Effects of Different Types of Investment Activities on
Stock Returns: from the Perspective of Company Life
Cycles
Institution: Amsterdam Business School, University of
Amsterdam
Programme: MSc Business Economics, Finance track
Type: Master Thesis
Supervisor: Tolga CASKURLU
Rui LIU
2016.07
Statement of Originality
This document is written by Student Rui Liu who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is 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.
Acknowledgements
On the completion of my thesis, I would like to express my deepest gratitude to all those whose kindness and suggestions have made this paper possible and better. First, I am greatly indebted to my supervisor Tolga Carkurlu who gave me valuable and patient instructions. His effective advice, shrewd comments have kept the thesis in the right direction. Then I am deeply grateful to the professors who have taught me over the past year of study. Finally, thanks to my lovely friends and classmates who have discussed with me about my thesis and given some useful advice on my thesis.
Abstract
This paper mainly focus on analyzing the effects of three types of investment activities on subsequent stock performance from the perspective of corporate life cycles. The three types are research and development investments, capital expenditures for property, plant and equipment, and merger and acquisition expenditures. I use Fama-Macbeth two-step regression to test my hypotheses and get the conclusions for this research. Finally I come to the conclusions that research and development investments are positively related to future stock returns; internal and external expenditures are negatively related to stock returns; positive influences of research and development investments are larger in young corporates; negative influences of internal expenditures are larger in mature corporates; and negative influences of external expenditures are larger in young corporates.
Table of Contents
1. Introduction ... 1
2. Literature Review ... 3
3. Hypotheses and Methodology ... 6
4. Data and Descriptive Statistics ... 9
5. Empirical Results and Robustness Checks ... 15
6. Conclusion ... 35
1. Introduction
This thesis aims to discuss the effects of various kinds of investment activities on company future stock performance. Furthermore, this issue will be mainly discussed from company life cycles angle. This study aims to discuss the direction and level of the effects of the three kinds of investment activities, particularly to study the impacts under various company life cycles.
Firstly, I give definition to different kinds of investment activities. Based on literatures, there are three kinds of investment activities. Research and development investments as the first type are utilized to analyze and create intellectual property, patents and so forth. Capital expenditures as the second type refer to internal expenditure to buy equipment, property and plant for company operation and manufacturing. Merger and acquisition expenditures, as the third type, refer to external expenditures for merger or acquisition of other companies.
Later, I look for literature on the similar topic. Some literatures emphasize on the positive influences of innovation efficiency to measure research and development products based on corporate future performance. Another aspect about this topic is that there will be better future stock performance for lower asset growth rate of corporate. Merger and acquisition will be discussed in the final aspect of this study. Target company’s returns will be positively affected by merger and acquisition yet bidder company’s returns will be negatively affected by it.
In reference to the literatures, company future performance may be influenced by different types of investment activities and the company’s stock return is a direct approach to measure company performance. Therefore, I come up my first hypothesis, that future stock performance should be positively affected by research and development investments and future stock performance should be negatively affected by capital expenditures. Future stock performance should be affected by expenditures of merger and acquisition and the effects of merger and acquisition may be greater than the effects of the other two kinds of investments.
2
Some typical control variates for this topic are also provided in other literatures. The control variates are company capitalization, return on assets and book-to-market ratio. They may be associated with future stock performance. According to the finding of my research, company life cycles and industry environment might be important factors for stock returns. Therefore, the perspective of company life cycles will be emphasized in the next part of my thesis. Furthermore, I find papers on the grouping approach of company life cycles. Therefore, I propose my second hypothesis.
I ensure the variates are utilized to study the topic and gather data from the CRSP and WRDS databases after proposing the hypotheses. Later, the data are processed in Stata and only the qualified companies are included in my sample. Subsequently, I gain descriptive data for the variates and come up primary sub conclusions. Companies in different industries tend to have different preferences to investment activities in the industry environment. The computers and pharmaceuticals industries favor research and development investments. Extractive industries, transportation industries etc. favor internal expenditures. Furthermore, transportation industry likes to choose external expenditures and some other sectors. It is proved by the subsequent descriptive data that the investment activities are affected by book-to-market ratio and firm size. According to subsequent descriptive data, when analyzing the influences of investment activities on stock performance, life cycles are significant factors. This section offers the foundation of subsequent research, proving the rationality of my hypotheses.
Subsequently, in order to obtain the coefficients of control variates and independent variates, I run Fama-Macbeth regressions. According to the results of regression, hypotheses are generally reasonable. Future stock returns are positively affected by research and development investments. Internal and external expenditures are negatively associated with future stock returns. Furthermore, the influences of external expenditures are greater than the
influences of the other two kinds. In particular, research and development investments for young companies have great effects and are more important. Internal expenditures for mature companies are more important. External expenditures have greater effects on almost all the companies yet have more influences on young companies. The robustness checks are shown and the hypotheses are demonstrated to be rational by changing the normalization variate for independent variates.
In terms of my contribution to this topic, I emphasize on the study from the perspective of company life cycles, finding that mature and young companies are opposed to various degrees of influences from identical kind of investment activities. In addition, I further extend the concept of capital expenditures and categorize it into internal and external expenditures during the process of problem analysis.
Literature review is the second part. Hypotheses and methodology is the third part. Data and descriptive statistics is the fourth part. Regression results and robustness checks is the fifth part. Summary and discussion will be provided in the last part.
2. Literature Review
My thesis aims to discuss different kinds of investment activities and the effects of those activities on subsequent stock returns; therefore I search for literatures on these areas and on life cycles.
When analyzing the determinants of the cross section of future stock returns, Cooper, Gulen and Schill (2008) found some previously documented variables, including Book-to-market ratios, firm size, last-term returns and so on. Subsequently, they analyzed a determinant of new asset growth in future stock returns utilizing the data of nonfinancial firms in the U.S. market. At last, they came to a conclusion that corporates with lower asset growth rate often
4
earn higher subsequent risk-adjusted stock returns. In general, corporates acquire asset growth through capital expenditures and asset growth rates are economically and statistically predictors of future abnormal stock returns. The ability of asset growth to predict the trend of future stock returns is attributed to the roles that different investment or financing activity components play.
Hirshleifer, Hsu and Li (2013) analyzed the effect of another type of investment activity. They first found that companies with more investment in innovation tend to have better future operating performance. They utilized nonfinancial firm-level data for the U.S. market and controlled some firm characters like book-to-market ratio and firm capitalization. Thus, they came to the conclusion that innovation efficiency, patents or citations scaled by R&D investments is a strong and robust positive predictor for future stock returns. They also proved that the effect of innovation efficiency on future stock performance is supplement to existing determinants for stock returns and cannot be replaced by previous factors.
The preceding papers provide the first motivation for my paper. I am able to gain my major hypotheses on this topic that the various future stock returns are affected by different activities of investment. Majorly utilized to purchase equipment, plant and property, capital expenditures are likely to affect the company’s subsequent stock return negatively, when controlling book-to-market ratio, company capitalization and so forth. Nonetheless, when controlling the same company features, expenditures of research and development are likely to enhance the company’s future stock performance.
The influences of merger and acquisition on future stock performance are considered in my thesis. Over the past quarter century, as one kind of investment activities, merger and acquisition has been extensively utilized by companies, particularly by mature companies in developed markets. Thus, I will conduct more comprehensive study on the topic about stock returns are affected by different kinds of investment. Generally, in the cash flow statement,
acquisitions and merger should be included in the capital expenditure items. Herein, in this paper, I distinguish between external investments (which are merger and acquisition and are referred to as expenditures of merger and acquisition in this paper) and internal investments (mainly invested in buying equipment, plant and property for company and are referred to as capital expenditure in this paper).
Merger and acquisition (M&A) can be considered as a consolidation of two firms. The target company is usually paid by the bidder company based on the acquisition of the target company in order to earn profits on it in the future (Bacon and Gersdorff, 2009). According to Campa and Hernando (2004), target company’s stock usually earns higher positive abnormal returns in the period of merger and acquisition announcement. Nonetheless, the bidder corporate’s stock is usually suffering from negative stock returns. Upon the announcement of combination, buy-side and sell-side would be affected by merger or acquisition activities. According to Campa and Hernando(2004), merger and acquisition can affect subsequent stock returns via many factors including the asset base of bidder and so forth. According to them, the payment amount in merger or acquisition may be affected. Based on stock returns and merger or acquisition, I will verify the relationship between the payment amounts.
Therefore, this thesis will emphasize on the effects of different kinds of investment activities, including on stock returns, merger and acquisition expenditures, capital expenditures and research and development investment. Study was conducted by Cooper, Gulen and Schill (2008) to control book-to-market ratio and company size. The same corporate characters are utilized by Hirshleifer, Hsu, and Li (2013). Anderson and Garcia-Feijoo (2006) utilized the typical control variates as other reports on the same topic when analyzing the relationship between security returns and capital investment. Company capitalization, book-to-market ratio and return on asset are typical controls. Companies can be categorized into value companies with high book-to-market ratio and growth companies with low book-to-market ratio in the
6
process of controlling the size. Companies can be grouped by company capitalization when controlling for book-to-market ratio.
I will emphasize on the influences of various investment activities on stock returns in various company life cycles in order to study the topic in depth. Bena and Li (2013) concluded that young corporations tend to invest in a large amount of risky Research and Development projects in order to achieve rapid expansion and growth of business; those R&D projects are vital to the success and survival of young corporations. Conversely, as mature companies often have successful R&D projects and patents, new R&D investment is not so crucial to mature companies. Besides, mature companies are more likely to take merger and acquisition projects. This inspires me that one type of investment activity may have different influences on future stock returns in different corporate life cycles due to the complex firm and industry environment. Analyzing the effects of different investment activities on stock returns from the perspective of corporate life cycle is my main contribution to the topic in this thesis.
DeAngelo, DeAngelo and Stulz (2006) found that dividends paid by publicly traded firms are higher when retained earnings are a larger portion of total equity and they called it life-cycle theory of dividends. They prove the significantly positive relationship between the amount of paid dividends and the portion of not contributed equity, controlling the characteristics of some firm. In addition, they divided the life cycles using the proportion of retained earnings to total equity. Mature companies often have more retained earnings. Here in my thesis, I will use the method of dividing life cycles from this paper to conduct my further research.
3. Hypotheses and Methodology
activities on stock return from the company life cycle perspectives.
The below hypotheses are provided according to the findings and inspirations from literature:
Hypothesis 1: Subsequent risk-adjusted stock returns are negatively influenced by merger and acquisition expenditures as well as capital expenditures; future stock returns are affected by investments of research and development.
Hypothesis 2: In comparison to mature companies, young companies are more influenced by investments of research and development. When compared with young companies, mature companies are more influenced by merger and acquisition expenditure and capital expenditure in a negative way.
The following part is designed methodology.
I will use lagged dependent variable. Dependent variable is the subsequent monthly stock returns of U.S. listed companies excluding financial companies. Independent variable is one type of investment from the following three types: capital expenditure (internal expenditures for property, plant and equipment), merger and acquisition expenditures (external expenditures), as well as research and development investments. The independent variables will be normalized by dividing ending total assets. Control variables are from literatures and include Book-to-Market ratio, firm capitalization and return on asset. Here Book-to-market ratio is the year-end stockholders’ equity divided by total stock market capitalization and stock market capitalization of non-tradable shares can be calculated by the multiplying stock price of tradable shares. Firm capitalization is noted as size and calculated as the natural logarithm of company’s ending market value. Return on asset is expressed by net income divided by total assets. According to descriptive statistics, I also add industry dummy variables into the model. I will use three industry dummies in total and they are noted as ind1, ind2 and ind3. For the first one, if a company belongs to pharmaceuticals or computers industries, industry dummy 1
8
is equal to 1, otherwise it is equal to 0. For the second one, if a company belongs to transportation, mining and construction, utilities or extractive industries, industry dummy 2 is equal to 1, otherwise it is equal to 0. For the third one, if a company is in the food, pharmaceuticals, transportation or services industries, industry dummy 3 is equal to 1, otherwise it is equal to 0.
To analyze specific effects of independent variables under specific conditions, I will group the data by some standards and analyze each group separately. For instance, different industries may have different results; thus I will do the same research under various industries. All the data are from WRDS database and the sample period is the complete period taken from WRDS database from 1962 to 2016 in the U.S. market. It is worth mentioning that financial firms (sic code: 6000-6499) are excluded from the sample.
I will use Fama-Macbeth regression to work with multiple assets across time. The variables and descriptions are in table1:
Table 1
This table depicts all variables used in this thesis, and it shows the name, notation and description for each dependent variable, independent variable and control variable. The dependent variable is future stock returns in the subsequent year and it is noted as ER. The notations of three independent variables are RDAT, PPEAT and MAAT, and they orderly represent the research and development investments, internal expenditures and external expenditures. The control variables are book-to-market ratio, firm capitalization, return on asset and industry dummies.
Category Variable name Notation Description
Dependent variable
Future stock returns ER Monthly stock returns in the subsequent year Independent variables Research and development investments
RDAT R&D expenditure divided by ending total assets
Capital expenditures PPEAT Expenditures for Property, Plant and Equity divided by total assets Merger and
acquisition expenditures
MAAT Merger and Acquisition investment divided by total asset
Control variables
Book-to-Market ratio B/M Ending Book-to-Market ratio Firm capitalization size Natural logarithm of ending
market value
Return on Assets ROA Net income divided by total assets
computers industries
ind2 =1, if transportation, mining and construction, utilities and
extractive industries
ind3 =1,if food, pharmaceuticals, transportation and services industries
First of all, I should get descriptive statistics for independent variables and then utilize reasonable grouping methods through descriptive statistics or control variables. Subsequently, I will run Fama-Macbeth regressions separately under each group to verify my hypothesis 1. This is the foundation of the study in the next part.
When analyzing the effects of investment activities on stock returns from the perspective of company life cycle, the grouping method should be like this: if a corporate has retained earnings larger than the sample’s medium, it should be in the high-level group; if a corporate has retained earnings lower than the sample’s medium, it should be in the low-level group. Then, Fama-Macbeth regressions can be employed to test the second hypothesis.
4. Data and Descriptive Statistics
I use all the nonfinancial firms listed in U.S. stock market and select the sample period from 1962 to 2016 which is a complete sample period available in the WRDS database. I get the monthly stock returns from CRSP and annual fundamental statistics from Compustat. To ensure I have a reasonable sample data, I process the raw data from database first. According to Cooper, Gulen and Schill (2008), the firms in data set should have at least 5 consecutive years of fundamental data. Therefore, I exclude firms with no more than 5 consecutive years of accounting information and start all the analysis and regressions at the beginning of 1967. Fama and French (1993) concluded that to mitigate backfilling bias, before a company
10
was included in a sample, the company should be listed on Compustat for at least two years. According to Fama and French (1992), I get all my accounting variables, including the three types of investments normalized by total assets and the control variables at the end of year t, utilizing the fundamental data from Compustat at the end of year t-1.
In order to directly test the effects of three types of investment activities on stock performance and focus on the differences between industries, I exclude all the financial firms whose Standard Industrial Classification (SIC) code fall between 6000 and 6499 and then divide all other firms into 15 groups according to the SIC code. The 15 industry groups are agriculture, chemicals, computers, durable manufacturers, extractive industries, utilities, food, insurance and real estate, mining and construction, pharmaceuticals, retail, services, textiles and printing, transportation and other.
First of all, I get the descriptive statistics for independent variables. Table 2 shows the mean and median values of the three independent variables for all companies in the sample sorted by SIC code:
Table 2
This table shows the descriptive statistics of the three independent variables. Firms are grouped by their industries. The fifteen industries are agriculture, chemicals, computers, durables manufactures, extractive industries, utilities, food, insurance and real estate, mining and construction pharmaceuticals, retail, services, textile and printing, transportation and others. The two main indexes are average and medium. The three independent variables are research and development investment, internal expenditures and external expenditures.
Industry Index
(percent)
RDAT PPEAT MAAT
(by SIC code)
Agriculture Average - 0.058 0.001 Medium - 0.079 0 Chemicals Average 0.028 0.061 0.006 Medium 0.024 0.058 0 Computers Average 0.067 0.057 0.017 Medium 0.047 0.040 0
Durable Manufacturers Average 0.051 0.051 0.014
Medium 0.032 0.046 0
Medium 0.007 0.107 0
Utilities Average 0.021 0.084 0.005
Medium 0.019 0.069 0
Food Average 0.009 0.074 0.041
Medium 0.008 0.074 0.001
Insurance and Real Estate
Average 0.055 0.029 0.001
Medium 0 0.009 0
Mining and Construction Average 0.003 0.078 0.019
Medium 0.003 0.052 0 Other Average 0.121 0.054 0.006 Medium 0.038 0.028 0 Pharmaceuticals Average 0.170 0.071 0.029 Medium 0.100 0.065 0 Retail Average 0.001 0.064 0.006 Medium 0 0.042 0 Services Average 0.027 0.062 0.037 Medium 0.022 0.033 0
Textiles and Printing Average 0.021 0.070 0.005
Medium 0.021 0.058 0
Transportation Average 0.012 0.122 0.047
Medium 0.002 0.088 0
The raw data of research and development investments, capital expenditures for property, plant and equipment as well as merger and acquisition expenditures are given in millions. The three independent variables are those raw data normalized by dividing total asset and are also given in millions in the WRDS database.
Table 2 depicts that companies in different industries have different investment structures. Specifically, for research and development investments, the amount spent by the pharmaceuticals industry ranks the highest, and is far larger than that spent by other industries. Except for other unnamed industries in this paper, the computer industry makes investment in research and development at the second place. The proportion of research and development investments by the pharmaceuticals industry takes up 28.39% of the total research and development investment on average. Furthermore, the proportion by computer industry
12
expenditures in property, plant and equipment (internal expenditures), extractive industries, transportation industries, utilities industries, as well as mining and construction industries rank the top four in their proportion of internal expenditures. Furthermore, the proportions of the top four industries are 13.12%, 11.33%, 7.79% and 7.27% in turn. For merger and acquisition expenditures (external expenditures), the transportation industry ranks the highest in its proportion, followed by the food industry, services industry and pharmaceuticals industry. The proportions of their external expenditures to total external expenditures for all the industries are 19.90%, 17.53%, 15.66% and 12.53% in turn.
The above results show that any type of investment activities would be influenced by the industry where there is a firm. Thus, I will add the industry control variables in the regressions to make the model more accurate.
Investment activities are not only related to the industry features but also are influenced by a company’s own characteristics. I first divide firms into three groups according to their book-to-market ratio. If a firm’s book-to-market ratio is larger than 75 percentage of firms’ book-to-market ratio, the firm should go to the high level group. If a firm’s book-to-market ratio is lower than 25 percentage of firms’ book-to-market ratio, the firm should go to the low level group. If a firm’s book-to-market ratio is larger than 25 percentage and lower than 75 percentage of firms’ book-to-market ratio, the firm should belong to the medium group. The same division method is applied to grouping companies by their size. If a company’s capitalization falls above the third quartile of all companies, the company should be in the high level group. If a company’s capitalization falls below the first quartile of all companies, the company should belong to the low level group. If a company’s capitalization falls between the first and the third quartile of all companies, the company should be in the medium level group.
Later, I get the descriptive statistics for independent variables under the groups by book-to-market ratio and firm capitalization. Table 3, table 4 and table 5 show the mean and median
value of research and development investments, of capital expenditures (internal expenditures) and of merger and acquisition expenditures (external expenditures) under groups respectively.
Table 3
This table shows the descriptive statistics for research and development investment. The firms are grouped by firm size and their book-to-market ratio. The two statistic indexes are average and medium.
Group
Statistics
Book-to-market ratio
(by size) Low Medium High
Low Average 0.094 0.062 0.056 Medium 0.011 0.013 0.013 Medium Average 0.056 0.039 0.032 Medium 0.035 0.024 0.023 High Average 0.073 0.037 0.040 Medium 0.030 0.024 0.017 Table 4
This table shows the descriptive statistics for internal expenditures. The firms are grouped by firm size and their book-to-market ratio. The two statistic indexes are average and medium.
Group
Statistics
Book-to-market ratio
(by size) Low Medium High
Low Average 0.028 0.029 0.026 Medium 0.016 0.021 0.014 Medium Average 0.065 0.057 0.064 Medium 0.048 0.041 0.041 High Average 0.074 0.073 0.072 Medium 0.060 0.058 0.057 Table 5
This table shows the descriptive statistics for internal expenditures. The firms are grouped by firm size and their book-to-market ratio. The two statistic indexes are average and medium.
Group
Statistics
Book-to-market ratio
(by size) Low Medium High
Low Average 0.017 0.005 0.004 Medium 0 0 0 Medium Average 0.026 0.021 0.017 Medium 0 0 0 High Average 0.0234 0.017 0.008 Medium 0 0 0
14
low book-to-market ratio group are significantly more than those of high book-to-market ratio group. However, when compared with companies in the low size group, companies in the high size group invest more through research and development yet less through capital expenditures.
Firms in the high size group show huge difference in the preference between research and development expenditure and capital expenditure. Possible economic explanation could be that: in terminological innovation, small size companies tend to have large-scale research and development investment in order to enhance core competitiveness or to seek for opportunities acquired by large corporates. Oppositely, due to the risk of research and development, in comparison with investment activities of research and development, large companies prefer merge with small companies with successful research and mature technology (Bena and Li, 2013). Meanwhile, when large firms obtain new growth opportunities through merger or acquisition, they will increase capital expenditure to transfer growth opportunities to assets. Therefore, when compared with small companies, large companies have more research and development investments and less capital expenditure. In fact, for merger and acquisition expenditure, the results are quite similar to the results of capital expenditure. However, for large firms, if they have lower book-to-market ratio, they are more willing to invest in merger or acquisition.
From the perspective of life cycles, companies can be divided into two groups according to their retained earnings. If a company has retained earnings larger than the median of all companies, the company should be in the high level group. If a company has retained earnings below the median, the company should belong to the low level group. Table 6 shows the descriptive statistics of the three independent variables from the perspective of life cycles. Firms with low level retained earnings are named as young firms while firms with high level retained earnings are named as mature firms.
Table 6
This tables shows the descriptive statistics of the three independent variables from the perspective of company life cycles. The firms are grouped by their retained earnings and they are divided into young firms and mature firms. The two indexes are average and medium. The three independent variables are research and development investments, internal expenditures and external expenditures.
Life cycle Index
(percent) RDAT PPEAT MAAT
(by Retained Earnings)
Low level Average 0.063 0.069 0.014
Medium 0.028 0.049 0
High level Average 0.037 0.064 0.017
Medium 0.023 0.054 0
The table above depicts that young firms tend to have higher research and development investment while mature companies tend to have high external expenditures. Later in the regressions, I will run the regressions separately in the low level and the high level groups.
5. Empirical Results and Robustness Checks
In this section, I will run regressions to test my hypotheses and verify my inferences earlier in this paper. The dependent variable is firm’s future stock returns. The three independent variables are research and development investment, capital expenditures (internal expenditures) and merger and acquisition (external expenditures) respectively. The control variables are industry dummies, book-to-market ratio, firm capitalization and return on assets. I will employ Fama-Macbeth two-steps regression to analyze the relationship.
First of all, I regress the logarithm of the next period of stock returns on control variables and research and development investments with or without the relating dummies successively. The regression results for the thirteen specifications are shown in table 7.
16
Table 7
Fama-Macbeth regressions of stock returns on the three kinds of investments
This table describes the average slopes (in percent) and their t-statistics in parentheses from monthly Fama-macbeth cross-sectional regressions of individual stock excess returns for one period on various independent variables and control variables. The dependent variable is the logarithm of stock returns. RDAT, PPEAT and MAAT orderly denote the amount invested in research and development, in plant, property and equipment, and in merger or acquisition within one fiscal year, and the total ending assets in each year are used to normalize the three independent variables. The control variates are size, B/M and ROA. Size is the firm capitalization which can be calculated as the natural logarithm of company’s ending market value. B/M represents firm book-to-market ratio which is the year-end stockholders’ equity divided by total stock market capitalization and stock market capitalization of non-tradable shares can be calculated by the multiplying stock price of tradable shares. ROA is return on asset which is expressed by net income divided by year-end total assets. Ind1, ind2 and ind3 are industry dummy variables. If a company belongs to pharmaceuticals or computers industries, ind1 is equal to 1, otherwise it is equal to 0. If a company belongs to transportation, mining and construction, utilities or extractive industries, ind2 is equal to 1, otherwise it is equal to 0. If a company is in the food, pharmaceuticals, transportation or services industries, industry dummy 3 is equal to 1, otherwise it is equal to 0. All independent variables are normalized to zero mean and 1 standard deviation after winsorization at the 1% and 99% levels. The return data are from January of 1962 to March of 2016. The reported adjusted 𝑅2 of each monthly cross-sectional regression.
Table 7
Dependent variable: the logarithm of stock returns
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
RDAT 0.233*** 0.190*** 0.207*** 0.193*** (4.23) (3.37) (4.34) (4.84) PPEAT -0.027** -0.035*** -0.073*** -0.045*** (-1.87) (-1.96) (-2.36) (-2.07) MAAT size -0.113*** -0.116*** -0.134*** -0.087*** -0.078*** -0.083*** -0.079*** -0.079*** (-3.25) (-3.39) (-3.44) (-3.76) (-10.49) (-10.41) (-9.39) (-9.22) B/M 0.075** 0.086*** 0.077*** 0.127*** 0.097*** 0.099*** 0.103*** 0.110*** (1.82) (2.13) (1.98) (2.67) (5.85) (5.96) (11.01) (6.50) ROA -0.385 -0.221 -0.558 -0.747** -0.559*** -0.510*** -0.441 -0.505*** (-0.61) (-0.32) (-0.95) (-1.87) (-3.40) (-2.94) (-0.83) (-2.14) ind1 0.101*** 0.182*** (2.07) (6.36) ind2 0.186*** -0.214*** (3.58) (-5.93) ind3 0.063 0.040 (1.14) (1.36) C -2.082*** -2.101*** -1.915*** -2.231*** -2.331*** -2.312*** -2.305*** -2.316*** (-8.63) (-8.82) (-7.14) (-12.97) (-38.7) (-35.24) (-36.01) (-29.95) adj.R-sq 0.274 0.312 0.299 0.312 0.193 0.220 0.224 0.218
18
Table 7 Continued
Dependent variable: the logarithm of stock returns
Variables Model 9 Model 10 Model 11 Model 12 Model 13
RDAT 4.663*** (2.33) PPEAT -3.118*** (-2.77) MAAT -0.937*** -1.046*** -1.712*** -1.373*** -1.332*** (-2.08) (-1.98) (-2.10) (-2.12) (-2.10) size -0.077*** -0.079*** -0.076*** -0.071*** -0.169** (-10.87) (-10.68) (-8.89) (-8.58) (-1.74) B/M 0.099*** 0.100*** 0.099*** 0.089*** 0.161*** (5.97) (6.01) (4.74) (5.41) (2.46) ROA -0.391 -0.364 -0.517** -0.644*** -0.057 (-1.46) (-1.36) (-1.80) (-3.08) (-0.11) ind1 0.183*** (6.12) ind2 -0.093*** (-2.11) ind3 -0.1120*** (-1.96) C -2.341*** -2.339*** -2.355*** -2.405*** -1.578** (-41.84) (-39.99) (-32.34) (-35.93) (1.95) adj.R-sq 0.198 0.225 0.230 0.227 0.383
It is obviously that firm capitalization and book-to-market ratio are significant control variables for all the three independent variables. However, return on assets is not a good control variable for this topic. In addition, the industry classification is useful for analyzing the relationship between investment activities and future stock returns.
Specifications 1 to specification 4 are tested for the influence of research and development investments. First of all, from the coefficients and t-statistics of RDAT, it can be shown that research and development investment is a significantly positive indicator for future stock returns in the developed market. The coefficients are around 0.2 means that 1 million dollars of research and development investment will probably increase 0.2 million dollars of stock returns. In the section of descriptive statistics, I conclude that pharmaceuticals and computer industries tend to invest more in research and development projects. Here I continue to test the relationship between research and development investment and explain variable in pharmaceuticals and computers through industry dummy variable 1. The result shows that this kind of relationship is influenced by the firm’s industry and industry dummy 1 is a significantly positive indicator for research and development investment. In further regressions, I also conclude that in transportation, mining and construction, utilities and extractive industries, future stock returns are also significantly sensitive to research and development investments. However, in food, pharmaceuticals, transportation, services and other industries, industry influence is not so obvious in future stock returns.
Specifications 5 to 8 are used to test the relationship between internal expenditures and subsequent stock returns. The coefficients of PPEAT show that capital expenditures for property, plant and equipment are significantly negative indicators for subsequent stock performance in the developed market and this conclusion is consistent with the literatures. The magnitude of coefficients of PPEAT is much smaller than that of coefficients of RDAT and this proves that the negative influences from internal expenditures are much smaller than the
20
influences of positive research and development investments on stock returns. The section of data descriptive statistics shows that companies in transportation, mining and construction, utilities and extractive industries prefer internal expenditures. Specifications 6 and 7 indicate that not only industry dummy 2, but also industry dummy 1 are significantly related to future stock returns.
Specifications 9 to 12 are used to indicate the relationship between external expenditures and future stock returns. The coefficients of MAAT are around 1 and are much larger than coefficients of both RDAT and PPEAT. The reason for that is merger and acquisition is a big issue for a company, and merging or acquiring other companies may require a large amount of capital. Some shareholders may not be optimistic about the future of the company after merger or acquisition, thus the stock price will suffer a large plunge. The negative influence of merger and acquisition expenditure is significant, and all the three control variables including book-to-market ratio, return on assets and firm capitalization are significant controls for this independent variable. In addition to this, industry dummy 1, industry dummy 2 and industry dummy 3 are all significantly related to future stock returns. Actually, the significantly negative relationship between merger and acquisition and subsequent stock returns can be found in almost all the industries. Merger and acquisition have huge influences on future stock performance not only in quantity but also in scale.
Specification 13 includes all the three independent variables and controls yet excludes industry dummies. The result indicates that research and development, internal expenditures and external expenditures are significant indicators of subsequent stock performance as mentioned in hypothesis 1. In developed markets, research and development investments are positively related to future stock returns, and capital expenditures for property, plant and equipment; also, merger and acquisition is negatively correlated with future stock returns.
Subsequently, I will present robustness checks for the previous results. In the preceding regressions, I use total ending assets to normalize the three independent variables. Here in this section, referring to Chan, Lakonishok and Sougiannis (2001), I will use current operating income to normalize the three independent variables, and utilize Fama and Macbeth two-steps regressions to present robustness checks. Furthermore, the results for robustness checks are shown in table 8.
22
Table 8
Robustness checks for model 1-13
This table depicts the robustness checks for model 1-13. The table describes the average slopes (in percent) and their t-statistics in parentheses from monthly Fama-macbeth cross-sectional regressions of individual stock excess returns for one period on various independent variables and control variables. The dependent variable is the logarithm of stock returns. RDAT, PPEAT and MAAT orderly denote the amount invested in research and development, in plant, property and equipment, and in merger or acquisition within one fiscal year, and current operating incomes in each year are used to normalize the three independent variables. They are independent variables. The control variates are size, B/M and ROA. Size is the firm capitalization which can be calculated as the natural logarithm of company’s ending market value. B/M represents firm book-to-market ratio which is the year-end stockholders’ equity divided by total stock market capitalization and stock market capitalization of non-tradable shares can be calculated by the multiplying stock price of tradable shares. ROA is return on asset which is expressed by net income divided by year-end total assets. Ind1, ind2 and ind3 are industry dummy variables. If a company belongs to pharmaceuticals or computers industries, ind1 is equal to 1, otherwise it is equal to 0. If a company belongs to transportation, mining and construction, utilities or extractive industries, ind2 is equal to 1, otherwise it is equal to 0. If a company is in the food, pharmaceuticals, transportation or services industries, industry dummy 3 is equal to 1, otherwise it is equal to 0. All independent variables are normalized to zero mean and 1 standard deviation after winsorization at the 1% and 99% levels. The return data are from January of 1962 to March of 2016. The reported adjusted 𝑅2 of each monthly cross-sectional regression.
Table 8
Dependent variable: the logarithm of stock returns
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
RDAT 0.228*** 0.150*** 0.219*** 0.175*** (3.87) (3.26) (4.10) (3.99) PPEAT -0.020** -0.059*** -0.062*** -0.058*** (-1.90) (-1.96) (-2.83) (-2.10) MAAT size -0.105*** -0.110*** -0.119*** -0.109*** -0.078*** -0.085*** -0.094*** -0.083*** (-3.08) (-2.58) (-1.97) (-3.88) (-10.73) (-10.76) (-9.73) (-8.53) B/M 0.081** 0.081*** 0.088*** 0.129*** 0.096*** 0.108*** 0.107*** 0.121*** (1.93) (2.08) (1.99) (2.45) (5.36) (5.76) (10.64) (6.63) ROA -0.373 -0.231 -0.552 -0.736** -0.519*** -0.625*** -0.443 -0.575*** (-0.69) (-0.35) (-0.88) (-1.91) (-3.37) (-2.82) (-0.82) (-2.21) ind1 0.118*** 0.182*** (2.14) (6.36) ind2 0.179*** -0.324*** (3.40) (-5.53) ind3 0.074 0.051 (1.37) (1.34) C -2.085*** -2.300*** -1.918*** -2.581*** -2.470*** -2.321*** -2.215*** -2.288*** (-7.53) (-8.47) (-7.79) (-13.01) (-24.7) (-21.26) (-23.01) (-29.85) adj.R-sq 0.239 0.404 0.298 0.315 0.186 0.233 0.235 0.235
24
Table 8 Continued
Dependent variable: the logarithm of stock returns
Variables Model 9 Model 10 Model 11 Model 12 Model 13
RDAT 4.594*** (2.53) PPEAT -3.108*** (-2.46) MAAT -0.963*** -1.188*** -1.636*** -1.166*** -1.380*** (-2.19) (-1.97) (-2.46) (-2.32) (-2.67) size -0.068*** -0.084*** -0.063*** -0.029*** -0.190** (-9.52) (-9.78) (-8.25) (-8.34) (-1.82) B/M 0.081*** 0.121*** 0.094*** 0.084*** 0.155*** (5.63) (5.71) (3.86) (5.25) (2.63) ROA -0.387 -0.324 -0.492** -0.665*** -0.047 (-1.58) (-1.48) (-1.93) (-3.47) (-0.64) ind1 0.172*** (6.72) ind2 -0.086*** (-2.47) ind3 -0.165*** (-1.98) C -2.325*** -2.319*** -2.369*** -2.345*** -1.365** (-34.84) (-31.89) (-37.32) (-31.09) (1.84) adj.R-sq 0.236 0.356 0.235 0.236 0.348
The results above indicate that the robustness checks favor the testing results for my first hypothesis.
Later, the empirical results for life cycles will be present. I demonstrate my first hypothesis in the first part of regressions that subsequent stock returns are influenced by the three kinds of investment activities. Generally, future stock returns will be increased by investments of research and development. Nonetheless, the future stock returns will decrease due to internal and external expenditures. Particularly, when compared with capital expenditures for equipment, property and plan, stock performance are more influenced by expenditures of merger and acquisition in a negative way. In addition, the companies’ characteristics and industries are influenced by the influences of investment kinds. The three independent variates show important effects on dependent variates under the controlling of return on assets, book-to-market ratio, industry dummies and company capitalization. Therefore, I summarize that the hypothesis 1 is proved to be reasonable.
I will analyze this topic and emphasize on the influences of different kinds of investment activities from the life cycles perspective in this section. Furthermore, the independent variates are investments of research and development, expenditures of merger and acquisition as well as capital expenditures for equipment, property and plant, which can be normalized by dividing ending assets in total while the dependent variate is the logarithm value of future stock return. Control variates are return on assets, book-to-market ratio, industry dummies and firm size in the preceding section. I will not repeat the regressions which present the above unimportant relationship and will include the industry dummies that are most related to every independent variate.
Firstly, all the companies will be categorized based on their life cycles. A company with retained earnings greater than the median of sample in total is in the mature corporate group.
26
Also, a company with retained earnings less than the median of the sample belongs to the young company group. Tables 9 shows the regression results.
Table 9
Fama-Macbeth regressions of stock returns on the three kinds of investments under life cycles
This table describes the average slopes (in percent) and their t-statistics in parentheses from monthly Fama-macbeth cross-sectional regressions of individual stock excess returns for one period on various independent variables and control variables under the groups of high retained earnings or of low retained earnings. The dependent variable is the logarithm of stock returns. RDAT, PPEAT and MAAT orderly denote the amount invested in research and development, in plant, property and equipment, and in merger or acquisition within one fiscal year, and the total ending assets in each year are used to normalize the three independent variables. The control variates are size, B/M and ROA. Size is the firm capitalization which can be calculated as the natural logarithm of company’s ending market value. B/M represents firm book-to-market ratio which is the year-end stockholders’ equity divided by total stock market capitalization and stock market capitalization of non-tradable shares can be calculated by the multiplying stock price of tradable shares. ROA is return on asset which is expressed by net income divided by year-end total assets. Ind1, ind2 and ind3 are industry dummy variables. If a company belongs to pharmaceuticals or computers industries, ind1 is equal to 1, otherwise it is equal to 0. If a company belongs to transportation, mining and construction, utilities or extractive industries, ind2 is equal to 1, otherwise it is equal to 0. If a company is in the food, pharmaceuticals, transportation or services industries, industry dummy 3 is equal to 1, otherwise it is equal to 0. All independent variables are normalized to zero mean and 1 standard deviation after winsorization at the 1% and 99% levels. The return data are from January of 1962 to March of 2016. The reported adjusted 𝑅2 of each monthly cross-sectional regression.
28
Table 9
Dependent variable: the logarithm of stock returns
Variables Specification 14 Specification 15 Specification 14 Specification 15
Low group High group Low group High group Low group High group Low group High group
RDAT 0.132*** 0.161** 1.723*** 1.998** (2.56) (1.87) (3.02) (1.88) PPEAT -0.924 -0.790** -0.958 -0.680*** -1.69 -1.81 -1.67 -2.46 MAAT size -0.053*** -0.079** -0.068*** -0.078*** -0.084*** -0.070*** -0.095*** -0.047*** (-1.97) (-1.79) (-3.11) (-2.76) -2.28 -2.08 -5.54 -1.99 B/M 0.109 0.059 0.097*** 0.061*** 0.176** 0.150*** 0.158*** 0.123*** (1.16) (1.02) (2.06) (2.05) 1.7 3.05 3.91 2.44 ROA 1.128 0.912 0.604 0.944 2.827 2.070 4.065 3.388 (1.08) (1.23) (1.44) (1.27) 1.2 1.11 0.91 1.6 ind1 0.196** 0.108*** (1.75) (2.59) ind2 0.320*** 0.143** 5.85 1.94 ind3 C -2.534*** -2.086*** -2.325*** -2.088*** -2.358*** -2.994*** -2.279*** -2.716*** (-5.36) (-8.75) (-11.97) (-8.76) -10.55 -3.1 -16.31 -2.78 N 8558 13090 8558 13090 19158 19697 19158 19697 adj.R-sq 0.501 0.348 0.558 0.351 0.322 0.269 0.375 0.303
Table 9 Continued
Dependent variable: the logarithm of stock returns
Variables Specification 18 Specification 19 Specification 20
Low group High group Low group High group Low group High group
RDAT 0.582*** 0.705*** (2.16) (2.78) PPEAT -1.085** -0.799*** (-1.85) (-2.14) MAAT -1.303*** -1.086** -1.175** -0.976** -1.127** -0.891*** (-2.56) (-1.92) (-1.89) (-1.88) (-1.88) (-1.99) size -0.094** -0.066** -0.085*** -0.067** -0.020** -0.026*** (-1.95) (-1.91) (-3.99) (-1.93) (-1.76) (-2.35) B/M 0.144** 0.166*** 0.124*** 0.166*** 0.153*** 0.185*** (1.81) (2.93) (2.62) (2.92) (3.34) (2.38) ROA 0.958 0.177 0.127 0.169 0.192 3.482 (0.86) (0.56) (0.09) (0.53) (0.58) (1.24) ind1 ind2 ind3 0.017** 0.020*** (1.93) (2.12) C -2.881*** -2.491*** -2.327*** -2.490*** -2.633*** -2.988*** (-2.29) (-5.7) (-15.33) (-5.7) (-11.92) (-9.22) N 18126 18237 18126 18237 7718 11924 adj.R-sq 0.317 0.295 0.365 0.309 0.623 0.494
30
This is the same in the specification 1 via specification 13. Company capitalization and book-to-market ratio are important control variates for this study. However, return on asset is not an important control to test the hypothesis. The analysis has different results for young and mature companies.
According to specifications 14 and 15, investments of research and development have larger influences on young firms than mature firms. The reason for that may be the projects of research and development held by young companies are riskier and more significant for young firms than projects of R&D for mature firms. To be more specific, projects of research and development held by young companies can be forced to stop or change more easily by policy changes, industry environment or macroeconomic environment. Also, in comparison to mature companies, young companies are more likely to be affected by the failure of research and development projects The stock performance of young companies oppose to higher level of volatility due to the combined effect of higher risk and more important influences.
Opposite regression results to regressions of the first independent variate are shown in Specifications 16 and 17. Internal expenditures are more influenced by companies in low level retained-earning-proportion group. Nonetheless, the coefficients of internal expenditures in the low level group are not significant. Furthermore, in comparison to young companies, mature companies are likely to have more successful patents and projects of research and development via h long-term accumulation. Furthermore, they have more opportunities to make investment in equipment, property and plant to help mature companies have greater strength and become industrial standard in the sector due to relative more abundant capitals of mature companies. As consequence, compared to young companies, mature companies are likely to have more capital expenditures for equipment, property and plant, and the systematic risk of internal investments is more serious in mature companies than in young companies.
depicted in Specifications 18 and 19 separately. Effects from the investments of merger and acquisition are very high in young and mature companies. Also, when compared to companies in high level retained-earing-ratio group, companies in low level group are more influenced. Merger and acquisition play an important role in both young and mature companies. Nonetheless, young companies barely have sufficient capital for merge or acquire of other companies and the fate of a young company can be affected by the consequences of merger or acquisition. The future stock would have more fluctuations which show the pessimistic mood of shareholders for merger or acquisition when young firms invest in merger and acquisition.
All the independent variates are shown in Specification 20 where industry dummies are excluded. The coefficients of the former specifications conform to those of the three kinds of investment activities.
In summary, the influences of the three kinds of investment activities are featured by life cycles, and the hypothesis 2 is convincing. Therefore, Company life cycles should considered to provide clearer understanding on the consequences and improve investors’ confidence during the process of evaluating the effects of internal and external expenditures and research and development investments.
I will conduct robustness checks to demonstrate that the summaries are sound and reasonable in the final part of the study. In order to normalize the three independent variates, I will utilize the existing ending income rather than total ending assets in the robustness checks of the first hypothesis. Furthermore, Fama and Macbeth regression will be utilized to show the robustness checks. The results of robustness check are shown in tables 10.
32
Table 10
Robustness checks for model 14-20
This table depicts the robustness checks for model 14-20. The table describes the average slopes (in percent) and their t-statistics in parentheses from monthly Fama-macbeth cross-sectional regressions of individual stock excess returns for one period on various independent variables and control variables under the groups of high retained earnings or of low retained earnings. The dependent variable is the logarithm of stock returns. RDAT, PPEAT and MAAT orderly denote the amount invested in research and development, in plant, property and equipment, and in merger or acquisition within one fiscal year, and current operating incomes in each year are used to normalize the three independent variables. They are independent variables. The control variates are size, B/M and ROA. Size is the firm capitalization which can be calculated as the natural logarithm of company’s ending market value. B/M represents firm book-to-market ratio which is the year-end stockholders’ equity divided by total stock market capitalization and stock market capitalization of non-tradable shares can be calculated by the multiplying stock price of tradable shares. ROA is return on asset which is expressed by net income divided by year-end total assets. Ind1, ind2 and ind3 are industry dummy variables. If a company belongs to pharmaceuticals or computers industries, ind1 is equal to 1, otherwise it is equal to 0. If a company belongs to transportation, mining and construction, utilities or extractive industries, ind2 is equal to 1, otherwise it is equal to 0. If a company is in the food, pharmaceuticals, transportation or services industries, industry dummy 3 is equal to 1, otherwise it is equal to 0. All independent variables are normalized to zero mean and 1 standard deviation after winsorization at the 1% and 99% levels. The return data are from January of 1962 to March of 2016. The reported adjusted 𝑅2 of each monthly cross-sectional regression.
Table 10
Dependent variable: the logarithm of stock returns
Variables Specification 14 Specification 15 Specification 16 Specification 17
Low group High group Low group High group Low group High group Low group High group
RDAT 0.147*** 0.136** 1.737*** 1.862** (2.34) (1.79) (2.52) (1.86) PPEAT -0.974 -0.746** -0.826 -0.624*** (-1.63) (-1.85) (-1.52) (-2.57) MAAT size -0.068** -0.093** -0.073*** -0.097*** -0.146*** -0.047*** -0.074*** -0.045** (-1.92) (-1.82) (-3.01) (-2.25) (-2.35) (-2.09) (-3.65) (-1.94) B/M 0.147 0.046 0.087*** 0.093*** 0.146** 0.144*** 0.134*** 0.155*** (1.35) (1.02) (2.35) (2.87) (1.72) (3.43) (3.94) (2.42) ROA 1.255 0.945 0.643 0.834 2.454 2.246 4.265 3.464 (1.36) (1.21) (1.64) (1.47) (1.27) (1.34) (1.34) (1.62) ind1 0.147** 0.178*** (1.71) (2.19) ind2 0.407*** 0.131** (5.83) (1.84) ind3 C -2.354*** -2.436*** -2.676*** -2.128*** -2.438*** -2.943*** -2.364*** -2.835*** (-2.56) (-8.15) (-7.97) (-8.74) (-8.45) (-3.34) (-16.46) (-2.32) N 8558 13090 8558 13090 19158 19697 19158 19697 adj.R-sq 0.545 0.346 0.542 0.346 0.247 0.247 0.333 0.344
34
Table 10 Continued
Dependent variable: the logarithm of stock returns
Variables Specification 18 Specification 19 Specification 20
Low group High group Low group High group Low group High group
RDAT 0.543*** 0.743*** (2.46) (2.43) PPEAT -1.300** -0.880*** (-1.80) (-2.46) MAAT -1.475*** -1.022** -1.124** -1.136** -1.187** -0.842** (-2.45) (-1.82) (-1.78) (-1.74) (-1.72) (-1.83) size -0.828*** -0.025** -0.147*** -0.047*** -0.020** -0.025*** (-1.96) (-1.94) (-3.36) (-1.98) (-1.69) (-2.45) B/M 0.155** 0.148*** 0.142*** 0.144*** 0.143*** 0.144*** (1.83) (2.35) (2.46) (2.14) (3.35) (2.54) ROA 0.943 0.243 0.463 0.426 0.186 3.446 (0.43) (0.46) (0.46) (0.54) (0.45) (1.41) ind1 ind2 ind3 0.014*** 0.019*** (1.98) (2.45) C -2.848*** -2.496*** -2.325*** -2.45*** -2.583*** -2.546*** (-2.86) (-5.57) (-15.43) (-5.76) (-9.43) (-9.46) N 18126 18237 18126 18237 7718 11924 adj.R-sq 0.325 0.264 0.349 0.324 0.643 0.452
The tables above indicate that my hypothesis 2 is convincing. In general, in developed market, life cycles have important influences on future stock performance for each kind of investment activities in this paper.
6. Conclusion
Based on the foundation from the literature, I propose my hypotheses of research and use complete regressions to demonstrate its rationality. In my first hypothesis, future stock returns may be influenced by different kinds of investment activities. In particular, the investments of research and development are associated with future stock performance positively; future stock performance are negatively associated with the capital expenditures for plant, equipment and plant; and expenditures of merger and acquisition are associated with the following stock performance negatively yet the influences are greater than the influences of other two kinds of investment activities. In my second hypothesis, there may be different magnitudes of the influences of investment activities in different company life cycles. To be specific, the impacts of research and development investments for young companies are much greater than the impacts for mature companies; the influences of capital expenditures for equipment, property and plant for mature companies are much greater than the influences for young companies; and In spite of the great influences on mature companies, expenditures of merger and acquisition have greater influences on young companies.
First of all, I choose the appropriate methodology and later ensure the variables that can be utilized in order to verify the rationality of my hypotheses. Descriptive data enable me to gain some sub conclusions. In order to o complete the robustness checks, I utilize the Fame-Macbeth regressions in subsequent analysis and use ending company income rather than total ending assets. The rationality of my hypotheses can be generally proved by the Robustness
36
checks.
The below summaries can be obtained: firstly, the influences of investment activities are associated with the environment of industry. To be specific, capital expenditures for equipment, property and plant are more favorable in the mining and construction, utilities, transportation, and extractive industries; the investments of research and development are more favorable in the computer and pharmaceuticals industries; and the expenditures of merger and acquisition are more favorable by pharmaceuticals, food, transportation and services industries.
Secondly, companies with less company capitalization are likely to have less capital expenditure yet more investments of research and development. The reason for that is companies with small capital amount is likely to enhance their investment of research and development in order to get patents and to gain key competitiveness.
Thirdly, young companies are likely to invest more in research development; nonetheless, mature companies are likely to have more external and internal expenditures.
Fourthly, subsequent stock performance can be affected by the kinds of investment activities. To be specific, future stock returns are positively influenced by activities of research and development; future stock returns are negatively influenced by internal and external expenditures; furthermore, external expenditures tend to have greater effects than the other two kinds of expenditures.
Fifthly the company life cycles may affect the influences of investment activities on future stock performance. To be specific, the negative impacts of internal expenditures for mature companies are greater; the positive effects of the investments of research and development for young companies are greater; and though they are large enough for mature and young companies, the negative impacts of external expenditures are greater for young corporates.
This thesis has one limitation that companies can make technological breakthrough via merger or acquisition instead of research and development by the bidder companies. Therefore,
the only concern in this thesis is the format shown in fundamental reports published by the companies when grouping the activities of investment in this thesis.
38
Reference:
Anderson, C., Garcia-Feijoo, L., 2006. Empirical evidence on capital investment, growth options, and security returns. Journal of Finance, 2006, 61(1): 171-194
Chan, L., Lakonishok, J., Sougiannis, T., 2001. The Stock Market Valuation of Research and Development of Expenditures. The Journal of Finance, 56, 2431-2456.
Cooper, M., Gulen, H., Schill, M., 2008. Asset growth and cross-section of stock returns. Journal of Finance, Vol LXIII, No.4
DeAngelo, H., DeAngelo, L., Stulz, R., 2006. Dividend policy and the earned/contributed capital mix: a test of the life-cycle theory. Journal of Financial Economics 81 (2006) 227-254 Fama, F., French, R., 1993. Common Risk Factors in the Returns on Stocks and Bonds. Journal of Financial Economics, 33(1):3-56
Fama, F., French, R., 1992. The Cross-Section of Expected Stock Returns. Journal of Finance, 47(2): 427-465
Gersdorff, N., Bacon, F., 2009. U.S. Merger and Acquisition: A Test of Market Efficiency. Journal of Finance and Accountancy
Hirshleifer D., Hsu P., Li D., 2013. Innovative efficiency and stock returns. Journal of Financial Economics 107 (2013) 632-654
Shah, P., Arora, P., 2014. M&A Announcements and Their Effect on Return to Shareholders: An Event Study. Accounting and Finance Research
