Master Thesis
The effect of unexpected investment on future
stock performance of US airlines during
2005-2014
Danni Guo (10840907)
15
thAugust 2015
2
Statement of Originality
This document is written by Student [Danni Guo] 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.
Abstract
Investment growth effect shows that there is negative relationship between asset investment growth and future stock returns. Some academic research find that planned investment (expected investment) is positive related to subsequent stock returns. As part of total investment, this statement contradict with investment growth effect. In my study, I find that unexpected investment growth is negative correlated with future stock returns and has more predict power than expected investment growth. It indicates that investment growth effect comes from unexpected part instead of expected part of total investment. Firms that overinvest have more strong unexpected investment growth effect.
4
Contents
1. Introduction... 6
2. Literature review... 10
2.1.
Investment growth effect
... 72.2.
Planned investment
... 82.3.
Return anomalies: Accruals, Net share issuance and Return
momentum
... 10
2.4. Firm size and Book-to-Market equity
……….103. Methodolody and Data ... 13
3.1.
Methodology
... 113.2.
Data
... 154. Methodolody and Data ... 17
1. Introduction
A number of theoretical works predict a negative relationship between asset investment and subsequent stock returns (see Cochrane, 1991, 1996; Berk, Green, and Naik 1999; Gomes, Kogan, and Zhang, 2003; and Li, Livdan, Zhang, 2008). There are several explanations for this asset investment effect. According to Titman, Wei and Xie (2004), overinvestment could be one of them. Firms that overinvestment might have negative abnormal returns, which indicates the poor performance on their investment. It is supported by the evidence that there is a much more stronger negative relationship between investment and returns for firms with greater investment discretion. Besides, when firms’ future expected are lower, they tend to invest more, which makes the asset investment effect arises. In other words, because the required rate of return vary over time, firms tend to have more investment when the rate is lower (Berk, Green, and Naik, 1999; Anderson and Garcia-Feijoo, 2006). Besides, when firms are overvalued, they tend to invest more. If the market corrects their overvaluation subsequently, the asset investment effect will occur, which resulting in lower future returns. By using the level of discretionary accruals as a proxy for equity market mispricing, Polk and Sapienza (2009) find that firms do invest more when they are more overvalued.
Planned investment refers to the money a company plans to spend in the coming year on inventory and capital goods. As part of the asset investment, planned investment can be regarded as expected investment. However, some of academic studies find that planned investment have positive relationship with future stock returns, which is contrary to the negative relationship between asset investment and subsequent stock returns (Lamont, 2000; Bhana, 2008, McConnell et al., 1985). The rest of asset investment which is unexpected investment could be the breakthrough point to explain this contradiction. If unexpected investment has negative relationship with future stock returns and this relationship is stronger than the relation between expected investment and subsequent stock returns, asset growth effect can be regarded still effective. Thus, I assume the asset investment effect results from the unexpected
6
instead of the expected part of the asset investment which means unexpected investment has negative relationship with subsequent stock returns.
I choose to research the airline industry because the data of purchase commitment is available. In annual financial reports, airlines would report how much they have committed to purchase in the next few years, which could be used as a proxy for the planned investment. By collecting the amount of committed investments, we can generate a measure of firms’ expected investment. Moreover, according to Hanlon (2007), travel and tourism is regarded as one of the largest industries in the world and air transport accounts for an important part of this industry.
The sample of airline companies in the US during the period of 2005-2014 are used in this study. At first, I am going to use the Fama and MacBeth (1973) regression to test the relationship between unexpected investment growth rate and future stock returns and between expected investment growth rate and subsequent stock returns respectively after controlling two firm characteristics. Then, I will test the regression to see whether the relationship is influenced by the mispricing effect. After that, I will divide firms into two groups including firms with positive unexpected asset growth and with negative unexpected asset growth respectively and do the regressions again to check which group has a stronger relationship.
Consist with the hypothesis, the results show that unexpected investment growth is negatively related to future stock returns while expected investment growth seems do not have strong predictive power.. Also, I find that firms that overinvest have more strong unexpected investment growth effect than those who are underinvest.
This paper sheds lights on our understanding the implication of the predictive power of unexpected investment growth on future stock returns. The main contribution of this work is building the negative relationship between unexpected investment and future stock returns, which shows that the asset investment effect results from unexpected part of the total investment.
The rest of this paper is organized as follows. Section 2 provides a review of the existing literature. Section 3 describes the methodology and the dataset of this study. The results will be given in Section 4 and relate them to earlier hypothesis. Section 5 concludes the paper and offers some directions for further research.
2. Literature review
2.1 Asset investment growth effect
There are some theoretical works that predict a negative relationship between asset investment and subsequent stock returns (see Cochrane, 1991, 1996; Berk, Green, and Naik 1999; Gomes, Kogan, and Zhang, 2003; and Li, Livdan, Zhang, 2008). The firm asset growth rate has an economically and statistically important ability to forecast returns in both large and small capitalization stocks. In the cross-section of stock returns, the asset growth rate has large explanatory power with the respect to other determinants of the cross-section of returns such as size and book-to-market ratio. Moreover, according to Cooper, Gulen, and Schill’s (2008) research, he suppose that on June 39th of every year from 1968 to 2006 an investor sorted U.S stocks into ten equal portfolios based on the past year’s percentage change in aggregate assets. If the investor bought the stocks with the highest past growth, the mean annual portfolio return would have been 4%. If the investors bought the stocks with the lowest growth stock would have experienced a 22% return premium. This empirical fact refers as the asset investment growth effect.
2.2 Planned investment
In a survey organized by the Commerce Department of the US government, during the period from 1947 to 1993, firms are asked to report planned capital expenditure in the next few years. They find that investment plans account for more than three
8
fourths of the variation in real annual aggregate investment growth.1 Purchase commitment is a commitment to purchase a specific number of items from suppliers in the future at a fixed price. A committed purchase can’t be canceled without paying a fee. Except for the survey, there is no database containing information about planned investment, and the survey ends up at 1993. It is safe to assume that a firm that has committed to purchase $1 billion worth of equipment is planning more investment than a firm that has committed to purchase $0.1 billion. Thus, using purchase commitment as a proxy variable for total planned investment is reasonable. Leland et al. (1977) argue that the dedication of internal funds could be treated as an information of the project quality. They believe that the action of managers who are willing to invest contains information of the project quality, since managers have inside information on firms’ projects. Also, investment plans have substantial forecasting power for returns (Lamont, 2000). He argues that planned investment are positively related with current stock returns. Bhana (2008) states that planned investment activity may have implications for current and future earnings. He researches 378 cases of planned investment made by South African companies during the period 1995-2004 and finds significant positive excess returns surrounding capital spending announcements, which means markets treat the announcements of investment as a good signal. McConnell et al. (1985) study a couple of firms that announce their planned investment from 1975 to 1981 and find that announcements of firms’ planned investment contains information of firms’ value. They also state that an announcement of increases in planned investment increases the price of the firm. A reasonable inference could be made that the more managers plan to invest in the next few years, the more confidence they have in their projects. Due to the information asymmetry, the amount of committed investment included in firms’ annual reports indicates managers’ willingness to invest and the high planned investment may serve as a signal that the projects are of good quality. Lenders are more willing to lend money to high quality projects, which means that the firm
1 Real actual investment growth is calculated by the formula = −
commits to invest a lot in the next few years has more access to external funds. However, the positive relationship between planned investment and future stock returns is not consisted with the asset investment growth effect. Thus, I assume the rest part of the total asset might have more predict power. My hypothesis is the unexpected part of asset growth rate is negatively related to subsequent stock returns.
2.3 Return anomalies: accruals, net share issuance, and return momentum
There are several financial analysis concentrate on examining the accrual and cash flow parts of current earnings for the sake of predicting future earning (Graham et al., 1962; Brownlee et al., 1990; Bernstein, 1993; Haskins et al., 1993; Kieso and Weygandt, 1995). Since investors tend to focus on reported earnings, this kind of analysis can be used to detect mispriced stocks. In Sloan’s (1996) study, he investigates whether stock prices reflect information about future earnings contained in the accrual and cash flow part of current earnings. The results show that the earning performance of accrual has lower persistence effect than earning performance of cash flow. Stock prices act as if investors fail to identify correctly the different properties of these two components of earnings. As the consequence, firms with higher accruals experience lower future stock returns. Besides, Teoh, Welch, and Wong (1998) and Xie (2001) provide similar evidence that higher discretionary accruals predict lower stock returns by examining the specific component of total accruals.
The stock issuance includes equity offerings, share repurchase announcements, and stock mergers. Firms issue share when it is overvalued and retire share when it undervalued. Thus, whether long-run stock returns followed by stock issuance reflect mispricing has been discussed for many years. By using the method of Stephens and Weisbach (1998), Pontiff and Woodgate (2006) construct an annual issuance measurement for all stocks to test the relationship between stock issuance and future stock returns. Their results indicate that share issuance are strongly correlated with the subsequent stock returns. By investigating initial public offering (IPO), Daniel and
10
Titman (2006) find the same results as well. In addition, Fama and French (2008) show that market mispricing on accruals and net share issuance is pervasive, thus, I will use total accruals and net share issuance to proxy for equity mispricing and examine the misvaluation-based explanation on the investment effect.
In addition to these two anomalies, momentum can be regarded as a premier return anomaly. According to Jegadeesh and Titman’s (1993) finding, stocks with low returns over the last year tend to have low returns for the next few months and stocks with high past returns tend to have high future returns. Therefore, momentum will be served as control variable in my study.
2.5 Size and book-to-market equity
The Market-to-Book ratio of a firm is the ratio of a firm’s market value over its book value while firm size refers to the market value of equity. According to Banz (1981), firm size has a negative linear function of their expected returns on securities. He finds that average returns on small stocks (low market equity) are high while average returns on large stocks (high market equity) are low. The Sharpe-Lintner-Black (SLB) model plays a significant role in drawing attention to the relationship between risk and expected returns. Thus, Fama and French (1992) uses this model to test the effect of size on average returns by using ten size portfolios based on the NYSE breakpoint. As a result, size is proved that has a strong negative explanation power of average stock. In addition, Stattman (1980) and Rosenberg, Reid and Lanstein (1985) find that average returns on U.S stocks have positive relationship with the ratio of a firm’s book value of stocks to its market value. Chan Hamao, and Lakonishok (1991) find that book-to-market equity has strong the explore power of average returns on Japanese stocks. Moreover, Fama and French (1992) find that there is not only a positive relationship between book-to-market equity and returns, but also this book-to-market effect is even more powerful than the size effect. In summary, both firm size and book-to-market equity have a significant relationship with the average
return since these variables are proxies for time-varying systematic risk. Thus, I am going to use firm size and book-to-market equity to control for risk in the analysis.
3. Methodology
3.1 Model
At first, firms are divided into two groups according to whether their unexpected asset growth rates are positive (Unexg>0, overinvestment) or negative (Unexg<0, underinvestment). Then, the Fama and MacBeth (1973) regression is used in this study to investigated the relation between unexpected assets growth rate and future stock returns. The regression model is estimated year-by-year using the OLS method with all firms available in a given year:
. = + , + ∗ , + ( ), + ∗ ( ),
+ , + , + , + ,
where indexes firms, indexes the year. represents firms’ stock returns. and are the unexpected and expected parts of total asset growth respectively. SZ and BM are firm size and the book-to-market equity ratio, which are used to control firm characteristics. Ln is natural logarithm. and are accruals and net share issuance that are regarded as proxies for equity mispricing (Fama and Frech, 2008). MOM is used to control the momentum effect (Jegadeesh and Titman, 1993). u is error term. The regression is performed yearly and the averages of the yearly estimated coefficients are used to investigate the effects of the explanatory variables on stock returns. is the coefficient of unexpected asset growth rate which is expected to be indifferent from zero when equity mispricing is negative co-vary with stock returns. Firm size (Ln(SZ)), Accruals (TAC) and net share issuance (NS) are expected to be negative related to stock returns, while book-to-market equity ratio (Ln(BM)) and momentum (MOM) are expected to positive related to stock returns. In other words, , and are expected to be negative, while and are expected to be positive.
12
3.2 Data
The data of the US airline companies through 2005 to 2014 are gathered from the website of U.S. Securities and Exchange Commission (SEC)2. 174 airlines are included within the code of 4512 and 4522. Unexpected asset growth rate (Unexg) is
( ) ( )
( ) and expected asset growth rate
(Exg) is
) . Committed investment is used as a proxy of expected investment.
Publicly-traded airline companies typically report the amount of committed investment for the next few years directly in their 10-K filings. The airlines’ 10-K filings could be found from SEC-filings, the official websites of companies or Thomson Reuters Eikon. The data of committed investments for each year from 2005-2014 were hand-collected from the annual report (10-K) of airlines in the US for the fiscal year from 2005 to 2014. The data could be found in the footnote to financial statement called “commitments and contingencies” or some footnotes with similar titles, it is the amount of money used to purchase aircrafts (purchase commitment). It is composed of two parts which are aircraft purchase commitments and other purchase commitments. Airlines without reporting their purchase commitment are excluded from the sample. After collecting the purchase commitments (in million USD), 28 airlines are contained in the sample. The main variables and control variables could be found in the standard database COMPUSTAT, over the period 2004-2010. Actual investment is measured by capital expenditure. Firm size (SZ) refers to the market value of equity, measure as price times shares outstanding the end of June of year t. The Market-to-Book ratio of a firm (BM) is the ratio of a firm’s market value over its book value, which is using market value of equity plus book value of assets minus book value of equity minus deferred taxes, divided by book value of assets. The total accruals (TAC) are measured as the difference between earnings and cash flow from operations, scaled by average total assets over the period. Cash flow are measured as income before extraordinary items plus depreciation. Net stock issuance (NS) are measured as the natural log of the ratio of the split-adjusted shares outstanding at the
2
fiscal yearend in t-1 to the split-adjusted shares outstanding at the fiscal yearend in t-2. The split-adjusted shares outstanding are measured as the shares outstanding times the adjustment factor. Momentum (MOM) are measured as the cumulated continuously compounded stock return from month j-12 to month j-2, where j is the month of the forecasted return. All the variables are winsorized. This selection procedure finally leaves 190 observations of airlines during the period 2005-2014. After the selection procedure above, the sample period from 2005 to 2014 but with gaps, so the data I acquire are unbalanced panel data.
Table 1. Summary statistics for selected variables in the sample
This table describes the means (Mean), minimums (Min), maximums (Max) and standard deviations (sd) for selected variables.
Variable Mean Min Max sd Unexg 0.13 -0.70 4.62 0.29 Exg -0.04 -1.05 0.50 0.10 NS 0.03 -1.36 3.62 0.11 MOM 0.19 -0.99 38.31 0.65 Ln(SZ) 4.92 0.08 10.91 2.02 Ln(BM) -0.30 -3.52 3.82 0.81 N 190
4.Results and analysis
Table 2. The Fama-MacBeth regression results by adding two control variables
This table reports the results of the Fama-MacBeth regressions by adding two control variables Ln(SZ) and Ln(BM). Column 1 estimates the effect of unexpected investment growth rate and two firm characteristics (size and book-to-market equity ratio) on future stock returns. The column 2 are the results after adding firm and time fixed effect. In column 3, unexpected investment growth rate is replaced by expected investment growth. Both expected and unexpected investment growth rate are included in column 5. Column 4, 5 and 6 extend Fama and MacBeth regression results by adding two equity misevaluation proxies (total accruals and net stock issuance) and
14
a momentum effect. All variables are defined in Model and Data. All standard errors are clustered at the firm level and t-statistics are reported in parentheses.
(1) (2) (3) (4) (5) Intercept 2.05 2.35 2.15 2.45 2.44 (4.05) (5.35) (4.65) (5.80) (5.69) Unexg -0.53 -0.83 -0.86 (-5.23) (-6.85) (-7.05) Exg -0.58 -0.70 -0.78 (-1.01) (-1.62) (-1.80) Ln(SZ) -0.10 -0.15 -0.11 -0.16 -0.15 (-1.66) (-2.96) (-2.03) (-3.14) (-3.07) Ln(BM) 0.22 0.28 0.20 0.24 0.23 (1.96) (2.95) (1.41) (2.37) (2.33)
Firm fixed effect Time fixed effect
R No No 0.3075 Yes Yes 0.3811 No No 0.2517 Yes Yes 0.2653 Yes Yes 0.4102
In this table, the regression coefficient on Unexg in column 1 is -0.53 with a t-statistic of 5.23 in column 1 while the coefficient is changed to -0.83 with a t-statistic of 6.85 in column 2 after adding firm and time fixed effect, which indicates that the fixed effects make coefficient more significant, and there is negative relationship between unexpected investment growth rate and dependent variable subsequent stock returns. After regressing expected investment growth rate, Ln(SZ) and Ln(BM) on future stock returns in column 3, I add fixed effects to the regression in column 4 as well. The regression coefficient on Exg in column 4 is -0.70 with t-statistic of -1.62, which is not significant.
In column 5 where both Unexg and Exg are included in the regression, the coefficient on Exg becomes significant at 10% level, while the coefficient on Unexg remains highly significant. It indicates that unexpected investment growth rate has a significant negative effect on future stock returns, while expected investment growth rate does not in the control of size and book-to-market effects. Besides, the coefficient on Ln(SZ) and Ln(BM) are -0.15 with t-statistic of -3.07 and 0.23 with t-statistic of 2.33, which are both significant. Thus, the results consist with findings of Fama and
French (1992) that size has a strong negative explanation power of average stock as well as Stattman (1980) and Rosenberg, Reid and Lanstein’s (1985) findings that average returns on U.S stocks have positive relationship with the ratio of a firm’s book value of stocks to its market value.
Table 3. The extended Fama-MacBeth regression results by adding three return anomalies (accruals, net share issuance, and return momentum)
This table reports the extended results of the Fama-MacBeth regressions by adding three return anomalies (accruals, net share issuance, and return momentum). Column 1 estimates the effect of unexpected investment growth rate, two equity misevaluation proxies (total accruals and net stock issuance) and a momentum effect on future stock returns. The column 2 are the results after adding firm and time fixed effect. In column 3, unexpected investment growth rate is replaced by expected investment growth. Both expected and unexpected investment growth rate are included in column 5. All variables are defined in Model and Data. All standard errors are clustered at the firm level and t-statistics are reported in parentheses.
(1) (2) (3) (4) (5) Intercept 2.03 2.28 2.11 2.27 2.27 (5.54) (5.62) (5.68) (5.70) (5.6) Unexg -0.62 -0.68 -0.69 (-5.02) (-5.17) (-5.025) Exg 0.21 0.24 0.05 (0.60) (0.63) (0.13) Ln(SZ) -0.11 -0.17 -0.14 -0.17 -0.17 (-3.26) (-3.48) (-3.24) (-3.55) (-3.55) Ln(BM) 0.21 0.29 0.29 0.31 0.31 (3.01) (3.25) (3.34) (3.48) (3.47) TAC -1.43 -1.52 -1.85 -1.90 -1.61 (-5.01) (-5.33) (-6.71) (-6.89) (-5.81) NS -0.21 -0.23 -0.79 -0.82 -0.23 (-0.65) (-0.78) (-3.01) (-3.10) (-0.80) MOM 0.41 (2..98) 0.50 (3.08) 0.48 (3.01) 0.50 (3.08) 0.50 (3.05)
16
The regression coefficient on Unexg is significant negative in column 2 and 5 (-0.68 with t-statistic -5.71 and -0.69 with t-statistic -5.025), while the regression coefficient on Exg is insignificant in column 4 and 5 (0.24 with t-statistic 0.63 and 0.05 with t-statistic 0.13). It shows that unexpected investment growth rate still has a strong negative relationship with future stock returns, while expected cannot predict future stock performance. After controlling both misevaluation proxies and risk, unexpected investment growth effect still works. Besides, the coefficient of TAC and MOM are -1.61 with t-statistics of -5.81 and 0.50 with t-statistics of 3.05 respectively, which are both significant. It is consistent with findings of Teoh, Welch, and Wong (1998) and Xie (2001) that higher discretionary accruals predict lower stock returns. In summary, results from the Fama and MacBeth regressions indicate that asset investment effect results from unexpected rather than expected parts.
Table 4. The Fama-MacBeth regression results on the overinvestment effect on stock returns
This table reports the results of the Fama-MacBeth regressions by dividing firms into two groups. The samples in column 1, 2 and 3 are firms that have positive unexpected investment growth rate and in column 4, 5 and 6, the firms have negative unexpected investment growth rate. Column 1 estimates the effect of unexpected investment growth rate and two firm characteristics (size and book-to-market equity ratio) on future stock returns. In column 2, unexpected investment growth rate is replaced by expected investment growth. These two variables are both included in column 3. Column 4, 5 and 6 extend Fama and MacBeth regression results by adding two equity misevaluation proxies (total accruals and net stock issuance) and a momentum effect. All variables are defined in Model and Data. All standard errors are clustered at the firm level and t-statistics are reported in parentheses.
R Firm fixed effect Time fixed effect
No No 0.3251 Yes Yes 0.3982 No No 0.285 Yes Yes 0.3065 Yes Yes 0.4856
Firms with Unexg>0 Firms with Unexg<0 (1) (2) (3) (4) (5) (6) Intercept 2.02 1.89 1.93 1.89 1.94 1.93 (5.21) (5.08) (5.27) (5.28) (5.10) (5.25) Unexg -0.40 -0.43 -0.56 -0.47 (5.08) (-2.68) (-1.57) (-1.22) Exg 0.25 0.40 0.34 -0.11 (0.42) (0.66) (0.84) (0.26) Ln(SZ) -0.13 -0.13 -0.13 -0.11 -0.12 -0.11 (-2.75) (-2.75) (-2.80) (-2.52) (-2.67) (-2.65) Ln(BM) 0.37 0.37 0.41 0.24 0.27 0.27 (3.55) (3.55) (3.90) (2.97) (3.15) (3.15) TAC -1.75 -1.75 -1.31 -1.35 -1.67 -1.48 (-3.35) (-4.13) (-3.98) (-4.17) (-4.81) (-4.65) NS -0.45 -0.78 -0.47 -0.76 -0.76 -0.76 (-1.58) (-2.79) (-1.62) (-1.71) (-1.75) (-1.75) MOM
Firm fixed effect Time fixed effect
R 0.64 (3.47) Yes Yes 0.4011 0.63 (3.44) Yes Yes 0.3583 0.62 (3.44) Yes Yes 0.4652 0.52 (3.39) Yes Yes 0.3858 0.53 (3.41) Yes Yes 0.3102 0.52 (3.40) Yes Yes 0.3968
It can be seen from the table that the coefficient of Unexg is -0.40 in column 1 and -0.43 in column 3, and both are significant at 1% level. It indicates that there is strong overinvestment effect after controlling for risk and equity mispricing proxies in overinvestment firms. In addition, the coefficient of Exg in both column 3 and 6 is statistically indifferent from zero. It shows that underinvestment effect on the future stock returns does not exist. Thus, it can be concluded that the unexpected investment effect only occurs in firms that overinvest and this overinvestment effect sustains strong under the effect of equity misevaluation and risk proxies.
5.Conclusion
In this paper, Fama-MacBeth model is used to studying the relationship between unexpected investment growth and subsequent stock returns. At the same time, 28 airlines in the US are chose as samples and the data of purchase commitments of these companies from 2005 to 2014 are collected by hand. A new measurement of planned investment (expected
18 investment) is designed by using purchase commitment as a proxy. Then I use the capital expenditures as the total investment to get the data of unexpected investment by deducting expected investment.
In the testing part, unexpected investment growth rate and expected investment growth rate
are used as independent variables to investigate their effect on future stock returns. As a result, I find there is a strong negative effect between unexpected investment growth rate and subsequent stock returns. However, as part of the total asset investment, expected investment growth rate seems does not have strong predict power of future stock returns, which is different from the opinions of McConnell et al. (1985) and Bhana (2008) who believe expected investment have positive effect on subsequent stock returns. Although, I can still make a conclusion that the asset investment effect comes from the unexpected part of total asset investment. According to the empirical results, this unexpected investment effect still works after controlling firm size and book-to-market ratio, which indicates neither risk nor equity mispricing is able to fully explain this effect. Moreover, I also find that this effect is strong only in the groups of firms that invest more than it should be according to investment opportunities and financial constraints. This indicates that unexpected investment effect largely results from overinvestment rather than underinvestment.
Looking back to my paper, lack of samples is the limitation of my estimations. Since there is not a database containing the data of purchase commitment, I have to collect those data by hand from each airlines’ annual report, which limits the range of samples. I believe if there are sufficient samples, my results will be more significant.
Reference
1. Anderson, C. W. and L. Garcia-Feijoó, 2006, “Empirical evidence on capital investment, growth options, and security returns,” Journal of Finance 61, 171-194.
2. Albuquerque, R., and Neng Wang. “Agency conflicts, investment, and asset pricing.” Journal of Finance 63 (2008), 1-40.
3. Baz, Rolf W., 1981, The relationship between return and market value of common stocks, Journal of Financial Economics 9, 3-18.
4. Berk, Jonathan B., Richard C. Green, and Vasant Naik, 1999, “Optimal investment, growth options and security returns,” Journal of Finance 54, 1153-1607.
5. Bhana, N. (2008). The market reaction to capital expenditure announcements.
Investment Analysts Journal, (68), 57-68.
6. Cochrane, John H., 1991, “Production-Based Asset Pricing and the Link Between Stock Returns and Economic Fluctuations,” Journal of Finance, Vol. XLVI, No. 1, pp. 209-237.
7. Chan, Louis K., Yasushi Hamao, and Josef Lakonishok, 1991, Fundamentals and stock returns in Japan, Journal of Finance 46. 1739-1789
8. Carlson, Murray, Adlai Fisher, and Ron Giammarino, 2004, “Corporate investment and asset price dynamics: Implications for the cross-section of returns,”
Journal of Finance 59, 2577-2603.
9. Chen, Long, and Lu Zhang, 2009, “A better three-factor model that explains more anomalies,” Journal of Finance, forthcoming.
10. Cochrane, John H., 1996, A cross-sectional test of an investment-based asset pricing model, Journal of Political Economy 104, 572-621.
11. Daniel, Kent D., and Sheridan Titman, 1997, “Evidence on the characteristics of cross-sectional variation in common stock returns,” Journal of Finance 52, 1-33. 12. Daniel, Kent D., Sheridan Titman, and K.C. John Wei, 2001, “Explaining the
cross-section of stock returns in Japan: factors or characteristics?” Journal of
20
13. Daniel, Kent, and Sheridan Titman, 2006, Market reactions to tangible and intangible information, Journal of Finance 61, 1605-1643.
14. Fama, Eugene F. and J. MacBeth, 1973, “Risk, return and equilibrium: Empirical tests,” Journal of Political Economy, 81, 607-636.
15. Fama, Eugene F., and Kenneth R. French, 1993, Common risk factors in the returns, Journal of Finance 47, 427-466.
16. Fama, Eugene F., and Kenneth R. French, 2008, Journal of Finance vol. 63, no. 4:1653–1678
17. Fazzari, S.M., R.G. Hubbard and B.C. Petersen, 1988, “Financing constraints and corporate investment,” Brookings Papers on Economic Activity 141-195.
18. Frankel, R. and C. M. C. Lee, 1998, "Accounting Valuation, Market Expectation, and Cross-sectional Stock Returns." Journal of Accounting and Economics, 25, 283-319.
19. Gomes, Joao, Leonid Kogan, and Lu Zhang, 2003, “Equilibrium cross-section of returns,” Journal of Political Economy 111, 693-732.
20. Hubbard, R.G., 1998, “Capital-market imperfections and investment,” Journal of
Economic Literature 36, 193-225.
21. Hubbard, R.G., A.K. Kashyap and T.M. Whited, 1995, “Internal finance and firm investment,” Journal of Money, Credit, and Banking 27, 683-701.
22. Jegadeesh, Narasimhan, and Sheridan Titman, 1993, Returns to buying winners and selling losers: Implications for stock market efficiency, Journal of Finance 48, 65-91.
23. Lamont, O. (1997). Cash flow and investment: Evidence from internal capital markets. The Journal of Finance, 52(1), 83-109.
24. Lamont, O. A. (2000). Investment plans and stock returns. The Journal of
Finance, 55(6), 2719-2745.
25. Li, Erica X. N., Dimitry Livdan, and Lu Zhang, 2008, “Anomalies,” Review of
Financial Studies 22, 2973-3004.
26. McConnell, J. J., & Muscarella, C. J. (1985). Corporate capital expenditure decisions and the market value of the firm. Journal of financial economics, 14(3),
399-422.
27. Michael J. Cooper, Huseyin Gulen, and Michael J. Schill, 2008, Asset growth and the cross-section of stock returns, Journal of Finance, VOL. LXIII, NO.4
28. Pontiff, Jeffrey, and Artemiza Woodgate, 2008, Share issuance and cross-sectional returns, Journal of Finance, Vol LXIII, NO.2.
29. Polk, Christopher, and Paola Sapienza, 2009, “The stock market and corporate investment: A test of catering theory,” Review of Financial Studies 22, 187-217. 30. Stattman, Dennis, 1980, Book values and stock returns, The Chicago MBA: A
Journal of Selected Papers 4, 25-45.
31. Sloan, Richard G., 1996, Do stock prices fully reflect information in accruals and cash flows about future earnings?, The Accounting Review 71, 289-315.
32. Stephens, Clifford P., and Michael S. Weisbach, 1998, Actual share acquisitions in open-market repurchase programs, Journal of Finance 52, 313–333.
33. Titman, Sheridan, K.C. John Wei and Feixue Xie, 2004, “Capital investments and stock returns,” Journal of Financial and Quantitative Analysis 39, 677-700. 34. Wei, K.C. John and Feixue Xie, 2008, “Accruals, capital investments, and stock
returns,” Financial Analysts Journal 64 (Number 5), 34-44.
35. Xie, H., 2001, “The mispricing of unexpected accruals,” The Accounting Review 76, 357-373.
22
Appendix
Table 5. Variables definition
The variables used in this paper are listed below with Compustat data items in parenthesis. Most of variables are measured as in Fama and French (2008).
Actual investment Measured as capital expenditures
Exg Expected investment growth rate. Measured as
) , expected investment is measured as purchase commitments
Unexg Unexpected investment growth rate. Measured as
( ) ( )
( )
SZ size or market value of equity, measured as price times shares outstanding at the end of June of year t.
BM book-to-market equity, measured as the ratio of the book value of equity to the market value of equity. The book value of equity is total assets (data 6), minus liabilities (data 181), plus balance sheet deferred taxes and investment tax credit (data 35) if available, minus preferred stock liquidating value (data 10) if available, or redemption value (data 56) if available, or carrying value (data 130), in that order. The market value of equity is measured at the end of December of year t.
TAC Total accruals, measured as the difference between earnings (data 172) and cash flow from operations (data 308), scaled by average total assets over the period. Since data item 308 is not available prior to 1987, we use the funds flow from operations (data 110) minus current accruals to measure cash flow from operations.
NS Net stock issuance, measured as the natural log of the ratio of the split-adjusted shares outstanding at the fiscal yearend in t-1 to the split-adjusted shares outstanding at the fiscal yearend in t-2. The split-adjusted shares outstanding are measured as the shares outstanding (data 25) times the adjustment factor (data 27). NS is set to zero if information needed to compute NS is missing.
MOM Momentum, measured as the cumulated continuously compounded stock return from month j-12 to month j-2, where j is the month of the forecasted return.
Table 6. Correlation analysis
This table reports correlations of variables.
variable TAC NS MOM Ln(SZ) Ln(BM)
TAC 1.00 0.04 -0.06 -0.09 0.004 NS 0.004 1.0000 -0.07 0.15 -0.22 MOM -0.06 -0.02 1.0000 0.07 0.06 Ln(SZ) -0.07 0.06 0.01 1.0000 -0.48 Ln(BM) 0.21 -0.11 0.03 -0.48 1.0000 Figure I
Actual and plan growth rate
Figure I shows the information about the growth rate changes of plan and actual investment during 2006 to 2014. It can be seen from the graph that the actual investment experiences a sharp decrease from 2007 and starts to recovery from 2011. It might result from the financial crisis start from the end of 2007. Due to the crisis, companies suffer from significant loss and confront with the capital deficiency. As discussed above, investment and cash flow have a sensitive relationship. Thus, when there are financial constraints, firms lessen their investment. However, planned investments decrease gradually from 2008. It is because companies always make their investment decision in advance, which make a lagged reaction to the crisis.
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1 2006 2007 2008 2009 2010 2011 2012 2013 2014 Year
图表标题
Plan CAPX24
Figure II
The growth rate of the difference between actual and plan investment and stock returns
Figure II represents the information about the growth rate of the difference between actual and plan investment and the growth rate of the stock price. It can be seen from the figure that when the growth rate of stock returns below zero, the area of the growth rate of difference is negative as well. In general, these two rates covary together, and both reach the bottom point at 3 (2008) resulting from the financial crisis. It proves the empirical results, which is the difference between actual and plan investment have a forecasting power of stock returns. Thus, firms ending up investment more than plan can be regarded as an indicator for better future stock performance.
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1 1.2 -2 0 2 4 6 8 10 12 1 2 3 4 5 6 7 8 9 St o ck r et u rn s D IF g Year
