CHANGES IN INSTITUTIONAL HOLDINGS
To exam the possibility that institutional investors tend to sell their shares instead of taking direct actions to influence corporate decisions through top management replacement because they believe under concentrated ownership structure it too costly.
I begin the analysis by computing industry-adjusted changes in the fraction of shares held by institutional investors for two years around top management turnovers. To test whether ownership structure affects institutional investors’ behavior in corporate decision intervention, by design, this analysis focuses on the forced chairman of board/CEO turnovers in family-controlled and nonfamily-controlled firms. In addition, due to small portion of institutional holdings in Taiwanese market, the daily variance is unlikely be significant. Therefore, I add up daily industry-adjusted residuals to
quarters to examine the institutional trading behavior around the chairman and CEO turnovers. Panel A in Table 3 reports the mean industry-adjusted institutions holding changes, Panel B in Table 3 reports the mean industry-adjusted foreign investors holding changes, Panel C in Table 3 reports the mean industry-adjusted securities investment trust companies holding changes and Panel D in Table 3 reports the mean industry-adjusted dealers holding changes in forced turnovers in family-controlled versus nonfamily-controlled firms including a year prior the turnovers and a year following the turnovers. Quarter 0 defined as the day in which the replacement is announced on M.O.P.S. T-statistics for test of difference between mean value of family-controlled samples and nonfamily-controlled samples are reported below. The mean values include all institutional holdings data available during 2002-2007.
Result of Panel A in Table 3 shows in the first quarter prior the forced turnover, family-controlled firms experience negative institutional selling of -0.51%
(statistically significant at 1% level), on average. This result consists with Parrino et al. (2003) showing declines holdings in institutional holdings preceding the CEO turnovers. In contrast, institutional investors increase their holdings of 0.02%
(although not statistically significant) instead of selling in nonfamily firms. This result supports H1, which states that institutional investors are more likely to sell their
Table 3
Industry-adjusted changes in institutions ownership in two years around chairman/CEO turnovers
This table shows mean change in percentage ownership by all institutional investors of forced turnover in family control firms and forced turnover in nonfamily control firms (Panel A), forced outside succession in family control firms and forced inside succession in family control firms (Panel E). Panel B, Panel C and Panel D report mean change in percentage ownership by foreign investors, securities investment trust companies and dealers separately.
T-statistics provided in the third and seventh rows of each panel test difference between mean value of family control samples and nonfamily control samples or outside succession samples and inside succession samples. T-statistics reported in parentheses test the hull hypothesis that the mean value do not differ from zero. The turnovers occurs in quarter t=0.
Quarters
[-4,0] [-4,-3] [-3,-2] [-2,-1] [-1,0] [0,1] [1,2] [2,3] [3,4] [0,4]
Panel A : Mean change in percentage institutional ownership (%)
Forced turnover in family control firms -1.83 -0.74 -0.41 -0.17 -0.51 -0.23 -0.32 -0.34 -0.51 -1.40 (-5.78) (-3.84) (-2.57) (-0.97) (-2.83) (-1.17) (-1.83) (-2.04) (-3.40) (-4.44)
*** *** *** *** * ** *** ***
Forced turnover in nonfamily control firms -0.72 -0.62 -0.29 0.17 0.02 -0.62 -1.14 -0.55 -0.30 -2.61 (-0.77) (-2.22) (-0.90) (0.39) (0.02) (-0.95) (-3.72) (-1.45) (-1.07) (-2.91)
** *** ***
t-statistic(H0: forcedfamily = forcednonfamily) -1.38 -0.29 -0.32 -0.79 -1 0.77 2.06 0.54 -0.62 1.54
**
[-4,0] [-4,-3] [-3,-2] [-2,-1] [-1,0] [0,1] [1,2] [2,3] [3,4] [0,4]
Panel B : Mean change in percentage foreign investors ownership (%)
Forced outside succession in family control firms -1.54 -0.6 -0.32 -0.3 -0.32 -0.13 -0.26 -0.31 -0.36 -1.05
*** *** ** ** ** * ** *** ***
(-5.42) (-3.61) (-2.42) (-2.18) (-1.94) (-0.74) (-1.63) (-2.38) (-2.72) (-3.54) Forced inside succession in family control firms -1 -0.63 -0.41 -0.04 0.09 -0.2 -0.79 -0.6 -0.12 -1.71
*** *** * **
(-1.18) (-2.86) (-1.54) (-0.17) (0.12) (-0.35) (-3.82) (-1.94) (-0.52) (-2.05) t-statistic(H0: forcedoutsider = forcedinsider) -0.75 0.07 0.28 -0.81 -0.82 0.17 1.54 0.93 -0.81 0.89
Panel C : Mean change in percentage securities investment trust companies ownership (%)
Forced outside succession in family control firms -0.32 -0.17 -0.14 0.1 -0.12 -0.14 0.07 0 -0.13 -0.34
** * * ** ** ***
(-2.35) (-1.07) (-1.73) (1.07) (-1.24) (-1.96) (-1.08) (-0.05) (-2.09) (-3.24) Forced inside succession in family control firms 0.17 0.02 0.09 0.15 -0.09 -0.33 -0.32 -0.03 -0.14 -0.81
**
(0.47) (0.13) (0.37) (0.49) (-0.36) (-1.06) (-1.44) (-0.12) (-1) (-2.2) t-statistic(H0: forcedoutsider = forcedinsider) -1.46 -0.84 -1.12 -0.2 -0.11 0.9 1.48 0.1 0.06 1.66
*
[-4,0] [-4,-3] [-3,-2] [-2,-1] [-1,0] [0,1] [1,2] [2,3] [3,4] [0,4]
Panel D : Mean change in percentage deals ownership (%)
Forced outside succession in family control firms 0.03 0.03 0.05 0.02 -0.07 0.04 0 -0.03 -0.02 -0.01
**
(0.64) (0.86) (1.61) (0.82) (-2.06) (1.28) (0.17) (-0.94) (-0.62) (-0.13) Forced inside succession in family control firms 0.11 -0.01 0.03 0.06 0.02 -0.09 -0.03 0.07 -0.04 -0.1
(1.27) (-0.18) (0.92) (1.1) (0.42) (-2.08) (-0.67) (1.01) (-0.87) (-1.12) t-statistic(H0: forcedoutsider = forcedinsider) -0.72 0.5 0.28 -0.54 -1.19 1.94 0.58 -1.35 0.28 0.81
**
Panel E : Mean change in percentage institutional ownership (%)
Forced outside succession in family control firms -2.60 -1.11 -0.62 0.17 -1.03 0.97 -0.03 -1.09 -0.22 -0.38 (-2.74) (-1.85) (-1.84) (0.37) (-2.94) (1.24) (-0.07) (-2.98) (-0.70) (-0.48)
*** * * *** ***
Forced inside succession in family control firms -1.67 -0.67 -0.37 -0.25 -0.39 -0.48 -0.38 -0.18 -0.57 -1.61 (-5.10) (-3.37) (-2.03) (-1.26) (-1.94) (-2.94) (-2.07) (-0.97) (-3.39) (-4.74)
*** *** ** ** *** ** *** ***
t-statistic(H0: forcedoutsider = forcedinsider) -1.12 -0.88 -0.61 0.88 -1.37 2.87 0.75 -2.1 0.89 1.49
*** **
*Indicates statistical significance at the 10% level; **at 5% level; ***at 1% level.
shares prior forced top management turnovers because of the possibility that direct involvement in corporate decision under concentrated corporate structure making it too costly. However, even though institutional holdings dose not decrease in nonfamily-controlled samples, there is limited evidence regards to the active participation in nonfamily-controlled firms prior forced turnovers.
Consistent with the pattern of shareholding changes a quarter prior forced turnovers, institutional holdings declines, on average, 1.83% (statistically significant at 1% level) a year prior forced turnover in family-controlled. Further investigating institutional holding changes a year after forced turnovers, I find that institutional holdings continue to decline, on average, 1.4% (statistically significant at 1% level) in family-controlled firms. On the other hand, shareholding changes in nonfamily-controlled firms are reversed and start showing negative changes of -2.61% (statistically significant at 1% level). This result provides no support for the intention of institutional selling prior forced top management turnovers. If institutional investors tend to sell their shares to influence the corporate decisions by replacing top managers, they shall regain their shares following the turnovers for the fulfilled purpose. But neither family-controlled nor nonfamily-controlled firms provide the evidence.
THE INTENTION OF CHANGES IN INSTITUTIONAL HOLDINGS
By reexamining all 562 forced turnovers samples on date of announcement, name and resume of replaced chairman/CEO, name and resume of new chairman/CEO, reason for change, I surprisingly find that there are substantial successors, especially in family-controlled firms, either employed at firm prior the turnovers or related to replaced managers. By reclassifying the forced turnovers samples in family-controlled firms as outside succession or inside succession post chairman/CEO turnovers, Table 4 shows the percentage of forced outside succession following chairman/CEO turnovers in family-controlled firms over 2002 to 2007. The evidence in this table reports less than 20% of family-controlled firms experience an outside succession. Thus, there is a large chance that negative changes in institutional holdings after forced turnovers is caused by unhappy result from unchanged ultimate controller even after management replacement. Since outside successors are better able to change the direction of a firm (Parrino, 1997), institutional investors would better content with outside succession if they have intention to influence corporate decisions by top managers replacement.
To further investigate, I assume if the selling behavior prior top management turnovers is to influence corporate decisions, institutional investors shall behave differently on outside appointments which appear to be more likely to substantially alter firm policies (H2). Panel E in Table 3 reports the mean industry-adjusted
Table 4
The number and frequency of forced outside succession after chairman/CEO turnovers in listed family control firms, 2002-2007.
Number of
institutions holdings changes of the outside succession and inside succession after forced top management turnovers in family-controlled firms including a year prior the turnovers and a year following the turnovers. In the first quarter post turnovers, the institutional holdings stop declining, instead, increase 0.97% (but not significant), on average. As to inside succession firms, institutional investors are still selling their shares of 0.48% (statistically significant at 1% level). Moreover, the t-statistics for tests of differences in mean holdings changes between outside succession and inside succession reported below are statistically significant at 1% level. This result indicates
the intention of institutional investors to influence corporate decisions by selling their shares prior forced turnovers. Even though changes of institutional holdings a year post the turnovers are still negative in both outside and inside succession, only significant in insider succession firms.
ROBUST TEST
Furthermore, this section testifies the robustness of the relation between ownership structure and changes in institutional ownership prior forced turnovers by regressing the changes in institutional holdings during a year preceding the turnovers on forced turnover samples in family-controlled firms (Model 1) and in nonfamily-controlled firms (Model 2) while controlling other governance variables:
outside succession dummy and the fraction of shares held by managers, see Table 5.
Additional variables are included in the regression to control for firm specific characters including industry-adjusted stock return, operating performance (EBIT/assets) and firm size (the log nature of assets).
Results in table 5 show that family-controlled firms experience statistically significant declines in institutional holdings (at 5% level) prior forced chairman/CEO turnovers. Instead, institutional investors increase their shares in nonfamily-controlled firms (statistically significant at 5% level) prior forced turnovers. Overall, the evidence consists with H1, which indicates that institutional investors under
Table 5
Regression of changes in institutional ownership on governance variables and other variables This table reports results from OLS regressions of changes in institutional ownership in a year prior to forced turnover on governance characteristics, including outside succession dummy (one if the successor is outsider) and the fraction of shares held by managers. The change in institutional ownership represented by industry-adjusted changes in institutions ownership during a year prior the forced turnover in family control firms (Model 1) and in nonfamily control firms (Model 2). The firm characteristics include forced family dummy variable (one if the family control firm experience forced turnovers), nonforced family dummy variable (one if the nonfamily control firm experience forced turnovers), industry-adjusted stock return a year prior turnovers, operating performance (EBIT/assets) a year prior turnovers and firm size (the log nature of assets). T-statistics are reported in parentheses.
Explanatory variable Dependent variable
Forced outside succession dummy -1.25 -1.47
(-1.61) (-1.89) *
*Indicates statistical significance at the 10% level; **at 5% level; ***at 1% level.
consideration of ownership structure tend to sell their shares forced chairman and CEO turnovers.
CORPORATE GOVERNANCE AFTER TOP MANAGEMENT TURNOVERS
This section investigates whether institutional selling prior to top management turnovers can lead to improved corporate governance (H3). The indicators of corporate governance used here are focused on the board efficiency including the percentage of outside directors and supervisors on the board, board size and board holdings. These indicators help determine the influence to the functioning of the board with respect to its monitoring ability. For example, evidence suggests that the board effectiveness in its monitoring function is determined by its independence, size, and composition ( John and Senbet, 1998). In order to investigate how the changes in institutional ownerships can influence firm’s corporate governance in terms of board efficiency through top management turnovers, I estimate frequencies of the percentage of outside directors and supervisors changes, board size changes and board holdings changes while institutional holdings decline prior to forced turnovers, forced turnovers in family-controlled firms and turnovers with outside succession in family-controlled firms (Table 6). In this table, even though the evidence shows higher frequencies of percentage of outside directors and supervisors increase, board size decrease and board holdings increase, only board size is significantly decreased
Table 6
Frequencies of board efficiency indicators including the percentage of outside directors and supervisors on the board, board size and board holdings during a year post top management turnovers. Turnovers are classified as forced turnovers, forced turnovers in family-controlled firms and turnovers with outside succession in family-controlled firms.
All
*Indicates statistical significance at the 10% level; **at 5% level; ***at 1% level.
when institutional holdings declined prior forced turnovers and forced turnovers in family-controlled firms; marginal significant in board holdings increase following forced turnovers in family-controlled firms. As a result, there is only limited support for the H3 that institutional selling can influence firms toward better corporate governance through top management turnovers. Even though institutional investors have intention to influence corporate activities through top management turnovers, the power of institutional selling is not strong enough to affect governance attribute after all.