
48
10 0
50 100 150 200 250 300
France Germany Netherlands Belgium SEO Frequent Issues
might be biased downwards given the number of frequent issuers included in the sample. Furthermore, the issue clustering might amplify the negative effect.