Master Thesis
MSc Finance
The Equity Share in Capital Issues and Aggregate Stock Returns: an
example from the United Kingdom (2003-2013)
by Pieter Zandee
Abstract
This paper shows that the share of equity issues in total capital issues has no predictive power in estimating future market returns in the United Kingdom from 2003 to 2013. Neither do I find evidence that firms tend to issue relatively more equity after periods of high market returns. The results suggest that the ability of financial managers to time stock market return is overestimated.
JEL classifications: G02, G14, G32
Author: Pieter Zandee
E-mail: zandee.pieter@gmail.com
Phone: +31650276545 Student number: S1781421 Supervisor: Dr. L. Dam1
2
I analyze whether financial managers time the market, by investigating whether firms issue relatively more equity prior to periods of low stock returns and vice versa. Equity market timing strategies refer to the practice of issuing shares at high prices and repurchasing at low prices (Baker and Wurgler, 2002). In this study I focus only on the practice of issuing equity at high prices.
The rationale behind the market timing strategy is as follows. Financial managers have a certain view about the firm value that is implied by the current share price. When they believe the market temporarily overvalues their stock, they can time their stock by issuing equity before they expect the share price to drop again and the overvaluation disappears. In case of overvaluation, the cost of equity is at that moment relatively low. Hence, firms will prefer to issue equity when they believe their stock is overvalued and issue debt otherwise.
When the share of equity issues in total capital issues is relatively high, the market timing theory attributes this partly to managers timing their stock. Therefore, a high equity share could serve as a sign for investors that market perception is temporarily over-optimistic. This has two consequences. First, firms will take advantage of it by issuing relatively cheap equity as a source of new financing. Second, subsequent expected market returns will be lower.
This is in line with the findings of Baker and Wurgler (2000). They find in their study based on US data between 1928 and 1997 that a high equity share in total capital issuance predicts significant lower market returns. They conclude that managerial timing of an inefficient equity market is the most credible explanation for their results. Hereby they implicitly assume that a high equity share in capital issues is largely caused by timing practices of managers. The main question of interest in this paper is whether the equity share in total capital issuance in the United Kingdom has any predictive power in forecasting returns of the FTSE All Share Index.
I use a dataset from the Bank of England on capital issuance from 2003 to 2013. Baker and Wurgler (2000) investigate the predictive power of the equity share only on one-year-ahead market returns. Assuming that financial managers are indeed trying to time the market, I believe that the results of their potential timing ability should be visible in less than one year. Therefore I examine both six and twelve months lags of the equity share in testing its predictive ability.
3
In addition to the initial gross series, I make several adjustments in defining the equity share. I adjust the equity share for repurchases and I add commercial paper issuance to bonds issuance in determining the debt component. These adjusted definitions of the equity share also predict negative six-months-ahead market returns, but again the results are not statistically significant. The only definition that has marginally significant predictive power (significant at the 10% level) is the series that is adjusted for repurchases and includes commercial paper issuance. A one standard deviation increase in this adjusted equity share variable predicts a 15.6 percent lower market return six months ahead.
Unlike the findings of Baker and Wurgler (2000) based on U.S. data, I do not find any significant evidence that the equity share in total capital issues has any predictive power in estimating future market returns in the United Kingdom from 2003 to 2013. Moreover, considering an increase in the equity share as a sign of market timing seems impetuous. There may be many other motives for firms to issue more equity, for instance high costs of alternative funding.
The article proceeds as follows. Section II presents an overview of the literature. Section III describes the data that is used. Section IV examines the predictive ability of the equity share in capital issuance and presents the results. Section V concludes this paper and mentions limitations.
II. Literature review
When firms want to attract additional external financing, the first question is always whether to issue debt or equity. This financing decision has been a debate since Modigliani and Miller‟s (1958) original work on capital structure. They show that in an efficient market where security pricing follows a random walk, and in the absence of taxes, bankruptcy costs, agency costs and asymmetric information, the value of the firm is unaffected by the way it is financed.
Since then, a large strand of literature addresses this financing decision. The static trade-off theory, which originates from the work of Kraus and Litzenberger (1973), states that a firm will borrow up to the point where marginal tax benefits equal the costs associated with the increased probability of financial distress.
4 Market timing theory
In addition to the above-mentioned traditional theories on capital structure, Taggart (1977) already touched upon the possibility of market timing strategies by firms. However, he also stated that further research was required. Marsh (1982) was one of the first to show that companies are influenced by market conditions and the past history of security prices in choosing between debt and equity. More specifically, he finds that management‟s appetite for equity is greater after strong recent stock market performance. Hovakimian et al. (2001) find similar results: firms that experience large stock price increases are more likely to issue equity than firms that experience stock price declines.
Baker and Wurgler (2000) also recognize the relationship between equity issuance and stock pricing. They claim that stock prices periodically diverge from fundamental values, and that managers and investment bankers take advantage of overpricing by selling stocks to overly optimistic investors. The assumption that securities can be (temporary) under- or overvalued is, as Schultz (2003) states, “an anathema to those who believe markets are efficient.” Given the possibility for these deviations from their fundamental values, firms may want to take advantage of this when determining their capital structure. They may want to issue equity before the (temporary) overvaluation of their equity disappears. The intention is to exploit temporary fluctuations in the cost of equity relative to the cost of other forms of capital (Baker and Wurgler, 2002). Hence, firms will issue equity when they perceive the relative cost of equity as low, and issue debt when they perceive the relative cost of equity as high (Huang and Ritter, 2004). Strategies like these are referred to as market timing strategies. Baker and Wurgler (2002) define equity market timing as the practice of issuing shares at high prices and repurchasing at low prices. Nelson (1999A) states that after firms issue equity their stock tends to do poorly and after firms repurchase equity their stock tends to do well. Subsequently, Nelson (1999B) finds supporting evidence for the explanation that firms are exploiting their superior knowledge about the value of their stock by buying it when it is undervalued and selling it when it is overvalued. Baker and Wurgler (2002) also claim that it is well-documented that firms issue equity when their market values are high. According to the market timing theory, market returns should subsequently be lower because the large share of equity issues is partly caused by stock timing of financial managers.
5
A survey by Graham and Harvey (2001) confirms the existence of this incentive in corporate finance practice. In an anonymous survey amongst 392 chief financial officers (CFO‟s) of public corporations in the United States, they find that two-thirds of the respondents agree that the magnitude of equity undervaluation/overvaluation is an important aspect in the decision to issue equity. Moreover, the survey shows that recent stock price performance is the third most popular factor affecting equity-issuance decisions.
The survey of Graham and Harvey (2001) indicates that financial managers will prefer to issue equity over debt when they believe their stock is overvalued because in this case the cost of equity is relatively cheap. The findings of Baker and Wurgler (2000) serve as a confirmation that managers try to time the stock market. Their financing variable is the share of equity issues in total equity and debt issues. They find that the equity share sometimes predicts significantly negative market returns. Since expected returns on the market are likely to be positive in any rational model, they eventually conclude that managerial timing of an inefficient equity market is the most credible explanation for their results. They state that the stock market as a whole may be inefficient and that managers exploit this inefficiency within their financing decisions. Their findings are in line with the characteristics of equity market timing. In a study on share buyback behavior, Brockman et al. (2001) also claim that their results show that managers exhibit substantial timing ability.
In contrast, Schultz (2003) states that managers have no timing ability. He argues that if firms are indeed able to time their stock to benefit from overvaluation, one would expect managers to also gain from this when they are trading their own shares in the company. However, Lee (1997) shows that when top executives sell their own shares in a secondary seasoned offering, their performance is not significantly different from that of similar firms. Other studies of insider trading (Lakonishok and Lee 2001; Jenter, 2005) also question the power of managers‟ timing ability. According to Schultz (2003), the findings by Lee (1997) contradict the assertion of behavioralists that managers are able to time the market with equity issues. Schultz (2003) alters the definition of market timing and refers to it as pseudo market timing. He states that the premise of the pseudo market timing hypothesis is that the more firms can receive for their equity, the more likely they are to issue stock, even if the market is efficient and managers have no timing ability. Hence, pseudo market timing refers to the tendency of firms to issue equity following high returns. Schultz (2003) shows how the long-run underperformance of IPOs and SEOs fits the pseudo market timing hypothesis.
6
the equity share based on the remaining 93% observations of the sample disappears. Hence, they question the timing ability of managers. They claim that the predictability comes from the negative relation during those two economic shocks. Moreover, Butler et al. (2005) also find that a model using the equity share as a predictor of future market returns does not outperform an unconditional model that includes only a constant term. Hence, consistent with the results of Schultz‟ pseudo-market timing theory, they claim that the equity share has no out-of-sample predictive power.
In addition, Butler et al. (2005) also looked at the predictive power of the change in the equity share rather than the absolute levels of the equity share. Again, they find no evidence of out-of-sample predictability.
Jenter et al. (2011) study the market timing hypothesis in a put option setting. They claim that management‟s desire to exploit perceived mispricing is a highly plausible motive for put option sales, whereas other motives seem suspect. They find that managers are able to identify undervalued equity and successfully time the market with their put option sales: in the 100 trading days following a put option issue, there is an abnormal stock return of nearly 5%.
In total, the role of market timing as an explanation for the financing decision continues to be an unresolved item. There appears to be consensus about the attempts of managers to time their stock, but their ability to do so is severely questioned. To the best of my knowledge there exists no recent research which explores the predictive ability of the equity share in total capital issuance for the United Kingdom. Since the Bank of England has been collecting data on capital issuance in the United Kingdom intensively since 2003, I will examine this dataset using a similar methodology as Baker and Wurgler (2000).
III. Data
The focus of this study is on financing that is provided through equity and bonds issuance. To determine the equity share in total capital issuance, figures on both equity issuance and bonds issuance are required. I define the equity share as
(1)
where is the equity share in total capital issuance, is total equity issuance and is total debt issuance at time t.
7
compared to bonds in their financing decision. Nonetheless, for robustness I also present the results when the equity and bonds issues are adjusted for repurchases / retirements.
Since data on other funding alternatives such as internal financing or bank loans is not easily available, I use bonds issuance as an indicator of how firms favor debt compared to equity. Bond issuance in this study reflects corporate bonds, meaning Private Non-Financial Companies (PNFC) only. These are private corporations that produce goods and services for the market and do not, as primary activity, deal in financial assets and liabilities.2
Commercial paper issuance is not included in the baseline analysis either. Commercial paper is generally used to finance working capital or to manage liquidity with maturities rarely exceeding 270 days. Hence, it tells little about how firms favor equity compared to debt with respect to their long-term external financing decisions. However, as a robustness check I include the results when commercial paper issuance is included in the total debt issuance.
The capital issuance I consider in this study is based in the United Kingdom and is comprised of sterling and foreign currency issuance by UK residents, and by sterling issuance of residents outside the UK. Including foreign currency issuance is important because capital is nowadays increasingly issued in foreign currencies, mainly for firms to access a wider investor base and to match the currency risk associated with their business activities.3
All capital issuance by (i) Monetary Financial Institutions (MFI), such as the central bank (Bank of England) and (ii) Other Financial Corporations (OFC), also known as financial quasi corporations such as insurance brokers and loan brokers, are neglected because their financing needs are heavily influenced by ongoing changes in financial regulations (i.e. specific capital requirements stipulated by Basel Accords). Capital issuance for public administration, defense, education, health and other social services are not included either.
The dependent variable in this study is the simple return on the FTSE All Share total return index. I use this index since it comprises all listed firms in the United Kingdom and therefore it is a good proxy to determine the market returns of the firms included in the dataset from the Bank of England. A. Aggregate Equity and Debt Issues Data, 2003-2013
The Bank of England has reported monthly levels of equity issues (ordinary and preference shares) and debt issues (bonds and commercial paper) since 2003. The statistics are derived from information
2
http://www.ons.gov.uk/ons/rel/naa1-rd/united-kingdom-national-accounts/the-blue-book--2013-edition/chapter-03--non-financial-corporations.html
8
provided by Issuing and Paying Agents (IPAs), who act on behalf of companies issuing debt on the UK capital markets, and by the London Stock Exchange (LSE).4 The IPA submits monthly information on each bond issue and on commercial paper transactions. The LSE submits similar information on each equity issue.
B. Issuance overview
Figure 1 shows that equity issuance was rising until 2006 when it started to stagnate. In 2009, the absolute volume of equity issuance rose drastically. According to the 2011 Q4 Quarterly Bulletin from the Bank of England, the increase in equity issuance was mainly driven in order to reduce leverage rather than to finance new projects. 5 Since 2011, total new equity issuance has been low. The appetite of investors for new equity issues might be lower due to poor returns on equity following the subprime mortgage crisis that tore down global stock indices.
Bonds issuance follow a similar pattern as equity issuance, although annual absolute levels are less volatile compared to equity issuance. According to the 2011 Q4 Quarterly Bulletin from the Bank of England, the spike in 2009 was mainly driven by new issuers whom intended to repay maturing bank loans. 4
C. Equity share in total new issues
Panel C shows the development of the equity share in total new issues. The equity and bond issuance levels are not adjusted for repurchases. Table I reports the summary statistics for these series. In the first half of the sample (02/2003 to 06/2008) absolute levels of equity and debt issuance are lower, as well as their standard deviations. The equity share is higher (0.27) in the first half of the sample compared to 0.22 in the second half of the sample (07/2008 to 12/2013). Throughout the entire sample, the deviations in debt issuance levels are relatively lower compared to equity issuance levels. The returns on the FTSE All Share Index also vary throughout the sample. In the first half, total returns are higher with lower standard deviations compared to the second half of the sample. This was likely due to the uncertain economic environment after the subprime mortgage crisis had emerged and the ensuing recession that lasted for several years.
9 PANEL A: Equity issue volume (e) in ₤M
PANEL B: Bond issue volume (e) in ₤M
PANEL C: The equity share in total new issues [ES = e/(e+b)]
Figure 1. Equity and bond issues, annual basis, 2003-2013. Equity and debt issue volumes are from the Bank
10 Table I
Monthly Equity and Bond Issues for the U.K., 02/2003 – 12/2013.
Monthly means and standard deviations for equity and bond issues as well as for equity market returns. Panel A: Equity and debt issue monthly volumes are from the Bank of England. The equity series (e) includes both common and preferred equity issues. The bond series (b) consists of bond issues. The equity share in total new issues (ES) is the fraction of equity issues in total new issues [ES = e/(e+b)]. The equity market returns on the FTSE All Share Index are monthly total simple returns. The amount of observations equals N = 119 months.
02/2003 – 12/2013 02/2003 – 06/2008 07/2008 – 12/2013
Mean SD Mean SD Mean SD
Panel A: Equity and debt issuance
Equity issues (e) 901 1309 846 1008 954 1553
Bond issues (b) 2805 1912 2526 1874 3076 1923
ES = e/(e+b) 0.24 0.19 0.27 0.19 0.22 0.18
Panel B: Equity market returns
11 IV. Results
This section explores the predictive power of the equity share in new issues on market returns. A. Preliminary analysis
Figure 2 and Figure 3 rank the levels of equity share from low to high and subsequently divides them into four quartiles. Figure 2 shows the mean annual returns of the FTSE All Share Index six months after the corresponding equity share level. The mean annual returns are 4.5 percent if the prior six months equity share ranks in the first quartile (equity share between 0.003 and 0.087). The mean annual returns are 35.1 percent if the prior six months equity share ranks in the second quartile (equity share between 0.094 and 0.205). Altogether, there seems to be no relationship between the level of equity share in new issues and the market returns six months later.
Figure 2. Mean equity returns by prior six months equity share in new issues, 2003-2013.
Mean annual return on the FTSE All Share Index by quartile of the prior six months share of equity issues in total equity and bond issues.
12
Figure 3. Mean equity returns by prior year equity share in new issues, 2003-2013. Mean annual return on
the FTSE All Share Index by quartile of the prior year share of equity issues in total equity and bond issues.
In Figure 4 and Figure 5 the sample is split into two, based on the level of equity share. All months with an equity share below the median of 0.204 are considered low equity months, whereas the other half above the median are considered high equity months.
Panel A in Figure 4 shows that low equity months tend to occur after periods of high return and are also followed by periods of high return. Panel B shows that high equity months tend to occur after periods of low return. This is counterintuitive as one would expect that firms are more interested in issuing equity after a period of high returns. This intuition is confirmed by the findings of Baker and Wurgler (2000), who state that firms lean toward equity after a year of high returns and lean away from equity after a year of low market returns. I test this hypothesis in Appendix A and do not find any evidence for this statement.
13
Panel A: Mean annual returns on the FTSE All Share Index around low equity years
Panel B: Mean annual returns on the FTSE All Share Index around high equity years
Figure 4: Mean past and future equity returns by equity share in new issues, 2003-2013. Figure 4A plots
mean past and future returns index for below-median equity share years. Figure 4B plots mean returns around above-median equity share years.
14
Panel A: Mean annual returns on the FTSE All Share Index around low equity years
Panel B: Mean annual returns on the FTSE All Share Index around high equity years
Figure 5: Mean past and future equity returns by equity share in new issues, 2003-2013. Figure 4A plots
mean past and future returns index for below-median equity share years. Figure 4B plots mean returns around above-median equity share years.
Table II tabulates Figure 4 and 5 more formally. I compare the mean returns six months and twelve months after low- and high equity months. After high equity months, mean annual returns tend to be lower, in particular for six and twelve months after the mean high equity month.
15 Table II
Past and future equity returns around high and low equity share years, 02/2003 – 12/2013
Market returns are mean annual total returns on the FTSE All Share index. The low share and high share are determined by whether the year zero share of equity issues in total new equity and bond issues [ES = e / (e + b)] was above or below the median of 0.204. Table includes figures where ES is lagging 6 and 12 months behind the respective market returns. t-statistics are shown in brackets and test the hypothesis of no difference between average returns.
FTSE All Share index
Years relative Low share High share Difference t-statistic Years relative Low share High share Difference t-statistic
16 B. Univariate regressions
To test the predictive power of the equity share in capital issuance I use ordinary least squares (OLS) time series regressions. The linear regression has the following form:
(2)
where is the total return between time t and k on the FTSE All Share Index and is the
equity share in total capital issuance at time t.
The equity share is standardized such that a one standard deviation increase in predicts a percent higher return. The null hypothesis is that and a rejection of this hypothesis would imply that the equity share in total issues has significant power in predicting market returns. I estimate equation (2) over the full sample period and over the first and second halves of the sample.
Table III shows the OLS regressions of six and twelve months ahead stock returns ( on the equity
share in new issues (ES). The results indicate that the six months ahead stock returns are on average 7.4 percent lower when the equity share increases by one standard deviation. The signs for are negative in both halves of the sample for the equity share lagging six ( months behind the
observed market returns. This is not the case for mean returns twelve months after the corresponding equity share level ( . However, none of the results are significant. Hence, I cannot reject the null
hypothesis that the equity share in total issues has no significant power in predicting the market. After adjusting the market returns for the risk-free rate by using the Bank of England‟s 3 months LIBOR rate, the results change very marginally and do not provide any additional insights (not reported in a table).
The OLS regression assumes the estimated equation to be linear. I use the Ramsey RESET test to check whether non-linear combinations of the fitted value help to explain the dependent variable in (2). In both cases the non-linear fitted values are not significant.6 Hence, I cannot reject the null hypothesis that the linear form implied by (2) is correct.
6
17 Table III
Predictive power of equity share in new issues for subsequent stock returns for the UK, 02/2003 – 12/2013.
OLS regression of annual total market return on the equity share in new issues ( ):
where denotes percentage returns on the FTSE All Share total return index and denotes the equity share in total new equity and bond issues [ES =
e / (e + b)]. is standardized to have zero mean and unit variance. The table contains results for with lags of 6 and 12 months. t-statistics are shown in brackets.
* significant at 5% level, **significant at 10% level
R² ²
Panel A: 02/2003 – 12/2013 returns
FTSE All Share 1.27 -7.60 0.01 0.71 1.16 0.00
(2.06)* (-0.85) (1.10) (0.13)
Panel B: 02/2003 – 06/2008 returns
FTSE All Share 1.43 -9.70 0.01 0.83 2.81 0.00
(2.44)* (-0.74) (1.31) (0.20)
Panel C: 07/2008 – 12/2013 returns
FTSE All Share 1.87 -9.84 0.01 1.51 -7.12 0.01
18 C. Alternative definitions of the equity share
In this section I consider other measures for calculating the equity share in capital issuance. Table IV shows the results.
The first row repeats the results of the unadjusted equity share as a predictor of the FTSE All Share Index. In the subsequent row, the equity and bonds issue volumes are adjusted for repurchases. Consequently, this row shows the results when using monthly net figures on equity and bonds issuance. The third row includes the issuance of commercial paper in addition to bond issues. These two then comprise the total debt issuance, whereas equity issuance is unadjusted in this row. The fourth row adjusts the third row for repurchases. Comparable to the second row, the fourth row reports net capital issuance rather than gross capital issuance.
Since net equity issues can be negative, this can cause problems. For instance, in January 2005 net equity issuance was -£1.181 billion and net bonds issuance was £1.504 billion. Hence, following the same approach as earlier in this paper, the net equity share would be -3.66. Net equity issuance is negative in 95 of 132 months. In order to have meaningful equity share series, I set all negative numbers for net equity shares to zero.
The third column shows the correlation of the adjusted series with the original series B. When adjusted for repurchases, the correlation with the original series drops to 0.18. After adding gross commercial paper issuance, the correlation is 0.77. If this series is adjusted for repurchases, it drops to 0.12.
19 Table IV
Univariate OLS regressions for predicting one-year ahead market returns
OLS regression of annual total market return on the equity share in new issues ( ):
where denotes percentage returns on the FTSE All Share total return index and denotes the equity share in total new equity and bond issues . The
definition of equity (e) is adjusted to account for repurchases. The definition of bond issues (b) is adjusted to account for repurchases, gross commercial paper issuance and net commercial paper issuances. For all adjustments, is standardized to have zero mean and unit variance. The table contains results for with lags of 6 and 12 months. t-statistics are shown in brackets.
Alternative definitions of ES Correlation with
unadjusted ES
e b t( ) R² t( ) R²
No adjustment No adjustment 1.00 -7.60 (-0.85) 0.01 1.16 (0.13) 0.00
Subtract repurchases Subtract repurchases 0.18 -6.69 (-0.74) 0.00 0.50 (0.54) 0.00
No adjustment Add commercial paper issuance 0.77 -8.87 (-0.99) 0.01 8.98 (0.98) 0.01
20 V. Conclusion
I find no significant evidence that the equity share in total capital issues has any predictive power in estimating future market returns in the United Kingdom from 2003 to 2013. Neither do I find evidence that firms tend to issue relatively more equity after periods of high market returns.
Hence, one might question the true predictive nature of the equity share in forecasting future market returns. The paper by Baker and Wurgler (2000) relies heavily on two economic shocks. Moreover, it is based on 70 annual observations and therefore it might suffer from small-sample bias. Schultz (2003) addresses this bias for studies which use managerial decision variables to predict market returns. Butler et al. (2005) highlight the potential for this bias in the predictive regressions based on the equity share in capital issues. They state that the predictive power stems from pseudo-market timing and not from managers‟ abnormal ability to time the equity markets. Baker et al. (2006) claim that the predictability of several managerial decision variables in forecasting market returns is too strong to attribute to small-sample bias, but they acknowledge that more research on this issue is needed.
Another explanation for the insignificant results of this paper could be the properties of the dataset that I use. The dependent variable is characterized by low variation in the long-term returns (see Appendix B). Moreover, the dataset includes 119 observations. A longer horizon of observations will contribute to making more accurate statements about the true predictive power of the equity share.
Lastly, labeling an increase in the equity share as a sign of „market timing‟ might be too optimistic. There can be many more motives that serve as a determinant in the financing decision other than market timing, such as the cost of alternative sources of external funding.
Limitations
21 REFERENCES
Baker, M., Taliaferro, R. and Wurgler, J. (2006), “Predicting returns with managerial decision variables: is there a small-sample bias?”, Journal of Finance, 61.
Baker, M., and Wurgler, J. (2000), “The equity share in new issues and aggregate stock returns”,
Journal of Finance, 55, 2219-2257.
Baker, M., and Wurgler, J. (2002), “Market timing and capital structure”, Journal of Finance, 57, 1–32. Baker, M., and Wurgler, J. (2011), “Behavioral Corporate Finance: An Updated Survey”, Working paper,
National Bureau of Economic Research (NBER).
Brockman, P. and Chung, D. Y. (2001), “Managerial timing and corporate liquidity: Evidence from actual share repurchases, Journal of Financial Economics, 61, 417-448.
Butler, A. W., Grullon, G. and Weston, J. P. (2005), “Can managers forecast aggregate market returns?”,
Journal of Finance, 60.
Graham, J.R. and Harvey, C.R. (2001), “The theory and practice of corporate finance: evidence from the field”,
Journal of Financial Economics, 60,187-243.
Hovakimian, A., Opler, T., and Titman, S. (2001), “The debt-equity choice”, Journal of
Financial and Quantitative Analysis 36, 1–24.
Huang, R., and Ritter, J. R. (2004), “Testing the market timing theory of capital structure”, Working paper,
University of Florida.
Jenter, D. (2005), “Market timing and managerial portfolio decisions”, Journal of Finance, 60, 1903-1949. Jenter, D., Lewellen, K., and Warner, J. B. (2011), “Security issue timing: what do managers know, and when do they know it?”, Journal of Finance,66, 413-443.
Kraus, A., and Litzenberger, R.H. (1973), "A State-Preference Model of Optimal Financial Leverage", Journal
of Finance, 911-922.
Lakonishok, J. and Lee, I. (2001), “Are insider trades informative?”, Review of Financial Studies, 14, 79-112 Lee, I., (1997), “Do firms knowingly sell overvalued equity?”, Journal of Finance, 52, 1439-1466.
Marsh, P. (1982), “The choice between equity and debt: an empirical study”, Journal of Finance, 37, 121 – 144. Modigliani, F., Miller, M.H. (1958), “The cost of capital, corporate finance and the theory of
investment,” American Economic Review, 49, 665-669.
Myers, S.C. (1984), “The capital structure puzzle,” Journal of Finance, 39, 575-592.
Nelson, W. R., (1999A), “The aggregate change in shares and the level of stock prices”, Working paper,
Federal Reserve Board.
Nelson, W. R., (1999B), “Why does the change in shares predict stock returns?”, Working paper, Federal
Reserve Board.
Appendix A. Market returns as a predictor of the equity share in capital issuance.
Baker and Wurgler (2000) state that firms lean away from equity after a year of low market returns, and that firms lean toward equity after a year of high returns. To test this predictive power of past market returns in forecasting the equity share in capital issues I use ordinary least squares (OLS) time series regressions. The linear regression has the following form:(3)
23
Table A1: Univariate OLS regressions for predicting equity share in new issues
OLS regression of the equity share in new issues ( ) on annual total market return (
where denotes the equity share in total new equity and bond issues [ES = e / (e + b)] and
denotes the percentage returns on the FTSE All Share total return index. is standardized to have zero mean and unit variance. The table contains results for with lags from one to twelve months. t-statistics are shown in brackets.
24
Appendix B. FTSE All Share Total Return Index
Figure B1. FTSE All Share Total Return Index, 2003-2013. Total return index data is retrieved from
