Visit https://ebookfinal.

com to download the full version and

explore more ebooks

Corporate Governance and Finance Volume 8

Advances in Financial Economics Volume 8 1st
Edition K. John

_____ Click the link below to download _____

https://ebookfinal.com/download/corporate-governance-
and-finance-volume-8-advances-in-financial-economics-
volume-8-1st-edition-k-john/

Explore and download more ebooks at ebookfinal.com

Here are some suggested products you might be interested in.
Click the link to download

Corporate Governance Volume 9 Advances in Financial

Economics 1st Edition Mark Hirschey

https://ebookfinal.com/download/corporate-governance-
volume-9-advances-in-financial-economics-1st-edition-mark-hirschey/

Berserk Volume 8 Kentaro Miura

https://ebookfinal.com/download/berserk-volume-8-kentaro-miura/

Gender Perspectives on Reproduction and Sexuality Volume 8

Advances in Gender Research 1st Edition Marcia Texler
Segal
https://ebookfinal.com/download/gender-perspectives-on-reproduction-
and-sexuality-volume-8-advances-in-gender-research-1st-edition-marcia-
texler-segal/

The Larkin Papers Volume 8 1848 1851

https://ebookfinal.com/download/the-larkin-papers-volume-8-1848-1851/
Africa Yearbook Volume 8 1st Edition Andreas Mehler

https://ebookfinal.com/download/africa-yearbook-volume-8-1st-edition-
andreas-mehler/

The New Palgrave Dictionary of Economics 8 Volume Set 2nd

Edition Steven N. Durlauf

https://ebookfinal.com/download/the-new-palgrave-dictionary-of-
economics-8-volume-set-2nd-edition-steven-n-durlauf/

The New Palgrave Dictionary of Economics Second Edition

Volume 8 Steven N. Durlauf And Lawrence E. Blume

https://ebookfinal.com/download/the-new-palgrave-dictionary-of-
economics-second-edition-volume-8-steven-n-durlauf-and-lawrence-e-
blume/

Explaining Financial Scandals Corporate Governance

Structured Finance and the Enlightened Sovereign Control
Paradigm Vincenzo Bavoso
https://ebookfinal.com/download/explaining-financial-scandals-
corporate-governance-structured-finance-and-the-enlightened-sovereign-
control-paradigm-vincenzo-bavoso/

Papers of Thomas Jefferson Volume 8 The Papers of Thomas

Jefferson Retirement Series Volume 8 1 October 1814 to 31
August 1815 Thomas Jefferson (Editor)
https://ebookfinal.com/download/papers-of-thomas-jefferson-
volume-8-the-papers-of-thomas-jefferson-retirement-series-
volume-8-1-october-1814-to-31-august-1815-thomas-jefferson-editor/
Corporate Governance and Finance Volume 8 Advances
in Financial Economics Volume 8 1st Edition K. John
Digital Instant Download
Author(s): K. John, A. Makhija, Mark Hirschey
ISBN(s): 9780762310272, 0762310278
Edition: 1
File Details: PDF, 1.12 MB
Year: 2003
Language: english
LIST OF CONTRIBUTORS

Christopher W. School of Business, University of Kansas, USA

Anderson

Gurmeet S. Bhabra University of Otago, New Zealand

Terry L. Campbell II College of Business and Economics, University of

Delaware, USA

Stephen P. Ferris Department of Finance, University of

Missouri-Columbia, USA

Kathleen P. Fuller Department of Banking and Finance, University

of Georgia and University of Michigan Business
School, USA

Michael B. Glatzer University of Georgia, USA

Mark Hirschey School of Business, University of Kansas, USA

Narayanan Jayaraman DuPree College of Management, Georgia Institute

of Technology, USA

Teresa A. John Stern School of Business, New York University,

USA

Gershon N. Mandelker Katz Graduate School of Business, University of

Pittsburgh, USA

John D. Martin Hankamer School of Business, Baylor University,

USA

Jeffry M. Netter Department of Banking and Finance, University of

Georgia, USA

Annette B. Poulsen Department of Banking and Finance,

University of Georgia, USA
vii
viii

Akin Sayrak Joseph M. Katz Graduate School of Business,

University of Pittsburgh, USA

Susan Scholz School of Business, University of Kansas, USA

Nilanjan Sen Nanyang Technological University, Singapore

Gopala K. Vasudevan School of Management, Northeastern University,

USA

Peng Peck Yen Nanyang Technological University, Singapore

BANK MONITORING, FIRM
PERFORMANCE, AND TOP
MANAGEMENT TURNOVER IN JAPAN

Christopher W. Anderson, Terry L. Campbell II,

Narayanan Jayaraman and Gershon N. Mandelker

ABSTRACT
An inverse relation between performance and managerial turnover at
Japanese firms suggests that bank monitoring substitutes for other gov-
ernance mechanisms (Kaplan, 1994; Kang & Shivdasani, 1995). Morck
and Nakamura (1999), however, report that Japanese banks protect their
self-interests as creditors rather than the interests of shareholders when
appointing corporate directors. We re-examine data on top management
changes at Japanese firms and find results consistent with this latter notion.
Specifically, management turnover is conditionally related to a firm’s ability
to meet its short-term obligations rather than profitability or stock returns.
Bank monitoring is therefore not a substitute for mechanisms that directly
serve shareholders’ interests.

1. INTRODUCTION
U.S.-style mechanisms for corporate control are relatively weak or non-existent
in Japan. For example, the ownership of many Japanese firms is characterized
by cross holding of shares by industrial and financial groups, and the market for

Corporate Governance and Finance

Advances in Financial Economics, Volume 8, 1–27
© 2003 Published by Elsevier Science Ltd.
ISSN: 1569-3732/PII: S1569373203080010
1
2 CHRISTOPHER W. ANDERSON ET AL.

takeovers is relatively inactive (Aoki, 1992; Kester, 1991). Outside directorships

are also rare in Japan (Ballon & Tomita, 1988; Kaplan, 1994), as are individual
block holders (Nishiyama, 1984; Prowse, 1992). An inactive takeover market,
the prevalence of firm grouping, and the absence of individual blockholders
imply that management is insulated from monitoring and control by shareholders
(Nishiyama, 1984). On the other hand, it has been argued that monitoring by com-
mercial banks substitutes for shareholder governance (Aoki, 1990, 1992; Sheard,
1989). Specifically, the intimacy of bank-client relationships, coupled with the
bank’s authority to intervene during periods of financial distress, may serve as a
mechanism that monitors firm performance and disciplines management. Kaplan
(1994, 1997), Kaplan and Minton (1994), and Kang and Shivdasani (1995) report
an inverse relation between Japanese firm performance and managerial turnover
that is seemingly consistent with this notion.
Since the collapse of asset prices in Japan in the early 1990s, the ensuing reces-
sion, and the onset of the banking crisis, the role of Japanese banks in the affairs
of their client firms has been critically re-examined. Kang and Stulz (2000) find
that bank-dependent Japanese firms experience worse stock price performance
than other firms during the 1990s. Gibson (1995) provides evidence that troubled
Japanese banks inefficiently ration credit among their borrowers. Kang, Shivdasani
and Yamada (2000) report that benefits to bank monitoring of investment decisions
by client firms appear to dissipate in the early 1990s when banks themselves
became troubled. Finally, Morck and Nakamura (1999) suggest that bank inter-
vention by means of appointments to boards of directors of client firms is in the
short-term self-interest of the bank rather than the interests of shareholders of such
firms. These studies question whether monitoring by Japanese banks effectively
substitutes for mechanisms of corporate control that directly safeguard shareholder
wealth.
The purpose of this study is to examine these two conflicting views of Japanese
corporate governance mechanisms by investigating the determinants of top man-
agement change. Specifically, we investigate the extent to which measures that are
directly related to creditors’ interest, such as liquidity and leverage, explain top
management changes at Japanese firms as opposed to profitability or stock returns.
We examine data on top management changes and performance at 207 Japanese
firms from 1984 to 1989, a period that precedes the collapse of asset prices and
ensuing recession in the 1990s but which matches sample periods of earlier studies
of management incentives in Japan (Kaplan, 1994; Kang & Shivdasani, 1995).
We first confirm the findings of these earlier studies that report an inverse relation
between profitability and managerial turnover for similar sample periods. We
then focus on measures of a firm’s ability to meet its short-term obligations and
the degree to which they predict top management changes. Our findings suggest
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 3

that firms with difficulties meeting short-term obligations display a higher rate
of top management change. Furthermore, we find that liquidity measures largely
dominate unconditional measures of profitability when examined simultaneously.
Specifically, while measures of liquidity and indebtedness are correlated with
profitability, when we isolate susceptibility to creditor monitoring in particular
from poor earnings performance in general, we find that it is the former notion that
matters most in influencing the likelihood of top management changes. For exam-
ple, independent of industry-adjusted earnings, firms with low interest coverage
ratios have significantly higher rates of top management change than firms with
high interest coverage ratios. In contrast, firms with high interest coverage ratios
but with relatively poor earnings performance display rates of managerial change
similar to those of profitable firms. These results suggest that top management
change at Japanese firms is triggered specifically by endangerment of creditors’
interests rather than poor profitability in general. Conversely, managers of firms
with sufficient liquidity to appease their creditors are not disciplined for poor firm
performance. Bank monitoring, therefore, does not substitute for other governance
mechanisms that directly safeguard the interests of shareholders. Instead, our
findings are consistent with the hypothesis that the Japanese corporate governance
system lacks mechanisms that vigorously protect shareholder interests.
The remainder of this paper is organized as follows. Section 2 briefly reviews
features of the Japanese economy and their implications for corporate governance.
Section 3 discusses our data and methodology. Section 4 presents our results, and
Section 5 concludes.

2. CORPORATE GOVERNANCE OF JAPANESE FIRMS

Japanese corporations seem to lack several governance mechanisms that function
in the U.S. to mitigate problems associated with the separation of ownership
from management. First, large individual or family shareholders are rare in Japan
(Kaplan, 1994; Prowse, 1992), and large corporate shareholders, especially banks
and insurance companies, are themselves diffusely held (Nishiyama, 1984).
Second, outside directors are observed relatively infrequently; Ballon and Tomita
(1988) claim that 43.5% of large Japanese companies have no outside directors,
while Kaplan (1994) cites 60% for his sample of 119 firms. Coupled with an
absence of independent corporate boards and powerful individual shareholders
is an inactive takeover market.1 One obstacle to takeovers among large Japanese
firms is corporate grouping or keiretsu. A keiretsu is a group of affiliated
companies linked by crossholding of shares and the predominance of a common
large bank in the financial dealings of member firms. The stable cumulative
4 CHRISTOPHER W. ANDERSON ET AL.

shareholdings of keiretsu members, including the so-called main bank, are large
enough to render hostile takeovers impossible (Aoki, 1990).
These features of the Japanese corporate system have led some observers to
question whether Japanese managers face any effective control mechanisms
(Nishiyama, 1984). In contrast, many authors cite the critical role played by banks
in Japan and suggest that creditor monitoring is an effective alternative mechanism
that promotes corporate efficiency (Aoki, 1990, 1992; Sheard, 1989). Close ties
between firms and banks may help resolve information asymmetries and conflicts
of interest between firms and suppliers of capital (Horiuchi, Packer & Fukada,
1988; Hoshi, Kashyap & Scharfstein, 1990a, 1991; Prowse, 1990), and provide
an efficient means for resolving financial distress (Hoshi, Kashyap & Scharfstein,
1990b; Suzuki & Wright, 1985). Most importantly, it is argued that monitoring
by financial institutions effectively substitutes for the missing takeover market
as a mechanism to discipline top management. In the case of keiretsu firms, for
example, Sheard (1989) characterizes the main bank as exploiting its informational
assets derived from being a primary supplier of capital to the firm to serve as a
specialized monitor of managerial actions. In situations where corporate perfor-
mance is inadequate, the bank may intervene by stipulating policy adjustments,
limiting managerial discretion, lending managerial talent, and even calling for top
management changes.2 Evidence on changes in Japanese boards of directors also
suggests that banks place directors representing their interest on corporate boards in
response to poor performance (Kaplan & Minton, 1994). Studies by Kaplan (1994,
1997) and Kang and Shivdasani (1995) find an inverse relation between managerial
turnover and stock returns or accounting income.3 On the basis of this empirical
relation, Kaplan and Ramseyer (1996) conclude that the notion of Japanese
managers acting without regard for shareholders is “another fable for the academy.”
The role of creditor monitoring in Japan, as discussed by Aoki (1990, 1992),
for example, raises several questions regarding the extent to which monitoring
by banks substitutes for mechanisms that directly protect shareholders’ welfare.
First, performance benchmarks for managers may be those that are of keen interest
to creditors such as the firm’s ability to meet its interest and principal payment
schedules. Second, the link between corporate performance and top management
change may be weak for firms performing above a certain threshold if only firms in
financial distress experience creditor-induced managerial turnover. Finally, firms
with a high exposure to creditor monitoring, i.e. firms which are highly leveraged
in general and heavily indebted to banks in particular, may display the strongest
relation between performance and turnover.
Morck and Nakamura (1999) test several implications of this perspective
by examining appointments of directors to Japanese boards. They report that
appointments of directors by banks are primarily related to the short-term
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 5

interests of the bank, measured, for instance, by liquidity. Kaplan and Minton
(1994) suggest that such appointments presage executive turnover, but Morck
and Nakamura do not directly investigate management turnover. The objective
of our study is to examine the hypothesis that exposure to creditor monitoring
triggered by poor liquidity better predicts top management changes rather than
unconditional corporate performance.

3. SAMPLE AND DATA DESCRIPTION

We construct a sample of Japanese firms by cross-referencing Japanese company
listings in Moody’s International with companies listed on the University of Rhode
Island’s Pacific Basin Capital Markets Database (PACAP) for fiscal years from
1983 through 1989. This sample comprises 207 firms, although for any year the
number of firms represented is less than 207 due to data deficiencies in Moody’s.
By nature of the selection criteria for coverage by Moody’s, the sample firms are
among the largest and best known firms in Japan. The sample contains many firms
that are smaller than the 119 firms composing the sample of a similar study by
Kaplan (1994). For fiscal year 1986, the 207 sample firms have mean (median)
sales of $3.74 billion ($1.14 billion), total assets of $2.73 billion ($1.27 billion),
and market value of equity of $2.06 billion ($1.22 billion).

3.1. Top Management Change

We identify executives with the title chairman, president, or chief executive

officer as listed in Moody’s for 1983 through 1989. We classify the firm as having
experienced top management change if any previously listed top executive ceases
to be listed as a top executive in the following year.4 A mere shuffling of titles
among executives without any executive ceasing to be listed is not classified as a
management change. This classification scheme is similar to that of Warner, Watts
and Wruck (1988). The resulting sample of firm years totals 1,086 observations.
Table 1 reports management changes by year, industry, and keiretsu versus
non-keiretsu classification. Panel A reports the frequency of management changes
by fiscal year. The annual frequency is relatively constant, with only 1986 having
an unusually high number of management changes. The average annual frequency
of 13.4% is similar to the turnover frequencies reported for samples of U.S.
firms by various authors as well as the 14.5% rate cited by Kaplan (1994) and
the 12.9% reported by Kang and Shivdasani (1995) for samples of Japanese
firms. Panel B demonstrates that the sample is well distributed across industry
6 CHRISTOPHER W. ANDERSON ET AL.

Table 1. Observed Frequency of Top Management Change.a

Fiscal-year Management Annual
Observations Changes Frequency

A. By fiscal year
1984 164 19 11.6%
1985 181 20 11.0%
1986 186 36 19.4%
1987 185 25 13.5%
1988 189 26 13.8%
1989 181 20 11.0%
All years 1,086 146 13.4%
B. By industry
Textiles 63 9 14.3%
Pulp & paper 35 6 17.1%
Chemicals 168 22 13.1%
Petroleum & rubber 23 2 8.7%
Glass & ceramics 42 10 23.8%
Iron & steel 34 5 14.7%
Other metals 47 8 17.0%
Machinery 108 10 9.3%
Electrical machinery 173 25 14.5%
Transportation equipment 112 17 15.2%
Precision equipment 52 6 11.5%
Other manufacturing 72 2 2.8%
Wholesale 50 11 22.0%
Retail 79 6 7.6%
Transportation & shipping 28 7 25.0%
C. By keiretsu classification
Keiretsu (69 firms) 385 53 13.8%
Non-keiretsu (138 firms) 701 93 13.3%
a This table reports observed managerial change for 207 firms composing a sample of 1,086 fiscal years

from 1984 to 1989. Management change is defined as the delisting of any top executives (chairman,
president, or CEO) previously listed as such by Moody’s International. The number of observations
per fiscal year reported in panel A varies according to data availability for sample firms in Moody’s.
Industry classification for panel B is according to the PACAP Database for Japan. Keiretsu classification
in panel C is by Nakatani (1984).

groups, but it also suggests considerable industry variation in the incidence of

top management change. Analysis of these patterns rejects the hypothesis that
turnover likelihood is identical across all industries. All analyses were reproduced
after including industry-specific dummy variables with results similar to those
reported. For brevity, further discussion of industry patterns is omitted. Panel C
of Table 1 compares top management changes in the keiretsu sample to that of the
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 7

non-keiretsu firm sample.5 The keiretsu and non-keiretsu firm samples comprise
69 and 138 firms, with 385 and 701 firm fiscal years, respectively. The rates of
management change are similar (13.8% for keiretsu members compared to 13.3%
for non-keiretsu firms), and a chi-squared test fails to reject the hypothesis that
they are identical.

3.2. Firm Performance

We compute performance measures for sample firms using the PACAP Database
for Japan. First, we calculate return on assets (ROA), defined as earnings before
interest and taxes divided by total assets. We then subtract the equally-weighted
matching industry average return on assets from this measure for each observation
and derive an industry-adjusted return on assets referred to subsequently as
IROA. Weisbach (1988), Barro and Barro (1990), and Kaplan (1994) also use
performance measures based on return on assets in studies concerning top
management incentives. We also compute profit margin on sales (IPM), defined
as operating income scaled by sales, adjusted by industry averages. Scaling
earnings by sales may be more appropriate to the extent that land and financial
asset appreciation, accelerated depreciation, and other factors distort book value
of assets.
Stock returns are the performance measures of choice for a host of studies
focusing on U.S. firms (Coughlan & Schmidt, 1985; Gibbons & Murphy, 1990;
Jensen & Murphy, 1990; Warner, Watts & Wruck, 1988). We calculate a number
of alternative stock return-based performance measures, including raw returns,
multiple-period cumulative raw returns, returns net of value weighted or equally
weighted return indices, and returns net of industry average returns. Among
these, the measures with the strongest relation to management changes are raw
returns. We report results using raw returns, cumulated over the two years prior
to the observation year (RET2) and the four years prior to the observation year
(RET4).
We examine whether variables that measure exposure to creditor pressure better
explain the incidence of top management change in Japan than profitability or stock
returns. This hypothesis is motivated by descriptive characterizations of Japanese
corporate governance (e.g. Aoki, 1992) and the findings of Morck and Nakamura
(1999) with respect to director appointments.6 We utilize several alternative mea-
sures of corporate liquidity and leverage. First, we calculate the interest coverage
ratio (COV) as operating income scaled by interest expense. Hoshi, Kashyap and
Scharfstein (1990b) use such a measure as an indicator of financial distress. We also
compute a firm’s quick ratio (QUICK) as current assets minus inventory divided
8 CHRISTOPHER W. ANDERSON ET AL.

Table 2. Summary Statistics for Firm-specific Variables.a

Variable Mean Median Std. Dev.

Sales (Yen billion) 944 282 2,429

Assets (Yen billion) 534 242 802
Market Value Equity (Yen billion) 422 215 589
ROA – return on assets 0.079 0.071 0.041
IROA – industry adjusted return on assets 0.009 0.003 0.037
PM – profit margin on sales 0.057 0.050 0.053
IPM – industry adjusted profit margin 0.010 0.004 0.048
RET2 – stock return, prior two years 0.488 0.344 0.673
RET4 – stock return, prior four years 1.057 0.715 1.357
COVb – operating income to interest expense 6.287b 3.642 6.103b
QUICK – current assets minus inventory 1.311 1.039 0.876
to current liabilities
DTV – total debt to value (debt + market equity) 0.428 0.399 0.216
LTV – loans to value (debt + market equity) 0.143 0.087 0.156
a This table provides the mean, median, and standard deviation of key variables for a sample of 1,086
fiscal years from 207 firms over 1984–1989. All data is from the PACAP Database for Japan. Industry
adjusted variables are calculated by subtracting the fiscal-year industry mean value of a given variable
from the raw value. Industry classification is by the PACAP two-digit code and the entire PACAP
universe of firms is sampled to derive industry means.
b To reduce distortions by extreme outliers, interest coverage observations above the 5% tail of the

distribution have been set to the 95% level of the distribution, or about 20.0. Without truncation the
mean and standard deviation of COV are 128.3 and 1050.9, respectively.

by current liabilities. The quick ratio is a stock measure of liquidity as opposed to

a flow measure such as the interest coverage ratio. We hypothesize that less liquid
firms will experience higher rates of management change.
In addition to these measures of liquidity, we also calculate two measures of
leverage: debt to firm value (DTV), defined as book value of debt divided by total
firm value (market value of equity plus book value of debt), and loans to value
(LTV), defined as total loans outstanding divided by firm value. Firms that have
high leverage and especially high bank loan leverage are more likely to be actively
monitored by creditors (Aoki, 1992; Hodder & Tschoegl, 1992). In contrast, less
highly leveraged firms are unlikely candidates for active monitoring and direct
intervention.
Table 2 provides the means, medians, and standard deviations for these
performance variables, as well as for sales, total assets, and market value of equity.
Table 3 reports correlations among variables. As discussed above, the average
size of sample firms is relatively large, as mean (median) sales, assets, and market
value of equity are 944 (282) billion yen, 534 (242) billion yen, and 422 (215)
billion yen, respectively. Measures of earnings performance are well behaved, with
 Bank Monitoring, Firm Performance, and Top Management Turnover in Japan
Table 3. Correlation Matrix for Firm-specific Variables.
Sales Assets MVE ROA IROA PM IPM RET2 RET4 COV QUIK DTV LTV

Sales – 0.736 0.298 −0.098 −0.065 −0.197 −0.133 −0.054 −0.068 −0.120 −0.107 0.311 0.228
Assets 0.736 – 0.588 −0.153 −0.092 −0.191 −0.126 −0.032 −0.059 −0.179 −0.188 0.372 0.284
MVE 0.298 0.588 – 0.119 0.157 0.086 0.083 0.188 0.217 0.152 −0.014 −0.217 −0.191
ROA −0.098 −0.153 0.119 – 0.941 0.806 0.773 0.170 0.288 0.722 0.305 −0.529 −0.411
IROA −0.065 −0.092 0.157 0.941 – 0.778 0.820 0.135 0.255 0.693 0.360 −0.504 −0.376
PM −0.197 −0.191 0.086 0.806 0.778 – 0.955 0.208 0.318 0.593 0.376 −0.497 −0.330
IPM −0.133 −0.126 0.083 0.773 0.820 0.955 – 0.164 0.278 0.572 0.388 −0.423 −0.265
RET2 −0.054 −0.032 0.188 0.170 0.135 0.208 0.164 – 0.720 0.105 −0.081 −0.233 −0.063
RET4 −0.068 −0.059 0.217 0.288 0.255 0.318 0.278 0.720 – 0.163 −0.044 −0.319 −0.132
COV −0.120 −0.179 0.152 0.722 0.693 0.593 0.572 0.105 0.163 – 0.566 −0.686 −0.619
QUIK −0.107 −0.188 −0.014 0.305 0.360 0.376 0.388 −0.081 −0.044 0.566 – −0.515 −0.457
DTV 0.311 0.372 −0.217 −0.529 −0.504 −0.497 −0.423 −0.233 −0.319 −0.686 −0.515 – 0.842
LTV 0.228 0.284 −0.191 −0.411 −0.376 −0.330 −0.265 −0.063 −0.132 −0.619 −0.457 0.842 –

9
10 CHRISTOPHER W. ANDERSON ET AL.

industry-adjusted return on assets (IROA) and profit margin (IPM) averaging close
to zero without excessively wide dispersion. The return measures presented (RET2
and RET4) seem quite large (i.e. median return for the two prior years is 34.4%,
median return for the four prior years is 71.5%). However, given that our sample
period includes the dramatic mid-to-late 1980s rise in the Japanese stock market,
the magnitude of these measures is not surprising. Variables which proxy for
exposure to creditor monitoring are also reasonably distributed with the exception
of interest coverage (COV), which has a skewed sample distribution due to a small
number of extreme, upper tail outliers. Due to the presence of these outliers (i.e. the
5% of the observations with interest coverage ratios which exceed about 20), we set
20.0 as the upper limit for the COV variable in all reported analyses. Not surpris-
ingly, the four measures for creditor monitoring are highly correlated as reported in
Table 3.

4. DETERMINANTS OF TOP MANAGEMENT CHANGE

We utilize the sample of firm-year management changes and corresponding firm
performance measures to estimate several permutations of a logit equation of the
form:


Pr(management change)
ln = a + X B + e (1)
1 − Pr(management change)
where Pr(·) indicates probability, a is a scalar, X is a vector of performance measures
and firm characteristics, B is the corresponding vector of coefficients, and e is a
zero mean error term. We use maximum likelihood techniques to estimate Eq. (1)
(see Judge, Griffiths, Hill, Lutkepohl & Lee, 1985). Results from an ordinary least
squares estimation of a linear probability model are similar to the reported logit
results. To illustrate the economic significance of the estimated coefficients from
Eq. (1), we estimate the implied probability of top management change conditional
on certain values of the independent variables via the following transformation of
the estimated form of Eq. (1):

exp(a + X  B)
Pr(management change/X) = (2)
1 + exp(a + X  B)
To gauge the economic significance of the estimated empirical relation we fre-
quently calculate the implied change in probability of top management change
that results from a one standard deviation change (plus or minus) in a performance
measure from its mean.
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 11

4.1. Impact of Profitability and Stock Returns on Top Management Change

Table 4 presents estimated logit equations of top management change conditional

on firm performance as measured by return on assets, profit margin, and stock
returns. The first two columns of Table 4 provide estimated logit equations
that explain the likelihood of top management change as a function of industry
adjusted return on assets (IROA) and industry adjusted profit margin (IPM).
These two measures are calculated for the fiscal year immediately prior to the
observation year. Similar results are obtained when observation-year measures
are used or when the measures are averaged across these two years. The results
in these columns indicate that top management changes in Japan are significantly
related to earnings performance, as the null hypotheses that management changes
are unrelated to IROA or IPM is rejected at the 1% level. The estimated relations
appear economically significant as well; the standardized coefficient estimates
imply that a one standard deviation increase (decrease) in IROA from the sample
mean results in a decrease (increase) in management change probability of 3.5%
(4.6%) from the 12.9% rate predicted for a firm with mean IROA performance.7
The analogous probabilities associated with a one standard deviation change in
IPM from the mean are 4.0% (5.5%) from a base level of 12.8% for a firm with
mean IPM.
In order to gauge the fit of the estimated logit models and the economic signifi-
cance of the observed relation between performance and top management change,
Table 5 presents actual and estimated incidence of top management changes across
performance quintiles. Panels A and B of Table 5, for example, present the actual
and predicted rates of top management change by IROA and IPM quintiles. A
comparison of actual and predicted rates of management change by performance
quintile indicates that the logit model estimations in columns (1) and (2) of Table 4
are not driven by outliers and, in fact, fit the data rather well. For IROA, the ob-
served frequency of management change increases monotonically as performance
falls. For IPM, the relation is nearly monotonic. Chi-squared tests for differences
across performance quintiles confirm these inferences. For both IROA and IPM,
the incidence of top management change is significantly different: (i) between the
best performing quintile and the worst performing quintile; (ii) between firms with
performance above the median and firms with below median performance; and (iii)
across all performance quintiles. The observed and model-implied differences in
the rate of management changes between the best and worst performing firms are
greater for IPM than for IROA, perhaps implying that scaling earnings by annual
sales captures relative performance better than scaling by total assets, but the high
degree of correlation between the variables impedes further investigation of this
question.
 12
Table 4. Likelihood of Top Management Change Based on Profitability and Stock Returns.a
Variable (1) (2) (3) (4) (5) (6) (7) (8)

INTERCEPT −1.821 −1.830 −1.787 −1.716 −1.780 −1.743 −1.799 −1.768

[−1.906] [−1.919] [−1.867] [−1.855] [−1.907] [−1.887] [−1.921] [−1.90]
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
IROA (industry −9.612 – – – −9.341 −8.776 – –
adjusted return [−0.359] [−0.347] [−0.324]
on assets) (0.0004) (0.0007) (0.0019)
IPM (industry – −8.832 – – – – −8.760 −8.400
adjusted profit [−0.421] [−0.418] [−0.398]
margin) (0.0001) (0.0001) (0.0002)
RET2 (stock – – −0.164 – −0.097 – −0.071 –
return, year [−0.111] [−0.065] [−0.048]

CHRISTOPHER W. ANDERSON ET AL.

t − 2 to t − 1) (0.2544) (0.5037) (0.6303)
RET4 (stock – – – −0.131 – −0.068 – −0.049
return, year [−0.178] [−0.092] [−0.067]
t − 4 to t − 1) (0.1049) (0.3957) (0.5355)
Chi-squared 13.53 17.67 1.37 3.01 13.78 13.30 18.21 17.56
(p-value) (0.0002) (0.0000) (0.2411) (0.0826) (0.0010) (0.0013) (0.0001) (0.0002)
Observations 1,086 1,086 1,079 1,062 1,079 1,062 1,079 1,062
a This table presents estimated logit equations where the dependent variable is equal to one when a top executive (chairman, president, or CEO) as
reported by Moody’s International ceases to be listed over 1984–1989. Independent variables are: IROA – industry adjusted return on assets in the
prior fiscal year; IPM – industry adjusted profit margin on sales in the prior fiscal year; RET2 – cumulative stock return in the two years prior to the
current fiscal year; and RET4 – cumulative stock return over the four prior years. Numbers in square brackets are the coefficients on standardized
independent variables. Numbers in parentheses are p-values associated with t-tests of the coefficient estimates.
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 13

Table 5. Predicted and Observed Managerial Change Based on Profitability and

Stock Returns.a
Quintile Median Management Change

Predicted Observed

A. Management change conditional on IROA

IROA Quintile
High 0.049 9.2% 10.1%
2 0.019 11.9% 10.6%
3 0.003 13.5% 13.4%
4 −0.009 15.0% 15.2%
Low −0.026 17.2% 18.0%
Chi-squared tests of observed turnover: high quintile = low: 5.67∗∗ ; above median = below: 6.24∗∗ ; high = 2 = 3 = 4
= low: 7.96∗
B. Management change conditional on IPM
IPM Quintile
High 0.058 8.8% 6.4%
2 0.022 11.7% 13.4%
3 0.004 13.4% 12.4%
4 −0.011 15.0% 16.1%
Low −0.030 17.3% 18.9%
Chi-squared tests of observed turnover: high quintile = low: 15.91∗∗∗ ; above median = below: 6.93∗∗∗ ; high = 2 = 3
= 4 = low: 17.62∗∗∗
C. Management change conditional on RET2
RET2 Quintile
High 1.264 12.0% 11.1%
2 0.656 13.1% 11.6%
3 0.344 13.7% 16.7%
4 0.093 14.2% 15.8%
Low −0.147 14.6% 12.0%
Chi-squared tests of observed turnover: high quintile = low: 0.09; above median = below: 1.32; high = 2 = 3 = 4 =
low: 4.92
D. Management change conditional on RET4
RET4 Quintile
High 2.520 11.4% 11.7%
2 1.226 13.3% 11.8%
3 0.715 14.1% 12.2%
4 0.329 14.7% 17.0%
Low −0.040 15.3% 15.6%
Chi-squared tests of observed turnover: high quintile = low: 1.33; above median = below: 3.53∗ ; high = 2 = 3 = 4 =
low: 4.23
∗∗∗ ∗∗ ∗
, , significantly different from zero at the 1%, 5%, and 10% levels, respectively.
a Each panel displays the sample distribution of a performance measure and turnover probabilities
across performance quintiles. Performance measures are: IROA – industry adjusted return on assets in
prior fiscal year; IPM – industry adjusted profit margin on sales in prior fiscal year; RET2 – cumulative
stock return in the two prior fiscal years; RET4 – cumulative stock return over the four prior fiscal
years. Predicted turnover is obtained by using the estimated logit equations in Table 4 conditional on
quintile median performance. Tests statistics for differences in observed turnover rates are chi-squared
tests with one, one, and four degrees of freedom, respectively.
14 CHRISTOPHER W. ANDERSON ET AL.

Columns (3) and (4) in Table 4 provide estimates of the relation between the
likelihood of top management change and stock return performance. In addition
to raw returns cumulated over the two-year (RET2) and four-year (RET4) periods
preceding an observation year, we also investigated market-adjusted returns,
industry-adjusted returns, and multiple-year adjusted returns. The only results that
approach statistical significance are those for raw returns. In Table 4, for example,
the coefficient estimate on raw returns cumulated over the four-year period prior to
an observation year (RET4) borders on statistical significance ( p-value = 10.5%).
The economic significance of the impact of stock returns on top management
change is low, however. The standardized coefficients on two-year returns (−0.111
as reported in column 3 of Table 4) and four-year returns (−0.178 as reported in
column 4) are smaller than the standardized coefficients for the earnings based
variables (e.g. −0.421 for IPM as reported in column 2). This implies that a one
standard deviation change in stock return has a much smaller impact on probability
of top management change than a one standard deviation change in IROA or IPM.
Second, when we examine observed top management change across performance
quintiles based on RET2 and RET4 (panels C and D of Table 5) we find that the
differences in rates of management change across quintiles are narrow, prone to
non-monotonicities, and, with the exception of the above versus below median
performance test for RET4 in panel D, statistically insignificant. Third, when we
include both an earnings based measure of performance and a stock return based
measure of performance (columns (5)–(8) of Table 4), estimated coefficients on
stock return measures, although of the hypothesized sign, are smaller in magnitude
and are statistically indistinguishable from zero. In short, we find little evidence
in support of a relation between stock returns and top management change.
These results on stock returns seem at first to contrast with those of Kaplan
(1994) and Kang and Shivdasani (1995), who find a significant relation between top
management change and raw stock returns for different samples of Japanese firms.
However, upon closer examination our results can be reconciled. First, Kaplan’s
sample is composed of the largest 119 Japanese firms with average annual sales
of $5.6 ($3.5) billion, while our sample comprises 207 firms with mean (median)
annual sales of $3.7 ($1.1) billion. When we restrict our sample to include only the
largest 119 firms in any given year or only firms with above sample medium sales,
we find that the coefficient on RET4, which borders on conventional significance
in Table 4, becomes statistically significant at the 10% level. Second, our results
are consistent with Kaplan’s in that only earnings-based variables have robust ex-
planatory power. Specifically, in both Kaplan’s tables and ours the magnitude of the
coefficients on earnings variables is much larger than that for stock return variables.
Furthermore, Kaplan finds that stock returns are not significant in a multivariate
framework in which earnings variables are also included. This latter finding is
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 15

consistent with columns (5)–(8) of Table 4. Finally, differences in definitions of

turnover and sample construction partially explain the difference in results. Kaplan
finds a significant relation between top management change and stock returns for
a very narrow definition of top management change – instances where a Japanese
president ceases to be president yet does not assume the chairmanship, a category
that contains only about one fourth of all of Kaplan’s observations of top manage-
ment change. When Kaplan utilizes a broader definition of turnover, top executive
change is not significantly related to stock returns. Our definition of top manage-
ment change is more refined than Kaplan’s broad definition, yet not as restrictive as
the “president does not become chairman” definition. For example, our definition
does not include instances where chairman/president relinquishes the presidency to
a newcomer but keeps the chairmanship; Kaplan’s broad definition regards this as
turnover, his narrow definition does not. In short, while we can replicate Kaplan’s
finding regarding stock returns on a restricted sample of larger firms and using
a particular return measure, our conclusion is that stock returns have marginal
explanatory power when compared to earnings-based measures of performance.
Kang and Shivdasani (1995) suggest that the weak relation between returns and
top management change may be partially due to imperfections in identification
of the turnover year due to reliance on Moody’s or failure to distinguish between
routine and non-routine managerial turnover. These suggestions do not explain,
however, why we fail to find economic and statistical significance for stock returns
yet find significant results for accounting-based measures. Second, while we
consider a departing chairman succeeded by a sitting president as top management
change, we do not identify a situation where a chair/president relinquishes only the
latter title as a management change. Consequently, the routine versus non-routine
distinction made by Kang and Shivdasani is somewhat, but not fully, reflected in
our definition of management change. Finally, Kang and Shivdasani themselves
do not report multivariate results. We suspect that such analysis would confirm
our finding that the role of earnings-based measures dwarfs that of stock returns.
Finally, the weak relation between stock returns and management change could
be peculiar to our sample period. Japanese security prices increased rapidly without
parallel growth in earnings during the late 1980s. Consequently, stock returns were
imprecise indicators of managerial performance during this period. This charac-
terization of stock return performance as a relatively noisy indicator of managerial
effort and decision-making quality may explain the important role seemingly
assumed by accounting measures such as industry adjusted return on assets or
profit margin. This explanation is consistent with Lambert and Larcker (1987),
who find that the relative importance of stock returns versus accounting measures
in evaluating and rewarding managerial performance in the U.S. is positively
related to the degree to which such measures are likely to be informative about
16 CHRISTOPHER W. ANDERSON ET AL.

managerial actions. In a similar vein, Jarrell and Dorkey (1992) find that accounting
measures of performance tend to be highly correlated with long-run stock return
performance. They argue for the use of such measures in managerial incentives
because they are less likely to be affected by macroeconomic “noise” such as that
coincident with the extraordinary run-up of Japanese stock prices in the late 1980s.

4.2. Top Management Change and Exposure to Creditor Monitoring

The prior section finds a relation between firm performance, especially when mea-
sured by accounting income, and managerial change. This finding is comparable to
those of Kaplan (1994) and Kang and Shivdasani (1995). As discussed in Section 2,
bank intervention during periods of poor performance may be a unique and
effective governance mechanism in the Japanese economy. Since active creditor
monitoring and intervention are most likely to occur when firms are highly in-
debted or experiencing severe liquidity problems, we next estimate the likelihood
of top management change as a function of alternative measures of liquidity
and leverage.
Table 6 presents estimated logit equations of top management change con-
ditional on variables that proxy for exposure to creditor monitoring: interest
coverage (COV), quick ratio (QUICK), debt to value (DTV), and loans to value
(LTV). Each coefficient has the hypothesized sign and is statistically significant
in the univariate equation estimates reported in columns (1)–(4) of Table 6.
The standardized coefficient estimates are all of a magnitude comparable to
those for accounting performance variables in Table 4. For example, a one
standard deviation increase (decrease) in COV results in a decreased (increased)
probability of top management change of 4.2% (5.8%) from the 12.7% rate of
top management change for firms with mean coverage ratios.8 The multivariate
equation estimates provided in columns (5) and (6) suggest that the interest
coverage ratio (COV) is the most important of these variables, as the debt to value
and loan to value coefficients are insignificant when COV enters the equation.
When either debt to value (DTV) or loans to value (LTV) enter the estimated
equations, QUICK is rendered insignificant (columns (7) and (8)).
Table 7 reports predicted and observed frequency of top management change
across sample quintiles based on liquidity and leverage. The actual incidence of top
management change is significantly related to each of the four variables examined.
Interest coverage (COV – panel A) is a particularly interesting variable. Firms
in the highest coverage quintile have annual management change of just 6.9%,
while firms in the middle three quintiles have turnover rates that are close to
the unconditional mean, and firms in the lowest coverage display an annual top
 Bank Monitoring, Firm Performance, and Top Management Turnover in Japan
Table 6. Likelihood of Top Management Change Based on Liquidity and Leverage.a
Variable (1) (2) (3) (4) (5) (6) (7) (8)

INTERCEPT –1.471 –1.427 –2.516 –2.117 –1.786 –1.552 –2.191 –1.769

[–1.928] [–1.893] [–1.898] [–1.886] [–1.928] [–1.926] [–1.915] [–1.898]
(0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000) (0.0000)
COV (operating –0.073 – – – –0.058 –0.067 – –
income over [–0.444] [−0.356] [–0.407]
interest) (0.0001) (0.0141) (0.0035)
QUICK (current – –0.355 – – – – –0.162 –0.223
assets – inv. to [–0.311] [–0.142] [–0.196]
current (0.0105) (0.2809) (0.1309)
liabilities)
DTV (total book – – 1.444 – 0.524 – 1.165 –
debt to firm [0.312] [0.113] [0.252]
value) (0.0003) (0.3319) (0.0140)
LTV (total loans – – – 1.615 – 0.310 – 1.149
to value) [0.251] [0.048] [0.179]
(0.0022) (0.6493) (0.0588)
Chi-squared 18.27 8.08 12.78 8.95 19.21 18.48 14.05 11.55
( p-value) (0.0000) (0.0045) (0.0003) (0.0028) (0.0001) (0.0001) (0.0009) (0.0031)
Observations 1,086 1,086 1,086 1,086 1,086 1,086 1,086 1,086
a This table presents estimated logit equations where the dependent variable is equal to one when a top executive (chairman, president, or CEO) listed
by Moody’s International ceases to be listed over 1984–1989. Independent variables are: COV – ratio of operating income divided by interest expense;
DTV – total book value of debt divided by sum of book debt and market value of equity; LTV – book value of loans divided by sum of book debt and
market value of equity; QUICK – total current assets minus inventory divided by total current liabilities. To improve model fit and reduce the influence
of extreme outliers, COV observations above the upper 5% tail have been set to the 95% level of the distribution. Numbers in square brackets are the
coefficients on standardized independent variables. Numbers in parentheses are p-values associated with t-tests of the coefficient estimates.

17
18 CHRISTOPHER W. ANDERSON ET AL.

Table 7. Predicted and Observed Managerial Change Based on Liquidity

and Leverage.a
Quintile Median Management Change

Predicted Observed

A. Management change conditional on COV

COV Quintile
High 18.770 5.6% 6.9%
2 7.100 12.1% 12.4%
3 3.640 15.0% 14.4%
4 2.040 16.5% 14.7%
Low 1.240 17.4% 18.9%
Chi-squared tests of observed turnover: high quintile = low: 14.47∗∗∗ ; above median = below: 12.80∗∗∗ ; high = 2 =
3 = 4 = low: 15.21∗∗∗
B. Management change conditional on QUICK
QUICK Quintile
High 2.283 9.6% 9.2%
2 1.363 12.9% 7.8%
3 1.039 14.2% 17.5%
4 0.835 15.1% 16.1%
Low 0.634 16.1% 16.6%
Chi-squared tests of observed turnover: high quintile = low: 4.62∗∗ ; above median = below: 12.80∗∗∗ ; high = 2 = 3 =
4 = low: 16.41∗∗∗
C. Management change conditional on DTV
DTV Quintile
High 0.166 9.3% 7.4%
2 0.286 10.9% 12.0%
3 0.399 12.6% 13.8%
4 0.534 14.9% 14.3%
Low 0.747 19.2% 19.7%
Chi-squared tests of observed turnover: high quintile = low: 14.63∗∗∗ ; above median = below: 5.37∗∗ ; low = 2 = 3 =
4 = high: 15.2∗∗∗
D. Management change conditional on LTV
LTV Quintile
High 0.000 10.8% 10.6%
2 0.028 11.2% 9.7%
3 0.087 12.2% 12.9%
4 0.181 13.9% 15.2%
Low 0.398 18.6% 18.8%
Chi-squared tests of observed turnover: high quintile = low: 5.91∗∗ ; above median = below: 4.58∗∗ ; low = 2 = 3 =
4 = high: 10.00∗∗
∗∗∗ ∗∗ ∗
, , significantly different from zero at the 1%, 5%, and 10% levels, respectively.
a Each panel displays the sample distribution of a performance measure and turnover probabilities
across quintiles for measures of liquidity and indebtedness. Measures are: COV – ratio of operating
income divided by interest expense; DTV – total book value of debt divided by sum of book debt and
market value of equity; LTV – book value of loans divided by sum of book debt and market value
of equity; QUICK – total current assets minus inventory divided by total current liabilities. Predicted
turnover is obtained by using the estimated logit equations in Table 6 conditional on quintile median
performance. Tests statistics for differences in observed turnover rates are chi-squared tests with one,
one, and four degrees of freedom, respectively.
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 19

management change rate of 18.9%. A similar pattern across quintiles is evident

for debt to value (DTV) in panel C. The results for quick ratio (QUICK – panel
B) indicate some non-monotonicities in top management change with respect to
this variable and a relatively poor model fit in comparison to the other variables.
Loans to value (LTV – panel D) displays a similar non-monotonicity, but only for
the second quintile.
The results in Tables 6 and 7 suggest that changes in Japanese top management
are significantly related to measures of liquidity and indebtedness. This suggests
that creditor monitoring is the mechanism that disciplines top management.
However, the evidence presented thus far does not distinguish creditor monitoring
per se from a simple hypothesis of performance-related managerial turnover. For
example, while there is an empirical relation between alternative measures of
leverage and observed top management change, these measures are correlated with
earnings performance (see Table 3). The relation between interest coverage and
alternative measures of corporate earnings such as IROA and IPM is obvious, but
the other leverage measures are also highly correlated with corporate profitability
(see Table 3). Prowse (1990), for example, documents that leverage for Japanese
firms is negatively related to past profitability. Consequently, the results attributed
to earnings performance per se may actually be due to exposure to creditor
monitoring, or the reverse. Furthermore, discussions of creditor monitoring in
Japan suggest that management is subject to active monitoring and potential dis-
cipline in times of financial distress. To further investigate the creditor-monitoring
hypothesis, we segment the sample by both firm profitability (measured by IROA
and IPM) and exposure to creditor monitoring (measured by COV, QUICK, DTV
and LTV). Specifically, in Table 8 we divide the sample into quadrants based
on the median profitability and median liquidity or leverage and investigate top
management change across the resulting four quadrants. Similar analyses were
conducted using return-based measures; consistent with our findings in Tables 4
and 5, conditioning top management change on the basis of returns had no
meaningful effect.
Panels A and B of Table 8 present persuasive evidence that it is exposure to
creditor monitoring per se and not mere earnings performance that influences
top management change in Japan. Panel A breaks the sample into four quadrants
based on industry-adjusted return on assets (IROA) and interest coverage (COV).
These two measures are highly correlated (their correlation coefficient, reported in
Table 3, is 0.69), yet there are still many observations with above median IROA yet
below median interest rate coverage (n = 142). Similarly, there are 142 firm-years
with below median IROA and above median coverage. For firms with above
median coverage (COV = high) there is no significant difference in the incidence
of top management change for firms with above median IROA (top management
change of 38/401 = 9.5%) and firms with below median IROA (top management
20 CHRISTOPHER W. ANDERSON ET AL.

Table 8. Management Change Based on Exposure to Creditor Monitoring

Versus Profitability.a
High Low Chi-squared

A. IROA × COV Quadrants

COV
IROA
High 9.5% (38/401) 14.8% (21/142) 2.89 {0.0894}
Low 10.6% (15/142) 18.0% (72/401) 4.58 {0.0323}
Chi-squared 0.14 {0.7097} 0.76 {0.3835}
B. IPM × COV Quadrants
COV
IPM
High 8.9% (32/361) 14.3% (26/182) 3.59 {0.0583}
Low 11.5% (21/182) 18.6% (67/361) 4.61 {0.0319}
Chi-squared 0.96 {0.3273} 1.60 {0.2065}
C. IROA × QUICK Quadrants
QUICK
IROA
High 7.7% (27/351) 16.7% (32/192) 9.86 {0.0017}
Low 13.5% (26/192) 17.4% (61/351) 1.39 {0.2388}
Chi-squared 4.64 {0.0312} 0.04 {0.8329}
D. IPM × QUICK Quadrants
QUICK
IPM
High 7.6% (25/331) 15.6% (33/212) 8.46 {0.0036}
Low 13.2% (28/212) 18.1% (60/331) 2.35 {0.1250}
Chi-squared 4.58 {0.0324} 0.60 {0.4375}
E. IROA × DTV Quadrants
DTV
IROA
High 14.0% (24/171) 9.4% (35/372) 2.49 {0.1145}
Low 16.7% (62/372) 14.6% (25/171) 0.37 {0.5432}
Chi-squared 0.62 {0.4312} 3.10 {0.0781}
F. IPM × DTV Quadrants
DTV
IPM
High 13.2% (24/182) 9.4% (34/361) 1.75{0.1859}
Low 17.2% (62/361) 14.3% (26/182) 0.76 {0.3844}
Chi-squared 1.48 {0.2237} 2.82 {0.0932}
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 21

Table 8. (Continued )
High Low Chi-squared

G. IROA × LTV Quadrants

LTV
IROA
High 11.9% (21/176) 10.4% (38/367) 0.30 {0.5830}
Low 17.4% (64/367) 13.1% (23/176) 1.74 {0.1873}
Chi-squared 2.84 {0.0918} 0.86 {0.3593}
H. IPM × LTV Quadrants
LTV
IPM
High 13.7% (29/213) 9.4% (31/330) 1.44 {0.2304}
Low 17.6% (58/330) 14.1% (30/213) 1.18 {0.2776}
Chi-squared 2.41 {0.1208} 2.80 {0.0942}
a The table provides incidence of top management change based on measures of profitability, liquidity,

and leverage. Each panel splits the sample of 1,086 observations into four quadrants based on a measure
of profitability, industry-adjusted return on assets (IROA) or industry-adjusted profit margin (IPM), and
then by measures of liquidity or indebtedness, i.e. interest coverage ratio (COV), quick ratio (QUICK),
debt to value ratio (DTV), or loans to value ratio (LTV). The notation “high” and “low” denotes above
median and below median, respectively. The rate of top management change per quadrant, the number
of instances of and sample size per quadrant (in parentheses), and chi-squared tests for equality of
turnover frequency across selected quadrants are provided. p-Values for the chi-squared statistics are
presented in scrolled brackets.

change of 15/142 = 10.6%). Similarly, for firms with below median coverage
ratios, there is no significant difference in top management change with respect
to IROA classification (above median rate of 21/142 = 14.8%, below median
rate of 18.0%, chi-squared statistic insignificant at 0.76). However, firms with
above median IROA but below median coverage have a significantly higher rate
of managerial change than firms with high IROA and high coverage (14.8% vs.
9.5%, chi-squared of 2.89). Similarly, firms with low coverage and low IROA have
significantly more management changes than firms with high coverage and low
IROA (18.0% vs. 10.6%, chi-squared of 4.58). The results using industry-adjusted
profit margin (IPM) to split the sample (panel B) parallel those of panel A. After
controlling for coverage there is no significant difference in management change
by IPM, yet after controlling for IPM there are significant differences by interest
coverage. These results suggest that top management change in Japanese firms
is related to exposure to creditor monitoring, measured by interest coverage, and
not to unconditional earnings performance. These findings buttress results from
Kang and Shivdasani (1995), who report that negative income (and presumably
negative cash flow and interest coverage) dramatically raises the odds of turnover,
22 CHRISTOPHER W. ANDERSON ET AL.

not merely below average performance.9 Panels C and D of Table 8, which

present results based on segmenting the sample by quick ratio and, alternatively,
IROA and IPM, provide further evidence in support of creditor monitoring. First,
firms with high levels of asset liquidity, i.e. above median quick ratios, display
significantly more management changes when profitability is low (13.5% vs.
7.7%, chi-square of 4.64, in panel C; 13.2% vs. 7.6%, chi-square of 4.58, in panel
D). However, the differences are even larger when high profit margin liquid firms
are compared to high profit illiquid (i.e. low quick ratio) firms (7.7% vs. 16.7% in
panel C; 7.6% vs. 15.6% in panel D). This suggests that a firm’s ability to meet its
short-term obligations is at least as important as relative profitability in influencing
top management change. Not surprisingly, the worst possible quadrant, the one
associated with low quick ratio and low profitability, displays the highest rate
of top management change (17.4% in panel C; 18.1% in panel D), but for firms
with poor profitability turnover does not significantly increase when QUICK
is below median.
Panels E–H of Table 8 suggest that the relation between profitability and top
management change is not conditional on aggregate leverage or bank leverage.
In all four panels, high performance, low leverage firms display the lowest
rates of managerial change. However, management changes increase when
performance declines for both high and low leverage samples, although these
differences are not always significant. These results do not suggest that high
leverage firms have a stronger turnover-performance relation than less leveraged
firms. We also estimate logit equations with the leverage measures, performance
measures, and interaction terms as independent variables. Consistent with
the breakdown in Table 8, we find that estimated coefficients on the leverage
measures and performance measures are statistically significant, but that the
coefficients on the interactions of these variables are not. For brevity we omit
these results.
In general, the results reported in Tables 6–8 indicate that liquidity variables
significantly affect the level of top management change at Japanese firms.
In particular, low interest coverage ratios seem more important in explaining
top management change than return of assets or profit margin. Specifically,
firms with below median coverage ratios have significantly higher rates of top
management change regardless of industry-adjusted profitability. A similar result
holds when we measure liquidity with the firm’s quick ratio. On the other hand,
top management change is not significantly higher for firms with below median
profitability but relatively high coverage ratios compared to firms with above
median profitability. These results suggest that management change occurs
disproportionately more often when the interests of creditors are threatened and
not when firms perform poorly yet have sufficient cash to appease creditors.
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 23

4.3. Additional Results on Keiretsu Firms Versus Non-keiretsu Firms

The role of Japanese banks in monitoring client firms and intervening in times of
financial distress is often hypothesized to be especially pronounced for members of
keiretsu corporate groups. The long-standing ties between lenders and borrowers,
extraordinary bank access to information on the borrower, and implicit control
rights afforded the so-called main bank under extraordinary circumstances suggest
that creditor monitoring is more likely to influence top management change for
keiretsu member firms. To test this implication we conduct a logit estimation
of top management change with a keiretsu dummy variable intercept and an
interactive dummy on measures of performance, liquidity, and indebtedness. A
negative coefficient estimate for the interactive term would imply that keiretsu
member firms display top management change that is more sensitive to firm
performance. The keiretsu specific intercept term is not significantly different
from zero in all specifications, confirming the prima facie evidence in Table 1 that
the incidence of managerial change is identical for keiretsu and non-keiretsu firms.
The coefficients on interactive dummy variables are also insignificant for all spec-
ifications, suggesting that the turnover-performance relations documented earlier
do not vary by keiretsu affiliation. The strongest statement we can make about
differential sensitivity of management changes to performance concerns four-year
cumulative stock returns (RET4): the relation between top management change
and stock returns is significantly negative for keiretsu firms and insignificantly
negative for non-member firms, but this difference itself is not significant. These
findings are similar to that reported by Morck and Nakamura (1999) with respect
to bank appointments to boards of directors. Additional specifications, involving
alternative variables and interactions of variables, also failed to distinguish
keiretsu from non-member firms. Consequently, we find no evidence that the
relation between top management change and Japanese firm performance differs
by keiretsu affiliation. For brevity, we omit full reporting of these results.

5. SUMMARY AND CONCLUSIONS

Recent empirical findings of an inverse relation between firm performance and
managerial turnover at Japanese firms are construed as evidence that bank moni-
toring substitutes for other corporate governance mechanisms (Kaplan, 1994; Kang
& Shivdasani, 1995). In contrast, Morck and Nakamura (1999) report that banks act
to protect their narrow self-interest as creditors when appointing corporate direc-
tors rather than the interests of shareholders. We examine these conflicting views
by investigating the extent to which top management changes at Japanese firms are
24 CHRISTOPHER W. ANDERSON ET AL.

explained by measures of liquidity and leverage that are of direct concern to credi-
tors as opposed to measures of interest to shareholders such as profitability or stock
returns. Specifically, we examine the relation between top management change and
profitability, stock-price performance, and variables that reflect a firm’s ability to
meet its short-term obligations for a sample of 207 Japanese firms in the 1980s.
First, we find that top management changes are significantly related to account-
ing measures of firm performance but are less sensitive to stock-price performance
measures. These results are consistent with evidence on the performance-turnover
relation as reported by Kaplan (1994) and Kang and Shivdasani (1995). Next, we
find that top management changes are significantly related to measures of liquidity
and leverage, and these measures largely dominate profitability per se. For instance,
we investigate management changes for firms with high (low) industry adjusted
rates of profitability but differential exposure to potential creditor monitoring as
measured by liquidity and leverage variables. Firms with high industry-adjusted
profitability but with low liquidity experience significantly higher rates of
management change than similarly profitable yet liquid firms. Similarly, rela-
tively unprofitable yet liquid firms have fewer management changes than other
unprofitable firms. In short, liquidity measures of keen interest to creditors seem
to drive top management changes to a greater degree than earnings or stock return
performance. Consistent with the evidence on appointments to corporate boards
reported by Morck and Nakamura (1999), our findings counter prior inferences
that a performance-turnover relation is evidence of effective corporate governance
by means of bank monitoring (Kaplan, 1994; Kang & Shivdasani, 1995; Kaplan
& Ramseyer, 1996). Instead, Japanese corporate governance mechanisms appear
skewed to protect creditors’ interests rather than those of shareholders. Specifically,
our results suggest that monitoring by creditors is driven by their own short-term
interests and is therefore an imperfect substitute for other mechanisms of corporate
control that directly protect the interests of shareholders of Japanese firms.

NOTES
1. Kester (1991) describes the Japanese takeover market as confined to only small,
privately negotiated deals, with the purchase price for publicly announced mergers and
acquisitions between 1982 and 1987 averaging less than the equivalent of $4 million.
2. Pascale and Rohlen (1983) illustrate how such main bank intervention facilitated the
turnaround of the Mazda Corporation in the 1970s.
3. Popular press discussion of Japanese restructuring in the face of the most recent
recession suggests that Japanese stakeholders hold managers accountable for poor
performance. See, for instance, The Wall Street Journal, July 8, 1993, “Japanese CEOs
Find Life is Getting Tougher at the Top.”
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 25

4. Lacking biographical data or the official reason for executive departure, our measure
of management change probably includes executive deaths in addition to departures and
retirements. This biases against a relation between firm performance and management
change. Kaplan (1994) indicates that the annual rate of executive death is less than 1%, or
approximately about one out of every 15 observed top management changes, so the effects
are unlikely to be significant.
5. We utilize Nakatani (1984) to classify sample firms as keiretsu or non-keiretsu firms.
Nakatani identifies a firm as a keiretsu member if one of the following conditions is met: (1)
a keiretsu’s main bank has been the largest lender to the firm for the prior three years and
the equity share held by keiretsu member firms exceeds 20%; (2) a keiretsu main bank has
been the lender of at least 40% of the firm’s debt for the prior three years; (3) the firm has
been historically identified as a member of a particular keiretsu. Nakatani also provides a
list of ostensibly independent firms. We identify only 18 sample firms (87 firm fiscal years)
as independent firms. The results do not differ significantly when we compare keiretsu firms
to these ostensibly independent firms.
6. Gilson (1989, 1990) also reports that financial distress tends to increase the likelihood
of management turnover at U.S. firms.
7. These probabilities are derived using Eq. (2) and the standardized coefficient estimates
from Table 4 column (1). The predicted rate of management change for firms with mean
IROA is calculated as exp(−1.906)/[1 + exp(−1.906)] = 12.9%. The implied probability
for a firm with a IROA one standard deviation above the mean is exp(−1.906−0.359)/[1 +
exp(−1.906 − 0.359)] = 9.4%. The implied probability for a firm with IROA one standard
deviation below the mean is exp(−1.906 + 0.359)X/[1 + exp(−1.906 + 0.359)] = 17.6%.
Analogous calculations lead to the implied probabilities of management change associated
with variation in IPM.
8. These implied probabilities are derived from Eq. (2) using the standardized coefficient
estimates from column (1) of Table 6. An example of such a calculation is presented in note 7.
9. In Kang and Shivdasani’s Table 6 turnover for negative income firm years is 13%
for firms with bank ties and 8.3% for firms without such ties compared to 1.2 and 2.9%,
respectively, for firm years with positive income.

ACKNOWLEDGMENTS
We are grateful for the comments and suggestions on previous incarnations of this
study. In particular, we thank Peter Abken, Mark Hirschey, Chuan Yang Hwang,
Jonathon Karpoff, Ken Lehn, Anil Makhija, Bob Nachtmann, David Nachman,
Tom Noe, Steve Smith, Larry Wall, Jerold Warner, and workshop participants at
the University of Pittsburgh and Georgia Tech.

REFERENCES
Aoki, M. (1990). Toward an economic model of the Japanese firm. Journal of Economic Literature,
28, 1–27.
26 CHRISTOPHER W. ANDERSON ET AL.

Aoki, M. (1992). Ex post monitoring by the main bank and corporate governance structure in Japan.
Presentation at the Center for Economic Policy Research, Stanford University (May).
Ballon, R., & Tomita, I. (1988). The financial behavior of Japanese corporations. New York, NY:
Kodansha International.
Barro, J., & Barro, R. (1990). Pay, performance and turnover of bank CEOs. Journal of Labor
Economics, 8, 448–481.
Coughlan, A., & Schmidt, R. (1985). Executive compensation, management turnover, and firm
performance: An empirical investigation. Journal of Accounting and Economics, 7, 43–66.
Gibbons, R., & Murphy, K. (1990). Relative performance evaluation for chief executive officers.
Industrial and Labor Relations Review, 43, 30s–50s.
Gibson, M. (1995). Can bank health affect investment? Evidence from Japan. Journal of Business, 68,
281–308.
Gilson, S. (1989). Management turnover and financial distress. Journal of Financial Economics, 25,
241–262.
Gilson, S. (1990). Bankruptcy, boards, banks, and blockholders: Evidence on changes in corporate
ownership and control when firms default. Journal of Financial Economics, 27, 355–387.
Hodder, J., & Tschoegl, A. (1992). Corporate finance in Japan. Policy paper, Center for Economic
Policy Research, Stanford University.
Horiuchi, A., Packer, F., & Fukada, S. (1988). What role has the ‘main bank’ played in Japan? Journal
of the Japanese and International Economies, 2, 159–180.
Hoshi, T., Kashyap, A., & Scharfstein, D. (1990a). Bank monitoring and investment: Evidence from the
changing structure of Japanese corporate banking relationships. In: R. G. Hubbard (Ed.), Asym-
metric Information, Corporate Finance, and Investment. Chicago, IL: University of Chicago
Press.
Hoshi, T., Kashyap, A., & Scharfstein, D. (1990b). The role of banks in reducing financial distress in
Japan. Journal of Financial Economics, 27, 67–88.
Hoshi, T., Kashyap, A., & Scharfstein, D. (1991). Corporate structure, liquidity, and investment:
evidence from Japanese industrial groups. Quarterly Journal of Economics, 106, 33–60.
Jarrell, G., & Dorkey, F. (1992). The longer-term relation between accounting performance and stock
returns. University of Rochester. Working paper.
Jensen, M., & Murphy, K. (1990). Performance pay and top-management incentives. Journal of Political
Economy, 98, 225–264.
Judge, G., Griffiths, W., Hill, R., Lutkepohl, H., & Lee, T. (1985). The theory and practice of econo-
metrics. London: Cambridge University Press.
Kang, J., & Shivdasani, A. (1995). Firm performance, corporate governance, and top executive turnover
in Japan. Journal of Financial Economics, 38, 29–58.
Kang, J., Shivdasani, A., & Yamada, T. (2000). The effect of bank relations on investment decisions:
An investigation of Japanese takeover bids. Journal of Finance, 55, 2197–2218.
Kang, J., & Stulz, R. (2000). Do banking shocks affect borrowing firm performance? An analysis of
the Japanese experience. Journal of Business, 71, 1–23.
Kaplan, S. (1994).Top executive rewards and firm performance: A comparison of Japan and the U.S.
Journal of Political Economy, 102, 510–546.
Kaplan, S. (1997). Corporate governance and corporate performance: A comparison of Germany, Japan,
and the U.S. In: D. Chew (Ed.), Studies in International Corporate Finance and Governance
(pp. 251–258). Oxford: Oxford University Press.
Kaplan, S., & Minton, B. (1994). Appointments of outsiders to Japanese boards: Determinants and
implications for managers. Journal of Financial Economics, 36, 225–258.
Bank Monitoring, Firm Performance, and Top Management Turnover in Japan 27

Kaplan, S., & Ramseyer, M. (1996). Those Japanese firms with their disdain for shareholders: Another
fable for the academy. Washington University Law Quarterly, 74(2), 403–418.
Kester, W. (1991). Japanese takeovers: The global contest for corporate control. Boston, MA: Harvard
Business School Press.
Lambert, R., & Larcker, D. (1987). An analysis of the use of accounting and market measures of
performance in executive compensation contracts. Journal of Accounting Research, 25(s),
85–125.
Morck, R., & Nakamura, M. (1999). Banks and corporate control in Japan. Journal of Finance, 54,
319–339.
Nakatani, I. (1984). The economic role of financial corporate grouping. In: M. Aoki (Ed.), The Economic
Analysis of the Japanese Firm. New York, NY: Elsevier Science Publishers.
Nishiyama, T. (1984). The structure of managerial control: Who owns and controls Japanese businesses?
In: K. Sato and Y. Hoshino (Eds), The Anatomy of Japanese Business. New York, NY: M. E.
Sharpe.
Pascale, R., & Rohlen, T. (1983). The Mazda turnaround. Journal of Japanese Studies, 9, 219–263.
Prowse, S. (1990). Institutional investment patterns and corporate financial behavior in the United
States and Japan. Journal of Financial Economics, 27, 43–66.
Prowse, S. (1992). The structure of corporate ownership in Japan. Journal of Finance, 47, 1121–1140.
Sheard, P. (1989). The main bank system and corporate monitoring and control in Japan. Journal of
Economic Behavior and Organization, 11, 399–422.
Suzuki, S., & Wright, R. (1985). Financial structure and bankruptcy risk in Japanese companies. Journal
of International Business Studies, 9, 97–110.
Warner, J., Watts, R., & Wruck, K. (1988). Stock prices and top management changes. Journal of
Financial Economics, 20, 461–492.
Weisbach, M. (1988). Outside directors and CEO turnover. Journal of Financial Economics, 20,
431–460.
CORPORATE GOVERNANCE IN
SINGAPORE: THE IMPACT OF
DIRECTORS’ EQUITY OWNERSHIP

Gurmeet S. Bhabra, Stephen P. Ferris, Nilanjan Sen

and Peng Peck Yen

ABSTRACT
We examine whether the curvilinear relationship between directors’ equity
ownership and firm performance exists in a non-Western economy such as
Singapore. We find that it does, although the inflection points are much higher
than that generally cited for U.S. firms. We then compare this relationship
across two kinds of firms that are not common to the U.S. marketplace. We
observe for founder-controlled firms that the impact of director ownership
is insignificant. We also examine government-linked corporations and in
spite of the presence of a government blockholder, find that the pattern of
alignment, entrenchment and then alignment remains operative.

1. INTRODUCTION

An important characteristic of the modern corporation is the divergence in iden-

tities between its managers and owners. Because of this separation, an agency
relationship develops between the firm’s owners and managers. In the corporate
business organization, managers who initiate and implement important decisions

Corporate Governance and Finance

Advances in Financial Economics, Volume 8, 29–46
Copyright © 2003 by Elsevier Science Ltd.
All rights of reproduction in any form reserved
ISSN: 1569-3732/PII: S1569373203080022
29
30 GURMEET S. BHABRA ET AL.

are not the major residual claimants. Therefore, they do not bear a major share of
the wealth effects of their decisions. Hence, Jensen and Meckling (1976) observe
that managers are likely to take actions that deviate from the wealth-maximizing
interests of their shareholders.
Among the many mechanisms designed to alleviate this agency conflict has
been the granting of equity ownership to the firm’s managers and directors. Jensen
and Meckling (1976) develop a formal model that establishes the relation between
corporate value and managerial equity ownership. Morck, Shleifer and Vishny
(1988), McConnell and Servaes (1990) and Short and Keasey (1998) show that the
relationship between managerial equity ownership and firm value is non-linear.
This non-linearity is the result of two conflicting influences resulting from equity
ownership: the convergence of interest and managerial entrenchment effects.
The convergence of interest effect (Denis, Denis & Sarin, 1997; Jensen &
Meckling, 1976; McConnell & Servaes, 1990; Morck, Shleifer & Vishny, 1988)
recognizes that large equity holdings by corporate insiders such as managers or
directors can be associated with a high market valuation for the firm. This is
due to the incentive effect of such ownership. As managers hold more equity,
their personal wealth is more directly tied to the performance of the firm’s stock.
Consequently, managers are more likely to make those decisions that enhance
shareholder value and are less likely to dissipate corporate value by shirking or
perquisite consumption.
The entrenchment effect (Demsetz, 1983; Fama & Jensen, 1983) occurs when
managers who hold a large portion of the firm’s equity possess sufficient voting
power to ensure that their internal positions are secure. Consequently, they
become insulated from external disciplining forces such as hostile takeovers or the
managerial labor market. Such managers have adequate voting power to guarantee
their continued employment with the firm and to maintain high levels of com-
pensation. It is also likely that these managers will spend increased time shirking
and perquisite consumption rather than engaged in value-maximizing activities.
Stulz (1988) develops a model consistent with managerial entrenchment whereby
high ownership by managers effectively precludes the possibility of an unfriendly
takeover. Consistent with this, Weston (1979) finds that firms reporting managerial
ownership in excess of 30% have never been acquired in a hostile takeover.
Morck, Shleifer and Vishny (1988) examine the relationship between managerial
ownership and firm value for a set of Fortune 500 firms. They find that firm value
increases as managerial ownership rises between 0% and 5%. Beyond the 5%
ownership level, increases in managerial ownership appears to be associated with
conditions conductive to managerial entrenchment. However, as board ownership
rises beyond the 25% level, firm value also increases in relation to rising managerial
equity ownership.
Corporate Governance in Singapore 31

McConnell and Servaes (1990) also find that the relation between managerial
equity ownership and firm value is curvilinear. They report that Tobin’s q, their
proxy for firm value, initially increases, then decreases as share ownership becomes
concentrated among corporate managers and board directors. This relationship
attains its maximum prior to a 50% level of managerial ownership.
Short and Keasey (1998) examine the international nature of the relationship
between firm value and managerial ownership by studying a set of U.K. firms.
They likewise find evidence that management moves from alignment, to entrench-
ment, and back to alignment as its ownership in the firm increases. They also
report that U.K. managers require higher levels of equity ownership than their
U.S. counterparts before they become entrenched.
The findings of Short and Keasey (1998) demonstrate that national culture and
governance systems influence the threshold at which managerial entrenchment
occurs. Hence, it suggests that an international examination of the relationship
between firm value and managerial ownership would provide new insights. In this
study we analyze the effect that equity ownership by the board of directors has
on firm value for a set of Singaporean corporations. Recently, the government of
Singapore has actively promoted the country as a gateway into the Asian economy
for U.S., European, and other international firms.1 Since it is at the crossroads of
capital inflows from so many different investors, we believe that Singapore is an
interesting market in which to study the relationship between insider ownership
and firm value.
Additionally, Singaporean firms demonstrate a diversity in ownership structure
that is lacking in many Western economies. In addition to the diffusely-held cor-
poration, the Singapore economy contains a number of founder-controlled firms
and government-private hybrids referred to as government-linked corporations.
Our analysis of the relationship between firm value and insider ownership is made
richer by the presence of these other kinds of firms in our sample.
Because Singapore differs from the West in terms of regulation and culture,
entrenchment might or might not occur. If entrenchment is found in Singapore,
then it might occur at a threshold different from that of Western firms. Harmony
and cooperation are preferred over disagreement and competition in Asian society.
There are also strict rules in Singapore to prevent raiders from accumulating large
ownership interests in a target without adequate disclosure. Under the Singapore
Companies Act, investors holding 5% or more of a firm’s outstanding shares must
disclose all equity transactions within two days of the transaction. Consequently,
the takeover market in Singapore is largely inactive. The threat of takeover as
a control mechanism is essentially non-existent. These regulatory and cultural
differences suggest that entrenchment might occur at a different level in Singapore
relative to other markets.
32 GURMEET S. BHABRA ET AL.

The contribution of this study resides in three areas. First, we examine whether
the curvilinear relationship between managerial equity ownership and firm per-
formance documented for U.S. and U.K. firms, exists in a non-Western economy
such as Singapore. Second, if such a relationship exists, does the threshold at
which managerial entrenchment occurs differ from that reported in the literature
for other nations. These results will provide insights into the international nature
of agency costs and how corporate governance mechanisms can be designed to
better overcome them. Third, we exploit the ownership diversity present in the
Singaporean economy by comparing the relationship between firm performance
and insider equity holdings across different kinds of firms. This analysis will allow
us to better understand the interplay between the equity ownership structure and
firm performance by studying firm types generally absent from Western economies
or relatively uncommon.

2. INSTITUTIONAL BACKGROUND

In this section we provide a discussion of several important aspects of Singaporean

finance and business practices that are different from that prevailing in the West.
These characteristics are important for understanding international differences
regarding the impact that director equity ownership might exert on firm perfor-
mance. Further, we discuss the three general types of business ownership that
exist for firms listing on the Singapore Exchange (SGX). The SGX represents
the 1999 merger of the Stock Exchange of Singapore (SGX) and the Singapore
International Monetary Exchange.

2.1. Differences Between Western and Singapore Business Practices

2.1.1. Institutional Equity Ownership

In Singapore, the Singapore Banking Act of 1970 precludes banks from holding
large amounts of common stock. Neither mutual nor pension funds are major
shareholders in Singaporean equities. This occurs despite the presence of a large
public pension fund in Singapore known as the Central Provident Fund (CPF).2
Nor are international equity mutual funds widely invested in Singapore, perhaps
due to the small capitalization of the SGX. Therefore, institutional investors have
only a limited role to play in the Singaporean equity market.

2.1.2. Shareholder Voting

U.S. pension funds are governed by the 1974 Employee Retirement Income
Security Act that legally obliges the funds to exercise their voting rights. Hence,
such funds are typically more active than their international counterparts.
Corporate Governance in Singapore 33

Singaporean businesses, like many Asian firms, place a significant importance on

external appearances. Hence, open confrontation is rare. Shareholders generally
exercise their power in private meetings with the firm’s management. Singaporean
shareholders typically fail to exercise their rights to vote at company general
meetings.

2.1.3. Disclosure of Equity Holdings

The U.S. has strict rules to prevent raiders from accumulating large stakes without
disclosing their holdings to the target firm. Existing U.S. law requires that an
announcement is made within 10 days following a 5% or greater acquisition.
Thereafter, any subsequent trading must be promptly disclosed.
In Singapore, the disclosure of a 5% or greater holding must be disclosed within
two days. Singapore listed companies are also subjected to a mandatory bid thresh-
old of 30% of outstanding equity. Any shareholder breaching the 30% threshold is
required to make an offer to all shareholders at the highest price paid by the offerer
during the last twelve months.

2.2. The Ownership Structure of Singaporean Firms

There are three general types of ownership structure that differentiate firms that
list on the SGX.

2.2.1. Founder
In these firms, one or a few individuals hold a substantial portion of the outstanding
equity. These individuals are the founders of the company and they provide the
daily operational management for the firm. There is little, if any, difference between
the managers and owners of these firms. The CEO is almost always the chairman
of the board of directors.

2.2.2. Government-Linked Corporations (GLC)

Shortly after gaining independence, the Singaporean government adopted an
industrialization strategy to reduce unemployment and to diversify the economy.
As part of this approach, Singapore established companies with a high level
of government equity ownership in select industries. These firms are referred
to as government-linked corporations or GLCs. During the 1980s, GLCs were
established in higher value-added industries in response to changing international
economic opportunities.
Our sample of twenty GLCs has a mean (median) level of government equity
ownership of 52.63% (60.59%). The maximum government ownership for our
sample firms is 75.5% of the outstanding equity while the minimum is 9.6%.
34 GURMEET S. BHABRA ET AL.

2.2.3. Corporate
These firms resemble the typical publicly-traded firm in the U.S. The ownership of
these firms is typically dispersed and managers provide the supervision necessary
for the firm to satisfy its operational requirements. The board chair and CEO
position are typically separate for these firms.

3. SAMPLE CONSTRUCTION AND DATA DESCRIPTION

3.1. Sample Construction

We begin construction of our sample by selecting the 100 largest and 50 smallest
firms listed on the SGX as of 2001. This represents nearly 40% of the 384 firms
listed on the SGX. We use the value of the firm’s equity market capitalization as
our measure for size. Our selection procedure is driven by a desire to construct a
sample that contains corporate, GLC, and founder firms. The largest firms consist
mainly of GLCs and corporate firms. The smallest firms are mostly founder-owned
firms and reflect their generally lower level of capitalization.
From these 150 firms, we exclude 15 financial institutions because of their
different accounting procedures for measuring income relative to firms in the
manufacturing and service sectors. We eliminate ten foreign firms from our sample
because of unavailability of their annual reports. Finally, 37 firms are dropped
because of data insufficiency or unavailability on Datastream International. We are
left with a final sample of 88 firms. We then create three sub-samples consisting
of 28 founder-controlled firms, 40 corporate firms and 20 GLC firms.
We extract the financial data for our sample firms from Datastream International.
We obtain all other information from company annual reports for the years ending
1998, 1999 and 2000.

3.2. Data Description

3.2.1. Measurement of Firm Value

Tobin’s q is used to measure the value of a firm. Consistent with Morck, Shleifer
and Vishny (1988) and McConnell and Servaes (1990), we calculate q as follows:
MVE + PS + BORROWING
q=
TA
The market value of equity (MVE) is the product of a firm’s share price and
the number of common shares outstanding. Preferred stock value (PS) is the book
value of the firm’s outstanding preferred shares while BORROWING is the value of
the firm’s short term liabilities net of its short term assets, plus the book value of the
firm’s long term debt. Total assets (TA) is the book value of the firm’s total assets.
Corporate Governance in Singapore 35

3.2.2. Measurement of Insiders’ Equity Ownership

We measure insiders’ equity ownership as the percent of the total shares outstand-
ing held by the directors of the firm. This measure of ownership is consistent with
Morck, Shleifer and Vishny (1988) and Short and Keasey (1998). Equity owner-
ship data in Singapore however is restricted to individuals who hold the position of
corporate director or hold a 5% or greater stake in the firm. The equity holdings of
senior executives are not routinely disclosed as is common in other nations. Hence,
we limit our analysis to the equity controlled by the firm’s board of directors.
We manually extract all director ownership data from the annual reports of our
sample firms.
The degree of director equity ownership in Singapore clusters at two extremes.
At one end, we observe that 41.4% of our sample (36 firms) report mean board
ownership of 53%. Not surprisingly, these firms are mostly founder-controlled
enterprises. At the other extreme, we find that GLCs constitute the majority of the
44 firms (50% of the sample) having a mean director equity ownership of 1.5% or
less. Corporate firms report a mean directorship ownership of 14.5%.

3.2.3. Measurement of the Control Variables

We include several additional variables in our analysis to control for other potential
influences on firm value. These variables are firm size, firm growth and the level
of firm debt.
Two advantages due to firm size can positively affect firm value. The first
occurs because larger firms tend to have greater access to external capital markets.
This allows them to raise the capital necessary to finance positive NPV projects.
Second, larger firm size provides economies of scale that can either generate
additional corporate profitability or serve as a barrier to entry for potential
competitors. We measure firm size (Size) as the natural logarithm of the firm’s
equity market capitalization.
Firm value will also be influenced by its growth in earnings. We anticipate that
firms with a stronger growth in sales and earnings will generate higher market
valuations. This, in turn, produces greater q values. Hence, we include a variable,
Growth, in our model to capture this possibility. We define Growth as the mean
annual change in sales over the years 1998 through 2000.
Debt, which we define as the book value of total debt divided by the firm’s
total assets, is included to control for a number of factors. First, it controls
for the possibility that the creditors exert an influence over the decisions and
operations of the firm. Stiglitz (1985), for instance, argues that control over
managerial actions is effectively exercised, not by shareholders, but by lenders,
particularly banks. Second, as suggested by Grossman and Hart (1982) and
Jensen (1986), debt might be used by management to signal that they have
bonded themselves to achieve the levels of cash flow necessary to meet the
36 GURMEET S. BHABRA ET AL.

debt repayments. Debt might therefore be used to resolve agency costs and en-
hance equity values since it reduces management’s ability to consume excessive
perquisites.

3.2.4. Comparison Across Sub-samples

As shown in Table 1, the founder sub-sample has the highest mean (median)
level of director equity ownership with 53.50% (55.67%). The GLC sub-sample
exhibits the lowest percent of director ownership with a mean (median) of 1.5%
(0.08%). The corporate sub-sample falls in between the two extremes with an
insider shareholding mean (median) of 14.90% (1.88%).

Table 1. Descriptive Statistics Across Samples.

Variable Mean Median Standard Minimum Maximum
Deviation

Panel A: Aggregate sample

Dir 24.14% 6.94% 27.19% 0% 82.64%
Size 12.22 12.77 1.99 9.26 17.54
Tobin’s q 0.74 0.56 0.64 0 3.30
Growth 8.17% 0.47% 43.06% (46.46%) 312.14%
Debt 0.25 0.24 0.19 0 0.86
Panel B: Government-linked corporations sub-sample
Dir 1.50% 0.08% 4.95% 0% 22.00%
Size 14.16 13.78 1.41 12.37 17.54
Tobin’s q 1.05 0.68 0.89 0.13 3.30
Growth 15.61% 7.48% 38.83% (34.34%) 133.05%
Debt 0.25 0.21 0.22 0 0.86
Panel C: Founder controlled sub-sample
Dir 53.5% 55.67% 15.32% 21.22% 75.75%
Size 10.49 10.38 0.95 9.26 13.37
Tobin’s q 0.63 0.56 0.43 0.08 2.29
Growth 13.58% (1.99%) 63.4% (46.46%) 312.14%
Debt 0.22 0.20 0.18 0 0.60
Panel D: Corporate sub-sample
Dir 14.9% 1.88% 22.49% 0% 82.64%
Size 12.46 13.01 1.77 9.78 15.68
Tobin’s q 0.65 0.51 0.57 0 3.02
Growth 0.65% (2.49%) 22.6% (41.85%) 85.26%
Debt 0.27 0.28 0.17 0 0.75

Note: DIR is the directors’ shareholding as a percentage of the total shares outstanding. Size is the log
of the firm’s equity market capitalization; GROWTH is the average annual sales growth from
1998 to 2000; DEBT is the total debt of the firm divided by the book value of total assets.
Corporate Governance in Singapore 37

As expected, GLCs are the largest firms as measured by market capitalization.

They are followed in size by the corporate enterprises. As anticipated, the founder-
controlled firms are the smallest companies in our sample. This sequencing reflects
the fact that GLCs enjoy government support and investment, corporate enterprises
have a longer operating history while the founder-controlled firms are usually a
first generation establishment.
The GLC sub-sample possesses the highest mean (median) Tobin’s q with a
value of 1.05 (0.68). The mean (median) q values for the founder and corporate
sub-samples are comparable in magnitudes, suggesting that these firms are equally
valued. The high q values for the GLC sub-sample might be attributable to the fact
they enjoy a financial stability resulting from government investment.
Consistent with its higher q values, the GLC sub-sample demonstrates the high-
est average annual sales growth with a mean (median) of 15.61% (7.48%). This is
followed by the founder sub-sample, with a mean (median) sales growth of 13.58%
(−1.99%). The corporate sub-sample has the lowest annual sales growth with a
mean (median) of 0.65% (−2.49%).

4. HYPOTHESIS DEVELOPMENT
4.1. Firm Value and Directors’ Equity Ownership

Shleifer and Vishny (1986) observe that after an initial period of interest alignment,
increasing managerial equity ownership results in entrenchment. As managers
increase their level of equity ownership, the threat of discharge correspondingly
decreases and managers are more capable of consuming agency perquisites. But
at still higher level of ownership however, the incentive effect re-establishes itself,
with managers now having to pay a larger share of their indulgence in non-value
maximizing activities. For our aggregate sample of Singaporean firms, we test
whether this pattern is present with our first hypothesis:

H1. There is a non-linear relation between firm value and the level of director
equity ownership in Singaporean firms.

We test this hypothesis by estimating the following model and examining the
coefficients on the director equity ownership variables:

Performance = ␣ + ␤1 DIR + ␤2 DIR2 + ␤3 DIR3

3
+ ␤i Control Variables (1)
i=1
38 GURMEET S. BHABRA ET AL.

We define the model’s variables as follows: DIR is the percentage of the outstanding
shares owned by the firm’s directors while DIR2 (DIR3 ) is the corresponding square
(cube) value. The control variables are the firm’s level of debt, its growth rate in
sales, and its equity market capitalization.
The level of managerial ownership where entrenchment occurs will be computed
by differentiating Tobin’s q with respect to directors’ equity ownership, assuming
that all other variables are constant. We then let (dq)/(d DIR) = 0 and solve for q.
To determine whether this is a maximum or minimum inflection point, the value of
(d2 q)/(d DIR2 ) is also calculated. If (d2 q)/(d DIR2 ) > 0, then the inflection point is
determined to be a maximum. Alternatively, if (d2 q)/(d DIR2 ) < 0, we determine
that the inflection point is a minimum.

4.1.1. Director Equity Ownership and Founder-Controlled Firms

In a founder-controlled firm, one or a few individuals control a substantial percent
of the firm’s outstanding equity. Given the high level of director ownership in
such firms, the costs associated with any non-maximizing behavior due to en-
trenchment will correspondingly fall more heavily on the directors. Consequently,
the relationship between firm value and director equity ownership might be
different for these firms relative to our aggregate sample. Thus, we hypothesize
that the relation between director ownership and corporate value is linear for these
firms, with the convergence of interest effect dominating any tendency towards
entrenchment:
H2. The value of a founder-controlled company is linearly related to the per-
centage of equity held by its directors.

4.1.2. Director Equity Ownership and GLC Firms

GLCs are firms owned by the investing arm of Singapore government. Director
equity ownership in such firms is generally low. As indicated in Section 2.2, the
mean level of government ownership in these firms is nearly 53%. The firm is
subject to government intervention and corporate business strategies can be influ-
enced by the government’s choice of social or political goals. These might conflict
with the firm’s desire for profit maximization, thus having a negative impact on
corporate values.
But government ownership might enhance investor confidence in the firm
since it provides monitoring comparable to that of a large bock-holder. Mikkelson
and Ruback (1985), Shleifer and Vishny (1986) and Brickley, Lease and
Smith (1988) report evidence consistent with greater monitoring by block-
holders and consequent positive influences on firm value. Thus, government
monitoring might exert a positive impact on firm value and neutralize whatever
Exploring the Variety of Random
Documents with Different Content
Neither Tony nor Mr. Holden slept a wink that night, for there was
always a chance that some vessel might come towing the Sea-Lark
back to port.
 CHAPTER XIV
CASTAWAYS

A s the wild swirl of water rushed over Jack, he clung desperately

to the handle of the sloop’s pump. The vessel staggered under
her load, but righted herself bravely.
“Are you there?” the boy spluttered, as soon as he could breathe
once more. It was too dark to see anything.
“I—I think so,” came back the mate’s voice. “I didn’t expect to be,
though. We don’t want much more of that stuff!”
The hours dragged along with leaden heels. Twice again during
the night the sloop almost foundered beneath a terrible blow, but
each time managed to right herself. It seemed to the distressed lads
almost a week of darkness must have passed before a faint blur of
light appeared in the eastern sky. When dawn began to approach
Jack had arrived at that state of physical exhaustion when further
effort was almost intolerable, but the sight of returning daylight, with
the possibilities it brought of being sighted, filled him anew with life.
Then, a little later, his eyes opened wide with blank surprise.
“Why—why, where are we?” he exclaimed.
The boys stared into the half-gloom away to the west. The roar of
surf was distinct above the rushing wind, and as the light increased it
was possible for the lads to make out a line of broken water less
than half a mile away.
“That isn’t the mainland,” George declared presently. “It’s only a
little bit of a place. It surely can’t be Lobster Island! That’s forty miles
or more away from Greenport.”
“Well, how far do you think we’ve drifted in the last twenty hours or
so?” asked the other. “I haven’t the least notion where this place is,
but I shouldn’t be surprised if we had gone as far as Lobster Island.
It’s a mercy we didn’t bump up against it during the night. We’d have
been broken to splinters in that surf.”
“I guess it is Lobster Island,” said George. “There isn’t any other
place it could be. Does anybody live on it?”
“I don’t believe so,” replied the captain. “But even so, I’m going to
dump this packet on the shore.”
“You can’t, Jack, without wrecking her.”
“Maybe, and maybe not. Anyway, it’s about the only chance I see
sticking up at present. Wait till we drift more to leeward of the island;
then I’m going to make it or bust.”
This sudden appearance of land was the most welcome sight Jack
could have imagined, but there remained a good deal of deep water
between them and it; and he was by no means certain that, in the
sloop’s badly crippled condition, she could be urged under the lee
shore. Meanwhile, as the Sea-Lark drifted, the boy made ready to
hoist the throat of the mainsail, and when the sloop was slowly going
past he hauled up part of the jib.
The sloop shipped a heavy sea during this operation, and when
the canvas bellied she was almost awash. But, on reaching the
comparatively smooth water under the lee of the little island she
became more tractable and by dint of delicate handling the boy was
able to run her ashore on a sheltered, sandy beach.
The moment the keel grounded the two lads, dripping wet though
they were—worn out with the hardest and longest spell of toil they
had ever known, and hungry as hunters—looked at each other and
laughed.
“Gee! but that was what I call a narrow squeak!” commented
George. “If you’d told me an hour ago that we should be safely
ashore by now, I should have thought you’d gone crazy.”
“It didn’t look much like it!”
“Have you cut yourself?” the mate asked, seeing something dark
drop from his chum’s hand.
 “That’s nothing,” replied Jack, dipping his hand over the side in the
water. “Just a bit of a blister that burst. Let’s look at your hands. Mine
hurt, I don’t mind admitting, now.”
George displayed the palms of his hands, which were in no better
condition than those of his friend.
Suddenly Jack sprang from his seat and, opening the door of the
companionway, dived into the cabin. A moment later he emerged
with a dry package of crackers and a bottle of water.
“You think I’ve had my mind fixed on saving the sloop, all the night,
don’t you?” he asked, proffering the package to the mate, and
stuffing a cracker into his own mouth. “But you’re wrong. I kept
remembering those crackers, but we’d both have been drowned as
sure as eggs are eggs if we’d opened that door and shipped a sea.
This stuff has been in a locker for almost a week, and I’d forgotten
about it.”
“If ever you see me turn up my nose at a cracker after this,” said
George, munching away, “I give you full permission to kick me from
one end of Greenport to another.”
“We’re not in Greenport yet,” replied Jack. “Oh, my back’s nearly
broken! I don’t think I could have gone on pumping for another hour
if my life had depended on it.”
Though their position was dismal enough, stranded as they were
on a barren beach, with their boat half full of water, the lads were
now strangely happy. The strain of their recent unnerving experience
had been greater than they realized, and now it was over, the sheer
joy of being alive, and of knowing that death was not likely to
overtake them any minute, was more than enough to compensate for
the fact that they were still far from being out of the wood. The sloop
was resting firmly enough on the sand, the receding tide leaving her
high and dry. There remained a great deal of water to be pumped out
of her bilge, but that could stay where it was for the present. No sign
of any human habitation was visible on the island, but after resting a
few minutes longer the boys went ashore and explored. They found
nothing much to reward them. The island was little more than a
barren rock, with sparse, coarse grass growing in places, and also a
few low, straggling bushes. It was less than a thousand feet long,
and only about two hundred feet across at its widest point. Possibly
no human foot had stepped ashore there for years. Still, it offered a
secure haven, and on that account the boys were thankful enough.
By eating very sparingly of their slender supply of crackers they
would at least be able to keep alive for the present.
“I don’t remember the geography of this part of the coast awfully
well,” said Jack, after they had made a cursory examination of the
place, “but if this is Lobster Island it can’t be so far off the mainland.
The wind certainly isn’t quite so strong now. I believe the worst of the
gale is over. I’m going to climb up to the top of that rock and see if I
can spot the coast.”
The rock in question was not easy to scale, as it offered no secure
foothold, but its summit was a full twenty feet above the level of the
ocean, and before long the captain of the Sea-Lark was perched
precariously on the top.
“Hooray!” he cried, shouting down to his chum, and pointing away
to the north-west. “We’re all right, George. I can see the coast
plainly. And there are two ships in sight—schooners, I think. That
must be Bristow harbor over there.”
 “We’re all right, George. I can see the coast plainly”

In his excitement he descended from his lofty perch a little too

rapidly. Some distance from the bottom he slipped and came
perilously near to breaking a bone or two as he rolled heavily to the
spot where George stood. He barked his shins, and bruised one of
his elbows, but was otherwise unhurt, and after rubbing the sore
places for a few moments almost forgot about them, in view of the
important discovery he had just made.
“If I can get that eye-bolt fixed in the top of the mast again,” he
said, “we’ll be away from here within the next few hours.”
“There’s a biggish sea running,” George cautioned, his eyes roving
the tumbling surface.
“It isn’t so bad,” replied the skipper. “Why, hang it! the wind isn’t
half as strong as it was during the night. In another six hours or so
we shall be able to slip across to Bristow in no time. You don’t want
another night of it, do you?”
“Go ahead,” said George. “I’m just crazy about sailing in lovely
weather like this. And the pump is the best part of it, too, isn’t it? It
seems years since I used a pump. Guess I must have forgotten how
to work the thing by now. If I have forgotten, Jack, I hope you’ll do
any little bit of pumping that might be necessary,” he added with a
laugh.
They had walked back to the Sea-Lark, and Jack was now
standing on her deck, surveying the damage aloft. He soon realized
that to replace the eye-bolt as it had been was a task beyond him,
but, equipping himself with a few yards of spare manila rope, he
climbed the mast and set about making temporary repairs. It had to
be a clumsy job at best, but elegance was of less importance than
strength; and before long he slipped down to the deck, convinced
that the gear would hold.
“The tide will float us again in about another eight hours,” he
declared. “If it’s safe to make a start by then, we shall have two or
three hours’ daylight to make the run across.”
Toward noon, when the ebb tide had ceased, and the water was
coming in the direction of the sloop once more, Jack fished the entire
commissariat supply out of the locker again. It consisted of exactly
five crackers and about half a pint of luke-warm water at the bottom
of the bottle. The wind had by now dropped considerably, and there
was every prospect of the lads being able to start on the journey to
Bristow as soon as the Sea-Lark floated.
“Two crackers and a piece of one for you,” said the captain,
dividing them out equally. “After we’ve eaten this we’ve got to starve
to death or eat sand. Gee! it’s funny how small a cracker is when
you’ve only got two and a half of them for dinner! If we’d only thought
to lash down that bluefish of yours!”
George, having eaten his share of the lunch, yawned. It was more
than thirty hours since he had been asleep.
“It’ll be hours before the sloop’s afloat again,” he said. “I’m going to
turn in and have a snooze.”
He went into the cabin, and stretching in his bunk just as he was,
fell asleep instantly. Jack sat on the deck, with his back against the
deck-house. He did not remember ever having been so sleepy and
tired. Presently his head nodded. He raised it with a jerk and then
lowered his chin to his chest once more, while leaden weights
seemed to be dragging his eyelids down. Just a short nap, he
reflected lazily, would make him feel much fresher. A moment later
he, too, was sound asleep, and when he awoke a puzzled
expression swept over his face. The water was lapping the side of
the sloop. It must have been that which awoke him. He had been
asleep for hours.
“Come on, George!” he shouted, leaping to his feet. “We’ll be
afloat soon.”
Rubbing his eyes, the mate emerged from the cabin.
“Why, the wind’s gone right down,” he said. “That’s fine!”
The sloop, which had been canting over on her side while ashore,
now lay almost on an even keel. Jack, armed with the boat-hook,
and George with a pole which had been picked up from the beach in
anticipation of this moment, leaned over the side when the Sea-Lark
began to rock slightly, and without much difficulty they got her afloat
once more. Up went the mainsail and jib, and away the sloop ran, in
the direction of the shore which Jack had observed from the top of
the rock. In half an hour they raised land, and not long after that their
craft was nuzzling one of the pile wharves in Bristow harbor.
“I’ve got just a dollar and fifteen cents,” said Jack, turning out his
pockets. “How much have you?”
“A dime.”
“We’ll manage all right. First of all we ought to telephone to
Greenport. You’d better speak to your mother.”
They found a telephone booth near the harbor, and presently with
Jack standing by his side, George was talking over the wire.
“Hello,” he said. “Is that you, Mother?... We’re at Bristow.... No, we
haven’t got drowned yet.... Yes, we’re all right. You might tell Mr.
Holden.... I’m jolly hungry, that’s all.... Well, it was a bit rough but you
needn’t have worried.... No, everything’s all right, really.... We’re
going to sleep on the boat here to-night, and if the weather holds
good we’ll sail back in the morning.... Yes, thanks, a dollar and a
quarter between us, and we sha’n’t need any more.... In the paper!
Our pictures! Oh, crickey!... All right, Mother. See you in the morning.
G’by.
“Oh, splash!” he exclaimed, hanging up the receiver. “That’s done
it! Jack, we’re dead and drowned and given up for lost, and in the
‘Greenport Gazette’ as corpses, and half the town’s almost in
mourning for us already.”
“Two eggs,” Jack was saying, already hurrying his chum away by
the arm; “no, three eggs, and bacon. Lots of bacon. And toast and
butter. And coffee, and—”
“Stop it! you’re making my mouth water,” protested George.
“Here’s a place to get eats.”
Two of the hungriest boys who had ever sat down in the restaurant
were soon giving their orders and appealing to the waitress to hurry;
and it took every cent of their available cash to settle the bill.
After supper they strolled around the little town for an hour or two,
and then returned to the sloop for a good night’s rest.
 “We ought to get some one to fix that gear for us properly before
we start off,” said Jack, when they had turned in. “It’s all right as it is
if we don’t strike any more bad weather, but we don’t want another
time like the last.”
“We’ll find somebody to do it,” replied George, sleepily, from his
bunk; and a few moments later the two young adventurers were lost
in slumber.
The next thing Jack knew, he was sitting bolt upright in bed.
Footsteps on the deck had awakened him.
“George!” he said.
“Eh? Eh? What’s wrong?” asked the mate, in the darkness.
Already Jack was out of his bunk.
“There’s somebody prowling about,” replied the skipper. “Listen!
He’s in the cockpit now.”
The cabin door opened, and Jack threw himself into a defensive
attitude.
A light flared up in the doorway as a match was struck, and both
boys burst into laughter.
“Well, you two!” said the familiar voice of Tony Santo.
“Dad!” George exclaimed. “How did you get here?”
“By train, of course,” replied Tony. “I just thought I’d run down and
see that you were all right.”
“It’s awfully good of you,” said George. “But I told Mother there
was nothing wrong.”
“Oh, no! Nothing at all wrong!” replied Tony, sarcastically. “What
happened?”
“Some of the gear gave way and we couldn’t use our sails,” Jack
explained.
“Well, my boys,” said Tony, “I’m glad it’s no worse. You certainly
did throw a scare into us all, but it wasn’t your fault. Go back to bed
now. There’s a hotel near here, and I’ll find a bed there. I’ll take you
ashore for breakfast, and as soon as we get her overhauled we’ll be
off for home.”
Next morning Tony strengthened the temporary repairs which Jack
had made, and, with the breeze still favorable, the Sea-Lark headed
for Greenport while most of the good folk of Bristow were still in bed.
The summer gale had vanished as quickly as it had come, and the
sloop had an easy passage back.
 CHAPTER XV
JACK LOSES COMMAND

J ack and his chum, on returning to Greenport, found themselves

overwhelmed with congratulations. They were stopped on the
street and plied with questions, and the reporter of the “Greenport
Gazette” who had, two days before, firmly believed he was writing
the boys’ obituary notices, wrung their hands warmly in the hope of
extracting a good “story.” But neither the captain nor the mate of the
Sea-Lark cared for publicity of this sort.
“You don’t want to print anything more about us in your paper,”
said Jack. “Everybody knows we got back all right, and there’s
nothing else to it.”
More than that he refused to say to the journalist, but to Cap’n
Crumbie he opened his heart, and the watchman nodded
understandingly as the boy recounted their adventures. The reporter,
knowing there were more ways of killing a dog than by drowning it,
awaited a favorable opportunity of tackling Cap’n Crumbie, and that
worthy, without the slightest hesitation, told the reporter all he
wanted to know. It was, therefore, with something of a shock that the
boys found two columns of the local paper filled with a thrilling
account of their narrow escape. Cap’n Crumbie, who at all times was
inclined to make a little go a long way when he was telling a story,
had polished up the high lights and introduced a few bright ideas of
his own; and the reporter, who was none too particular about facts
when it came to turning out exciting “copy,” had let himself go. The
combined result was a truly harrowing yarn, which made Jack and
his friend roar with laughter, but which also had the effect of swelling
the number of ferry patrons.
Mr. Farnham, who was a business man to his finger-tips, stood on
the hotel landing with his wife, watching the Sea-Lark discharging an
unusually large load, and he laughed softly.
 “Sweet are the uses of advertisement!” he misquoted with a
glance at his wife.
“You mean their adversity has proved an advertisement,” replied
Mrs. Farnham. “Yes, but I should be sorry if they had another
advertisement of the same kind.”
“Surely,” replied the man of business, “but give Jack due credit.
Lots of chaps would have taken a day or two off to rest, after going
through all that. His hands were so sore at first that he could hardly
hold the wheel. But instead of lying back and listening to
congratulations, he got on the job while the rush lasted.”
“He certainly has worked hard this summer.”
“He has,” replied Mr. Farnham, thoughtfully, “and fellows with as
much grit as that aren’t any too plenty. He ought to go a good long
way in this world. But it won’t be as a ferryman.”
“It won’t?”
“No”; and Mr. Farnham smiled. “I have a notion that by the time he
gets through High School he’ll be the sort of chap I shall find very
useful in my office in New York. But there’s time enough to think
about that.
“By the way,” he said to Jack, stepping down on to the landing as
soon as the last of the passengers had gone, “I have just got a new
dinghy in place of the one I’ve been using as a tender to the power-
boat. I have no use for the old dinghy now, so if you’d like to hitch it
up behind the Sea-Lark, you can have it as a tender.”
“Why—why,” began Jack, who knew the dinghy well enough, and
would have liked nothing better than to own her, but felt that Mr.
Farnham had given him quite enough as it was, “that would be
splendid, only—”
“Wait a minute,” put in Mr. Farnham, quickly discerning what was
in the lad’s mind. “I’m not going to make a gift of her to you, exactly.
Let’s put the thing on a proper business footing, eh?”
Jack smiled. “I’d be glad to,” he said; “But how?”
 “You can take her on condition that whenever I want to use your
ferry during the rest of my vacation this year, I’m allowed to travel
without paying my fare.”
“But you hardly ever come across,” protested Jack.
“Well, well, I’ll have to make a few special journeys to work off the
price of the dinghy. Not another word, now. She’s yours. Rod will
hand her over to you to-day.”
“Thank you ever so much,” Jack called after Mr. Farnham, who
had already turned and was walking away toward his bungalow.
The proprietor of Holden’s Ferry had but little time for gossip with
his friend the watchman for several days after his return from
Bristow, but early one morning, while Jack was preparing the sloop
for the day’s work, Cap’n Crumbie descended from the wharf and sat
on the deck-house watching the lad use the swab.
“There’s one thing I forgot to tell you,” said the watchman. “I’ve got
a notion that p’r’aps we’ve been misjudging those two fellers Hegan
and Martin.”
“Misjudging them?”
“Well, I dunno,” replied the Cap’n. “P’r’aps it was only me that did
the misjudging, but I surely did think it was either one or the other o’
them that tried to brain you with a bar o’ steel that night.”
“Well?” said Jack, curiously.
“Well, ’tain’t reasonable to think so now. If they’d wanted to do you
an injury they wouldn’t have acted like they did when we all thought
you was getting drownded out there.”
Jack put down the swab.
“How do you mean?” he asked.
“I watched ’em, watched ’em close, too, when they heard you’d
been blown out to sea,” said Cap’n Crumbie. “An’ if I ever seed a
case of genooine sorrow in a feller’s face, it was then.”
 “Really!” said Jack, a little puzzled. He still had a painfully vivid
memory of having been held down to the floor of the cabin by the
throat and almost choked to death. “You can’t always go much by
looks, though.”
“It wasn’t only their looks,” said the Cap’n, shaking his head
solemnly. “It was Hegan’s idea to start a subscription to pay Barker
for the hire of his old tug to go and save you. And Martin offered to
chip in, too. They meant it, all right. In another minute or two we’d ha’
been handing that shark Barker the thirty dollars he asked for before
he’d send the tug out. But just then Tony came along and paid
Barker out of his own pocket.”
“How funny!” said Jack, with a perplexed frown. “I’m glad you told
me. Next time they come along I must thank them for it. They were
both down on the wharf yesterday, but I was pretty busy and they
didn’t speak.”
“It just shows you,” observed the watchman, “how you can be
mistaken in folks.”
“Ye-es,” said Jack, a trifle doubtfully. Then, “Hello, here they
come,” he added.
Hegan and Martin strolled to the edge of the wharf and looked
down on the deck of the Sea-Lark.
“Good morning,” said Jack. “Cap’n Crumbie has been telling me
about your being kind enough to start a subscription for the tug when
we were blown out to sea. It was awfully kind of you.”
“Subscription?” said Hegan. “Oh, yes, I’d forgotten. That’s nothing.
Forget it! You can’t stand by and see a friend drown, can you?”
“I’m glad it didn’t cost you anything, after all,” said Jack, lightly. “I
don’t mind admitting we should have been mighty glad to see that
tug, and I’m much obliged to you, all the same.”
“Say, to-morrow’s Sunday. You don’t run the ferry Sundays, do
you?”
“No.”
 “I was just sayin’ to my friend Martin that p’r’aps we might
persuade you to take us for a sail. We’re both going back to New
York to-morrow night, and I’d like one good run in the sloop afore I
go. What d’you say?”
“Why, I’ll be glad to,” said the boy, graciously, feeling that was the
least he could do to repay them for their generous offer of
assistance. “As early as you like. What about seven o’clock?”
“Fine,” replied Hegan. “We’ll be here.”
As George had promised to visit some friends with his mother next
day, Jack arranged with Rod to accompany him on the trip in the
sloop, promising to pick him up at the hotel landing as they sailed.
The men kept their appointment punctually enough. As a matter of
fact, they arrived at the wharf immediately after Jack and George left
the vessel to go home for breakfast; and, finding the cabin door
locked, they asked Cap’n Crumbie where they could get the key.
“I guess Jack must have it,” replied the Cap’n; and he remained
there, chatting with them, until the skipper of the Sea-Lark returned.
Sailing across to the landing, they found Rod awaiting them, and
then the sloop’s bow was turned toward the sea.
“Now, which way do you want to go?” asked Jack. “The water is
dead calm.”
“How about a run down the coast as far as Penley?” Martin
suggested, glancing sideways at Hegan.
“It’s all the same to me,” replied Hegan, airily. “So long as I’m
afloat with a good cigar in my mouth, it don’t make any odds whether
we go north, south, east, or west.”
“All right,” said Jack.
Soon after they got clear of the harbor and round the end of the
breakwater, however, Hegan, for some unaccountable reason,
changed his mind.
“Let’s run up the coast, as far as Indian Head,” he said.
 “I thought you didn’t care where you were so long as you were
afloat,” replied Jack, laughing. “We mightn’t be able to get back if this
bit of a breeze dropped, because of the tide.”
“Oh, come on,” said Hegan, with rough good humor. “Let’s take a
chance. I’d like to see the coast around that way, and this wind ain’t
goin’ to drop.”
“Well, if you really want to,” agreed Jack. “But don’t blame me if
you miss your train through not getting back on time.”
“That’s all right,” said Hegan. “I want to see Indian Head from the
ocean. It’s years since I was off there. How far is it to Baymouth from
the Head?”
The question was put with such curious intentness that Jack
glanced at the man before replying.
“Thinking of swimming it?” he asked. “About a couple of miles as
the crab walks.”
“I thought it was about that,” replied Hegan; and then he strolled
forward to where Martin was leaning against the mast. The two men
talked for some time in low voices, watching the coast-line as the
sloop slid slowly past, but neither Jack nor Rodney took much notice
of them. Presently, however, Hegan turned round and shouted aft to
the captain.
“Couldn’t you keep her a bit farther out?” he asked casually. “We
don’t want to hug the shore all the way up.”
Jack waved a hand in reply, and gave a slight turn to the wheel, in
response to which the Sea-Lark headed farther east, and before long
a considerable distance separated the sloop from the shore.
“I guess we had better not go much farther,” he called out then. “It
looks kind of hazy over there.”
“Why, we can’t be so far off Indian Head now, are we?” Martin
queried. “Both of us wanted to have a look at it.”
“There it is,” replied Jack, pointing off on the port bow to a blur on
the coast which was rendered vague by the slight haze.
 “All right. You don’t mind going up as far as that, do you?”
Jack hesitated a moment. The wind was so light now that it would
barely be sufficient to carry them back over the tide, and Greenport
harbor was fully seven or eight miles off.
“This is no power-boat, you know,” he said, endeavoring to meet
the wishes of the men in good part. “And I don’t like that haze, either.
It wouldn’t surprise me a bit if there was a regular fog soon. I think
we’ll turn back.”
Hegan walked aft with his companion at his heels.
“Nothing doing!” he said in a tone which astounded the skipper.
“Keep her going just as she is till you get orders from me.”
“Orders!” Jack repeated. “If you talk like that I’ll dump you both out
on the nearest beach and leave you to get back as well as you can.”
“No you won’t,” said Hegan with an ugly expression, drawing a
small but wicked-looking revolver from his coat pocket and pointing it
at Jack.
“Put that thing down and stop your nonsense!” said Jack, furious
at such a liberty being taken. Rodney, taken aback for a moment by
the suddenness of the men’s change of front, recovered his self-
possession and quietly reached down to the mast rail for one of the
belaying-pins.
 “Put that thing down and stop your nonsense”

“Drop that!”
 The words came from Martin like the crack of a whip as he swung
around, and Rodney saw that he, too, was armed.
“Is this a joke?” Jack demanded, white to the lips. He was more
than half inclined to let go the wheel and with one quick step forward
push Hegan over the rail into the sea. But there was something
about the man’s manner that showed he meant to fire if he were not
obeyed.
“Yes, just our little joke!” Hegan replied. “All the same, you won’t
see any fun in it if you don’t do as you’re told.”
“There won’t be any fun in for you either, soon,” replied Jack,
glancing over his shoulder. “Look at this fog-bank drifting up. We’re
going to be in a nice fix.”
“Just what I want,” replied Hegan. “Now, take it calm, and p’r’aps
you won’t get hurt. I don’t know that it wouldn’t be best to thump you
both on the top of the head and drop you overboard. Nice time
you’ve given us! ain’t you?”
“Given you?”
“Never mind about that!” snapped Hegan. “The least said the
soonest mended. Here, give me that wheel, and get for’ard. Keep
’em covered, Martin. This feller looks as though he was going to try
to give us a bit o’ trouble. What d’you say? Shall we make ’em swim
for it? A two-mile swim on a day like this is good for any one.” He
laughed evilly.
“You stick to the program, Hegan,” replied Martin. “No killin’; that’s
what we agreed on.”
The edge of the fog-bank was already enveloping the sloop, and
the coast-line was now hidden from view.
“But a nice little swim—” Hegan began.
“Shut up!” Martin snarled, losing his temper.
“All right,” replied Hegan. “You always was a chicken-livered cuss,
huh? Now, Captain, oblige me and my friend by steppin’ for’ard up
Welcome to our website – the ideal destination for book lovers and
knowledge seekers. With a mission to inspire endlessly, we offer a
vast collection of books, ranging from classic literary works to
specialized publications, self-development books, and children's
literature. Each book is a new journey of discovery, expanding
knowledge and enriching the soul of the reade

Our website is not just a platform for buying books, but a bridge
connecting readers to the timeless values of culture and wisdom. With
an elegant, user-friendly interface and an intelligent search system,
we are committed to providing a quick and convenient shopping
experience. Additionally, our special promotions and home delivery
services ensure that you save time and fully enjoy the joy of reading.

Let us accompany you on the journey of exploring knowledge and

personal growth!

ebookfinal.com