Internal governance and corporate acquisition activities

Abstract

The literature shows that the performance of bidders has been historically poor both in the short run and in the long run. Agency theory tells us that unless CEO owns 100% of the firm, the decision-making process would deviate from shareholder value maximization. Building upon the theory of internal governance by Acharya et al. (J Financ 66(3):689–720, 2011), this study documents the salutary effect of the novel governance mechanism on corporate acquisition activities. Internal governance is optimal if neither the CEO nor her subordinates are dominating. The curvilinear relationship suggests that when power and responsibility sharing is balanced in acquiring firms, myopic CEOs would have lower acquisition propensity, lower likelihood of targeting public firms, higher deal completion rate, and stronger short term gains. A system of regression equations is applied to mitigate the concerns about potential endogeneity and selection bias. Moreover, the empirical evidence demonstrates that good internal governance has strong predictive power for long term performance in the post-acquisition period, which sheds light on the influential role of subordinates in post-deal integration.

Appendices

Appendix 1

Variable	Description
Firm level data
$\delta$	The internal governance measure is the fraction of executive titles held by a CEO, a proxy for the degree of delegation between CEO and immediate subordinates. The metric is calculated as the number of executive titles of the CEO ($f$) scaled by the total number of titles carried by the top management team ($f + g$), i.e., $\delta = {f \mathord{\left/ {\vphantom {f {(f + g)}}} \right. \kern-\nulldelimiterspace} {(f + g)}}$
OtherPower	An index variable measures the overall level of other aspects of CEO power relative to industry peers. It is defined as the sum of a sequence of dummies that identifies the other three aspects of CEO power in addition to structural power: prestige power, expert power and ownership power. Specifically, Prestige Power is a dummy variable that takes the value of 1 if the CEO is the vice chair. Expert Power is the sum of two dummies: Director takes the value of 1 if the CEO is the only executive director on the board; and Expert takes the value of 1 if the number of business segments of the firm is higher than the industry median. Ownership Power is the sum of two dummies: Share takes the value of 1 if the percentage of shares owned by the CEO is higher than the industry median; and Pay takes the value of 1 if the pay slice of the CEO is higher than the industry median. Industries are defined at the 2-digit SIC level
#Deals	The number of acquisition deals announced in a fiscal year
Intensity	Acquisition intensity is defined as the total value of deals announced in a fiscal year scaled by book value of assets at the beginning of the fiscal year
Failure	Failure rate is defined as the number of unsuccessful deals (Classified by SDC Item: STATUS) in a fiscal year divided by the total number of deals in the fiscal year. To be conservative, a deal is considered failed if it is not clearly coded “Completed” [The results are qualitatively similar by alternative classifications of deal failure (e.g. missing effective/unconditional dates by SDC Item: DATEEFFUNCON)]
Public	The number of deals targeting public firms announced in a fiscal year scaled by the total number of deals in the fiscal year
Deal	A dummy variable that takes the value of unity if an acquiring firm is involved in at least one deal in a fiscal year, and zero otherwise
AggDollar	The aggregate dollar gain (in millions) of acquirer is the sum of dollar gains for all the deals in a fiscal year. The dollar gain is the change in the market capitalization of an acquirer from 2 days before (− 2) the deal announcement to 1 day after (+ 1)
AvgReturn	The average return gain of acquirer is the deal value weighted average return rate for all the deals in a fiscal year. The return rate is the change in the market capitalization of an acquirer from 2 days before (− 2) the deal announcement to 1 day after (+ 1) divided by the market capitalization of an acquirer 2 days before (− 2) the deal announcement
EAbReturn	The abnormal return gain of acquirer is the deal value weighted average abnormal holding period return over the (− 2, + 1) event window for all the deals in a fiscal year. The abnormal return is defined as a model-free holding period return in excess of the equally weighted market portfolio
VAbReturn	The abnormal return gain of acquirer is the deal value weighted average abnormal holding period return over the (− 2, + 1) event window for all the deals in a fiscal year. The abnormal return is defined as a model-free holding period return in excess of the value weighted market portfolio
M/B	The industry adjusted market-to-book ratio is defined as the firm’s market-to-book ratio minus the industry’s median market-to-book ratio. The median level is calculated at the two-digit SIC industry-year level using the Compustat universe
Leverage	The long term debt and debt in current liabilities divided by book value of assets at the beginning of the fiscal year
Size	The natural log of book value of assets
R&D	The research and development expenditures divided by book value of assets at the beginning of the fiscal year
FirmAge	The number of years that a firm has data available in Compustat
Directors	The total number of directors serving on the board
Outsiders	The percentage of outsider directors
Cash	The book value of cash and short-term investments divided by book value of assets at the beginning of the fiscal year
PPE	The book value of property, plant and equipment divided by book value of assets at the beginning of the fiscal year
IndCapx	The natural log of industry capital expenditures as measured by the sum of capital expenditures in industries identified by 2-digit SIC codes
SReturn	The annual stock return during the fiscal year
PPS	The pay performance sensitivity measured as CEO total portfolio delta (in thousands)
CPS	The total CEO compensation divided by the total compensation for the whole top management team of five executives
Duality	A dummy variable takes the value of 1 if the CEO is also the chair of the board
Founder	A dummy variable takes the value of 1 if the CEO is also the founder of the firm
Ownership	The percentage of shares owned by the CEO
Tenure	The number of years since the CEO took office
Deal level data
Subsidiary	A dummy variable takes the value of unity if the target is a subsidiary firm and zero otherwise
Private	A dummy variable takes the value of unity if the target is a private firm, and zero otherwise
Focus	A dummy variable takes the value of unity if the target has the same two-digit SIC code as the acquirer, and zero otherwise
Cross	A dummy variable takes the value of unity if the target is a non-US firm, and zero otherwise
Tender	A dummy variable takes the value of unity if the deal involves a tender offer, and zero otherwise
Friendly	A dummy variable indicates the deal attitude. It takes the value of unity if the offer is friendly, and zero otherwise
ACap	The acquirer’s market capitalization one month before the deal announcement

Appendix 2

Below analytical process is by no means a full-fledged development of econometric modeling but an intuitive reasoning in support of my empirical strategy. I will start with the most general form of a multiple linear regression model,

$$y = X\beta + \varepsilon$$

(6)

in which, $\begin{gathered} y = \left( {\begin{array}{*{20}c} {y_{1} } \\ {y_{2} } \\ \vdots \\ {y_{N} } \\ \end{array} } \right)_{N \times 1} , \, X = \left( {\begin{array}{*{20}c} {x_{11} } & {x_{12} } & \cdots & {x_{1K} } \\ {x_{21} } & {x_{22} } & \cdots & {x_{2K} } \\ \vdots & \vdots & \ddots & \vdots \\ {x_{N1} } & {x_{N2} } & \cdots & {x_{NK} } \\ \end{array} } \right)_{N \times K} = \left( {\begin{array}{*{20}c} {{\text{x}}_{1} } & {{\text{x}}_{2} } & \cdots & {{\text{x}}_{K} } \\ \end{array} } \right)_{N \times K} \hfill \\ \hfill \\ \end{gathered}$, and ${\text{x}}_{i} = \left( {\begin{array}{*{20}c} {{\text{x}}_{1i} } \\ {{\text{x}}_{2i} } \\ \vdots \\ {{\text{x}}_{Ni} } \\ \end{array} } \right)_{N \times 1} i = 1,2, \ldots K, \, \beta { = }\left( {\begin{array}{*{20}c} {\beta_{1} } \\ {\beta_{2} } \\ \vdots \\ {\beta_{K} } \\ \end{array} } \right)_{K \times 1} , \, \varepsilon { = }\left( {\begin{array}{*{20}c} {\varepsilon_{1} } \\ {\varepsilon_{2} } \\ \vdots \\ {\varepsilon_{N} } \\ \end{array} } \right)_{N \times 1}$.

To simplify the derivation, I standardize $y$ and ${\text{x}}_{i}$ so that $\widetilde{y} = \frac{{y - \overline{y}1_{N \times 1} }}{{S_{y} }}, \, \widetilde{{\text{x}}}_{k} = \frac{{{\text{x}}_{k} - \overline{x}_{k} 1_{N \times 1} }}{{S_{{{\text{x}}_{k} }} }}$, where $1_{N \times 1}$ is a vector of ones, and $S_{{\{ y,x_{k} \} }}$ represents corresponding standard deviations for dependent and independent variables. To be concise, I still denote the standardized $\widetilde{y}$ and $\widetilde{{\text{x}}}_{k}$ as $y$ and ${\text{x}}_{i}$. Moreover, for the purpose of this derivation, I assume no endogeneity and assume homogeneous standard errors that are not clustered at any level. The standard OLS estimator is as follows,

$$\widehat{\beta } = (X^{T} X)^{ - 1} XY = \left( {\frac{1}{N}X^{T} X} \right)^{ - 1} \frac{1}{N}X^{T} Y$$

(7)

with the variance of the parameter estimates:

$$\begin{aligned} Var\left( {\widehat{\beta }|X} \right) & = E\left( {\left( {\widehat{\beta } - E\left( {\widehat{\beta }|X} \right)} \right)\left( {\widehat{\beta } - E\left( {\widehat{\beta }|X} \right)} \right)^{T} |X} \right) \\ & = \left( {X^{T} X} \right)^{ - 1} X^{T} E\left( {\varepsilon^{T} \varepsilon |X} \right)X^{T} \left( {X^{T} X} \right)^{ - 1} \\ & = \sigma^{2} \left( {X^{T} X} \right)^{ - 1} \\ & = \frac{1}{N}\sigma^{2} \left( {\frac{1}{N}X^{T} X} \right)^{ - 1} . \\ \end{aligned}$$

(8)

If $K = 2, \, \frac{1}{N}X^{T} X = \left( {\begin{array}{*{20}c} {\frac{{1^{T} 1}}{N}} & {\frac{1}{N}x_{i}^{T} 1} \\ {\frac{1}{N}x_{i}^{T} 1} & {\frac{1}{N}x_{i}^{T} x_{i} } \\ \end{array} } \right) = \left( {\begin{array}{*{20}c} {\frac{1}{N}N} & {\frac{1}{N}\sum {x_{i} } } \\ {\frac{1}{N}\sum {x_{i} } } & {\frac{1}{N}\sum {x_{i}^{2} } } \\ \end{array} } \right)$. For the sake of tractability, I further assume that ${\text{x}}_{i}$ is orthogonal to each other, i.e.${\text{x}}_{i}^{T} {\text{x}}_{j} = 0$ for $i \ne j$. Notice that for standardized ${\text{x}}_{i}$, such a restriction is roughly equivalent to no severe multicollinearity among independent variables. As such,

$$\frac{1}{N}X^{T} X = \left( {\begin{array}{*{20}c} {\frac{{1^{T} 1}}{N}} & {\frac{{{\text{x}}_{2}^{T} 1}}{N}} & \cdots & {\frac{{{\text{x}}_{K}^{T} 1}}{N}} \\ {\frac{{{\text{x}}_{2}^{T} 1}}{N}} & {\frac{{{\text{x}}_{2}^{T} {\text{x}}_{2} }}{N}} & \cdots & 0 \\ \vdots & \vdots & \ddots & 0 \\ {\frac{{{\text{x}}_{K}^{T} 1}}{N}} & 0 & 0 & {\frac{{{\text{x}}_{K}^{T} {\text{x}}_{K} }}{N}} \\ \end{array} } \right).$$

(9)

As $N$ is sufficiently large (large sample analysis),

$$\begin{gathered} \frac{1}{N}X^{\prime}X = \left( {\begin{array}{*{20}c} 1 & {\overline{x}_{1} } & \cdots & {\overline{x}_{K} } \\ {\overline{x}_{1} } & {\hat{\sigma }_{{x_{1} }}^{2} } & \cdots & 0 \\ \vdots & \vdots & \ddots & 0 \\ {\overline{x}_{K} } & 0 & 0 & {\hat{\sigma }_{{x_{k} }}^{2} } \\ \end{array} } \right) \to \left( {\begin{array}{*{20}c} 1 & 0 & \cdots & 0 \\ 0 & {\sigma_{{x_{1} }}^{2} } & \cdots & 0 \\ \vdots & \vdots & \ddots & 0 \\ 0 & 0 & 0 & {\sigma_{{x_{k} }}^{2} } \\ \end{array} } \right){. } \hfill \\ \, \hfill \\ \end{gathered}$$

(10)

Provided that $y$ and ${\text{x}}_{i}$ are standardized, I could derive the following expression,

$$\left( {\frac{1}{N}X^{\prime}X} \right)^{ - 1} = \left( {\begin{array}{*{20}c} 1 & 0 & \cdots & 0 \\ 0 & {\frac{1}{{\sigma_{{x_{1} }}^{2} }}} & \cdots & 0 \\ \vdots & \vdots & \ddots & 0 \\ 0 & 0 & 0 & {\frac{1}{{\sigma_{{x_{1} }}^{2} }}} \\ \end{array} } \right) = \left( {\begin{array}{*{20}c} 1 & 0 & \cdots & 0 \\ 0 & 1 & \cdots & 0 \\ \vdots & \vdots & \ddots & 0 \\ 0 & 0 & 0 & 1 \\ \end{array} } \right) = I_{N} .$$

(11)

Analogously, when $N$ is sufficiently large

$$\frac{1}{N}X^{\prime}Y = \left( {\begin{array}{*{20}c} {\frac{1}{N}{\text{x}}_{1} ^{\prime}y} \\ {\frac{1}{N}{\text{x}}_{2} ^{\prime}y} \\ \vdots \\ {\frac{1}{N}{\text{x}}_{K} ^{\prime}y} \\ \end{array} } \right) \to \left( {\begin{array}{*{20}c} {\sigma_{{{\text{x}}_{1} ,y}} } \\ {\sigma_{{{\text{x}}_{2} ,y}} } \\ \vdots \\ {\sigma_{{{\text{x}}_{K} ,y}} } \\ \end{array} } \right).$$

(12)

With the all the above setting in place, we could examine the statistical significance of $\widehat{\beta }_{i}$ using the following equation.

$$\left| {t_{{\widehat{\beta }_{i} }} } \right| = \left| {\frac{{\left[ {\widehat{\beta }} \right]_{i,1} - \left[ \beta \right]_{i,1} }}{{\sqrt {\left[ {Var(\widehat{\beta }|X)} \right]_{i,i} } }}} \right| \to \left| {\frac{{Var({\text{x}}_{i} )^{ - 1} \times Cov({\text{x}}_{i} ,y) - \beta_{i} }}{{\sqrt {Var({\text{x}}_{i} )^{ - 1} \left( {\frac{1}{N}\sigma_{\varepsilon }^{2} } \right)} }}} \right| = \left| {\frac{{\sigma_{{{\text{x}}_{i} ,y}}^{2} - \beta_{i} }}{{\sqrt {\frac{1}{N}\sigma_{\varepsilon }^{2} } }}} \right|$$

(13)

Provided that the variable of interest ${\text{x}}_{i}$ ($\delta$), does not change within the same fiscal year and the acquisition policy is very sticky, the components of equation (11) shall change accordingly as follows: $N \to N \times M$, $\frac{1}{N}\sigma_{\varepsilon }^{2} \to \frac{1}{N \times M}\sigma_{\varepsilon }^{2} , \, \sigma_{{{\text{x}}_{i} ,y}}^{2} \to M \times \sigma_{{{\text{x}}_{i} ,y}}^{2}$, where $M$ is average number of acquisitions per fiscal period. Therefore, asymptomatically, the magnitude of statistical significance ($\left| {t_{{\widehat{\beta }_{i} }} } \right|$), will increase dramatically in $M$.⁹ Therefore, taking into the consideration of the data and the research goals, I use firm-year as the unit of this study.