Printer Friendly

Junior-for-senior announcements a study of the role of inside ownership.

Robert M. Hull (*)

Abstract

In this paper, we investigate the role of inside ownership in explaining announcement period stock returns for 455 junior-for-senior transactions. We find that returns are more negative for firms with higher inside ownership percentages. Returns become even more negative for firms in which insiders are expected to be decreasing their ownership percentages. [C] 2002 Elsevier Science Inc. All rights reserved.

JEL classification: D820; G140; G320

Keywords: Inside ownership; Junior-for-senior; Announcement period return; Security type; Firm size

1. Introduction

Empirical researchers (Cornett & Travlos, 1989; Finnerty, 1985; Hull, 1994; Masulis, 1983; Shah, 1994) document statistically significant negative stock returns at the announcement of pure leverage decreases, consisting of junior-for-senior transactions. (1) They propose an information explanation for the significant negative market response. For example, leverage decrease announcements can signal negative information stemming from insider behavior (Leland & Pyle, 1977), the alteration in fixed obligations (Ross, 1977), adverse selection (Myers, 1984), and firm size (Bhushan, 1989). Other explanations for the negative response include those associated with lost corporate tax shield (Modigliani & Miller, 1963), issue costs (Hull & Fortin, 1993/1994; Hull & Kerchner, 1996), and agency-based costs (Galai & Masulis, 1976; Jensen, 1986; Jensen & Meckling, 1976).

In a separate line of empirical work, researchers investigate the effect of inside ownership on the market's reaction to the announcements of various corporate decisions, including stock repurchases (Vermaelen, 1981), dividend initiations (Born, 1988), sell-offs (Hirschey & Zaima, 1989), and stock splits (Han & Suk, 1998).(2) They find that firms with higher inside holdings experience greater announcement period returns. Han and Suk (1998) offer further insight into the role of inside ownership by documenting that the significant relationship between inside ownership and announcement returns is caused by small firms. To date, researchers have not examined how the percentage (or level) of inside ownership influences the announcement period returns for pure leverage decreases.

In this paper, we extend pure leverage decrease and inside ownership research by testing whether announcement period stock returns for junior-for-senior announcements are more negative for firms with higher levels of inside ownership. We also examine whether returns are more negative when insiders are expected to have greater decreases in their ownership levels.

We begin our analysis by dividing our sample of 455 junior-for-senior transactions into three portfolios formed according to the percentage of inside ownership. The average announcement period stock returns are - 1.39%, - 2.13%, and - 3.16% for the portfolios with the lowest, middle, and highest inside ownerships, respectively. The differences in returns among the three portfolios generate significant statistics for various tests. These significant differences support the notion that junior-for-senior announcements by firms with higher inside ownership have a more negative market response. Correlation and regression tests offer further support for the important role played by inside ownership. We show that both the level of inside ownership and the expected change in the level influence the market's reaction.

We organize the remainder of the paper as follows. Section 2 develops testable hypotheses and Section 3 presents data and announcement period results. We provide correlation and regression results in Section 4 and offer concluding statements in Section 5.

2. Testable hypotheses

Myers and Majluf (1984) present a model in which asymmetry of information can lead a manager (who acts in the interests of all shareholders) to be unwilling to issue junior securities if those securities are undervalued in the public market. One implication of this model is that the market will react negatively to the announcement of the issue of a junior security. The market realizes that a manager has a reservation level of undervaluation below which the manager will refuse to issue the junior security. The decision to issue, therefore, informs the market that firm value is not above this reservation value. Thus, the distribution of possible firm values shifts downwards, leading to a lower expected value for the firm.

Myers and Majluf (1984) assume that the manager wants to issue securities to raise cash needed to undertake an investment opportunity. In addition, they effectively assume that there is no agency problem between the manager and the shareholders. By relaxing both of these assumptions, we can generate a hypothesis regarding the effect of inside ownership on the announcement of junior-for-senior transactions.

Once we allow for agency problems between the manager and the shareholders, the manager's reservation level of undervaluation may increase, if not disappear altogether. The manager may prefer, for example, to have less debt and more equity in order to be less subject to the discipline that debt imposes. If the manager is totally unconcerned with the cost that issuing undervalued equity imposes on the shareholders, then there is no level of undervaluation at which the manager will not be willing to exchange equity for debt. In this circumstance, the decision to make such an exchange does not signal anything regarding the manager's private information about the true value of the firm.

This extreme outcome stems from the fact that the manager receives benefits from the exchange but bears no costs. If a manager owns equity in the firm, then the costs borne by the shareholders are also borne, in part, by the manager and the greater the manager's ownership, the greater the costs imposed by the exchange. To the extent that the cost to the manager of a junior-for-senior exchange eventually outweighs the benefit, a reservation level of undervaluation is restored to the model. Ceteris paribus, the manager's reservation level of undervaluation will be lower the greater is the manager's ownership of the firm. Thus, the greater is the manager's ownership, the stronger is the negative signal associated with a willingness to exchange junior-for-senior securities.

The above argument leads to our primary research hypothesis (Hypothesis 1), which asserts:

Hypothesis 1: The stock returns for junior-for-senior announcements will be more negative for firms with higher levels of inside ownership.

Besides the level of inside ownership, the change in this level should also influence stock returns. Leland and Pyle (1977) advocate that stock returns are positively related to the change in the level of inside ownership. Because insiders are privy to information about the firm's earnings prospects, the market fears that they may try to decrease their ownership percentages when prospects are poor. Thus, our secondary hypothesis (Hypothesis 2) can be formulated:

Hypothesis 2: The stock returns for junior-for-senior announcements will be more negative for firms in which insiders are expected to undergo greater decreases in their levels of ownership.

3. Data and announcement period results

3.1. Data

We identify junior-for-senior announcement dates from The Wall Street Journal (WSJ) and Investment Dealers' Digest (IDD) and gather inside ownership data from the last proxy statement prior to the announcement. We obtain the number of primary and secondary shares from the announcements published in WSJ and IDD. The offering value is computed by multiplying the planned number of shares times the price on the day before the announcement as provided by the CRSP Daily Price Files. When computing firm size values, we rely heavily on data from Compustat Annual Files, Moody's Industrial Manuals, and CRSP Daily Price Files.

We utilize the OLS market model procedure described by Brown and Warner (1985) to compute cumulative abnormal returns (CARs), using return data from the CRSP Daily Return Files. When computing alpha and beta parameters, we use the equal-weighted CRSP NASDAQ and CRSP AMEX/NYSE indices for respective OTC and AMEX/NYSE observations. The comparison period includes the 200 trading days from days -240 to -41 before the announcement day (day 0). (3)

For inclusion in tests, an observation must satisfy the following five screens. First, it must be a pure leverage decrease consisting of a junior-for-senior transaction. Second, it must have available stock return data during its announcement and comparison periods. Third, it must not have confounding news announcements for a 5-day window (days -2 to + 2) surrounding the announcement day. Fourth, it must not use current assets in excess of 30% of the offering value when retiring senior borrowings. (4) Fifth, it must have available data to compute values for variables used in empirical tests, including values for inside ownership percentage, the expected decrease in inside ownership, the relative size of the offering, and firm size.

After applying these screens, we have 455 junior-for-senior transactions remaining. These include 350 common stock-for-non-convertible debt transactions and 105 other transactions that involve at least one security type other than common stock or nonconvertible debt. These 105 transactions consist of 36 convertible debt-for-nonconvertible debt, 36 convertible preferred-for-nonconvertible debt, 18 common-for-convertible debt, 9 nonconvertible preferred-for-nonconvertible debt, 4 convertible preferred-for-convertible debt, and 2 common-for-nonconvertible preferred.

Table 1 provides descriptive statistics for our sample of 455 junior-for-senior transactions. The time period statistics reveal that fewer observations occur for the 1985-1989 period. This can be at least partly explained by the fact that our sources are less likely to report the purpose of the offering during this period. Without this information, we cannot learn whether the junior security offering is used to reduce a senior form of borrowing.

Table 1 next reports statistics for five key variables. The mean (median) inside ownership is 12.37% (5.10%), while the mean (median) primary offering value is US$46 million (US$24 million). For common stock value and firm value, the means (medians) are US$908 million (US$335 million) and US$1879 million (US$617 million), respectively. The mean (median) is 7.97% (5.31%) for primary offering value as a percentage of firm value.

3.2. Announcement period results

To investigate whether returns are more negative for firms with higher levels of inside ownership, we partition our junior-for-senior sample into three portfolios. Portfolio 1 contains firms with the lowest inside ownership percentages. Portfolios 2 and 3 include firms with the middle and highest inside ownership percentages, respectively.

Table 2 provides 3-day announcement period returns for days - 1 to + 1. The first row in Panel A reports that the 3-day CAR is - 2.23% (t = - 10.34) for the total sample. The last three rows in Panel A provide CAR results for the three portfolios. Consistent with Hypothesis 1, these rows reveal that CARs decrease monotonically as the percentage of inside ownership increases. The least negative CAR is - 1.39% (t = - 4.65) for Portfolio 1, in which inside ownership ranges from 0.10% to 2.40%. For Portfolio 2, the CAR decreases to - 2.13% (t = - 6.43), with inside ownership ranging from 2.50% to 12.30%.

For Portfolio 3, the CAR further decreases to - 3.16% (t= - 6.90), with inside ownership ranging from 12.40% to 81.00%.

Panel B in Table 2 reports statistical results when comparing portfolio CARs. The first row provides analysis of variance results when testing whether CARs are identical across portfolios. The F statistic of 5.77 reveals that CARs are significantly different across portfolios. Although not presented in table format, comparable results are found when the sample is partitioned into either four or five portfolios.

The last three rows in Panel B report nonpaired one-tailed parametric t statistics when testing the null hypothesis that the portfolio with lower inside ownership has a CAR that is more negative or equal to that for the portfolio with higher inside ownership. Comparisons between any two portfolios reject the null and support Hypothesis 1 by generating differences in CARs that are significant at the 5% level or better.

The results in both panels of Table 2 offer evidence that the wealth effects differ across firms with dissimilar inside ownership percentages. Consistent with Hypothesis 1, junior-for-senior announcements by companies with higher inside ownership are associated with a more negative market response. Thus, investors in firms undergoing pure leverage decreases can expect a more negative market response as the level of inside ownership increases.

4. Correlation and regression analysis

4.1. Description of variables

We now perform correlation and regression tests to investigate, not only the role of inside ownership, but also that of other determinants of the market's reaction to junior-for-senior announcements. The six variables used in these tests are defined below.

CAR = 3-day average CAR expressed in percentage form.

INS = inside ownership, as measured by the percentage of outstanding common shares owned by insiders at the time of the announcement.

ECH = expected decrease in inside ownership, as measured by secondary shares offered as a percentage of outstanding common shares.

RSZ = relative size of the offering, as measured by the primary offering value as a percentage of firm value.

TYP = 0 if common stock-for-nonconvertible debt, else 1.

LFS = firm size, as measured by the logarithm of firm value. (5)

CAR is the dependent variable used in regression tests. INS tests Hypothesis 1 by examining whether the level of inside ownership influences the market response to the junior-for-senior announcement. As discussed previously, we expect the response to become more negative as the level of inside ownership increases. Thus, we predict a negative coefficient for INS.

ECH examines Hypothesis 2 by testing whether the expected decrease in inside ownership is related to announcement period returns. ECH estimates the decrease in inside ownership based on the following consideration. There are 51 primary offerings for which a secondary offering is attached. For these 51 combination offerings, insiders have an opportunity to decrease their ownership proportion by selling their shares through the secondary offering. When mentioned, the financial press refers to those selling as "principal" or "major" shareholders. Thus, it is reasonable for the market to anticipate that insiders are selling their shares through the secondary offerings. With this in mind, we create the variable ECH. It is the number of secondary shares offered as a percentage of the number of outstanding shares. If there are no secondary shares then ECH is zero. Leland and Pyle (1977) predict a negative coefficient for ECH because greater expected selling of shares by insiders will signal more negative news to the market.

RSZ controls for the relative size of the offering and can capture various wealth effects. Under the assumption that insiders are not increasing their ownership by buying shares from the junior offering, greater positive values for RSZ indicate greater decreases in inside ownership. Thus, Leland and Pyle (1977) predict a negative coefficient because greater decreases in inside ownership (represented by greater RSZ values) convey more negative news. Ross (1977) posits a negative coefficient because a greater reduction in fixed obligations signals a smaller capacity to service these obligations in the future. Modigliani and Miller (1963) predict a negative coefficient because a greater reduction in leverage implies a greater tax shield loss. Prior pure leverage decrease studies (Cornett & Travlos, 1989; Hull & Hull & Kerchner, 2000; Moellenberndt, 1994) generally find an insignificant relation between the relative size of the offering and the abnormal stock returns.

TYP controls for the types of securities issued and retired. Adverse selection signaling theory based in Myers (1984) suggests a positive coefficient for TYP because the market response worsens whenever a more junior security is issued to retire a more senior security. (6) A positive coefficient is also consistent with the notion that issuing common stock involves greater flotation costs than other securities. Prior researchers find mixed results when examining the types of securities issued and retired. Cornett and Travlos (1989) report an insignificant coefficient, while Hull and Kerchner (1997) find a significant coefficient.

LFS controls for firm size. Differential information theory (Bhushan, 1989; Verrechia, 1980) predicts that small firms reveal more information because their security offering announcements are less expected. Thus, we predict a positive coefficient for LFS because junior-for-senior announcements for small firms will convey more negative information. Researchers (Hull, Mazachek, & Ockree, 1998; Hull & Pinches, 1995) find that small firms have a more negative response to pure leverage decrease announcements. They note that a small firm effect may proxy for an issue costs effect in regression tests because issue costs increase as firm size decreases.

4.2. Correlation and regression results

Before regressing independent variables against CARs, we do a correlation analysis to help interpret regression tests. Table 3 provides the correlation results. This table reports significant correlation coefficients ([rho]'s) between CAR and each of the five independent variables.

Table 3 reveals there are pairs of independent variables that have large [rho]'s and thus can produce potential colinearity problems for regression tests. For example, four independent variables (INS, ECH, RSZ, and LFS) are significantly correlated with one another, with Pearson and Spearman [rho]'s ranging from .21 to -.77. The greatest correlation involves LFS with both INS and RSZ. Thus, the most difficult task involves distinguishing the wealth effects associated with firm size from those associated with the level of inside ownership and the relative size of the offering. TYP is the only variable that does not exhibit much potential colinearity with other variables.

Like previous research (Han & Suk, 1998; Hull & Michelson, 1999), we assume a linear functional form for the regression equation. The equation is:

CAR = [a.sub.0] + [a.sub.1]INS + [a.sub.2]ECH + [a.sub.3]RSZ + [a.sub.4]TYP + [a.sub.5]LFS + [member of] where is an error term. Using this equation, OLS regression tests generate coefficients, t statistics, F values, and [R.sup.2] values. We compute one-tailed t statistics for explanatory variables because each has a definite prediction concerning the sign of its coefficient. Although not reported, the regression results are robust to other alternate functional forms, including the White (1980) correction for heteroskedasticity.

Given the serious colinearity that can exist between independent variables, regression tests are not only performed with all variables included but are also repeated four times. With each repetition, one of the following variables is deleted: INS, ECH, RSZ, and LFS. Other tests are performed with various combinations of variables deleted to insure that results are not explained by colinearity. We omit the details of these latter tests because they add little to what is provided in Table 4. (7)

Panel A in Table 4 reports that coefficients for all independent variables exhibit (in varying degrees) evidence of significance. Consistent with H1, the coefficient of INS is negative and significant at the 1% level for all tests. Thus, the valuation effect at the announcement of pure leverage decreases is more negative for firms that have higher levels of inside ownership.

Because of the strong correlation between inside ownership (INS) and firm size (LFS), we regress INS on LFS to create a data set of residuals not correlated with LFS. When tested with ECH, RSZ, and TYP, the variable from this data set is significant at the 1% level with t= -2.38. As reported in Table 4, INS has t= -2.91 when tested with these three variables. We repeat the procedure when LFS is regressed on INS to create a data set of residuals not correlated with INS. The variable from this data set is not significant when tested with ECH, RSZ, and TYP As reported in Table 4, LFS is significant at the 5% level when tested with these three variables. These additional regression tests suggest that the significant findings for INS in Table 4 stem from the level of inside ownership and not from its correlation with firm size. These tests also indicate that the significance of LFS may be explained by its capacity to proxy for the level of inside ownership.

As predicted by Hypothesis 2 and shown by the significant statistics for ECH, the market response is more negative when there are greater expected decreases in inside holdings. RSZ is only significant if we omit LFS from the test. The significant statistics for TYP support a wealth effect associated with two general classes of transactions, while the significant statistics for LFS support an asymmetric information effect as captured by firm size. However, when we include RSZ in the test, the t statistic for LFS is not significant. We attribute this result to the colinearity between RSZ and LFS, which makes it difficult to distinguish between the effect from the relative size of the offering and that from firm size.

To further understand the roles of the relative size of the offering (RSZ) and firm size (LFS), we regress RSZ on LFS to create a data set of residuals not correlated with LFS. When tested with INS, ECH, and TYP, the variable from this data set is not significant. When INS is omitted, it is still insignificant. We repeat the procedure when LFS is regressed on RSZ to create a data set of residuals not correlated with RSZ. The variable from this data set is insignificant when tested with INS, ECH, and TYP. However, when INS is omitted, it is significant at the 5% level. This test reinforces the previously stated notion that the significant findings for LFS can be explained by its capacity to capture a wealth effect stemming from inside ownership. The support for RSZ in Table 4 appears to result from its capacity to capture wealth effects attributed to factors other than the relative size of the offering.

5. Summary

This study extends the pure leverage decrease and inside ownership research by examining the role of inside ownership when firms announce junior-for-senior transactions. Tests of portfolios, formed according to the percentage of inside ownership, show that portfolios with higher levels of inside ownership have stock returns significantly more negative than those with lower levels of inside ownership. Correlation and regression tests offer further evidence that the level of insider ownership influences the market response. These tests also show that the market response becomes more negative when insiders are expected to undergo greater decreases in their ownership. We find that the market reaction continues to become more negative if less junior securities are issued or more senior borrowings are retired. Variables representing the relative size of the offering and firm size appear to be proxying for other effects.

Practically speaking, our findings can guide managers in terms of what to expect when announcing a pure leverage decrease. Most important, managers of firms with high inside ownership can anticipate a more negative market response. This negative response can be mitigated if insiders are willing to maintain or increase their percentages of ownership. Managers can further lessen the negative response if they can issue a security other than common stock or retire a security other than nonconvertible debt.

Acknowledgments

We would like to thank the editor, Diane Denis, and the anonymous referee for their helpful comments.

(*) Corresponding author. Tel.: +1-785-231-1010; fax: +1-785-231-1063.

E-mail address: zzhull@washburn.edu (R.M. Hull).

(1.) Pure leverage decrease studies enable empirical tests to focus on wealth effects resulting from capital structure changes and not from asset structure changes.

(2.) Section 16 of the Securities Exchange Act of 1934 defines insiders as officers, directors, and beneficial owners. "Officer" refers to president, vice president, secretary, treasurer or financial officer, comptroller, principle accounting officer, or any person who does such managerial functions. "Director" refers to an individual who is a member of the board of directors. "Beneficial owner" refers to a large shareholder who owns 10% or more of a class of registered equity shares in the corporation. Officers and directors with fiduciary duties toward the corporation and its shareholders routinely come across material private information during their business functions within the corporation. Beneficial owners, by virtue of their partial control over activities of the corporation through a nominee director, often become privy to material private information.

(3.) Like Han and Suk (1998), we examine other methodological variations and obtain CAR results similar to those we report in this paper.

(4.) Prior research uses this screen (see Shah, 1994).

(5.) Firm value is defined in Table 1. We express firm value in millions of dollars before taking the logarithm.

(6.) Hull and Kerchner (1997) perform an extensive analysis of the variable TYP and conclude that an adverse selection effect is more likely than other wealth effects including signaling effects predicated on decreases in fractional inside holdings or relative debt levels.

(7.) Additionally, we analyzed variance inflation factors to detect whether multicolinearity is a problem (see Kennedy, 1986). Nothing was indicated by this analysis.

References

Bhushan, R. (1989). Firm characteristics and analyst following. Journal of Accounting and Economics, 11, 255-274.

Born, J. (1988). Inside ownership and signals: evidence from dividend initiation announcement effects. Financial Management, 17, 38-45.

Brown, S., & Warner, J. (1985). Using daily stock returns: the case of event studies. Journal of Financial Economics, 14, 1-31.

Cornett, M., & Travlos, N. (1989). Information effects associated with debt for equity and equity for debt exchange offers. Journal of Finance, 44, 451-468.

Finnerty, J. (1985). Stock-for-debt swaps and shareholder returns. Financial Management, 14, 5-17.

Galai, D., & Masulis, R. (1976). The option pricing model and the risk factor of stock. Journal of Financial Economics, 3, 53-81.

Han, K., & Suk, D. (1998). Inside ownership and signals: evidence from stock split announcement effects. Financial Review, 33, 1-18.

Hirschey, M., & Zaima, J. (1989). Insider trading, ownership structure, and the market assessment of corporate sell-offs. Journal of Finance, 44, 971-990.

Hull, R. (1994). Stock price behavior of pure capital structure issuance and cancellation announcements. Journal of Financial Research, 42, 439-448.

Hull, R., & Fortin, R. (1993/1994). Issuance expenses and common stock offerings for over-the-counter firms. Journal of Small Business Finance, 3, 1-17.

Hull, R., & Kerchner, R. (1996). Issue costs and common stock offerings. Financial Management, 25, 54-66.

Hull, R., & Kerchner, R. (1997). Pure leverage decreases: a study of two junior-for-senior groups. Quarterly Journal of Business and Economics, 36, 51-68.

Hull, R., & Kerchner, R. (2000). An ex ante model to estimate issue costs. Journal of Research in Finance, 2, 131-168.

Hull, R., Mazachek, J., & Ockree, K. (1998). Firm size, common stock offerings, and announcement period returns. Quarterly Journal of Business and Economics, 37, 3-24.

Hull, R., & Michelson, S. (1999). The information contents of senior offerings that reduce junior securities. Quarterly Review of Economics and Finance, 39, 419-438.

Hull, R., & Moellenberndt, R. (1994). Bank debt reduction announcements and negative signaling. Financial Management, 23, 21-30.

Hull, R., & Pinches, G. (1994/1995). Firm size and the information contents of over-the-counter common stock offerings. Journal of Small Business Finance, 4, 31-56.

Jensen, M. (1986). Agency costs of free cash flow, corporate finance, and takeovers. American Economic Review, 76, 323-329.

Jensen, M., & Meckling, W. (1976). Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305-360.

Kennedy, P. (1986). A guide to econometrics. Cambridge, MA: MIT Press.

Leland, H., & Pyle, D. (1977). Informational asymmetries, financial structure, and financial intermediation. Journal of Finance, 22, 371-381.

Masulis, R. (1983). The impact of capital structure change on firm value: some estimates. Journal of Finance, 38, 107-126.

Modigliani, F., & Miller, M. (1963). Corporate income taxes and the cost of capital: a correction. American Economics Review, 53, 433-443.

Myers, S. (1984). The capital structure puzzle. Journal of Finance, 39, 575-592.

Myers, S., & Majluf, N. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, 187-221.

Ross, S. (1977). The determination of financial structure: the incentive-signaling approach. Bell Journal of Economics, 8, 23-40.

Shah, K. (1994). The nature of information conveyed by pure capital structure changes. Journal of Financial Economics, 36, 89-127.

Vermaelen, T. (1981). Common stock repurchases and market signaling: an empirical study. Journal of Financial Economics, 9, 139-183.

Verrechia, R. (1980). The rapidity of price adjustments to information. Journal of Accounting and Economics, 2, 63-92.

White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica, 48, 817-838.
Table 1

Descriptive statistics

Time period  Number of observation  Percent of total

1979 - 1981  141                    31
1982 - 1984  253                    56
1985 - 1989   61                    13
Key variable                 Mean       Median

Inside ownership             12.37%     5.10%
Primary offering value       US$46 M    US$24 M
Common stock value           US$908 M   US$335 M
Firm value                   US$1879 M  US$617 M
Primary offering value as a   7.97%     5.31%
 percentage of firm value

This table reports descriptive statistics for the sample of 455
junior-for-senior transactions. M represents millions. Inside ownership
is the percentage of outstanding shares owned by insiders (e.g.,
officers, directors, and other beneficial owners) at the time of the
announcement. Primary offering value is the price the day before the
announcement times the planned number of primary shares. Common stock
value is the stock price the day before the announcement period times
the number of shares outstanding. Firm value includes common stock
value, the liquidation value of preferred stock (if applicable), and the
book value of long-term debt obligations and current liabilities.
Table 2

CAR results for days - 1 to + 1

Panel A: Announcement return results by portfolio

Sample tested   Number of observations  Inside ownership percentages

All portfolios  455                      0.10-81.00%
Portfolio 1     152                      0.10-2.40%
Portfolio 2     152                      2.50-12.30%
Portfolio 3     151                     12.40-81.00%

Sample tested   CARs (%)  t statistic

All portfolios   -2.23    -l0.34 (***)
Portfolio 1      -1.39     -4.65 (***)
Portfolio 2      -2.13     -6.43 (***)
Portfolio 3      -3.16     -6.90 (***)
Panel B: Tests for equality of portfolio returns

Test    Portfolios compared  Test statistic

F test  1, 2, and 3          5.77 (***)
t test  1 and 2              1.66 (**)
t test  2 and 3              1.81 (**)
t test  1 and 3              3.23 (***)

This table reports CAR results for the sample of 455 junior-for- senior
transactions. Panel A reports results for the total sample and three
portfolios when testing whether the CAR is equal to zero. The three
portfolios are formed according to inside ownership. Portfolio 1
contains firms with the lowest inside ownership. Portfolios 2 and 3
include firms with the middle and highest inside ownership,
respectively. As before, inside ownership is the percentage of
outstanding shares owned by insiders at the time of the announcement.
The first row of Panel B reports analysis of variance results when
testing the null hypothesis that CARs are equal across portfolios. The
last three rows report nonpaired one-tailed parametric t statistic when
testing the null hypothesis that the CAR for the portfolio with lower
inside ownership is more negative or equal to the CAR for the portfolio
with higher inside ownership. The symbols ** and *** indicate
statistical significance at the 5% and 1% levels, respectively.
Table 3

Correlation results

     CAR         INS         ECH         RSZ         TYP

CAR  -           -.13 (***)  -.12 (***)  -.15 (***)   .09 (**)
INS  -.21 (***)  -            .39 (***)   .45 (***)   .01
ECH  -.18 (***)   .37 (***)  -            .26 (***)  -.19 (***)
RSZ  -.15 (***)   .30 (***)   .21 (***)  -            .01
TYP   .11 (**)   -.02        -.12 (***)  -.01        -
LFS   .18 (***)  -.47 (***)  -.27 (***)  -.66 (***)   .02

     LFS

CAR   .15 (***)
INS  -.57 (***)
ECH  -.31 (***)
RSZ  -.77 (***)
TYP   .04
LFS  -

This table reports correlation results for the sample of 455
junior-for-senior transactions. The six variables used in the
correlation tests are defined below.

CAR = 3-day average CAR expressed in percentage form. INS = inside
ownership, as measured by the percentage of outstanding common shares
owned by insiders at the time of the announcement. ECH = expected
decrease in inside ownership, as measured by secondary shares offered as
a percentage of outstanding common shares. RSZ = relative size of the
offering, as measured by the primary offering value as a percentage of
firm value. TYP = 0 if common stock-for- nonconvertible debt, els 1. LFS
= firm size, as measured by the logarithm of firm value. The Pearson
correlation coefficients are reported in the lower left- hand half of
the table. The Spearman correlation coefficients are provided in the
upper right- hand half of the table. The symbols (**) and (***) indicate
statistical significance at the 5% and 1% levels, respectively.
Table 4

Regression results

Constant [a.sub.0]    INS [a.sub.1]         ECH [a.sub.2]

-2.5954, -1.78 (*)    -0.0378, -2.45 (***)  -0.1639, -1.81 (**)
-4.0603, -3.04 (***)  --                    -0.2274, -2.60 (***)
-2.7178, -1.86 (*)    -0.0457, -3.09 (***)  --
-3.5616, -3.33 (***)  -0.0373, -2.42 (***)  -0.1684, -1.86 (**)
-1.4911, -4.62 (***)  -0.0420, -2.91 (***)  -0.1679, -1.86 (**)

Constant [a.sub.0]    RSZ [a.sub.3]        TYP [a.sub.4]

-2.5954, -1.78 (*)    -0.0304, -0.98       1.0316, 2.07 (**)
-4.0603, -3.04 (***)  -0.0280, -0.90       1.0017, 2.00 (**)
-2.7178, -1.86 (*)    -0.0332, -1.07       1.1439, 2.31 (***)
-3.5616, -3.33 (***)  --                   1.0255, 2.06 (**)
-1.4911, -4.62 (***)  -0.0452, -1.86 (**)  1.0352, 2.08 (**)

Constant [a.sub.0]    LFS [a.sub.5]      F value     [R.sup.2] values

-2.5954, -1.78 (*)    0.1458, 0.78       6.98 (***)  0.072, (0.062)
-4.0603, -3.04 (***)  0.3055, 1.73 (**)  7.14 (***)  0.060, (0.051)
-2.7178, -1.86 (*)    0.1652, 0.88       7.87 (***)  0.065, (0.057)
-3.5616, -3.33 (***)  0.2587, 1.75 (**)  8.48 (***)  0.070, (0.062)
-1.4911, -4.62 (***)  --                 8.58 (***)  0.071, (0.063)

This table reports regression results for the sample of 455
junior-for-senior transactions. The regression model is:
CAR = [a.sub.0] + [a.sub.1]INS + [a.sub.2] ECH + [a.sub.3]RSZ +
[a.sub.4] TYP + [a.sub.5]LFS + [epsilon], where [epsilon] is the error
term. CAR = 3-day average CAR expressed in percentage form.
INS = inside ownership, as measured by the percentage of outstanding
common shares owned by insiders at the time of the announcement.
ECH = expected decrease in inside ownership, as measured by secondary
shares offered as a percentage of outstanding common shares.
RSZ = relative size of the offering, as measured by the primary offering
value as a percentage of firm value. TYP = 0 if common
stock-for-nonconvertible debt, else 1. LFS = firm size, as measured by
the logarithm of firm value.

For the first six coloumns, we report estimated coefficients and OLS
t statistics. Except for the "constant" column, the t test is
one-tailed because each explanatoy variable has a definite prediction
concerning the sign of its coefficient. The next to the last column
provides F test results and the final column reports unadjusted
[R.sup.2] results with adjusted [R.sup.2] results in parentheses. The
symbols (*), (**) and (***) indicate statistical significance at the
10%, 5%, and 1% levels, respectively.
COPYRIGHT 2001 JAI Press, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hull, Robert M.; Mazachek, JuliAnn
Publication:Review of Financial Economics
Article Type:Statistical Data Included
Geographic Code:4EUUK
Date:Jun 22, 2001
Words:5659
Previous Article:Twenty-five years of corporate governance research ... and counting.
Next Article:Maturity and exercise price of executive stock options.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters