[PPT] - Kotaro Inoue Columbia Business School Motivation: What is real PowerPoint Presentation

SLIDE 1

Kotaro Inoue

February 16, 2018 Japan Economic Seminar Columbia Business School

Tokyo I nstitute of Technology School of Engineering Professor

Enjoying The Quiet Life: Corporate Decision- Making By Entrenced Managers with N. Ikeda and S. Watanabe

SLIDE 2

Motivation: What is real costs of cross-shareholding?

Cross shareholding “Mochiai”

– Interlocking ownership among business counterparties (keiretsu), lender banks, and member firms of corporate group (zaibatsu)

Stable shareholding “Antei-Kabunushi”

– Long-term and stable shareholding by business counterparties and others who have close relation with the firms

Both were historically introduced to protect management

from hostile takeover attempts in Japan

– Still works as protection against activist hedge funds and against pressure from proxy advisory firms – Managers of firms with high cross shareholding should feel safe compared to those without such shield

2

SLIDE 3

Potential outcome of cross-shareholding

Japan is suffering from low corporate profitability,

low economic growth, and poor stock market performance for the latest two decades, in spite

f unprecedented monetary easing policy
Avoidance of risky investments and restructurings
f unprofitable businesses are often pointed out

as a cause of the low profitability and growth of Japanese firms

Under shield of cross-shareholding, managers

might not take optimal risk for the firm value and growth, but simply enjoy their quiet life

3

SLIDE 4

Risk taking and Tobin’s q Arikawa, Inoue, and Saito (2017DP)

4

SLIDE 5

A hypothesis of inactive management

Quiet life hypothesis

– Hicks (1935) and Hart (1983) argue that the managers

f the monopolistic enterprises which are insulated

from competition of the product market may not make adequate efforts simply because the negligence increases their utility – Bertrand and Mullainathan (2003JPE) shows the supporting evidences of this hypothesis from the US data that managers lower their effort level when they are protected from disciplinary effects from stock market – It seems possible explanation for behavior of Japanese firms in last two decades

5

SLIDE 6

An evidences of managerial quiet life

Bertrand and Mullainathan (2003)

– Analyzed effects of the State Takeover Laws on decision-making of US firms using plant-level data – Introduction of the law into state where the corporate headquarter was located, the company’s plant construction and closure decreased, wages of workers increased, and the corporate performance turned worse – Managers avoid difficult decisions when they are protected from disciplinary effects from stock market

6

SLIDE 7

Cross-shareholding

Development and recent trend of cross-shareholding

– Developed from late 1960s to serve as a deterrent to foreign takeovers (Hoshi and Kashyap 2001, p126) – In particular, banks and corporate group member firms are the center of interlocking shareholding arrangements – Cross-shareholding decreased in the period between 1997 and 2004 mainly due to selling shares by banks under banking crisis (Miyajima and Nitta, 2011) – Japanese firms still intentionally maintain the interlocking relation in the subsequent period

Many firms sustain cross-shareholding and this works as the strong

protection against shareholder power (Tanaka and Xu, 2005)

Japanese firms start to reduce cross-shareholding recently to respond

to Japanese Corporate Governance Code introduced in 2015

7

SLIDE 8

Historical ownership structure of Japanese Firms

This figure is based on the figures provided in Table 1 of Miyajima and Nitta (2011)

8

0% 5% 10% 15% 20% 25% 30% 35% 1987 1991 1996 2001 2006 2008 Cross Shareholding Stable Shareholding Institutional Investors Individual Investors

SLIDE 9

Effects of cross- and stable- shareholding

Two distinctive views to the effects

– Positive effects from long-term commitment

Main banks, Keiretsu, and business counterparties provide

effective monitoring and provide capital in flexible manner

– Hoshi, Kashyap, and Scharfstein (1990, 1991)

This view seems to be more relevant in 1970s and 80s

– Negative effects from management entrenchment

Whether they have positive or negative effects in

current Japanese corporations is an empirical question

9

SLIDE 10

This study

To test the quiet life hypothesis in Japan, this

study analyzes effects of cross- and stable shareholding on “difficult” corporate decisions

– Large capital investments, M&A, R&D, and restructurings – panel data of listed firms on TSE1 from 2004 to 2014

We find cross-shareholding allows managers to

avoid difficult decisions, and this seems to result in lower risk taking by the firms

– Institutional Investors show opposite effects – The results are consistent with quiet life hypothesis

10

SLIDE 11

Hypothesis 1

Under the quiet life hypothesis, managers tend to

decrease investments even when they have positive net present value to avoid the managerial efforts associated with large investments

H1: When a company has high cross- and stable-

shareholding ratio, the level of capital expenditure, M&A investments, and R&D expenses are low

11

SLIDE 12

Hypothesis 2

Under the quiet life hypothesis, even if the

company has an unprofitable business, the protected managers are likely to postpone the decision to restructure the unprofitable business since it requires a significant managerial efforts in Japan

H2: When a company has high cross- and stable

shareholding ratio, the probability of executing restructuring is low

12

SLIDE 13

Alternative hypotheses relating to the effects of cross shareholding

Career concern hypothesis

– Protected managers do not have to concern their own career and will increase risky investments such as R&D

Aghion, Van Reenen, and Zingales, 2013

– Predicts opposite effects on risky investments such as R&D from cross shareholding

Over investments by entrenched managers

– Entrenchment can result in overinvestment which also provides private benefit to managers

Jensen (1986), Harford (1999)

– Predicts opposite effects on investments from cross shareholding

13

SLIDE 14

Data and Sample

Data

– Governance related data: Nikkei Needs Cges – Other financial data: Nikkei Needs Financial Quest – M&A related data: Recof M&A database

Sample

– Listed firms on Tokyo Stock Exchange 1st Section – Panel data from FY2004 to 2014

24,506 firm-year

14

SLIDE 15

Independent Variables of Interest

Cross shareholding

– Nikkei data captures interlocking shareholding more than 1% of outstanding shares – It is reports that even when it includes less than 1% but within the largest 30 shareholders, the ratio increase by

nly 2%
Stable shareholding

– Ratio of stable shareholders excluding cross-shareholding and director ownership – The relation between stable shareholders and the firms are less clear due to lack of interlocking ownership relation

15

SLIDE 16

Dependent Variables

Investment behaviors

– CAPEX: Capital expenditure / Total Assets – M&A: Annual M&A investments / Total Assets – R&D: R&D expenses / Total Assets

Restructurings

– Sub dummy: dummy variable which takes one when the firm sells its subsidiaries – Reducing Seg dummy: dummy variable which takes

ne when the firm reduce its number of business

segment (test in multiple business segment subsample)

16

SLIDE 17

Descriptive Statistics

17

N Mean S.D Min Median Max CAPEX[t] Ratio of Capital expenditure to total assets 22362 0.0418 0.0438 0.0000 0.0307 0.9429 M&A[t] Ratio of M&A value to total assets 19811 0.0068 0.0464 0.0000 0.0000 1.7675 R&D[t] Ratiof of R&D expenses to total assets 14801 0.0227 0.0381 0.0000 0.0127 2.2593 Sub dummy[t] Dummy variable of subsidiary divesture 22363 0.1110 0.3141 0.0000 0.0000 1.0000 Reducing Seg dummy[t] Dummy variable of business segment reduction 13952 0.0673 0.2506 0.0000 0.0000 1.0000 Risk-taking[t] Industry and year adjusted standard deviation of ROA 18271 0.0274 0.0322 0.0015 0.0195 0.7211 Cross[t-1] Ratio of cross shareholders 21689 0.0786 0.0858 0.0000 0.0537 0.5630 Stable.exc.Cross[t-1] Ratio of stable shareholders excluding cross shareholders and directors 19593 0.2711 0.1866 0.0000 0.2267 0.9289 Inst[t-1] Institutional investor ownership 21608 0.1806 0.1572 0.0000 0.1392 0.8560 Ind[t-1] Independent director ratio 21820 0.0615 0.1148 0.0000 0.0000 0.8889 Dir[t-1] Director ownership 21697 0.0644 0.1175 0.0000 0.0082 0.9970 SO[t-1] Dummy variable of introduction of stock option 21718 0.3458 0.4756 0.0000 0.0000 1.0000 MtoB[t-1] Ratio of Market capitalization to net assets 21910 1.5066 2.1474 0.0032 1.0033 47.9082 Cash[t-1] Ratio of Cash and short-term securities to total assets 21998 0.1706 0.1428 0.0002 0.1314 0.9816 Lev[t-1] Ratio of interest-bearing debt to total assets 21998 0.2026 0.1980 0.0000 0.1611 8.6923 Assets[t-1] Total Assets (million Yen) 22008 306,864 1,196,878 52,941 35,500,000 ROA[t-1] ROA 21984 0.0546 0.1322

7.4231

0.0577 11.7428

SLIDE 18

Ownership structure in the period

We exclude FY2004 from analysis using stable

shareholding due to change of definition of the stable shareholding in the database

– It adds treasury shares and corporate block shareholding of more than 3% from FY2005

18

SLIDE 19

Regression Approach

Random effects panel regression

– The mean and median of yearly change of the cross shareholding are -0.06% and 0.00% respectively – Since there is little change over time in the cross- shareholding for most firms, the inclusion of fixed effects would force identification of the cross-shareholding coefficient from only those small changes – In the linear regressions, we cluster standard errors by firm to adjust for cross-sectional heteroskedasticity and within- firm serial correlation of error terms

As robustness check, we additionally employ both

propensity score matching analysis and 2SLS

19

SLIDE 20

Main results; Test of Hypothesis 1

20

Cross[t-1]

0.016 ***
0.022 ***
0.011 **
0.015 ***
0.005 **
0.015 ***

(-3.29) (-3.93) (-2.40) (-2.61) (-2.09) (-4.08) Stable.exc.Cross[t-1]

0.008 **
0.006
0.010 ***

(-2.44) (-1.32) (-3.17) MtoB[t-1] 0.001 *** 0.001 *** 0.001 *** 0.001 ***

0.000 *
0.001 **

(3.80) (3.88) (4.12) (3.77) (-1.83) (-2.35) Cash[t-1] 0.001 0.001 0.016 *** 0.014 ** 0.002 0.002 (0.21) (0.09) (2.65) (1.98) (0.29) (0.31) LnAssets[t-1] 0.001 * 0.001 ** 0.000 0.000

0.001
0.001

(1.91) (1.97) (0.04) (0.16) (-0.91) (-0.71) Lev[t-1]

0.028 ***
0.026 ***
0.006
0.006
0.009 **
0.009 **

(-7.25) (-6.63) (-1.62) (-1.46) (-2.46) (-2.41) ROA[t-1] 0.010 0.009

0.018 **
0.014 *
0.012
0.008

(1.43) (1.37) (-2.34) (-1.82) (-1.57) (-1.51) (Intercept) 0.040 *** 0.042 *** 0.006 0.007 0.026 *** 0.028 *** (6.74) (6.50) (0.78) (0.79) (3.25) (3.38) Industry Yes Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes Yes N 21111 19068 18727 17099 14219 12813 R sq. : within 0.045 0.045 0.008 0.008 0.018 0.024 R sq. : bitween 0.171 0.181 0.068 0.057 0.338 0.327 R sq. : overall 0.139 0.145 0.021 0.019 0.319 0.325 CAPEX M&A R&D

SLIDE 21

Main results; Test of Hypothesis 2

We employ random

effect Probit model for the divesture if subsidiary dummy and the reducing of business segment dummy

21

Cross[t-1]

0.963 ***
1.238 ***
0.995 ***
1.105 ***

(-3.65) (-4.31) (-4.21) (-4.32) Stable.exc.Cross[t-1]

0.231 *
0.202 *

(-1.93) (-1.81) MtoB[t-1] 0.003 0.003 0.022 *** 0.022 ** (0.31) (0.31) (2.67) (2.46) Cash[t-1] 0.427 ** 0.296 * 0.229 0.194 (2.55) (1.67) (1.30) (1.04) LnAssets[t-1] 0.263 *** 0.254 *** 0.027 ** 0.022 * (17.39) (16.25) (2.33) (1.76) Lev[t-1] 1.178 *** 1.086 *** 0.598 *** 0.560 *** (9.26) (8.09) (5.17) (4.55) ROA[t-1]

0.827 ***
0.901 ***

0.112 0.125 (-3.99) (-4.15) (1.19) (1.29) (Intercept)

4.224 ***
3.981 ***
1.946 ***
1.820 ***

(-22.95) (-19.85) (-12.41) (-10.25) Industry Yes Yes Yes Yes Year Yes Yes Yes Yes N 21112 19068 13845 12587 Sub dummy Reducing Seg dummy

SLIDE 22

Effects of Cross-shareholding

Firms protected by higher cross-shareholding invest less

and conduct less restructuring

– Consistent with H1 and H2 – Not consistent with career concern hypothesis and

verinvestment hypothesis under entrenchment

– Stable shareholding plays mostly similar effects, but is statistically weaker than effects of cross-shareholding

The effects of cross shareholding are common in both

manufacturers and non-manufacturers

ROA and MtoB ratio of previous year are controlled

– Above results are not likely to be caused by a possible selection bias that profitable and growing firms have previously dissolved cross- and stable shareholding – ROA and cross shareholding do not show high correlation (-0.02)

22

SLIDE 23

Economic Condition and Managerial Quiet Life

Are the negative effects of cross-shareholding on firm

investments observed in both economic expansion and depression?

We included the interaction between GDP growth and

cross-shareholding in the presented regression (Not presented in the tables)

The main effect of cross-shareholding remains

significant and negative, and the interaction term is significant and negative only for CAPEX

– The negative influence of cross-shareholding on firm investments becomes more obvious in the periods of economic expansion

23

SLIDE 24

An outcome of managerial quiet life

Risk taking

– SD of Industry-year adjusted ROA of latest 10 years – John, Litov, and Yeung(2008) shows that this variable is correlated with firm growth

Cross shareholding decreases

risk-taking

Observed lower investments
f the firms with high cross

shareholding indicate lower future growth

24

SLIDE 25

Is corporate governance effective to mitigate the problem?

Can enhanced corporate governance mitigate the

costs of managerial quiet life

– Monitoring

Monitors outside the firms: Institutional ownership (Int)
Monitors inside the firms: Independent director ratio (Ind)

– Alignment of wealth of management and shareholders

Director ownership (Dir) and introduction of stock option

(SO: dummy variable which takes one if the firm introduce stock option)

25

SLIDE 26

Cross[t-1]

0.041 ***
0.023 ***
0.007 **
1.426 ***
1.637 ***
0.020 ***

(-6.02) (-3.68) (-2.44) (-3.75) (-4.54) (-4.68) Cross[t-1]*Inst[t-1] 0.161 *** 0.069 ** 0.012 2.675 * 2.211 0.042 ** (5.49) (2.18) (0.99) (1.84) (1.62) (2.23) Cross[t-1]*Ind[t-1] 0.062 ** 0.077 0.021 * 1.695 6.602 ***

0.007

(2.32) (1.00) (1.69) (1.01) (4.11) (-0.57) Cross[t-1]*SO[t-1]

0.004

0.012 * 0.000 0.250

0.167

0.004 * (-0.69) (1.86) (0.05) (0.66) (-0.42) (1.86) Cross[t-1]*Dir[t-1]

0.040
0.122 **
0.059
5.762
7.438
0.053

(-0.53) (-2.52) (-1.34) (-1.27) (-1.48) (-0.84) Risk-taking CAPEX M&A R&D Sub dummy Reducing Seg dummy

Effects of Enhanced Monitoring

Institutional ownership (Inst) mitigates the negative effects of cross-

shareholding on CAPEX, M&A, Sell-off of subsidiaries, and Risk-taking

Independent director ratio (Ind) mitigates the negative effects of cross-

shareholding on CAPEX, R&D, and Reduction of business segments

26

SLIDE 27

Cross[t-1]

0.041 ***
0.023 ***
0.007 **
1.426 ***
1.637 ***
0.020 ***

(-6.02) (-3.68) (-2.44) (-3.75) (-4.54) (-4.68) Cross[t-1]*Inst[t-1] 0.161 *** 0.069 ** 0.012 2.675 * 2.211 0.042 ** (5.49) (2.18) (0.99) (1.84) (1.62) (2.23) Cross[t-1]*Ind[t-1] 0.062 ** 0.077 0.021 * 1.695 6.602 ***

0.007

(2.32) (1.00) (1.69) (1.01) (4.11) (-0.57) Cross[t-1]*SO[t-1]

0.004

0.012 * 0.000 0.250

0.167

0.004 * (-0.69) (1.86) (0.05) (0.66) (-0.42) (1.86) Cross[t-1]*Dir[t-1]

0.040
0.122 **
0.059
5.762
7.438
0.053

(-0.53) (-2.52) (-1.34) (-1.27) (-1.48) (-0.84) Risk-taking CAPEX M&A R&D Sub dummy Reducing Seg dummy

Effects of Alignment

Alignments are less effective to mitigate the negative effects
f cross-shareholding on both investments and

restructurings than enhanced monitoring of management

Quiet life might be more important for Japanese managers

than money

27

SLIDE 28

Robustness Tests

A potential concern is that banks dissolved their

interlocking shareholder arrangements with firms

f which stock prices are relatively high, and poor

performing firms maintain cross shareholding (Miyajima and Nitta, 2011)

– So far, we include both M/B ratio and ROA in the independent variables to control the potential bias

Further, we employ propensity matching and

two-stage-least-square regression as robustness tests

28

SLIDE 29

PSM of firms with high cross-shareholding

Estimation results of cross-shareholding used in

PSM

– Firs size: + – Cash holding: – – Leverage: insignificant – ROA: insignificant – Tobin’s q: negative and significant – High in paper, chemical, increasing in steel, and decreasing in automobile – Low in electronics, telecom, retail and services

29

SLIDE 30

Propensity Score Matching

We test the differences between treatment firms ( 1st tertile

firms of cross-shareholding) and control firms (nearest matching firms in 2nd and 3rd tertiles)

Sub dummy and Reducing Seg dummy of the treatment firms are

lower than their control firms in McNemar's chi-squared tests (statistically significant in 1% and 10% level respectively)

30

variables treatment control CAPEX Mean 0.0380 0.0393

2.149 **

Median 0.0301 0.0304

1.958 *

M&A Mean 0.0037 0.0049

1.674 *

Median 0.0000 0.0000

2.664 ***

R&D Mean 0.0181 0.0214

5.379 ***

Median 0.0117 0.0129

3.668 ***

Risk-taking Mean 0.0194 0.0237

14.234 ***

Median 0.0162 0.0187

13.658 ***

t-stat or z-stat

SLIDE 31

2SLS

Most of the currently
bserved cross-

shareholdings are legacy of 20th Century

We employ natural

logarithm of firm age as IV in the first stage regression to predict cross-shareholding

In the first stage

regression, t-statistics

f IV is 37.80
We find consistent

results with H1 and H2

31

Cross[t-1]

0.152 ***
0.025
0.173 ***
1.132 ***
0.336 ***
0.219 ***

(-6.01) (-1.41) (-5.18) (-6.33) (-3.28) (-10.61) MtoB[t-1] 0.001 *** 0.001 ***

0.001 ***
0.002 *

0.002 * 0.001 *** (7.95) (7.05) (-6.21) (-1.85) (1.69) (10.08) Cash[t-1]

0.007 **

0.014 *** 0.002 0.000

0.003

0.007 *** (-2.31) (3.86) (0.78) (0.01) (-0.11) (3.43) LnAssets[t-1] 0.001 *** 0.000

0.000

0.045 *** 0.003 **

0.009 ***

(4.08) (0.43) (-0.29) (17.70) (2.42) (-28.72) Lev[t-1]

0.023 ***
0.006 **
0.005 ***

0.199 *** 0.070 *** 0.010 *** (-9.51) (-2.50) (-2.58) (9.87) (5.04) (5.87) ROA[t-1] 0.011 ***

0.019 ***
0.014 ***
0.037 **

0.034 ** 0.011 *** (5.82) (-4.06) (-6.26) (-1.96) (2.08) (11.43) (Intercept) 0.041 *** 0.006 * 0.025 ***

0.292 ***

0.033 * 0.140 *** (10.57) (1.87) (6.28) (-10.07) (1.71) (43.68) N 21111 18727 14219 21112 13845 18147 R sq. : within 0.025 0.007 0.001 0.001 0.102 0.030 R sq. : bitween 0.151 0.070 0.281 0.097 0.051 0.254 R sq. : overall 0.114 0.021 0.255 0.039 0.095 0.206 CAPEX M&A R&D Risk-taking Sub dummy Reducing Seg dummy

SLIDE 32

Analysis with more detailed industry classification

Cross[t-1]

0.026 ***
0.014 **
0.013 ***
1.060 ***
1.033 ***
0.016 ***

(-4.53) (-2.36) (-3.77) (-3.61) (-3.76) (-4.07) Stable.exc.Cross[t-1]

0.009 ***
0.005
0.010 ***
0.205 *
0.189
0.006 *

(-2.78) (-1.20) (-3.09) (-1.69) (-1.61) (-1.65) MtoB[t-1] 0.001 *** 0.001 ***

0.001 **

0.002 0.020 ** 0.001 ** (3.80) (3.66) (-2.37) (0.25) (2.27) (2.20) Cash[t-1] 0.001 0.014 * 0.002 0.240 0.203 0.004 (0.14) (1.82) (0.26) (1.36) (1.04) (0.69) LnAssets[t-1] 0.001

0.000
0.001

0.239 *** 0.016

0.013 ***

(1.15) (-0.59) (-0.85) (14.21) (1.16) (-5.59) Lev[t-1]

0.026 ***
0.008
0.009 **

1.022 *** 0.561 *** 0.007 (-6.51) (-1.47) (-2.40) (7.48) (4.27) (1.24) ROA[t-1] 0.008

0.014 *
0.008
0.910 ***

0.137 0.011 ** (1.36) (-1.84) (-1.48) (-4.21) (1.40) (2.07) (Intercept) 0.026 *** 0.007 0.027 ***

4.978 ***
1.532 ***

0.155 *** (3.32) (0.63) (2.90) (-7.63) (-4.25) (6.65) Industry (small classification) Yes Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes Yes N 19068 17099 12813 19068 12587 16562 R sq. : within 0.046 0.009 0.024 0.128 R sq. : bitween 0.247 0.094 0.377 0.323 R sq. : overall 0.190 0.028 0.367 0.286 Sub dummy Risk-taking CAPEX M&A R&D Reducing Seg

We also employ the

Nikkei Industry Small Classification Codes which provide 139 codes instead of TSE 33 industry codes

– cross-shareholding might be heterogeneous across industries

When we use small

classification, we

btains mostly similar

results

32

SLIDE 33

Summary of results from robustness tests

PSM results show that observed negative effects
f cross-shareholding are not generated by other

character of firms than cross-shareholding

Results of PSM test show that the highest cross

shareholding tertile firms invest 0.5% less than control firms relative to the total assets

– This is not negligible since the total of the three investments (CAPEX, M&A, and R&D) are roughly 7%

f the total assets
The results of 2SLS test indicate that the effects of

cross-shareholding is robust

33

SLIDE 34

Who are the cross-shareholders?

Although the main purpose of cross-shareholding is to provide

protection to managers, different group of cross-shareholders might have different incentive to maintain the relationship

We attempt to identify cross-shareholders of individual firms

– Sample: firms which are ranked in top 33% with cross-shareholding ratio as of the year 2004, and in the top five companies in sales amount of TSE 33 industries respectively – 144 large firms which have significant cross-shareholding

Categorize the shareholders listed in top 10 shareholders list of the 144

firms into three groups – Zaibatsu group (“Six Major Zaibatsu”) : 35% – Industry group (other corporate shareholders)

Excluding firms of founding family and those which hold more than 15% shares

– Banks

34

SLIDE 35

Bank

0.013
0.027

0.021 1.890 1.137 (-0.35) (-1.58) (0.54) (0.91) (0.55) Zaibatsu group 0.001 0.009

0.057 **
0.886

2.328 (0.03) (0.78) (-2.32) (-0.60) (1.58) Industry group

0.021
0.028 **
0.011
0.482
1.959 *

(-0.92) (-2.50) (-0.57) (-0.43) (-1.67) Control & Yes Yes Yes Yes Yes Intercept Industry Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes N 1342 1280 988 1342 1230 R&D M&A CAPEX Sub dummy Reducing Seg dummy

Cross-shareholder composition and firm behavior

35

Results show that none of the three groups show outstanding effects
It indicates that the three groups have mostly similar effects on corporate

behaviors

– Industry group shows negative effects on M&A and restructuring of businesses

SLIDE 36

Conclusion

Although cross-shareholding in Japan decreases its

importance from late 1990s, remaining cross-shareholding shows significant negative effects on investments and restructuring of Japanese firms

Above findings are consistent with quiet life hypothesis of

entrenched managers under protection provided by cross shareholders

Long-term and stable shareholding do not necessarily have

positive effects on long-term risky investments such as R&D

Enhanced monitoring of management by institutional

shareholders and by outside directors can mitigate the quiet life problem caused by entrenched managers

36

SLIDE 37

Reference

Aghion, P., J. Van Reenen J, and L. Zingales (2013). Innovation and Institutional Ownership. American

Economic Review 103, 277-304.

Bertrand, M., and Mullainathan, S. (2003). Enjoying the Quiet Life? Corporate Governance and Managerial
Preferences. Journal of Political Economy 111, 1043-1075.
Harford, J. (1999). Corporate Cash Reserves and Acquisitions. Journal of Finance 54, 1969–1997.
Hart, O. D. (1983). The MarketMechanism as an Incentive Scheme. Bell Journal of Economics 14, 366-382.
Hicks, J. R. (1935). Annual Survey of Economic Theory: The Theory of Monopoly. Econometrica 3, 1-20.
Hoshi, T., Kashyap, A., and Scharfstein, D. (1990). The Role of Banks in Reducing the Costs of Financial

Distress in Japan. Journal of Financial Economics 27, 67-88.

Hoshi, T., Kashyap, A., and Scharfstein, D. (1991). Corporate Structure, Liquidity, and Investment: Evidence

from Japanese Industrial Groups. Quarterly Journal of Economics, 33-60.

Hoshi, T., and Kashyap, A. (2004). Corporate Financing and Governance in Japan: The Road to the Future.

MIT press.

John, K., Litov, L., and Yeung, B. (2008). Corporate Governance and Risk-Taking. Journal of Finance 63,

1679-1728.

Miyajima H. and Nitta, K. (2011), Diversification and Outlook of Ownership Structure: Unwinding and

‘Resurgence’ of Cross-Shareholding and the Role of Foreign Investors. in Miyajima, H. (ed.), Corporate Governance in Japan: Toward a Redesign and Restoration of Competitiveness. Toyo Keizai Inc. 105–149 . (in Japanese)

Tanaka, W. (2012). Corporate Acquisitions and Anti-takeover Measures. Shoji-houmu.(in Japanese)

37

SLIDE 38

Thank You

SLIDE 39

APPENDIX

SLIDE 40

Correlation Matrix of Variables

Only total assets and institutional ownership shows correlation exceeding 0.4

– We confirm our results in interest do not change even when we exclude firm size from control variables

Cross-shareholding and Stable shareholder shareholding (excl. Cross-share) are

negatively correlated (-0.26)

– These two might be substitute

Cross shareholding and stable shareholding do not have high correlation with ROA

– An interpretation is cross shareholding does not correlate with current profitability but with future growth opportunity

40

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (1) CAPEX[t] 1.000 (2) M&A[t]

0.002

1.000 (3) R&D[t] 0.045 0.030 1.000 (4) Sub dummy[t]

0.006

0.042

0.001

1.000 (5) Reducing Seg dummy[t]

0.021

0.026 0.007 0.059 1.000 (6) Risk-taking[t] 0.058 0.077 0.223 0.027 0.070 1.000 (7) Cross[t-1]

0.072
0.058
0.119
0.044
0.046
0.240

1.000 (8) Stable.exc.Cross[t-1] 0.003

0.014
0.061
0.047
0.009
0.051
0.262

1.000 (9) Inst[t-1] 0.104 0.012 0.120 0.113 0.005

0.083
0.036
0.315

1.000 (10) Ind[t-1] 0.027 0.040 0.105 0.061 0.061 0.141

0.124

0.003 0.086 1.000 (11) Dir[t-1] 0.073 0.048 0.053

0.034

0.023 0.225

0.287
0.288
0.184

0.006 1.000 (12) SO[t-1] 0.090 0.084 0.156 0.070 0.040 0.206

0.217
0.100

0.120 0.154 0.258 1.000 (13) MtoB[t-1] 0.132 0.113 0.102 0.016 0.022 0.249

0.179
0.028

0.021 0.082 0.182 0.189 1.000 (14) Cash[t-1]

0.046

0.089 0.242

0.031

0.022 0.312

0.248
0.106
0.020

0.119 0.362 0.246 0.221 1.000 (15) Lev[t-1] 0.046

0.027
0.106

0.111 0.051

0.015

0.033

0.078
0.069

0.007

0.023
0.035

0.053

0.375

1.000 (16) LnAssets[t-1] 0.020

0.044
0.047

0.172 0.001

0.348

0.160

0.083

0.617

0.012
0.399
0.102
0.164
0.375

0.157 1.000 (17) ROA[t-1] 0.133

0.019
0.143
0.023

0.023

0.045
0.020

0.039 0.083

0.004

0.023 0.003 0.042 0.080

0.233

0.077 1.000

ｆｆ

SLIDE 41

H1 and Manufacturing

41

Cross[t-1]

0.017 ***
0.020 **
0.016 ***
0.011
0.010 **
0.023 ***

(-2.85) (-2.13) (-3.99) (-1.07) (-2.26) (-4.02) Stable.exc.Cross[t-1]

0.001
0.010 **
0.004
0.006
0.007 *
0.016 ***

(-0.36) (-2.34) (-1.34) (-0.97) (-1.82) (-3.22) MtoB[t-1] 0.002 *** 0.001 ** 0.001 ** 0.001 ***

0.000
0.001 **

(4.04) (2.53) (2.04) (2.94) (-1.16) (-2.55) Cash[t-1]

0.012 **

0.007 0.006 0.017

0.010

0.019 ** (-2.02) (0.65) (1.12) (1.44) (-1.49) (2.16) LnAssets[t-1] 0.003 ***

0.002 **

0.001 ***

0.001
0.001
0.001

(7.41) (-1.99) (3.16) (-0.84) (-1.03) (-0.80) Lev[t-1]

0.036 ***
0.012 **
0.011 ***
0.002
0.010 ***
0.005

(-8.00) (-2.09) (-3.39) (-0.33) (-3.26) (-0.55) ROA[t-1] 0.032 *** 0.005

0.027 **
0.009
0.002
0.019 *

(4.18) (1.02) (-2.26) (-0.95) (-0.35) (-1.72) (Intercept) 0.003 0.066 ***

0.007

0.017 0.028 ** 0.028 ** (0.35) (6.65) (-1.22) (0.98) (2.19) (2.57) Industry Yes Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes Yes N 9574 9494 8799 8300 9252 3561 R sq. : within 0.097 0.022 0.007 0.010 0.033 0.034 R sq. : bitween 0.213 0.207 0.059 0.053 0.297 0.198 R sq. : overall 0.169 0.154 0.017 0.021 0.306 0.177 CAPEX M&A R&D

Non- manufacturing Non- manufacturing manufacturing manufacturing Non- manufacturing manufacturing

SLIDE 42

H2 and Manufacturing

42

SLIDE 43

Cross shareholding and TFP

According to

Bertrand and Mullainathan (2003), we estimate the following equation:

lnSalse[I,t] =αlnAssets[I,t-1] +

βlnLabor[I,t-1] + γlnMaterials[i,t-1] + u[i,t]

We define the

residuals calculated form the regression as TFP

43

Cross[t-1]

0.145 ***
0.125 ***
0.164 ***
0.098 ***

(-4.11) (-3.70) (-5.53) (-3.47) LnAssets[t-1] 0.222 *** 0.255 *** (19.21) (17.45) LnLabor[t-1] 0.076 *** 0.113 *** (8.05) (9.24) LnMaterials[t-1] 0.694 *** 0.629 *** (48.61) (34.86) (Intercept) 0.453 *** 0.552 *** 0.015 *** 0.105 *** (9.49) (11.43) (3.65) (11.11) Industry No Yes No Yes Year No Yes No Yes N 21150 21150 21150 21150 R sq. : within 0.453 0.502 0.001 0.092 R sq. : bitween 0.985 0.987 0.007 0.092 R sq. : overall 0.977 0.980 0.002 0.090 LnSales TFP

SLIDE 44

Definition of stable shareholding

Stable shareholding = Cross-shareholders + financial

institutions (property accounts) + nonfinancial corporate owners + parent company ownership +

thers

44

SLIDE 45

Reply to Konari’s Comments

February 16, 2018 JES Kotaro Inoue

45

SLIDE 46

Reply: PSM of firms with high cross-shareholding

Estimation results of cross-shareholding used in PSM

– Firs size: + – Cash holding: – – Leverage: insignificant – ROA: insignificant – Tobin’s q: negative and significant

This is against argument that firms with valuable human capital

are likely to protect them with cross-shareholding

– High in paper, chemical, increasing in steel, and decreasing in automobile

These are not industries that human capital is important

– Low in electronics, telecom, retail and services

46

SLIDE 47

Reply: GMM IV

Cross[t-1]

0.174 ***
0.026
0.182 ***
1.045 ***
0.336 ***
0.240 ***

(-6.12) (-0.89) (-10.86) (-5.17) (-3.05) (-8.20) MtoB[t-1] 0.002 *** 0.001 *** 0.000

0.001

0.002 0.003 *** (4.29) (3.69) (0.25) (-0.70) (1.32) (3.88) Cash[t-1]

0.040 ***

0.013 ** 0.033 *** 0.042

0.003

0.023 *** (-7.09) (2.45) (7.17) (1.36) (-0.10) (3.59) LnAssets[t-1] 0.001 ** 0.000 0.001 *** 0.046 *** 0.003 **

0.005 ***

(2.21) (0.26) (7.41) (13.86) (2.22) (-9.18) Lev[t-1]

0.002
0.006
0.006 ***

0.197 *** 0.070 *** 0.001 (-0.60) (-1.63) (-3.19) (8.06) (4.81) (0.39) ROA[t-1] 0.039 *

0.019 **
0.078 ***
0.091

0.034

0.006

(1.84) (-2.50) (-4.76) (-1.58) (1.02) (-0.29) (Intercept) 0.048 *** 0.006 0.005 *

0.315 ***

0.033 0.096 *** (8.79) (0.95) (1.91) (-8.57) (1.54) (14.74) Industry Yes Yes Yes Yes Yes Yes Year Yes Yes Yes Yes Yes Yes N 21111 18727 14219 21112 13845 18147

adj. R-sq

0.092 0.018 0.218 0.017 0.092 0.016 firm cluster CAPEX M&A R&D Sub dummy Reducing Seg dummy Risk-taking firm cluster firm cluster robust firm cluster firm cluster

Results of two-

step GMM (firm cluster) do not change from the presented results

47

SLIDE 48

Reply: “Quiet life?”

Quiet Life

– Predicts managers without disciplinary effects reduce their effort level – not necessarily consistent with “underinvest”

Although M&A is value increasing investments in Japan

(Inoue and Kato; 2006, Mitsuya, Ikeda, and Inoue; 2017, and other)

– Oki Electric Industry

Cross-shareholding: 3.2% vs industry median: 6.2%
Oki might have enterprising spirit since it has lower

cross-shareholding

48

SLIDE 49

Reply: “Quiet life?”

Protecting valuable human capital and long-term

relationship

– valuable human capital and relationship-based transaction stories are not consistent with low q or low R&D investments of cross-shareholding firms

Cross-shareholding also works negatively to TFP in non-

reported analysis

– Cross-shareholding is higher in industries where importance of human capital is relatively low

paper, chemical, steel

– I would like to see if in-house tournament incentive is associated with higher cross-shareholding

49

SLIDE 50

What happens in bubble period?

We cannot tell if our quiet life hypothesis can explain firm behavior

before bubble period

– This is our future research topic

Nakamura, Ikeda, and Inoue (mimeo) find that stable shareholders

(not cross) do not have significant effect on long-term investments

– We do not believe firms with cross-shareholding typically over-invest

Managerial commitment for risk-taking might be different

significantly between firms with and without cross-shareholding

– Sony acquired Columbia Pictures in 1989 and has developed entertainment business to one of the three core businesses – Matsushita Electric Industries also acquired MCA (Universal) in 1990 to catch up with Sony, but divested it in 1995 without creating synergy due to lack of managerial commitment for MCA – Cross-shareholding ratio in 2004

Sony: 0.02% vs Panasonic (MEI): 5.0%

50

SLIDE 51

Robustness of PSM

Logit: treatment firm =1, control firm =0
Sample: Treatment and control firms selected by

PSM

51

2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) h_cross_1h_cross_1h_cross_1h_cross_1h_cross_1h_cross_1h_cross_1h_cross_1h_cross_1h_cross_1 high_cross_1 LnAssets

0.011

0.010

0.000

0.025 0.044 0.022 0.005 0.002 0.037 0.041 (-0.27) (0.24) (-0.01) (0.62) (1.14) (0.60) (0.13) (0.05) (0.95) (1.07) Cash 0.469 0.474

0.545
0.558
0.122

0.810 0.213 0.375 0.410 0.370 (0.65) (0.65) (-0.75) (-0.77) (-0.18) (1.15) (0.33) (0.61) (0.64) (0.59) ROA

1.017
1.835
1.277
1.510

0.012 0.275 0.883

0.035

0.190

0.362

(-0.82) (-1.62) (-1.23) (-1.47) (0.01) (0.34) (0.83) (-0.03) (0.18) (-0.31) MtoB

0.041
0.076
0.053
0.095
0.069
0.041
0.093
0.074
0.189*
0.099

(-0.52) (-0.90) (-0.79) (-1.28) (-0.77) (-0.39) (-0.97) (-0.70) (-1.72) (-1.10) Leverege 0.393 0.249 0.180 0.224 0.264 0.038 0.348 0.383 0.388 0.304 (1.08) (0.65) (0.45) (0.55) (0.66) (0.11) (0.93) (0.99) (0.98) (0.77) Industry Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes N 1368 1431 1453 1496 1486 1460 1435 1421 1422 1458 chi2 19.331 26.207 20.492 30.781 20.220 14.733 15.948 12.199 20.738 23.438 p 0.993 0.885 0.976 0.755 0.989 1.000 0.998 1.000 0.980 0.932