Download Business Relationships, Corporate Governance, and Firm

Business Relationships, Corporate Governance, and Firm Performance:
Evidence from Private Placements of Common Stock†
Karen H. Wruck* and YiLin Wu**
March 2008
Abstract
Utilizing a large sample and unique data gathered directly from private placement contracts, we
conduct an in-depth exploration of two important questions that remain unresolved in the
literature to date. First, how are private placement investors involved in the affairs of the issuing
firm? Second, do relationships between issuers and investors, or a lack thereof, help explain the
stock price performance “puzzle” associated with private placements? Examining relationships
provides insight into the source and possible causes of poor post-placement stock price
performance among some firms. Placements associated with new relationships have positive
announcement returns and insignificant post-placement returns. Placements without new
relationships are non-events at announcement. Poor post-placement performance is confined to
placements lacking any type of relationship and placements to key business partners. This study
also provides new evidence on the association between relationships, deal structure and issuer
characteristics, and about the influence of private placements on issuer governance.
JEL classification: G39; G32; K22
Keywords: Private placement; private equity; equity issuance; relationship investing; relationship
investor; hierarchy of investors; agency theory; asymmetric information; entrenchment; specific
investment; relationship-specific investment, governance, blockholders; ownership
concentration.
* Corresponding Author: Karen Hopper Wruck, Professor, Fisher College of Business, The Ohio State University,
Fisher College of Business, Columbus, Ohio 43210, USA, email: [email protected].
** Yilin Wu, Assistant Professor, Department of Quantitative Finance, College of Technology Management,
National Tsing Hua University, Hsinchu, Taiwan, R.O.C., email: yilinwu@ mx.nthu.edu.tw.
† We would like to thank seminar participants at Chinese University of Hong Kong, City University of Hong Kong,
National Central University, National Cheng Kung University, National Chiao Tung University, National Tsing Hua
University, and Yuan-Ze University. This paper was presented at the 13th Conference on the Theories and Practices
of Securities and Financial Markets in Taiwan held in December 2005, the 2006 American Finance Association
Meetings, the 2007 American Enterprise Conference on Private Equity, the 2008 LeBow College of Business Center
of Corporate Governance Symposium, and the 2008 Journal of Corporate Finance Conference on Corporate Control,
Mergers and Acquisitions. We have benefited from the comments of Joseph P.H. Fan, Michael Hertzel, Peter Klein,
Annette Poulsen, Clemens Sialm, René M. Stulz and Tracie Woidtke. The authors are grateful for research support
from the Dice Center for Financial Research, the Fisher College of Business, The Ohio State University, and
research support from Taiwan, R.O.C., National Science Council Grant.
Business Relationships, Corporate Governance, and Firm Performance:
Evidence from Private Placements of Common Stock
Karen H. Wruck and YiLin Wu
Draft: March 2008
1. Introduction
Recent growth in private equity markets has outstripped growth in public markets, both
domestically and globally. Between 1995 and 2006, the number of private placements increased
from 127 to 2,719 annually, with aggregate proceeds soaring from $1.87 billion to
$88.0 billion—a forty-seven-fold increase over twelve years.1 This rapid growth increases the
importance of understanding the structure of private placements and their impact on issuer
performance.
Beyond the growing importance of private equity as a source of capital, private placements
provide an opportunity to study the impact of relationships between investors and issuers on
performance, deal structure and governance. Private placements involve the issuance of a block
of shares to a relatively small group of investors. These investors may have a relationship with
the firm prior to the placement and, in addition, may create a new relationship around the time of
issuance. A relationship could involve a business arrangement, an employment agreement,
and/or a governance affiliation.
The potential value of relationships between investors and issuers dates back at least to
Jensen and Meckling (1976) and Fama (1985). Much of the literature in this area focuses on
relationship banking (e.g., James (1987), Lummer and McConnell (1989), Rajan (1992),
Petersen and Rajan (1994) and La Porta, López-de-Silanes and Zamarripa (2003), an exception is
Bhagat, Black, and Blair (2004)). Broadly speaking, the literature shows that relationships create
value by helping borrowers overcome information problems and/or resolve agency problems.
Prior studies have tackled the topic of relationships in private placements of equity but, as
detailed below, their data allow only broad-brush classifications. Our primary data are gathered
from private placement contracts, not press reports or databases. By supplementing contract data
with other data sources, we are able to identify and classify both pre-existing and new postplacement relationships between issuers and investors along multiple dimensions. Thus, our data
allow us to conduct an in-depth exploration of two important questions that remain unresolved in
1
The source is Sagient Research (http://www.sagientresearch.com). Private placements include private investments
in public equity (PIPE), 144-A placements and Regulation S transactions.
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
the literature to date. First, how are private placement investors involved in the affairs of the
issuing firm? Second, do relationships between issuers and investors, or a lack thereof, help
explain the stock price performance “puzzle” associated with private placements? The stock
price performance puzzle referred to is, of course, the well-documented empirical result that
private placements are met with a positive stock price response at announcement, but are
followed by poor post-placement performance that more than offsets announcement period
gains.2
Prior research exploring the association between relationships and performance in private
equity transactions yields contradictory findings, and data limitations restrict the dimensionality
of the relationship variables used. Krishnamurthy, Spindt, Subramaniam and Woidtke (2005)
focus on affiliated versus unaffiliated investors; affiliated investors have a pre-placement
relationship with the issuer as reported in Lexis-Nexis. They find poorer mean long-run
performance following placements to unaffiliated investors than placements to affiliated
investors. They do not examine new post-placement relationships, nor do they examine
differences in performance across categories of affiliated investors.
Barclay, Holderness and Sheehan (2007) focus on post-placement investor activism.
However, they rely on press reports to identify active investors; investors not appearing in press
reports are classified as passive. Press reports identify very few active investors, leading these
authors to identify investor passivity as the leading cause of poor post-placement performance.
Other research focuses on investor type, e.g., individual, corporation, venture capital firm, which
is a proxy for the nature of the issuer-investor relationship. Meidan (2006) finds that after
controlling for placement characteristics, investor type is unrelated to post-placement
performance. Brophy, Ouimet and Sialm (2007) focus on hedge fund versus other investors.
They find evidence that firms issuing to hedge funds have poorer post-placement performance,
but for the sub-sample of their firms making traditional, unregistered private placements there is
no post-performance difference based on investor type.
Using richer data and a larger sample, we provide new insights into the sources and possible
causes of poor post-placement stock price performance among some firms. This moves us
toward a resolution of apparent contradictions and puzzles in prior research. Our data show that
2
Hertzel, Lemmon, Linck and Reese (2002) is the first study of which we are aware that documents this finding.
Subsequent papers confirming these findings with different samples include Barclay, Holderness and Sheehan
(2007), Brophy, Ouitmet and Sialm (2007), Chaplinsky and Haushalter (2007), Krishnamurthy, Spindt,
Subramaniam and Woidke (2005) and Meidan (2006).
2
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
relationships between issuers and investors are pervasive in private placements. Of the issuers in
our sample, 64% placed shares with investors with whom they had a pre-placement relationship:
managers, key business partners, prior blockholders and/or prior directors. In addition, many new
relationships are formed as part of, or following, a placement. The distinction between pre- and
post-placement relationships is new to this study. We find that new relationships play a critical
role in explaining announcement period and post-placement stock price performance.
Controlling for issuer and issuance characteristics, placements associated with relationships
have higher abnormal returns at announcement; placements to key business partners, directors
and/or blockholders have incremental abnormal returns ranging from 1.6% to 1.8%. Evidence
regarding the value of new relationships is also striking. At announcement, placements not
associated with the formation of a new relationship have insignificant abnormal returns; they are
viewed as “non-events” by the market. In contrast, placements associated with new relationships
experience significant positive abnormal returns at announcement. Controlling for firm and
issuance characteristics, placements in which a new block is formed have an incremental
abnormal return of 1.6% to 2.8%. Placements in which investors gain a directorship have a 0.8%
to 2.6% incremental abnormal return. These findings are consistent with relationship investors
(pre-existing and new) helping to resolve information and/or agency problems.
Taking relationships into account also lends insight into the post-placement performance
puzzle. In our sample, poor post-placement performance is confined to two sub-samples:
placements to key business partners and placements to outside investors who form no postplacement relationship with the issuer. The abnormal return for the key business partner subsample is -14.3% over six months, and -38.1% three years post-placement. The analogous
returns for the outsiders sub-sample are -10.31% and -38.9%. Interestingly, investors in these
sub-samples are in distinctly different positions with respect to information flow and access to
governance.
Key business partners have a pre-placement relationship with the firm that likely involves
some form of specific investment. These investors may be “tied to the mast” of the issuer
through their relationship. Even if the issuer faces bleak future prospects, participation in the
private placement may be rational, especially if the alternative is not to invest and “go down with
the ship.” In contrast, the outsiders have no relationship tie to the firm either before or after the
placement. They probably suffer an information disadvantage and, with no access to governance
processes, they have no power to intervene ex post. Placements to all other non-management
3
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
investors have insignificant post-placement abnormal returns. Placements to managers have
positive abnormal returns over the six months following announcement, but insignificant returns
over a three-year horizon.
Our data also allow us to conduct what, to our knowledge, are two unique analyses: first, the
association between pre-placement relationships and issuer and placement characteristics, and
second, the influence of private placements on the governance structure of the issuer. We find
that placements to pre-existing relationship investors are larger and more concentrated, are made
by issuers with higher future profitability and, with the exception of management investors, are
sold at a smaller price discount. Our evidence also suggests that issuers for whom innovation is
important are more likely to issue to relationship investors, while issuers with high specific risk
are less likely to do so. Perhaps surprisingly, issuers with CEOs of relatively long tenure tend to
make placements to outsiders.
Our measure of governance influence is the percentage of the issuing firm’s directors
appointed by investors post-placement. We find that, with the exception of those who are already
directors, relationship investors (pre-existing and new) are more likely to gain governance
influence than other investors. Issuers whose CEOs have relatively large shareholdings are more
likely to grant directorships to investors, which is inconsistent with entrenchment. Issuers in
“new economy” industries and with high specific risk grant investors more governance influence
than other issuers, suggesting that access to governance is especially valuable when information
asymmetries and/or specific investments are important. We also find that stronger governance
influence is associated with a larger price discount. This is consistent with compensation to
investors for participating in the governance process.
This paper is organized as follows. Section 2 presents our relationship classification
taxonomy and develops hypotheses. Section 3 describes the data collection process and presents
descriptive statistics. Section 4 analyzes issuer performance. Section 5 examines the association
between relationships, and placement and firm characteristics. Section 6 concludes.
2. Theory, hypotheses and prior research
Below, we develop predictions regarding the association between relationships and issuer
performance, between relationships and placement characteristics, and predictions regarding the
influence of relationships on issuers’ governance structure.
4
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
2.1. Relationships and Issuer Performance
2.1.1. Agency Costs and Specific Investments
Agency theory predicts that blockholdings, governance affiliations and other relationships
affect firm value either positively, by aligning incentives between investors and decision-makers,
or negatively through entrenchment (Jensen and Meckling (1976)).3 Theories of specific
investment predict that relationships encourage specific investment; relationships serve as
mechanisms to bond against ex post opportunistic behavior. Without such mechanisms, the
potential for hold-up results in too little investment in specific assets (Grossman and Hart (1986),
Hart and Moore (1990), and Klein, Crawford and Alchian (1978)).
Prior research suggests that private placements affect agency costs and facilitate relationshipspecific investments. The positive association between block size and changes in firm value is
consistent with a reduction in agency costs (Hertzel and Smith (1993) and Wruck (1989)) There
is also a suggestion of entrenchment.4 The earlier-referenced finding by Krishnamurthy, Spindt,
Subramaniam and Woidtke (2005) of stronger post-placement returns for firms issuing to
affiliated investors is consistent with these investors helping to reduce agency costs.
With regard to governance, findings by Allen and Phillips (2000) suggest that governance
affiliations don’t matter: in their sample, the announcement response to a private placement is
unrelated to the investor’s director status. In contrast, Barclay, Holderness and Sheehan (2007)’s
earlier-referenced finding of weak post-placement returns for firms issuing to passive investors is
consistent with the lack of a governance relationship exacerbating agency problems. Allen and
Phillips (2000) also find that placements involving an alliance or joint venture relationship are
associated with a more positive announcement return than other placements, suggesting that the
equity stake sold in the placement encourages specific investment.
Our data allow us to conduct a deeper analysis of the role relationships and relationship types
play in attenuating or exacerbating agency problems, and their role in facilitating relationshipspecific investments. If relationships in private placements align incentives and/or encourage
3
In addition, relationship investors who develop an anti-manager reputation may face fewer investment
opportunities in the future, and thus might demonstrate a supportive attitude toward managers even when doing so
decreases shareholder value. For example, several companies withdrew pension assets from Tiger Management
when Tiger ran a proxy contest for representation on the Cleveland-Cliffs board in 1991, despite Tiger’s strong
performance record (See Black (1992)).
4
This is consistent with the literature on ownership structure, which documents the positive valuation effect of
increases in ownership concentration over much of the ownership spectrum (e.g. Kahn and Winton (1998),
McConnell and Servaes (1990), and Shleifer and Vishny (1986)). Wruck (1989) documents a negative relation
between firm value and ownership concentration for placements resulting in 5% to 25% ownership by blockholders,
and for placements in which investors gain a controlling interest.
5
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
specific investments, the announcement response and longer-run performance will be higher for
placements to relationship investors than for placements to outsiders. It should also be the case
that, all else constant, private placements in which new relationships are formed are received
more positively than other placements. The weakest performance at announcement, and in the
longer-term, should be placements without relationships: specifically, placements to outside
investors in which no new relationships are formed. However, if relationship investors facilitate
entrenchment, placements to outside investors will be associated with stronger returns than
placements to relationship investors.
We also test one hypothesis regarding relationship specific investment and performance.
Placements to key business partners are likely to be those in which specific investment is most
important. If such specific investment is valuable, placements to key business partners and
placements in which new partnerships are formed should be associated with relatively strong
performance.
2.1.2. Information Asymmetry and Signaling
Prior research also shows that information problems affect the choice of equity issuance
method (Myers and Majluf (1984) and Hertzel and Smith (1993)). Most relevant here is the
finding that proxies for information asymmetry are positively associated with the likelihood that
a firm will issue equity via a private placement rather conducting a public offering (Cronqvist
and Nilson (2005) and Wu (2004)). These findings suggest that a single investor or small group
of investors can attenuate information problems. In such situations, information sharing between
the firm and private investors reduces the cost of asymmetric information, thus increasing the
availability of new capital and/or reducing its cost.
Among private placement investors, those with pre-placement relationships with the issuer
are in a better position to help the issuer resolve information problems than outside investors.
Through their prior relationship, they are already familiar with the firm and likely have access to
information not available to outsiders.5 This leads to the prediction that, all else constant, the
market response at announcement will be higher for placements to pre-existing relationship
investors than placements to other investors. If relationship investors participate in offerings by
higher quality firms, we also expect the longer-term performance of those issuers to be higher
than that of other issuers.
5
Boot (2000) identifies two elements that are critical in a relationship investing situation: (1) disclosure: investors
gather proprietary information and keep it confidential and (2) time: information gathering takes place over time
through multiple interactions.
6
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Across types of relationship investors (managers, key business partners and investors with a
governance affiliation), access to information will vary. Managers likely have the best access.
Key business partners and investors with a governance affiliation have better access than
outsiders, but it is not clear which group would have consistently better (or worse) access.
Neither is it clear whether the access given to each of these investor types would be consistent
across firms. Nonetheless, we examine the data to see if differences emerge.
Alternatively, outsiders might be better informed than relationship investors. For example,
managers and/or other relationship investors may have a positively biased view of the issuer’s
prospects (e.g., Heaton (2002), Malmendier and Gate (2005), and Wong and Zhang (2007)).
Outside investors may be more objective and/or generate superior information as part of their
due diligence processes. If outsiders are better informed and/or less positively biased than
relationship investors, placements to outside investors will be more favorably received at
announcement, and will be associated with stronger post-placement performance.
2.2. Relationships, Issuer Characteristics and Governance
Private placements involve a process of matching issuers with investors. Issuers with
different business and firm characteristics are likely to seek investors with different
characteristics. Because private placements are voluntary transactions, incumbent managers
influence their structure, including the “choice” of investors and the creation of new
relationships. If relationship investors facilitate incentive alignment, firms with weak governance
and/or poor managerial incentives will seek to avoid such investors. In contrast, if relationship
investors facilitate entrenchment, issuers with entrenched managers are more likely to issue to
relationship investors than issuers whose managers are not entrenched.
Thus far, we have focused on the incentive effects of ownership, ignoring any associated
risk. When firm-specific risk is high, it may be difficult or costly for a relationship investor to
diversify away specific risk, particularly if the investor is a manager or key business partner. In
such situations, efficiency dictates a separation of the management and risk-bearing functions.
Thus, we expect that issuers with high specific risk are less likely to issue to relationship
investors and more likely to issue to outsiders.
Our data also allow us to examine how variation in the magnitude of information problems
faced by issuers is associated with relationship characteristics. Among private placement firms,
those facing relatively severe information problems are more likely to issue to relationship
investors than to outsiders. In addition, the potential for value creation through relationship7
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
specific investments differs across firms. Raith (2003, 2005) suggests that when asymmetric
information about firm-specific investment technology is high, relationship investing is valuable.
This implies that among private placement firms, those where innovation is important are more
likely to issue to relationship investors.
An important dimension of investors’ influence is their participation in the issuer’s
governance process. The value and desirability of investor participation is likely to vary across
issuers and investors. For issuers facing information problems, with high firm-specific risk
and/or whose value is comprised largely of growth options, bundling the opportunity to invest
with a governance affiliation is likely to be valuable to investors because it provides access to
information flow and an opportunity to influence decision-making. Also, for investors with a
non-governance relationship with the issuer, a directorship could help align incentives and
complement the value of the non-governance relationship. For an investor lacking another
relationship with the issuer, a directorship may serve as the primary mechanism through which
the investor can access information and exert influence. However, if another type of relationship
allows investors sufficient influence, or if the investor prefers a passive investment, the cost of a
governance affiliation might be greater than its benefit.
2.3. Relationships and Issue Pricing
In private placements, shares are typically sold at a significant discount from the market
price. Theory suggests a number of hypotheses regarding relationships and price discounts. One
hypothesis is that the discount is compensation to the investor for costs incurred in monitoring, in
reducing agency costs, in facilitating specific investment, holding an illiquid investment, or
helping the firm resolve information problems. Given that investors with pre-existing
relationships are more likely to be more involved in and informed about the issuer’s activities,
this line of reasoning suggests they will receive a higher discount than other investors. If,
however, the placement is designed to entrench, the pricing implications are ambiguous. The
offer price might be highly discounted to reflect compensation for entrenchment, or mildly
discounted or undiscounted, with the investor gaining private benefits in return.
Alternatively, the discount could be a reflection of the issuer’s true value. Firms facing
problems of asymmetric information have difficulty communicating their true value to the
market. Pre-existing relationship investors, with superior access to information, will choose to
participate in a placement when they believe the firm’s prospects are better than the market’s
assessment. Here, we expect a smaller issue discount in placements to relationship investors than
8
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
in placements to outsiders. It may also be the case that investors with a pre-existing relationship
are willing to accept a smaller discount in order to avoid the perception of self-dealing.
3. Sample formation and variable construction
3.1. Sample construction
Our sample consists of 1,976 “traditional” (or Regulation D) private placements of common
stock made by U.S. public companies between 1980 and 1999.6 For each placement, we collect
detailed data on the investor and the relationship between the investor and the issuing firm. Of
our sample placements, 6% (128) are made by firms on the NYSE, 9% (182) by firms listed on
Amex, 83% (1,634) by firms listed on Nasdaq, and 2% (32) by firms listed on other exchanges.
There are 424 firms that make two or more private placements during the sample period. The
sample transactions involve 8,049 investors in total. Our sample period ends prior to the burst of
the Nasdaq bubble, so the structure of our placements and our findings should be untainted by
that major market event and any related fallout.
To construct our sample, we first collect a sample of completed private placements of
common stock from the company news file of the Lexis-Nexis database. These data are
supplemented by information from the Dow Jones Interactive database, FISonline, the S & P
corporate descriptions file of the Lexis-Nexis database, the Security Data Corporation database
and SEC filings (Forms 8-K, 10-Q and 10-K). We eliminate placements where the common
stock offered is not of the same class as the shares listed on CRSP and/or the firm’s stock is not
listed on CRSP at the announcement date. This yields a preliminary sample of 2,652.
To gather our unique data on investors and relationships, we exploit the fact that companies
filing with the SEC must file private placements exhibits that report the original placement
agreement or contract. Examining these contracts allows us to move beyond the relatively vague
description of investors given in a typical press release (e.g., issuance to “certain directors and
officers,” “certain institutional investors,” or “a private investment group”) to identify specific
investors. We collect information on the contracting date, the number of investors, the number of
shares purchased by each investor, purchase price, names of the investor(s), and relationship
6
Traditional private equity (Regulation D) market differs from the PIPE (private investment in public equity)
market. PIPEs eliminate Regulation D resale restrictions by requiring that the shares issued be registered with the
SEC within 30 days of closing. As a result, PIPE issues are more liquid than traditional private equity issues,
implying that the latter create a tighter relationship between issuer and investor. Structured PIPEs include clauses,
such as repricing rights, that protect investors against negative outcomes. Such clauses are not common even in the
PIPE market (resets clauses are present in only 7% of Brophy, Ouimet and Sialm (2007)’s sample and in only 5% of
Chaplinsky and Haushalter (2007)’s sample, and play little role in our sample.
9
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
between the issuer and investor(s). Placements for which we cannot find detailed information on
investors are excluded (676 of the initial 2,652), leaving us with a final sample of 1,976
placements.
3.2. Classifying relationships associated with private placements
3.2.1. Pre-placement relationships
We classify relationships based on the closeness of the investor to the firm’s business and
governance activities. Investors with no prior relationship with the issuer are outside investors
(OUT).7 Investors with a prior relationship with the issuer are assigned to one of three categories:
managers (MGT), key business partners (KEY), and governance-only investors (GOV).
Managers are managers of the issuing firm prior to the placement. Key business partners are
strategic alliance partners, customers and/or suppliers of the issuer prior to the placement.
Governance-only investors have only board representation and/or ownership of a 5% or greater
block prior to the placement. Managers and key business partners may also have board seats
and/or significant stock ownership. This is captured by other variables.
Relationship duration is defined as the log transformation of one plus the length of the
relationship in years. For MGT, we use years employed collected from career biographies in
Form 10-K and proxy statements. For KEY, we use the duration of the business relationship
collected from the related party transactions section of annual reports and proxy statements,
and/or the earliest press report of a business relationship in the company news file of the LexisNexis database.
For GOV, we use the duration of uninterrupted service as a director or
consecutive years as a 5% or greater blockholder. We also measure relationship scope, defined
as a count of different types of pre-placement relationships between the issuer and a relationship
investor. Scope includes four relationship types: employment, key business relationships, 5% or
greater blockholdings and whether or not a pre-placement directorship was held.
Equity ownership and directorship data are collected from the Compact Disclosure database,
proxy statements, Form 10-K, and insider-trading database including Thomson Financial Insiders
Filing, the Summary of Insider Transactions and the Insider file of the Lexis-Nexis database.
Recognizing the measurement error issues that arise in Compact Disclosure data, we use SEC
7
As referenced earlier, Brophy, Ouimet and Sialm (2007) focus on hedge fund versus non-hedge fund investors.
While hedge funds would generally by classified as outside investors under our scheme, as these authors also note,
they do not comprise the majority of outside investors. Brophy, Ouimet and Sialm (2007) show that hedge funds
account for only 16% of investments in traditional private placements. Percentages for our sample are smaller.
10
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
filings, when available, as our primary source of ownership data (Dlugosz, Fahlenbrach,
Gompers, and Metrick (2006)).
3.2.2. New post-placement relationships
To capture post-placement relationships, we collect data from the placement agreement and
follow investors and issuers for three years following completion to determine whether or not a
new relationship is formed. New relationships fall into 4 categories: i) managerial appointment
(investors or their affiliates become managers at the issuing firm), ii) new key business
partnership (investors become new key business partners), iii) director appointment (investors or
their assignees gain new seats on the issuer’s board) and/or iv) an investor’s purchase results in a
new block.8
3.3. Construction of other variables
In addition to the variables as described above, the following variables are employed in our
analysis. Following Berger, Ofek, and Yermack (1997), our two CEO incentive variables are
CEO tenure (measured as ln (1+years in CEO position)) and the CEO’s percentage share
ownership. Our specific investment hypotheses predict that issuers in new economy industries
and innovators are more likely to issue to relationship investors than other issuers. To test this,
we construct two proxies for innovation and technology intensity i) a new economy indicator
variable (New economy) equal to one for issuers in computer-related and telecommunications
industries and zero otherwise,9 and ii) the number of patent applications made by the firm over
the three years following the private placement (Patenty+3). Patenty+3 captures both industryspecific and time-varying components of technological advancement.
Our proxy for firm specific risk is the standard deviation of the residuals of an issuer’s daily
stock returns purged of market risk as measured by the Fama French three factor model (1993)
and of industry risk as measured by residuals of daily industry stock returns (based on two-digit
SIC code) regressed on Fama French’s three factors (1993). Some models include average
trading volume (VOL) over shares outstanding two years prior to the placement as a proxy for
the information environment. We use data on the number of investors and the number of shares
purchased by each investor to compute a placement Herfindahl index. This index captures
8
In the case where the definition of dominant investor is an investor who buys a 5% or more of the shares
outstanding as part of the placement, the creation of a new 5% or greater block characterizes situations in which the
dominant investor was not previously a 5% or greater blockholder. If the dominant investor was a 5% or greater
blockholder prior to the placement, we do not consider a new block to have been created.
9
The specific SIC codes are 3570, 3571, 3572, 3576, 3577, 3661, 3674, 4812, 4813, 5045, 5961, 7370, 7371, 7372,
or 7373 (following Murphy (2003)).
11
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
purchasing concentration among investors, squaring the fraction of a placement purchased by
each investor, then summing the squared fractions across investors.
We use three measures of issuer performance: announcement period abnormal returns,
longer-run post-placement stock returns, and industry-adjusted profitability two years postplacement. Announcement period abnormal returns are based on market model regressions of
firm stock returns on the CRSP equally weighted market index. The estimation period for the
market model is from day -260 through day -11 relative to announcement (day 0). Day 0 is the
Lexis-Nexis news date or, if that is unavailable, the private placement contract date. The
announcement period spans days -3 to 0. The variable is labeled CAR_MM day (-3,0).
For comparability with prior work, we present two measures of post-placement abnormal
returns, and in addition, compute several more as robustness checks. The measures we present
are CAR_MM days (1,120), which is the cumulative market model abnormal return for an
approximately 6 month period spanning days 1 to 120 relative to announcement. It is computed
as described in the preceding paragraph. We also compute matched-firm abnormal return, where
the match is based on industry, size, book-to-market and momentum (or prior returns). This
return is computed over 36 months following announcement. Our method follows that of Barber
and Lyon (1997), and is similar to the match-adjusted return methods applied in Brophy, Ouimet
and Sialm (2007), Chaplinsky and Haushalter (2007) and Hertzel, Lemmon, Linck and
Rees (2002). This variable is labeled CAR_ISBM (0,35).10 For robustness, we compute median
CAR_ISBM (0,35)s, two-year match-adjusted abnormal returns and implied two- and three-year
abnormal returns based on alphas from Fama-French three and four factor calendar time models.
Results are qualitatively similar to those presented here.
For industry-adjusted profitability, the industry benchmark is median operating profit for all
firms in the same three-digit Compustat SIC classification as the issuing firm. If fewer than 10
firms populate a three-digit SIC code for the relevant year, we use median operating profit for all
firms in the issuer’s two-digit SIC classification.
10
More specifically, we classify companies into the 48 Fama and French (1997) industries. When then rank
companies in each industry by their market capitalization measured at the end of the June prior to the private
placement, their book-to-market ratio (for private placement announcement dates between July of year t and June of
year t + 1 (the book value of equity is measured at the fiscal year-end of calendar year t − 1, and the market value of
equity calculated in December of year t − 1), and return during the prior 12 months. We match each sample firm
with another company in the same industry that did not have a private placement in the previous three years (since
we measure three years of post-placement performance). The matching firm is the firm with the smallest sum of the
absolute deviation of size, book-to-market, and prior return rankings relative to the sample firm. If size, book-tomarket, or prior returns are missing, we match the sample firm to another company in the same industry group with
missing size, book-to-market, or prior returns.
12
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
3.4. Characteristics of private placements and investors
Table 1 presents summary statistics. Panels A and B show that overall, the characteristics of
our sample are similar to those used in prior work. We have a larger sample, however, and a
larger number of small placements. In total, our sample firms raised $29.6 billion. Gross
proceeds average $15.3 million, with a much smaller median of $3.0 million (panel A). Average
block size is 11.4% of shares outstanding. Mean offer price represents an 11.3% discount from
market price ten days after announcement. Given the well-documented positive average response
to announcement (e.g., (Wruck (1989) and Hertzel and Smith (1993)), this measure overstates
the discount relative to a pre-announcement benchmark (Meidan (2006)). Measured relative to
closing price on the day prior to announcement, the average discount is 6.8%.
Table 1, panel B classifies private placements by the number of participating investors. Well
over half (65%) are sold to a single investor, with transactions involving a single investor raising
a total of $15.9 billion or 54% of sample gross proceeds. Roughly 85% of private placements are
sold to fewer than 5 investors and represent 66% of sample gross proceeds. Only 15% of
placements are sold to 5 or more investors. Prior studies also find that most private placements
are sold to a small number of investors.11
Table 1, panel C categorizes private placements based on our investor and relationship data.
We find that relationships between issuers and investors are common in private placements. The
vast majority of placements involve at least one type of relationship investor; only 36% of
placements involve an outsider. The remaining 64% involve only relationship investors. Eightysix percent of placements involve at least one relationship investor. This is surprising given that
prior work paints a picture of few pre-placement relationships. For example, in Krishnamurthy,
Spindt, Subramaniam and Woidtke (2005)’s sample, 33% of issuances (131 out of 397) are
classified as placements to affiliated investors.
Table 2 presents pre-placement relationship characteristics, purchasing patterns across
investor types, and post-placement relationship characteristics. Prior to the placement, across all
investor types, mean (median) pre-placement equity ownership is 4.8% (0.0%), with 12.3%
holding board seats pre-placement. Post-placement mean (median) equity ownership is 6.7%
(3.6%), with 7.0% gaining directorship rights as part of the placement agreement. Relationship
investors hold larger equity stakes both before and after private placements than outsiders.
11
For example, Wruck (1989) reports that 58% of private placements in her sample are sold to a single investor,
with very few placements sold to more than 6 investors.
13
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Among relationship investors, GOV investors hold the largest stakes, with 11.8% pre- and 14.9%
post-placement, on average. Outsiders hold only .24% on average pre- and 3.7% post-placement.
Not surprisingly, among relationship investors, managers have the longest duration and
greatest scope. Duration is 1.33 (2.8 years) on average for managers compared to 0.54 (0.7 years)
for key business partners, and 1.06 (1.9 years) for governance-only investors. The scope variable
shows that on average (at the median) managers have 2.1 (2.0) relationships with the firm, while
key business partners have 1.6 (1.0). GOV investors have 1.3 on average and 1.0 at the median,
suggesting that either a directorship or a block, but not both, is the typical governance-only
relationship.
Purchasing patterns suggest that relationship investors offer issuers the benefit of increased
access to capital. Relationship investors are associated with larger dollar purchases than outside
investors, particularly KEY ($11.8 million) and GOV ($5.3 million) versus OUT ($2.4 million).
Relationship investors also purchase a larger fraction of issuance and a larger percentage of the
total shares outstanding. Again, KEY is the largest both in terms of block size (8% on average)
and fraction of issued shares purchased (79% on average). This suggests a benefit analogous to
the benefits provided to debt issuers who borrow from banks with which they have an
established relationship. For example, Hoshi, Kashyap, and Scharfstein (1990) show that
relationship banks relax credit constraints for firms, and Petersen and Rajan (1994) and Berger
and Udell (1995) find that close ties to banks are associated with an increased availability of
financing.
4. Relationships and issuer performance
4.1. Defining dominant or lead relationship type
For the remainder of the paper, the placement serves as our unit of analysis. Recall that 65%
of our sample placements are made to a single investor (see table 1, panel B). For these
placements, analyzing the relationship between issuer and investor at a placement level is
straightforward. However, 35% of placements in our sample have multiple investors. These
investors may be of different types and further may have differing relationships with the issuer.
To conduct our work with the placement as the unit of analysis, we examine all investors
involved in each placement to determine a “dominant” or “lead” relationship type. We adopt
two definitions of dominant investor. First, a relationship type is dominant if that type purchases
more than 50% of the shares offered in the placement. There are 1,571 placements with a
dominant investor under this definition. Second, a relationship type is dominant if it purchases a
14
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
block of stock ! 5% of shares outstanding in the placement. There are 864 placements with a
dominant investor under this definition. Note that under both definitions, some placements do
not have a dominant investor. These placements are excluded in some subsequent analyses. Also
note that under the second definition, a single placement can have multiple dominant investors.
This occurs in less than 7% of placements. When it does, we assign the relevant placement to
each of the relevant investor categories.
Again, we find evidence that relationships between issuers and investors are more prevalent
than suggested by prior work. For example, Barclay, Holderness, and Sheehan (2007) classify
only 12% of their sample placements (70 of 594) as having “active” investors. In contrast, of the
investors involved in our sample transactions, 48% (880 of 1,818) form new post-placement
relationships with the issuer; 3% (52 of 1,818) involve a new management position or business
partnership, 22% (406 of 1,818) gain a new directorship, and 38% (689 of 1,818) become a new
5% or greater blockholder.
4.2. Relationships and Issuer Stock Returns
Table 3, panel A presents stock-price performance for all placements, whether or not they
have a dominant investor. The mean CAR_MM for days (-3,0) for all placements is 2.02%
(p=.00) and the mean CAR_MM for days (1,120) is -3.31% (p=.08). Both are similar to findings
in prior studies. The average three-year match-adjusted return is significantly negative at
-25.27% (p=.000), and is also similar to findings in prior studies. For example, the three-year
match-adjusted return is -23.78% in Hertzel, Lemmon, Linck and Rees (2002) and -38.39% in
Krishnamurthy, Spindt, Subramaniam and Woidtke (2005).12
Results based on the formation of new relationships are striking and confirm the validity of
our research design, which distinguishes between pre-existing and new relationships. Placements
without a new relationship are a non-event at announcement (CAR_MM (-3,0)=0.63, p=.15) and
are followed by significant, negative post-event abnormal returns (CAR_MM (1,120)=-6.2%,
p=.03, CAR_ISBM=--38.13%, p=0.00). Placements with a new relationship have a CAR_MM (3,0) of 3.51% (p=0.00) and insignificant post-event abnormal returns. These findings are
consistent with new relationships creating value by facilitating the resolution of information or
agency problems. These findings also suggest that investors not forming new relationships are
12
Other post-placement returns include Brophy, Ouimet and Sialm (2007) with -21.25% match-adjusted buy-andhold returns for day 6 to 500, Chaplinsky and Haushalter (2007) with -29.9% match-adjusted buy-and-hold returns
for month 0 to 24, Barclay, Holderness, and Sheehan (2005) with -9.4% CAR (-1, 120), and Meidan (2006) with
returns ranging from -3.2% to -33.8% over the six months following announcement depending upon the placement’s
price premium.
15
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
systematically disadvantaged; they invest in relatively weak firms whose problems surface postplacement.
For brevity, we do not tabulate the findings for placements with versus those without
dominant investors. Examining these placements, however, shows that announcement CARs and
post-placement abnormal returns are higher for placements with dominant investors than for
those without dominant investors. The pattern of abnormal returns for placements with and
without new relationships is qualitatively similar to that of the overall sample for placements
with and without dominant investors under both definitions of dominant investors.
Table 3, panel B reports abnormal returns for private placements in which over 50% of the
issue is sold to a single investor. Table 3, panel C reports results for placements in which a 5% or
greater block is issued to a pre-placement relationship type. Results are presented by preplacement relationship type and by the status of new relationships. In both panels, for all
relationship types, announcement period CARs are positive. The pattern of performance is
similar for both definitions of dominant investor, however, the abnormal returns are consistently
more positive under the ! 5% block definition. This is not surprising. In the median placement, a
block representing 7.2% of total shares outstanding is issued (see table 1, panel A). Thus, the
purchase of 50% of shares issued in the median placement represents a block of roughly 3.6%:
an event of lesser impact than the purchase of a 5% or greater block.
Mean CARs do not support hypotheses predicting a stronger announcement period response
for placements to relationship investors versus outsiders, and vice versa. The average
announcement CAR for placements to relationship and outside investors are 2.15% and 2.03% in
panel B, and 3.29% and 3.76% in panel C. The differences are not statistically significant.
Consistent with that notion that private placements help bond specific investments, placements to
key business partners have the largest announcement response (2.38% in panel B, and 4.11% in
panel C). The next largest response is to placements dominated by outside investors (2.03% in
panel B, and 3.76% in panel C). Placements to management investors have a marginally
significant CAR (-3,0) in panel B (1.87%, p=0.093) and an insignificant CAR (-3,0) in panel C
(2.92%, p=0.180).13 Similarly, announcement period CARs for placements to GOV investors are
13
This is consistent with Barclay, Holderness, and Sheehan (2007), who find that the average announcement period
CAR for placements to managers is insignificant.
16
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
insignificant in panel B and marginally significant in panel C (2.91%, p=0.068).14 Taken
together, the relatively positive CARs for outside investors are inconsistent with an information
advantage for relationship investors, with the exception of key business partners.
Examining the role of new relationships confirms the result that placements without new
relationships are “non-events.” Across all categories in Panel B, placements without new
relationships have an insignificant CAR (-3,0). In contrast, with the exception of placements to
managers, all placements associated with new relationships have significant positive abnormal
returns at announcement.
The relatively positive announcement response to placements associated with new
relationships is driven by placements in which the investor gains directorship rights and
placements that establish a new 5% or greater block. These findings are consistent with new
governance relationships reducing agency costs and thus increasing firm value. The findings also
suggest a variation on our information hypotheses, in which new rather than old, relationships
create informational advantage or send a positive signal to the market.
Finally, consistent with our prediction, the weakest announcement period CAR occurs in
placements to outside investors in which no new relationship is formed. The mean CAR (-3,0)
for that group is actually negative, but is not significant (-0.45%, p=0.67).
Table 3, panel B shows weak post-placement performance for several categories, including
relationship investors overall (CAR_MM (1,120)=-4.25%, CAR_ISBM=-25.01%), and outsiders
(CAR_MM (1,120)=-6.85%, CAR_ISBM=-26.82%). Panel C, however, shows that there is no
post-event “reversal” for placements with a dominant investor who purchases a 5% or greater
block.
Examining results by relationship type, shows that poor post-event performance persists only
for placements in which no new relationship is formed. Placements to relationship investors with
no new relationship experience a -6.50% six-month abnormal return and a -29.05% three-year
abnormal return. Outsiders experience significantly poor post-event performance as well. This is
consistent with an informational disadvantage and/or inability to access governance on the part
of outside investors.15 Key business partners with no new relationships also have particularly
14
This is consistent with Allen and Phillips (2000) who find no significant difference between the announcement
response to placements where the investors have a directorship and placements in which the investor has no
directorship.
15
A potential concern regarding our interpretation of these findings is that outside investors may protect themselves
from poor performance through contractual provisions such as price reset clauses. If this were the case, the negative
17
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
poor performance (CAR_MM (1,120)=-14.32%, CAR_ISBM=-38.05%). The latter finding is
consistent with over-optimism on the part of key business partners. It is also consistent with the
notion that key business partners are “tied to the mast,” in the sense that investing is a lesser evil
than not investing.16
Among related prior findings is Krishnamurthy Spindt, Subramaniam, and Woidtke (2005)’s
result that post-placement abnormal returns for placements to affiliated investors is -36.94%
(insignificant) and that the analogous return for unaffiliated investors is -47.05% (significant).
However, their data do not allow them to identify dominant investors as we have here. Thus,
their affiliated classification includes all placements in which Lexis-Nexis reports that an
affiliated investor participated, regardless of the size of that investor’s participation. Similarly,
their unaffiliated classification includes only placements in which all investors are reported to be
unaffiliated. This reduces the power of their tests. Additionally, they do not examine
performance differences across types of affiliations, which, as we show here, are critical.
Other related prior findings are found in Barclay, Holderness and Sheehan (2007). They
measure market model CARs from days -1 to 120, and find abnormal returns of 16.8% (p=0.14)
to management investors, 3.6% (p=0.34) to active investors and -13.2% (p=.00) to passive
investors. Notice, however, that our data and resulting relationship classifications allow for more
effective assessment of where and why poor performance occurs. For example, key business
partners are likely not best thought of as passive investors, but rather as active investors who
invest in firms that perform poorly ex post.
We also analyze announcement and post-announcement period abnormal returns for publicly
traded dominant investors (not tabulated here). The lack of significant abnormal returns around
announcement suggests either that the benefits associated with private placements are not shared
with investors (analogous to a target firm’s shareholders capturing the benefits from a merger or
acquisition (e.g., Jensen and Ruback (1983), Jarrell, Brickley, and Netter (1988), and Andrade,
Mitchell and Stafford (2001)), or that the benefits (or costs) that accrue to the investor are too
small to be measurable. Consistent with Allen and Phillips (2000), announcement period CARs
are insignificant. Further, there is no evidence of post-placement abnormal performance for
returns we observe would understate the actual returns to outside investors. As stated earlier, such provisions are
not part of the more traditional placement contracts that comprise our sample.
16
Panel C shows a significant negative CAR_ISBM for key business partners that form new relationships of
-35.54%. Exploration of this finding shows that it is driven by outliers. Further confirming this, when median
CAR_ISBMs are examined, this result disappears; there is no significant return to this group of investors and the
return to key business partners without new relationships is significantly worse.
18
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
publicly traded dominant investors. This result is consistent with the relationship lending
literature. For example, Bharath, Dahiya, Saunders, and Srinivasan (2007) find that the benefit
from relationship lending accrues largely to the borrowing firm, not the lender.
4.4. Relationships and post-placement profitability
Prior work documents weak profitability following both public offerings and private
placements (e.g., Loughran and Ritter (1997) and Hertzel, Lemmon, Linck, and Rees (2002)).
We are not aware of prior work that examines post-placement profitability based on
relationships. Thus, below we examine whether post-placement profitability differs based on the
relationship between the issuer and investors. Our profitability measure is industry-adjusted
operating profit two years following the placement.17
Table 4 presents median post-placement industry-adjusted operating profit. Results for the
overall sample are presented in panel A. Panel B presents results for placements to dominant
investors who purchase more than 50% of the shares issued. Panel C presents results for
dominant investors who buy a 5% or greater block in the placement. Consistent with prior
findings, median profitability is negative for the overall sample (panel A, -0.150, p=0.000). For
example, Hertzel, Lemmon, Linck and Rees (2002) find industry-adjusted profitability of -0.107
and -0.086 for post-placement years 1 and 2, respectively.
Firms making placements to relationship investors have stronger post-placement profitability
than firms making placements to outsiders (-0.123 vs. -0.165 in panel B, and -0.074 vs. -0.128 in
panel C). In panel B, post-placement profitability is significantly negative for all types of
relationship investors. In panel C, it is significantly negative for KEY and GOV investors, but
not for managers.
Consistent with our stock performance findings, placements with new relationships are more
profitable than placements without new relationships. While industry-adjusted profitability is
negative for all groups, in panel B, profitability is significantly more negative for placements
without new relationships. Panel C shows a fairly similar pattern (with the exception of MGT
investors) but the differences are not significant.
Loughran and Ritter (1997) interpret deteriorating operating performance following seasoned
equity offerings as consistent with the hypothesis that poor post-offering stock price performance
reflects the market’s disappointment about post-placement earnings. Our findings are broadly
17
We also conduct an analysis of the change in profitability for each sample firm from one year pre- to two years
post-placement. The findings from this analysis are qualitatively similar to those presented here.
19
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
consistent with this interpretation. Issuers making placements to management investors
experience both normal post-placement returns and industry-adjusted operating profit. This
provides support for the hypothesis that managers participating in private placements have an
information advantage. Issuers making placements that involve new relationships have no
unusual post-placement stock price performance and have industry-adjusted profitability that is
less negative than other issuers. Issuers making placements to governance only investors with no
new relationships constitute an exception. Such placements have insignificant average long-run
abnormal returns, but negative industry-adjusted profitability.
4.5. Cross-sectional analysis of abnormal returns at announcement
Table 5 presents results from regressions of CAR (-3,0) on relationship, placement and issuer
characteristics. Placements lacking information sufficient to classify them as having or not
having a dominant investor are excluded. Odd numbered models ((1), (3), (5) and (7)) include all
placements with sufficient investor information. Even numbered models ((2), (4), (6) and (8))
include only placements with a dominant investor under each of the two definitions (a single
investor type purchases >50% of shares issued (models (2) and (4)) and a single investor type
purchases a ! 5% block of shares in the placement (models (6) and (8)).
We estimate each model using OLS regression. We also use three-stage least-squares (3SLS)
because placement characteristics, such as directorship rights, pricing and announcement period
returns may be jointly determined. It is potentially problematic, therefore, to use price discount
and added directorships as explanatory variables in our models. For example, issuers and
investors might anticipate the market’s reaction to announcement and incorporate that into price
negotiations, so that price discount affects announcement returns and vice versa. There is, to our
knowledge, no generally accepted causal model linking private placement contracting and deal
terms with announcement effects. So, we start by assuming that the right to a board seat
influences pricing. In turn, the information content associated with directorship assignment rights
and price discount influences announcement period CARs. Our 3SLS approach is to first
estimate a model of added directorships using exogenous variables. Then we estimate a model of
the Winsorized price discount using predicted directorship rights as one of the explanatory
variables. The final step is to simultaneously estimate CAR regression models using predicted
added directorships and predicted Winsorized price discount as two of the explanatory variables,
20
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
and re-estimate directorship and Winsorized pricing models using seemingly unrelated
regression (Zellner (1962)).18
Table 5 presents our regression results. Models (1), (2), (5) and (6) are OLS models and the
remaining models are 3SLS. After controlling for firm and placement characteristics, differences
in the announcement response based on relationship characteristics are significant. Odd
numbered models show that the market responds more positively to placements with dominant
investors. The dominant investor coefficient is positive and significant in all odd numbered
models. The magnitude of the coefficients suggests a 2.5% to 5% higher CAR (-3,0) for
placements with dominant investors, and is consistent with these investors playing a role in
resolving information and/or incentive problems.
In addition, post-placement relationship variables have significant positive coefficients. For
models (1) and (3), the coefficients for variables indicating the purchase of a 5% or greater block
and added directorships are positive and significant. The magnitude of these coefficients
suggests a 1.7% to 2.3% higher CAR (-3,0) for purchases of a 5% or greater block and a .74% to
2% higher CAR (-3,0) for placements in which directorships are added. (The 5% or greater block
variable is not relevant for models (5) and (7).) The added directorship variable is significant and
positive in these models as well, with the coefficient ranging from 1.16% to 1.60%.
Even numbered models include only placements with dominant investors. In models (2), (4)
and (8) the coefficients for KEY investors are positive and significant. Their magnitude indicates
an approximately 1.6% to 5.9% higher CAR (-3,0) for placements made to this investor type. In
model (2), the coefficient for GOV investors is also positive and significant at 1.848. Neither of
these investor types is significant in model (6). In that model, however, the coefficient for
management investors is significantly negative. Further, the magnitude of the coefficient
indicates a 4.31% lower CAR (-3,0) for these placements. This is consistent with our univariate
findings. Finally, added directorship(s) have a significant, positive coefficient in all even
numbered models, with coefficients ranging from 0.75% to 2.64%.
To test for CEO incentive effects, we include CEO shareholdings and CEO tenure as
explanatory variables. CEO shareholdings are not significant in any of the models. CEO tenure is
insignificant in most models, but is significantly negative in models (6) and (8). Recall that these
18
We have also estimated these models using recursive median regressions in which we first estimate the director
assignment model, then use its predicted value in the pricing model and finally use both predicted values in our
CAR regression models. While results are not dissimilar to those presented here, the recursive estimation technique
implies a unidirectional (one-way) relationship among the three models and so does not take explicitly account for
interdependence among the three models. To conserve space, we do not present these models.
21
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
regressions include only placements with dominant investor types who purchase a 5% or greater
block. The negative coefficient for CEO tenure in these models is associated with a 1.23% to
1.29% smaller CAR (-3,0).
To test hypotheses regarding specific investment, we include the new economy and patent
variables. We also include our proxy for firm specific risk. The coefficient for the new economy
dummy variable is positive and significant in models (1), (2) and (5), but is significantly negative
in (8). The positive coefficients are consistent with private placements encouraging specific
investment. The negative coefficient in model (8), however, is not. Perhaps placements of 5% or
greater blocks for new economy firms signal future prospects that are sufficiently bleak to rule
out a public offering. The patent variable is not significant in any models. The coefficient for
firm specific risk is significantly positive for all models, with the exception of model (5). This
suggests that placements made by firms with high specific risk are relatively good news.
Coefficient estimates for other placement characteristics and control variables are of mixed
significance.
In addition to models presented above, we explored a variety of alternative specifications, not
reported here. They yield similar findings.19
5. Relationships, placement structure and issuer characteristics
5.1. Probit analysis of pre-placement relationships
To test hypotheses about how placement structure and firm characteristics are related to preplacement relationships, we estimate a series of probit models. The dependent variable in each
model equals one for a particular investor type and zero for an alternative type. Thus, the models
estimate the probability of “choosing” one investor type over another. We use the term
“choosing” broadly, understanding that the placement process is one of searching and matching
on both sides and that firms are not unconstrained in their options for issuing securities.
19
We classify dominant investors as relationship investors or outsiders and use this classification in our model. The
coefficient for relationship investors is significantly positive for definition 1 and insignificant for definition 2. We
replace investor types with scope and length of relationship. Consistent with the relationship banking literature
(Berlin and Mester (1999), and Degryse and van Cayseele (2000)), results show a positive relationship between
announcement return and scope, suggestive of a synergy effect of scope whereas the opposite holds for duration.
We replace the new 5% or greater block variable with three variables that separate changes in ownership into three
parts (less than 5%, between 5% and 25%, and greater than 25%). The results show a positive relation between
CAR (-3,0) and ownership concentration changes for investors with small pre-placement ownership (less than 5%)
and greater than 25% post-placement ownership. However, the CAR (0,3) is small if both the pre- and postplacement ownership are within the middle range of ownership (5% to 25%). We have also conducted median
linear regression analysis, which is robust to potential outliers and fat-tailed announcement period return
distributions. The results are qualitatively similar to those presented here.
22
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Table 6 presents five models in which dominant investor is defined as an investor purchasing
more than 50% of the shares issued in a placement. An indicator variable captures placements in
which a 5% or larger block of shares is purchased. Model (1) compares placements with and
without dominant investors. Model (2) compares placements with dominant relationship
investors to those with dominant outside investors. Models (3), (4) and (5) compare each
dominant relationship type (MGT, KEY and GOV) with dominant outside investors.
The models include three explanatory variables to capture post-placement relationships and
profitability: indicator variables for a new large block and for added directorships, and postplacement industry-adjusted profit. In models (2) through (5), the coefficient for new large
investor is significantly negative, while the coefficient for added directorships is significantly
positive; placements to relationship investors are less likely to be associated with a new block
and more likely to be associated with added directorships than placements to outsiders.
(Model (1) has a positive coefficient on new large investors, but that is mechanical.)
In models (1) to (4), the coefficient on future profitability is significantly positive, indicating
that relationship investors invest in issuers with higher future profitability. This is consistent with
the hypothesis that relationship investors have better access to information than outsiders, with
the superior information reflected in post-placement profitability. It is also consistent with our
univariate findings.
The models include two variables to test whether strong CEOs encourage or discourage
relationship investing: CEO share ownership and CEO tenure. Coefficients for these variables
are negative, but the coefficient for CEO share ownership is significant only in models (1)
and (5). The coefficient for CEO tenure is significant in models (2), (4) and (5). Recall that if
relationships entrench managers, managers with long tenure and/or high share ownership are
more likely to “choose” relationship investors. Negative coefficients are not only inconsistent
with this hypothesis; they indicate the opposite. Managers with long tenure and larger
shareholdings are more likely to issue to outsiders.
The models include five firm characteristics: an indicator variable for “new economy firms,”
a patent variable (Patenty+3), firm specific risk, average share trading volume as a percentage of
shares outstanding, and market-to-book. New economy, patent and market-to-book proxy for the
extent to which innovation and growth-options are important. They also serve as proxies for
information problems, as does volume. Some of these variables have significant coefficients. In
models (2) and (3), patent has a significant, positive coefficient. This is consistent with
23
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
innovative firms resolving information problems by “choosing” relationship over outside
investors.
Firm specific risk has a significant negative coefficient in model (4)—KEY versus OUT.
This is consistent with key business partners, who are already bearing firm specific risk, being a
less efficient “choice” than outsiders for firms with high specific risk. Volume has a significant
negative coefficient in model (5)—GOV versus OUT. This is consistent with firms facing
information problems choosing a governance-only relationship investor over an outsider.
Finally, the models include three placement characteristic variables: log of issue size, a
placement Herfindahl index and price discount. With the exception of model (3), the coefficient
for issue size is positive and significant. With the exception of managers, dominant relationship
investors participate in larger placements than outsiders. The significant positive coefficient for
placement Herfindahl in models (2) and (4) indicates that placements to relationship investors,
particularly key business partners, are more concentrated than placements to outsiders.
Consistent with information rather than agency-based hypotheses, shares are issued to
relationship investors at a smaller discount than shares issued to outsiders (models (2) and (4),
coefficients of -.002 and -.004, respectively). In contrast, management investors receive a higher
discount than outsiders (significant positive coefficient of .005).
In addition to models presented above, we explored a variety of alternative specifications, not
reported here. They yield similar findings.20
5.2. Relationships and influence on issuer governance
Table 7 presents models of the association between the governance rights granted to
placement investors, relationships and firm and placement characteristics. For 3SLS models,
numbers correspond to the model numbers in tables 5 and 8, indicating they are estimated
simultaneously. The dependent variable is ln ((1+ added directorships)/board size). Added
directorships is the number of director appointments following the placement. Board size is the
total number of directors on the board in the placement year. Again, odd numbered models
include all placements and even numbered models include only placements with dominant
investors.
20
We substitute ln(firm size) for ln(issue size). We use alternative measures for post-placement performance:
change in industry-adjusted operating profit, buy and hold returns and CARs. Change in operating profit has
positive, significant coefficients in most models. Post-placement stock returns are not significant in any
specifications. We include (CEO_Share)2 and (ln(issue))2 to allow for nonlinearities. We also replace CEO_Share
with ln(CEO_Share/(1- CEO_Share)) to convert a bounded variable to an unbounded one (Demsetz and Lehn
(1985)). None of these changes have a significant impact on our findings.
24
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Even numbered models consistently show that placements to MGT and KEY investors are
associated with stronger additional board influence than placements to other dominant investors.
Coefficients for MGT and KEY relationship variables are positive and highly significant. The
pre-existing directorships variable has a significant negative coefficient, indicating a lower
likelihood of additional directorships for this group of investors—perhaps because investors
already holding directorships do not desire additional board influence. The coefficient for preexisting large investors variable is significantly positive in most models, suggesting that the
placement helps blockholders gain additional governance influence. New post-placement
relationships (including new management positions, new key business partnerships and new 5%
or greater blocks) are positively associated with increased board influence, suggesting synergies
between different types of post-placement relationships.
Proxies for CEO incentives allow us to test whether strong CEOs encourage or discourage
increased board influence by dominant investors. The coefficients for CEO share ownership are
consistently positive and significant. This is not consistent with entrenchment. Rather, it suggests
that CEOs with higher share ownership encourage placement investors to take on governance
roles. CEO tenure is insignificant.
Issuers in new economy industries and with high firm specific risk are positively associated
with added directorships. This suggests that governance relationships between issuers and
investors are relatively valuable for these firms, perhaps due to the value of specific investments
or the value of access to governance when the potential for information problems is high.
Not surprisingly, block size is positively associated with added directorships in all models,
indicating that the larger the percentage of total shares issued the greater the likelihood of added
directorships. Added directorships are also positively associated with a price discount, consistent
with compensation for monitoring and/or reducing agency costs.
5.3. Pricing discounts and relationships
Prior work provides some evidence on the association between price discounts and investor
type. For example, Barclay, Holderness, and Sheehan (2007) find that price discounts are smaller
in placements to active versus passive investors (1.8% versus 20.8%). In contrast, Meidan (2006)
finds that investors who gain board representation receive a higher discount than other investors.
Below, we exploit our unique data to examine the association between price discount,
relationships and other placement and firm characteristics in further depth.
25
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Table 8 presents results of price discount regressions. For 3SLS, model numbers correspond
to the model numbers in tables 5 and 7, indicating they are estimated simultaneously. The
dependent variable is Winsorized price discount computed relative to closing price one day prior
to announcement. Again, odd numbered models include all placements and even numbered
models include only placements with dominant investors.
Among placements in which dominant investors purchase a 5% or greater block, managers
obtain a higher price discount. The magnitude of the coefficient is consistent with a 9% to 10%
additional discount for MGT investors. This result is not present when dominant investor is
defined as an investor purchasing more than 50% of shares issued. The higher discount could be
compensation to managers for bearing firm-specific risk or undertaking firm-specific
investments in human capital. It may also be the case that the price at which managers buy shares
is “too low.” The coefficient for KEY investors is negative and significant in all even numbered
models. Coefficient estimates range from -11.9% to -13.3%, showing that KEY investors receive
a significantly smaller discount than other investors types. The lower discount could be due to
the value of specific investments associated with the business relationship, or to avoid the
impression of self-dealing.
The new relationship variables are largely insignificant. The added directorships variable has
a significant positive coefficient in models (3), (4) and (7), but a significant negative coefficient
in model (8). CEO incentive variables are insignificant in all models.
New economy firms and firms with high specific risk sell shares at a larger discount
(coefficients are significant and positive in all models). For the new economy indicator,
coefficients range from 5.9% to 13.4%—a substantial increase in the discount. This is consistent
with extra discounting to compensate investors for helping resolve information problems,
engaging in monitoring or, in the case of specific risk, risk bearing. The coefficient on patents is
significantly negative in five of the eight models. In contrast to new economy firms, firms with a
relatively large number of patents issue shares at a smaller discount. Perhaps the patents signal
higher firm quality.
Larger dollar placements are associated with a smaller price discount in all models. Block
size is associated with a larger price discount in all models. Both these findings are consistent
with prior studies (e.g. Hertzel and Smith (1993), Barclay, Holderness, and Sheehan (2007), and
Meidan (2006)). The placement Herfindahl index has a significant negative coefficient in six of
26
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
the eight models, indicating that placements to a more concentrated investor group are sold at a
smaller discount.
We replace the investor types with relationship scope and duration variables. Our results are
mixed. Similar to Petersen and Rajan (1994) who find little impact of relationship length on loan
interest rate, we find an insignificant coefficient for both scope and duration in price discount
regressions. For investors purchasing more than 50% of shares issued, however, price discount
falls with duration and increases with scope. (Berlin and Mester (1999) and Degryse and van
Cayseele (2000)).
6. Conclusions
We began our paper with two important questions. First, how are private placement
investors involved in the affairs of the issuing firm? Second, why is the positive stock price
response at announcement followed by a larger post-placement “reversal”? Below, we
summarize the contribution of our findings toward the development of more complete answers to
these questions.
We present evidence that an overwhelming proportion of private placement investors are
actively involved as managers, key business partners and as participants in the issuer’s
governance processes. Many of those without pre-placement relationships form new
relationships following completion. Relationship investors in private placements play a role
analogous to that of relationship lenders. They provide access to more capital, gain greater
influence in the issuing firm’s governance than other investors, and participate in placements by
issuers for which information problems and specific investments are important.
Our analysis of performance shows that, for the most part, relationships (both pre-existing
and new) are associated with value creation—a positive stock price at announcement followed
by no unusual post-placement performance. In fact, the pattern of stock price performance
characterized by the second question is not representative of the pattern of performance for most
of our sample. In our sample, poor post-placement performance is confined to placements to true
outsiders, where there is no relationship, and placements to key business partners in which no
new relationship is formed. We hypothesize that the latter group is making a rational decision to
invest, despite the poor performance. However, digging more deeply into these transactions is an
interesting area for further exploration.
Current trends indicate continued expansion of the importance of private placements, which
are rapidly becoming a conventional source of financing for firms of all kinds, not just small firm
27
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
with information problems (see, e.g., Business Wire, October 3, 2006). Is a system developing
that is analogous to the powerful system of relationship banking in countries such as Germany
and Japan, only based on equity? Advocates of relationship banking assert that it has worked in
Germany and Japan, where companies benefit from long-term institutional holdings and are less
subject to short-term performance pressures. If private equity markets are moving in this
direction, continued exploration of the role of relationships in private equity markets will prove
an important area for future research.
28
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
References
Allen, Jeffrey W. and Gordon M. Phillips, 2000, Corporate equity ownership, strategic alliances,
and product market relationships, Journal of Finance 55, 2791-2815.
Barber, Brad M. and John D. Lyon, 1997, Detecting long-run abnormal stock returns: The
empirical power and specification of test statistics, Journal of Financial Economics 43, 341372.
Barclay, Michael J., Clifford G. Holderness and Denis P. Sheehan, 2007, Private placements and
managerial entrenchment, Journal of Corporate Finance 13, 461-484.
Berger, Allen N. and Gregory F. Udell, 1995, Relationship lending and lines of credit in small
firm finance, Journal of Business 68, 351-381.
Berger Philip G., Eli Ofek and David L. Yermack, 1997, Managerial entrenchment and capital
structure decisions, Journal of Finance 52, 1411-1438.
Berlin, M. and L. Mester, 1999, Deposits and relationship lending, Review of Financial Studies
12, 579-607.
Bhagat, Sanjai, Bernard S. Black and Margaret M. Blair, 2004, Relational investing and firm
performance, Journal of Financial Research 27, 1-30.
Bharath, Sreedhar T., Sandeep Dahiya, Anthony Saunders and Anand Srinivasan, 2007, So what
do I get? The bank’s view of lending relationships, Journal of Financial Economics 85, 368419.
Black Bernard S., 1992, Agents watching agents: The promise of institutional investor voice,
UCLA Law Review 39, 811-893.
Boot, Arnoud W. A., 2000, Relationship banking: What do we know? Journal of Financial
Intermediation 9, 7-25.
Brophy, David J., Paige P. Ouimet and Clemens Sialm, 2004, PIPE dreams? The performance of
companies issuing equity privately, NBER Working Paper.
Brophy, David J., Paige P. Ouimet and Clemens Sialm, 2007, Hedge funds as investors of last
resort? Review of Financial Studies, forthcoming.
Chaplinsky S. and David Haushalter, 2007, Financing under extreme uncertainty: Contract terms
and returns to private investments in public equity, Working Paper, University of Virginia.
Cronqvist, Henrik and Mattias Nilsson, 2005, The choice between rights offerings and private
equity placements, Journal of Financial Economics 78, 375-407.
Degryse, H. and P. van Cayseele, 2000, Relationship lending within a bank-based system:
Evidence from European small business data, Journal of Financial Intermediation 9, 90–
109.
29
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Demsetz, Harold and Kenneth Lehn, 1985, The structure of corporate ownership: Causes and
consequences, Journal of Political Economy 93, 1155-1177.
Dlugosz, Jennifer, Fahlenbrach R., Gompers Paul, and Metrick Andrew, 2006, Large blocks of
stock: Prevalence, size, and measurement, Journal of Corporate Finance 12, 594– 618.
Fama, Eugene F., 1985, What’s different about banks? Journal of Monetary Economics 15, 2939.
Fama, Eugene F. and Kenneth R. French, 1993, Common risk factors in the returns on stock and
bonds, Journal of Financial Economics 33, 3-56.
Grossman, Sanford J. and Oliver D. Hart, 1986, The costs and benefits of ownership: a theory of
vertical and lateral integration, Journal of Political Economy 94, 691-719.
Hart, Oliver D. and John Moore, 1990, Property rights and the nature of the firm. Journal of
Political Economy 98, 1119-1158.
Heaton, John C., 2002, Managerial optimism and corporate finance, Financial Management 31,
33-45.
Hertzel, Michael G. and Richard L. Smith, 1993, Market discounts and shareholder gains for
placing equity privately, Journal of Finance 48, 459-485.
Hertzel, Michael G., Michael Lemmon, James Linck and Lynn Rees, 2002, Long-run
performance following private placements of equity, Journal of Finance 57, 2595-2617.
Hoshi, Takeo, Anil Kashyap and David Scharfstein, 1990, The role of banks in reducing the
costs of financial distress in Japan, Journal of Financial Economics 27, 67-88.
James, Christopher, 1987, Some evidence on the uniqueness of bank loans, Journal of Financial
Economics 19, 217-235.
Jensen, M. C. and W. H. Meckling, 1976, Theory of the firm: managerial behavior, agency costs
and ownership structure, Journal of Financial Economics 3:305−360.
Kahn, Charles M. and Andrew Winton, 1998, Ownership structure, speculation, and shareholder
intervention, Journal of Finance 53, 99-129.
Klein Benjamin, Robert G. Crawford and Armen A. Alchian, 1978. Vertical integration,
appropriable rents, and the competitive contracting process. Journal of Law and Economics
21, 297-326.
Krishnamurthy, Srinivasan, Paul Spindt, Venkat Subramaniam and Tracie Woidtke, 2005, Does
investor identity matter in equity issues? Evidence from private placements, Journal of
Financial Intermediation 14, 210-238.
La Porta, R., Florencio López-de-Silanes and Guillermo Zamarripa, 2003, Related lending,
Quarterly Journal of Economics 118, 231-268.
Loughran, Timothy and Jay R. Ritter, 1997, The operating performance of firms conducting
seasoned equity offerings, Journal of Finance 52, 1823-1850.
30
Business Relationships, Corporate Governance, and Firm Performance
DRAFT: March 2008
Wruck and Wu
Lummer, Scott L. and John J. McConnell, 1989, Further evidence on the bank lending process
and the capital-market-response to bank loan agreements, Journal of Financial Economics
25, 99-122.
Malmendier, Ulrike and Geoffrey A. Tate, 2005, CEO overconfidence and corporate investment,
Journal of Finance 60, 2661-2700.
Meidan, Danny, 2006, The informativeness of offer characteristics versus investor identity in
pipe transactions, Unpublished working paper, Northwestern University.
McConnell, John J. and Henri Servaes, 1990, Additional evidence on equity ownership and
corporate value, Journal of Financial Economics 27, 595-612.
Murphy, Kevin J., 2003, Stock-based pay in new economy firms, Journal of Accounting and
Economics 34, 129-147.
Myers, Stewart C. and Nicholas Majluf, 1984, Corporate financing and investment decisions
when firms have information that investors do not have, Journal of Financial Economics13,
187-221.
Petersen, Mitchell A. and Raghuram G. Rajan, 1994, The benefits of lending relationships:
evidence from small business data, Journal of Finance 49, 3-37.
Raith, Michael, 2003, Competition, risk, and managerial incentives, American Economic Review
93, 1425-1436.
Raith, Michael, 2005, Specific knowledge and performance measurement, Unpublished working
paper, University of Rochester.
Rajan, Raghuram G., 1992, Insiders and outsiders: The choice between informed and arm'slength debt, Journal of Finance 47, 1367-1400.
Shleifer, Andrei and Robert W. Vishny, 1986, Large shareholders and corporate control, Journal
of Political Economy 94, 461-488.
Wruck, Karen H., 1989, Equity ownership concentration and firm value: evidence from private
equity financings, Journal of Financial Economics 23, 3-28.
Wong, M. H. Franco and Frank Zhang, 2007, Managerial optimism and analysts' forecast bias,
Unpublished working paper, University of Toronto.
Wu, YiLin, 2004, The choice of equity-selling mechanisms, Journal of Financial Economics 74,
93-119.
Zellner, Arnold, 1962, An efficient method of estimating seemingly unrelated regressions and
tests of aggregation bias, Journal of American Statistical Association 57, 348-368.
31
Table 1. Private placement characteristics.
Sample of 1,976 private placements of equity by U.S. public firms from 1980–1999. Block size is defined as shares
offered/(shares offered + outstanding shares), where outstanding shares are measured on the announcement date or, if
unreported, at the end of the quarter prior to the announcement. Issue discount is defined as (benchmark price less offering
price)/benchmark price. We use two benchmark prices: share price ten days after announcement and share price one day
prior to announcement. Types of pre-placement relationship are categorized as management relationship (MGT), key business
relationship (KEY), governance relationship (GOV), and outside relationship (OUT). Managers are individuals who are
managers of the firm prior to the placement. Key business relationship consists of strategic alliances, customers and suppliers.
Governance relationship is of directorship and/or blockholding (≧5%). Outside relationship refers to no documented preplacement employment, business, directorship, or blockholding relationship.
Panel A: Characteristics of private placement
Mean
Median
Standard deviation
N
Gross proceeds (millions)
$15.26
$3.00
$6.47
1,937
Block size (%)
11.39%
7.24%
13.40%
1,942
Issue discount (% relative to 10 days
post-announcement)
11.33%
10.96%
56.04%
1,854
Issue discount (% relative to 1 day preannouncement)
6.79%
7.80%
85.77%
1,854
Panel B: Number of investors participating in the private placement
1
2-4
5-14
More than 14
Number of transactions
(% of placements)
1,282
(65%)
398
(20%)
141
(7%)
155
(8%)
Number of investors
(% of investors)
1,282
(16%)
1,076
(13%)
1,046
(13%)
4,645
(58%)
$15,953
(54%)
$3,444
(12%)
$8,506
(29%)
$1,656
(5%)
Total gross proceeds (millions)
(% of sample gross proceeds)
Panel C: Types of investors participating in the private placement
Management
Key Business
Governance
Outside
Number of transactions
(% of placements)
376
(19%)
886
(45%)
439
(22%)
713
(36%)
Number of investors
(% of investors)
616
(8%)
913
(11%)
619
(8%)
5,901
(73%)
$1,698.145
(6%)
$10,622.992
(36%)
$3,219.148
(11%)
$14,018.235
(47%)
Total gross proceeds (millions)
(% of sample gross proceeds)
Table 2. Pre- and post-placement relationship characteristics and purchasing patterns across investor types.
Types of investors are categorized as management investors (MGT), key business partners (KEY), governance investors (GOV), and outside investors (OUT). Managers
(MGT) are individuals who are managers of the firm prior to the placement. Key business partners (KEY) are strategic alliance partners, customers and suppliers. Governance
investors (GOV) are existing directors and/or blockholders (≧5%). Outside investors (OUT) refer to investors with no documented pre-placement employment, business,
directorship, or blockholding relationship. We collect data on equity ownership, and directorships from Compact Disclosure, proxy statements, Form 10-K, and insider trading
database including Thomson Financial Insiders Filing, the Summary of Insider Transactions and the Inside file of the Lexis-Nexis database. Duration of a relationship is the
log transformation of one plus the length of the relationship in years. For MGT, we use years employed collected from career biographies in Form 10-K and proxy statements.
For KEY, we use the duration of the business relationship collected from the related party transactions section of annual reports and proxy statements, and/or the earliest press
report of a business relationship in the company news file of the Lexis-Nexis database. For GOV, we use the duration of uninterrupted service as a director or consecutive
years as a 5% or greater blockholder. Scope of a relationship is the number of existing relationships between the investor and the firm, including any employment relationship,
key business partnership, 5% or greater block owned pre-placement, and directorship(s) held pre-placement. Block size purchased is defined as shares offered/(shares offered +
outstanding shares), where outstanding shares are measured on the offering announcement date or, if unreported, at the end of the quarter prior to the announcement. Fraction
of total placement acquired is fraction of shares bought by investors in the indicated classification. Board seat right as part of placement agreement equals one for investors
with the right to appoint directors and zero otherwise. Added directorships is the number of directors appointments in the three years following completion by investors with
board seat rights. We use a t-test (Kruskal-Wallis test) to compare means (medians) between relationship investors and outsiders, and a one way ANOVA (Kruskal-Wallis
test) to compare means (medians) among MGT and KEY and GOV, and a chi-squared independence test to examine for independence between (among) type of investors and
the frequency counts in a two-way classification.
Pre-placement investor characteristics
Purchasing patterns of investors
Post-placement investor characteristics
Pre-issue
equity
ownership
(%)
Pre-existing
directorships
Duration
Mean
(Median)
[N]
Count
(Row %)
[N]
Mean
(Median)
[N]
Scope
Mean
(Median)
[N]
Amount
purchased
($ millions)
Block size
purchased
(%)
Fraction of Post-issue
total
equity
placement ownership
acquired
(%)
(%)
Mean
(Median)
[N]
Mean
(Median)
[N]
Mean
(Median)
[N]
Mean
(Median)
[N]
Board
seat
right as
part of
placement
agreement
Count
(Row %)
[N]
44.98
(33.33)
[6,512]
57.95
(61.79)
[2,148]
38.60
(14.00)
[4,364]
6.71
(3.63)
[6,588]
12.94
(6.71)
[2,148]
3.69
(1.50)
[4,440]
561
(6.97)
[8,049]
381
(17.74)
[2,148]
180
(3.05)
[5,901]
719
(8.93)
[8,049]
506
(23.56)
[2,148]
213
(3.61)
[5,901]
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
39.28
(19.50)
[616]
79.31
(100.00)
[913]
45.03
(30.20)
[619]
11.00
(3.60)
[616]
12.94
(6.25)
[913]
14.87
(10.43)
[619]
86
(13.96)
[616]
218
(23.88)
[913]
77
(12.44)
[619]
107
(17.37)
[616]
306
(33.52)
[913]
93
(15.02)
[619]
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Added
directorships
Count
(Row %)
[N]
Panel A: Pre- and post-placement characteristics and purchasing patterns of overall private placement investors
All investors
Relationship Investors (RI)
Outside Investors (OUT)
Test: (RI)=(OUT)
4.81
(0.00)
[3,878]
8.48
(1.55)
[2,148]
0.24
(0.00)
[1,730]
(0.000)
(0.000)
993
(12.34)
[8,049]
993
(46.23)
[2,148]
0
(0.00)
[5,901]
(0.000)
0.24
(0.00)
[8,007]
0.92
(0.69)
[2,106]
0.00
(0.00)
[5,901]
(0.000)
(0.000)
0.45
(1.00)
[8,049]
1.68
(1.00)
[2,148]
0.00
(0.00)
[5,901]
3.92
(0.73)
[8,049]
7.35
(0.88)
[2,148]
2.38
(0.53)
[5,901]
3.34
(2.03)
[6,588]
6.16
(2.48)
[2,148]
1.98
(1.59)
[4,440]
(0.000)
(0.000)
(P-value Mean or Count)
(P-value Median)
(0.000)
(0.000)
(0.000)
(0.000)
Panel B: Pre- and post-placement characteristics and purchasing patterns among relationship investors
Management Investors (MGT)
Key Business Partners (KEY)
Governance Investors (GOV)
Test: (MGT)=(KEY)=(GOV)
8.62
(1.64)
[616]
6.18
(0.00)
[913]
11.75
(7.57)
[619]
402
(65.26)
[616]
248
(27.16)
[913]
343
(55.41)
[619]
1.33
(1.39)
[587]
0.54
(0.00)
[900]
1.06
(1.10)
[619]
2.13
(2.00)
[616]
1.64
(1.00)
[913]
1.29
(1.00)
[619]
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
2.78
3.94
(0.15)
(0.80)
[616]
[616]
11.84
8.08
(2.09)
(4.06)
[913]
[913]
5.29
5.54
(0.79)
(2.67)
[619]
[619]
(P-value Mean or Count)
(P-value Median)
(0.000)
(0.000)
(0.000)
(0.000)
Table 3. Issuer returns around announcement and following announcement for private placements.
Dominant transactions (DOM) are defined as transactions with 1) an investor buying more than 50% of shares offered in the placement or 2) investors buying a block of
5% or more of shares outstanding in the placement. CAR_MM (%) are calculated using a market model regression of firm stock returns on the CRSP equally weighted
market index. The estimation window is (-260, -11), with day 0 being the announcement. CAR_ISBM is calculated using a matched control firm for each sample firm.
Sample firms are matched to a comparable company based on Fama French (1997) industry, market capitalization, book-to-market and momentum. Each sample firm is
matched to a firm in the same industry that did not issue a private placement during the previous three years. Within industry, matches are made based on the sum of the
absolute deviations of size, book-to-market, and momentum rankings. If size, book-to-market, or momentum returns are missing, a match is made with a company in the
same industry with missing size, book-to-market, or momentum returns. In addition to documenting pre-placement relationships, we include four non-mutuallyexclusive categories of new relationships: (1) Managerial appointment (investors or their affiliates become employees of the issuer within three years of the
placements), (2) New key business partnership (investors become key business partners of the issuer within three years of the placements), (3) Director appointment
(investors or their assignees become directors of the issuer within three years of the placements), and (4) Newly created block ( ! 5%). Superscripts *, **, and *** in the
second column indicate that the mean CARs between private placements without and with new relationships differ significantly at the 10%, 5%, and 1% levels,
respectively using two-tailed tests. P-values in square brackets are for two-tailed t-test that the mean CARs is equal to zero and for t-test that the mean CARs are equal
between placements to dominant relationship investors (RI-DOM) and placements to dominant outside investors (OUT-DOM). Statistics with p-values of .10 or lower
are highlighted in bold face type. — = not applicable.
All
Without
new postplacement
relationships
With new post-placement relationships
All
new relationships
Managerial
appointment
New key
business
partnership
Director
appointment
Newly created
block ! 5%
Panel A: Issuer returns around and following announcement for overall sample
All Transactions
CAR_MM days (-3,0)
2.02
0.63***
3.51
0.83
0.73
3.15
3.94
[P-value]
(N)
[0.000]
(1,818)
[0.147]
(938)
[0.000]
(880)
[0.652]
(45)
[0.796]
(7)
[0.000]
(406)
[0.000]
(689)
-3.31
-6.20*
-0.22
8.52
73.50
0.01
-0.89
[0.079]
(1,823)
[0.025]
(941)
[0.932]
(882)
[0.466]
(45)
[0.028]
(7)
[0.998]
(408)
[0.758]
(690)
-25.27
-38.13**
-11.98
-11.49
41.87
-21.35
-16.15
[0.000]
(1,509)
[0.000]
(767)
[0.205]
(742)
[0.612]
(39)
[0.546]
(6)
[0.148]
(339)
[0.147]
(585)
CAR_MM days (1, 120)
[P-value]
(N)
CAR_ISBM months (0,35)
[P-value]
(N)
All
Without
new postplacement
relationships
With new post-placement relationships
All
new relationships
Managerial
appointment
New key
business
partnership
Director
appointment
Newly created
block ! 5%
Panel B: Issuer returns around and following announcement for placements to definition 1 dominant investors
Dominant Relationship investors (RI-DOM)
CAR_MM days (-3,0)
2.15
[P-value]
[0.000]
(N)
(1,051)
CAR_MM days (1, 120)
-4.25
[P-value]
[0.083]
(N)
(1,054)
CAR_ISBM months (0,35)
-25.01
[P-value]
[0.007]
(N)
(915)
Dominant Management Investors (MGT-DOM)
CAR_MM days (-3,0)
1.87
[P-value]
[0.093]
(N)
(179)
CAR_MM days (1, 120)
4.56
[P-value]
[0.392]
(N)
(179)
CAR_ISBM months (0,35)
-1.78
[P-value]
[0.907]
(N)
(150)
Dominant Key Business Partners (KEY-DOM)
CAR_MM days (-3,0)
2.38
[P-value]
[0.000]
(N)
(649)
CAR_MM days (1, 120)
-9.15
[P-value]
[0.003]
(N)
(652)
CAR_ISBM months (0,35)
-33.01
[P-value]
[0.012]
(N)
(575)
Dominant Governance Investors (GOV-DOM)
CAR_MM days (-3,0)
1.72
[P-value]
[0.109]
(N)
(223)
CAR_MM days (1, 120)
3.02
[P-value]
[0.612]
(N)
(223)
CAR_ISBM months (0,35)
-19.16
[P-value]
[0.229]
(N)
(190)
Dominant Outside Investors (OUT-DOM)
CAR_MM days (-3,0)
2.03
[P-value]
[0.002]
(N)
(428)
CAR_MM days (1, 120)
-6.85
[P-value]
[0.052]
(N)
(427)
CAR_ISBM months (0,35)
-26.82
[P-value]
[0.012]
(N)
(354)
0.78***
[0.143]
(600)
-6.50
[0.062]
(601)
-29.05
[0.019]
(514)
3.95
[0.000]
(451)
-1.26
[0.706]
(453)
-19.84
[0.155]
(401)
-1.62
[0.360]
(27)
15.13
[0.246]
(27)
1.74
[0.954]
(26)
-3.58
—
(1)
29.72
—
(1)
86.97
—
(1)
3.20
[0.001]
(234)
-6.53
[0.178]
(235)
-43.62
[0.036]
(205)
4.71
[0.000]
(318)
-0.37
[0.923]
(320)
-22.29
[0.228]
(287)
1.39
[0.286]
(121)
-1.60*
[0.802]
(121)
-12.28*
[0.448]
(103)
2.88
[0.177]
(58)
17.41
[0.070]
(58)
21.24
[0.524]
(47)
—
—
—
—
—
—
4.61
[0.132]
(32)
5.11
[0.732]
(32)
-35.84
[0.330]
(23)
3.02
[0.271]
(38)
29.42
[0.006]
(38)
67.94
[0.090]
(32)
0.80***
[0.280]
(329)
-14.32*
[0.003]
(330)
-38.05
[0.050]
(286)
4.00
[0.000]
(320)
-3.86
[0.318]
(322)
-28.01
[0.117]
(289)
-1.91
[0.393]
(18)
20.60
[0.264]
(18)
24.86
[0.560]
(17)
—
3.45
[0.006]
(156)
-8.37
[0.128]
(157)
-49.84
[0.081]
(143)
4.28
[0.000]
(254)
-4.94
[0.258]
(256)
-38.49
[0.076]
(230)
0.32***
[0.754]
(150)
6.75
[0.381]
(150)
-22.27
[0.290]
(125)
4.59
[0.067]
(73)
-4.65
[0.609]
(73)
-13.19
[0.573]
(65)
-1.03
[0.741]
(9)
4.20
[0.780]
(9)
-41.93
[0.220]
(9)
-3.58
—
(1)
29.72
—
(1)
86.97
—
(1)
1.35
[0.248]
(46)
-8.35
[0.510]
(46)
-25.43
[0.309]
(39)
11.35
[0.092]
(26)
1.03
[0.932]
(26)
11.21
[0.812]
(25)
-0.45***
[0.668]
(147)
-10.31
[0.089]
(147)
-38.92
[0.028]
(118)
3.33
[0.000]
(281)
-5.03
[0.247]
(280)
-20.77
[0.121]
(236)
-1.72
[0.776]
(7)
15.41
[0.755]
(7)
15.92
[0.731]
(6)
1.88
[0.640]
(5)
60.15
[0.071]
(5)
64.20
[0.539]
(4)
3.32
[0.006]
(88)
0.50
[0.947]
(88)
-7.37
[0.714]
(74)
3.44
[0.000]
(274)
-5.22
[0.234]
(273)
-18.89
[0.168]
(229)
—
—
CAR_MM days (-3,0)
[0.878]
[0.298]
[0.567]
[0.982]
[0.581]
[0.941]
[0.337]
CAR_MM days (1,120)
RI-DOM =OUT-DOM
[0.544]
[0.584]
[0.489]
[0.994]
[0.641]
[0.427]
[0.405]
CAR_ISBM months (0,35)
RI-DOM =OUT-DOM
[0.898]
[0.645]
[0.961]
[0.794]
[0.920]
[0.209]
[0.882]
All
Without
new postplacement
relationships
With new post-placement relationships
All
new relationships
Managerial
appointment
New key
business
partnership
Director
appointment
Newly created
block ! 5%
Panel C: Issuer returns around and following announcement for placements to definition 2 dominant investors
Dominant Relationship investors (RI-DOM)
CAR_MM days (-3,0)
3.29
[P-value]
[0.000]
(N)
(507)
CAR_MM days (1, 120)
3.79
[P-value]
[0.263]
(N)
(509)
CAR_ISBM months (0,35)
-16.06
[P-value]
[0.217]
(N)
(447)
Dominant Management Investors (MGT-DOM)
CAR_MM days (-3,0)
2.92
[P-value]
[0.180]
(N)
(72)
CAR_MM days (1, 120)
18.55
[P-value]
[0.047]
(N)
(72)
CAR_ISBM months (0,35)
27.46
[P-value]
[0.307]
(N)
(58)
Dominant Key Business Partners (KEY-DOM)
CAR_MM days (-3,0)
4.11
[P-value]
[0.000]
(N)
(331)
CAR_MM days (1, 120)
-2.48
[P-value]
[0.526]
(N)
(333)
CAR_ISBM months (0,35)
-32.31
[P-value]
[0.057]
(N)
(300)
Dominant Governance Investors (GOV-DOM)
CAR_MM days (-3,0)
2.91
[P-value]
[0.068]
(N)
(136)
CAR_MM days (1, 120)
9.28
[P-value]
[0.223]
(N)
(136)
CAR_ISBM months (0,35)
13.99
[P-value]
[0.486]
(N)
(118)
Dominant Outside Investors (OUT-DOM)
CAR_MM days (-3,0)
3.76
[P-value]
[0.000]
(N)
(327)
CAR_MM days (1, 120)
1.61
[P-value]
[0.721]
(N)
(326)
CAR_ISBM months (0,35)
-5.68
[P-value]
[0.683]
(N)
(271)
0.66***
[0.586]
(129)
11.35
[0.131]
(129)
9.14
[0.637]
(116)
4.18
[0.000]
(378)
1.22
[0.745]
(380)
-24.85
[0.124]
(331)
-0.45
[0.858]
(18)
10.78
[0.529]
(18)
-35.50
[0.207]
(17)
-3.58
—
(1)
29.72
—
(1)
86.97
—
(1)
2.76
[0.013]
(174)
-3.18
[0.575]
(175)
-51.75
[0.046]
(155)
4.61
[0.000]
(315)
0.85
[0.835]
(317)
-25.87
[0.180]
(274)
4.70
[0.327]
(24)
2.47
[0.873]
(24)
-13.96*
[0.713]
(19)
2.10
[0.373]
(48)
26.60
[0.023]
(48)
47.63
[0.182]
(39)
—
—
—
—
—
—
5.61
[0.151]
(24)
23.21
[0.247]
(24)
50.34
[0.093]
(17)
1.62
[0.489]
(41)
31.66
[0.005]
(41)
52.37
[0.198]
(34)
1.30*
[0.642]
(46)
6.33
[0.527]
(46)
-11.89
[0.576]
(41)
4.60
[0.000]
(285)
-3.90
[0.360]
(287)
-35.54†
[0.066]
(259)
0.42
[0.896]
(14)
17.03
[0.422]
(14)
-26.55
[0.425]
(13)
—
3.37
[0.020]
(127)
-8.19
[0.191]
(128)
-65.78
[0.049]
(118)
4.58
[0.000]
(254)
-5.18
[0.243]
(256)
-38.68
[0.074]
(230)
1.16*
[0.461]
(79)
19.62*
[0.073]
(79)
28.81
[0.330]
(67)
5.34
[0.089]
(57)
-5.06
[0.614]
(57)
-5.47
[0.832]
(51)
-3.51
[0.250]
(4)
-11.10
[0.671]
(4)
-64.59
[0.303]
(4)
-3.58
—
(1)
29.72
—
(1)
86.97
—
(1)
1.42
[0.304]
(32)
-1.28
[0.934]
(32)
-20.27
[0.366]
(27)
10.75
[0.122]
(25)
-7.33
[0.538]
(25)
20.00
[0.683]
(24)
—
3.76
[0.000]
(327)
1.61
[0.721]
(326)
-5.68
[0.683]
(271)
2.41
[0.657]
(9)
29.00
[0.458]
(9)
18.94
[0.576]
(8)
2.61
[0.437]
(6)
53.36
[0.129]
(6)
-81.74
[0.634]
(4)
3.64
[0.002]
(98)
7.22
[0.308]
(98)
5.566
[0.786]
(81)
3.76
[0.000]
(327)
1.61
[0.721]
(326)
-5.68
[0.683]
(271)
—
—
—
—
CAR_MM days (-3,0)
[0.617]
—
[0.777]
[0.630]
[0.896]
[0.451]
[0.556]
CAR_MM days (1,120)
RI-DOM =OUT-DOM
[0.268]
—
[0.528]
[0.664]
[0.192]
[0.446]
[0.583]
CAR_ISBM months (0,35)
RI-DOM =OUT-DOM
[0.586]
—
[0.368]
[0.214]
[0.659]
[0.082]
[0.395]
† Exploration of this finding shows that it is driven by outliers. Further confirming this, when median CAR_ISBMs are examined, this result disappears; there is no
significant return to this group of investors and the return to key business partners without new relationships is significantly worse.
Table 4. Post-placement profitability (PROy+2) for private placements.
Dominant transactions (DOM) are defined as transactions with 1) an investor buying more than 50% of shares offered in the placement or 2) investors buying a block
of 5% or more of shares outstanding in the placement. We include four non-mutually-exclusive categories of new relationships: (1) Managerial appointment
(investors or their affiliates become employees of the issuer within three years of the placements), (2) New key business partnership (investors become key business
partners of the issuer within three years of the placements), (3) Director appointment (investors or assignees of the investors become directors of the issuer within
three years of the placements), and (4) Newly created block ( ! 5%). PROy+2 is the firm’s two-year industry-adjusted operating profitability (the ratio of income
before interest, depreciation and taxes scaled by lagged firm assets) two years after the placement. Our industry benchmark is the median operating profit for firms in
the same three-digit Compustat SIC. If there are fewer than 10 firms in a three-digit SIC code, we use median operating profit for firms in the same two-digit SIC
code. Superscripts *, **, and *** in the second column indicate that the median profitability between private placements with and without new relationships differ
significantly at the 10%, 5%, and 1% levels, respectively using two-tailed tests. P-values in square brackets are for one-sample Wilcoxon test that the median
profitability is equal to zero and for Kruskal-Wallis test that the median profitability are equal between placements to dominant relationship investors (RI-DOM) and
placements to dominant outside investors (OUT-DOM). Statistics with p-values of .10 or lower are highlighted in bold face type. — = not applicable.
All
Without
With new post-placement relationships
new postAll
Managerial
New key
Director
Newly created
placement
new
appointment
business
appointment
blocks ! 5%
relationships relationships
partnership
Panel A: Post-placement profitability for overall sample
All Transactions
PROy+2
[P-value]
(N)
-0.150
[0.000]
(1,680)
-0.197***
[0.000]
857
-0.098
[0.000]
(823)
-0.087
[0.016]
(46)
-0.151
[0.097]
(9)
-0.081
[0.000]
(391)
-0.104
[0.000]
(631)
Panel B: Post-placement profitability for firms making private placements to definition 1 dominant investors
Dominant Relationship investors (RI-DOM)
PROy+2
-0.123
[P-value]
[0.000]
(N)
(971)
-0.181***
[0.000]
(551)
-0.071
[0.001]
(420)
-0.057
[0.136]
(29)
0.148
—
(1)
-0.055
[0.033]
(225)
-0.079
[0.003]
(286)
Dominant Management Investors (MGT-DOM)
PROy+2
-0.081
[P-value]
[0.011]
(N)
(177)
-0.286*
[0.025]
(115]
-0.018
[0.253]
(62)
—
—
-0.008
[0.618]
(36)
-0.018
[0.268]
(40)
Dominant Key Business Partners (KEY-DOM)
PROy+2
-0.114
[P-value]
[0.000]
(N)
(599)
-0.178***
[0.000]
(306]
-0.081
[0.007]
(293)
0.032
[0.167]
[19)
—
-0.037
[0.132]
(143)
-0.089
[0.004]
(228)
Dominant Governance Investors (GOV-DOM)
PROy+2
-0.138
[P-value]
[0.000]
(N)
(195)
-0.182*
[0.000]
(130)
-0.124
[0.003]
(65)
-0.141
[0.032]
(10)
0.148
—
(1)
-0.124
[0.015]
(46)
-0.058
[0.151]
(18)
Dominant Outside Investors (OUT-DOM)
PROy+2
-0.165
[P-value]
[0.000]
(N)
(386)
-0.210*
[0.000]
(125)
-0.138
[0.000]
(261)
-0.259
[0.108]
(7)
-0.151
[0.108]
(7)
-0.087
[0.000]
(83)
-0.136
[0.000]
(253)
Test: RI-DOM =OUT-DOM
[0.232]
[0.013]
[0.145]
[0.127]
[0.445]
[0.046]
0.148
—
(1)
-0.067
[0.020]
(165)
-0.066
[0.007]
(278)
[0.058]
Panel C: Post-placement profitability for firms making private placements to definition 2 dominant investors
Dominant Relationship investors (RI-DOM)
PROy+2
-0.074
[P-value]
[0.000]
(N)
(452)
-0.084
[0.075]
(114)
-0.069
[0.003]
(338)
Dominant Management Investors (MGT-DOM)
PROy+2
-0.008
[P-value]
[0.644]
(N)
(75)
0.081
[0.678]
(23)
-0.018
[0.332]
(52)
—
—
-0.002
[1.000]
(27)
-0.017
[0.371]
(45)
Dominant Key Business Partners (KEY-DOM)
PROy+2
-0.062
[P-value]
[0.017]
(N)
(296)
-0.052
[0.618]
(36)
-0.068
[0.022]
(260)
0.032
[0.092]
(13)
—
-0.054
[0.093]
(115)
-0.074
[0.021]
(230)
Dominant Governance Investors (GOV-DOM)
PROy+2
-0.122
[P-value]
[0.000]
(N)
(115)
-0.093
[0.000]
(65)
-0.124
[0.007]
(50)
-0.124
[0.584]
(4)
0.148
—
(1)
-0.129
[0.010]
(30)
0.007
[0.255]
(20)
—
-0.128
[0.000]
(302)
-0.259
[0.097]
(9)
-0.208
[0.108]
(7)
-0.096
[0.000]
(94)
-0.128
[0.000]
(302)
—
[0.009]
[0.125]
[0.127]
[0.289]
[0.007]
Dominant Outside Investors (OUT-DOM)
PROy+2
-0.128
[P-value]
[0.000]
(N)
(302)
Test: RI-DOM =OUT-DOM
[0.008]
0.015
[0.332]
(17)
Table 5. Regressions of stock price response to announcement on relationship, characteristics of the private placement, and the issuing firm.
The regression models in this table estimate the association between the market response to announcement of a private placement (CAR (-3,0)) and the relationship
between the firm and investor(s), characteristics of the placement and issuing firm. Variables used in the model are defined as follows. DOM equals one in placements
with dominant investors and zero in placements without dominant investor. MGT-DOM equals one when a manager is a dominant investor and zero for placements with
other types of dominant investors. KEY-DOM equals one when a key business partner is the dominant investor and zero for placements with other types of dominant
investors. GOV-DOM equals one when the dominant investor is a director or 5% or greater shareholder prior to the placement and zero otherwise. New large investor(s)
equals one if a new 5% or greater blockholder is established as part of the placement and zero otherwise. Added directorship(s) variable equals the number of director
appointments in the following three years for investors with board seat rights. Price discount is measured relative to closing price one day prior to the private placement
announcement date and is Winsorized at 5% in either tail. Results presented are OLS regressions and the three-stage least-squares regressions in which we first estimate a
model of added directorships, then we estimate a model of Winsorized price discount using predicted directorship rights from the first model as one of the explanatory
variables, and the final step is to simultaneously estimate CAR regression models (using predicted added directorships and predicted Winsorized price discount as two of
the explanatory variables) and re-estimate directorship and Winsorized pricing models using seemingly unrelated regression (Zellner (1962)). P-values (in parentheses)
are based on White-corrected standard errors. Coefficients with p-values of .10 or lower are highlighted in bold face type.
Dependent variable: CAR (-3, 0) (%)
Definition 1
Definition 2
Investor purchases > 50% of shares issued
Investor purchases ! 5% of shares outstanding
OLS Regressions
Simultaneous Models
OLS Regressions
Simultaneous Models
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Intercept
1.705
5.179
-1.685
-1.858
-0.557
10.545
-0.532
-1.302
(0.213)
(0.020)
(0.418)
(0.678)
(0.827)
(0.000)
(0.786)
(0.838)
Pre-placement Relationship Variables
DOM
2.485
4.952
2.498
2.459
(0.066)
(0.017)
(0.007)
(0.008)
MGT-DOM
-0.565
(0.578)
1.543
(0.401)
-4.314
(0.010)
-3.980
(0.180)
KEY-DOM
1.608
(0.032)
5.858
(0.000)
0.720
(0.503)
4.252
(0.097)
GOV-DOM
1.848
(0.059)
1.590
(0.370)
0.383
(0.763)
0.477
(0.833)
Post-placement New Relationship Variables
New large investor(s)
1.656
(0.007)
1.607
(0.021)
2.306
(0.014)
2.843
(0.013)
—
—
—
—
0.736
(0.050)
0.749
(0.063)
2.003
(0.001)
2.643
(0.000)
1.160
(0.086)
0.959
(0.084)
1.597
(0.009)
1.757
(0.070)
-0.011
(0.663)
-0.039
(0.140)
0.023
(0.537)
-0.042
(0.395)
-0.001
(0.992)
-0.049
(0.219)
-0.001
(0.969)
-0.082
(0.270)
-0.047
(0.831)
-0.137
(0.548)
-0.082
(0.807)
-0.437
(0.294)
-0.293
(0.391)
-1.227
(0.044)
-0.322
(0.327)
-1.291
(0.037)
1.651
(0.040)
2.235
(0.010)
0.318
(0.797)
2.320
(0.158)
0.098
(0.094)
0.363
(0.791)
-0.186
(0.875)
-5.532
(0.054)
Patenty+3
0.010
(0.340)
0.011
(0.301)
0.003
(0.878)
0.021
(0.308)
0.001
(0.921)
0.010
(0.477)
0.002
(0.916)
-0.003
(0.894)
Firm specific risk (%)
0.199
(0.041)
0.305
(0.003)
0.489
(0.001)
0.635
(0.001)
0.401
(0.136)
0.369
(0.014)
0.402
(0.002)
1.009
(0.001)
0.048
(0.782)
0.197
(0.288)
1.206
(0.001)
1.124
(0.010)
0.297
(0.244)
0.466
(0.221)
0.990
(0.020)
0.992
(0.263)
Placement Herfindahl index
-3.644
(0.028)
-4.332
(0.021)
-5.059
(0.045)
-8.119
(0.008)
-0.782
(0.523)
-6.167
(0.001)
-0.775
(0.188)
-3.166
(0.018)
Winsorized price discount 5% tail (%)
-0.007
(0.464)
-0.007
(0.456)
-0.010
(0.004)
-0.009
(0.010)
-0.006
(0.153)
-0023
(0.141)
-0.009
(0.029)
-0.013
(0.099)
-1.021
(0.115)
-1.656
(0.015)
-1.768
(0.076)
-1.729
(0.122)
-1.678
(0.122)
-1.021
(0.256)
-1.697
(0.075)
-2.109
(0.198)
Added directorship(s)
Proxies for CEO Power and Entrenchment
CEO Share (%)
CEO Tenure
Proxies for Relationship-Specific Investment
New economy
Placement Characteristics
ln (issue) $ millions
Control Variables
Dummy for multiple placements in sample
Dummy for multiple dominant investors in a single placement
Year fixed effect
No.
P-value
R-squared
-0.952
(0.701)
Yes
1,170
(0.000)
0.042
Yes
955
(0.000)
0.063
Yes
1,170
(0.000)
0.044
Yes
955
(0.000)
0.049
Yes
1,137
(0.003)
0.037
Yes
500
(0.013)
0.058
-1.497
(0.735)
Yes
1,137
(0.000)
0.036
Yes
500
(0.000)
0.060
Table 6. Probit analysis of the pre-placement relationships.
Dominant transaction is defined as a single investor of one type who purchases over 50% of the shares offered in the private placement. Non-dominant transaction
is defined as transactions with no investor buying more than 50% of shares offered in the placement. Types of dominant transactions are categorized as dominated
by management investors, by key business partners, by governance investors, and by outside investors. Key business partners are strategic alliances, customers and
suppliers. Governance investors are existing directors and/or blockholders (≧5%). Outside investors refer to investors with no documented pre-placement
employment, business, directorship, or blockholding relationship. Dependent variables are defined as follows: Model (1) one for dominant transaction and zero for
non-dominant transaction, Model (2) one for transaction with dominant relationship investors and zero for transaction with dominant outside investors, Model (3)
one if management is the dominant investor and zero if an outside investor is dominant, Model (4) one if key business partner is dominant and zero if an outside
investor is dominant, and Model (5) one if a governance investor is the dominant investor and zero if an outside investor is dominant. Volume is defined as the
trading volume/the average of outstanding shares over the previous two years. Market to book is defined as market value of equity divided by its book value of
equity, measured at the fiscal year end prior to the placement announcement. P-values (in parentheses) are based on White-corrected standard errors. Coefficients
with p-values of .10 or lower are highlighted in bold face type.
(1)
(2)
(3)
(4)
(5)
For each dependent variable the first listed
Dominant
Dominant
Dominant
Dominant
Dominant
investor type is indicated by 1, the second by 0.
vs.
Relationship
Management
Key Business
Governance
Non Dominant
vs.
vs.
vs.
vs.
Dominant
Dominant
Dominant
Dominant
Outside
Outside
Outside
Outside
Intercept
-3.459
0.346
0.546
-1.048
1.149
(0.000)
(0.342)
(0.314)
(0.027)
(0.049)
Post-placement New Relationship Variables and Post-Placement Profitability
New large investor(s)
0.414
-0.931
-0.935
-0.671
-2.380
(0.045)
(0.000)
(0.000)
(0.000)
(0.000)
Added directorship(s)
0.185
(0.417)
0.380
(0.004)
0.479
(0.026)
0.382
(0.012)
0.654
(0.004)
PROy+2
0.351
(0.002)
0.172
(0.028)
0.195
(0.090)
0.310
(0.006)
-0.056
(0.645)
-0.012
(0.096)
-0.005
(0.239)
-0.003
(0.714)
-0.006
(0.237)
-0.017
(0.048)
-0.011
(0.894)
-0.077
(0.066)
-0.064
(0.325)
-0.095
(0.047)
-0.098
(0.072)
-0.161
(0.529)
0.025
(0.878)
-0.267
(0.386)
0.086
(0.669)
0.263
(0.336)
Patenty+3
0.002
(0.707)
0.006
(0.065)
0.017
(0.000)
0.003
(0.314)
0.009
(0.233)
Firm specific risk (%)
2.979
(0.415)
-2.570
(0.179)
-3.927
(0.201)
-6.845
(0.095)
-0.764
(0.786)
Volume
-0.017
(0.848)
-0.027
(0.519)
-0.016
(0.816)
-0.006
(0.904)
-0.203
(0.007)
Market to book
0.002
(0.616)
-0.001
(0.533)
-0.005
(0.497)
-0.000
(0.892)
-0.004
(0.471)
0.092
(0.078)
0.023
(0.096)
-0.146
(0.003)
0.073
(0.092)
0.119
(0.035)
Placement Herfindahl index
6.377
(0.000)
0.835
(0.005)
-0.333
(0.452)
1.899
(0.000)
-0.341
(0.489)
Price discount (% relative to 1 day prior to
announcement)
Control Variables
Year fixed effect
Dummy for multiple placements in model
0.001
(0.463)
-0.002
(0.055)
0.005
(0.068)
-0.004
(0.003)
0.003
(0.268)
Yes
0.199
(0.351)
Yes
0.304
(0.009)
Yes
0.105
(0.536)
Yes
0.425
(0.001)
Yes
-0.108
(0.551)
No.
P-value
Pseudo R2
992
(0.000)
0.772
803
(0.000)
0.150
337
(0.000)
0.253
561
(0.000)
0.169
341
(0.000)
0.370
Proxies for CEO Power and Entrenchment
CEO Share (%)
CEO Tenure
Firm Characteristics
New economy
Placement Characteristics
ln (issue) $ millions
Table 7. Regressions of added directorships as a percentage of board size on relationship, characteristics of the private placement, and the issuing firm.
The dependent variable is ln (1+ added directorships/ board size) where added directorships are the number of director appointments in the following
three years by investors with board seat rights and board size is the number of current board of directors. DOM equals one in placements with
dominant investors and zero in placements without dominant investor. MGT-DOM equals one when a manager is a dominant investor and zero for
placements with other types of dominant investors. KEY-DOM equals one when a key business partner is the dominant investor and zero for
placements with other types of dominant investors. GOV-DOM equals one when the dominant investor is a director or 5% or greater shareholder prior
to the placement and zero otherwise. Pre-existing directorship(s) variable equals one for investors who are existing directors. Pre-existing large
investor(s) variable equals one for investors who are existing shareholders with at least or greater than 5% of the shares outstanding. New nondirectorship relationship equals one when an investor establishes a new relationship (new management positions, key business partnerships, and 5%
or greater blocks) as part of the placement or over the three years following the placement and zero otherwise. Winsorized price discount 5% tail is
price discount measured relative to closing price one day prior to the private placement announcement date and is winsorized at 5% in either tail.
Results presented are OLS regressions and the three-stage least-squares regressions in which we first estimate a model of added directorships, then we
estimate a model of Winsorized price discount using predicted directorship rights from the first model as one of the explanatory variables, and the
final step is to simultaneously estimate CAR regression models (using predicted added directorships and predicted Winsorized price discount as two
of the explanatory variables) and re-estimate directorship and Winsorized pricing models using seemingly unrelated regression (Zellner (1962)). Pvalues (in parentheses) are based on White-corrected standard errors. Coefficients with p-values of .10 or lower are highlighted in bold face type.
Dependent variable: ln ((1+ added directorships)/ board size)
Definition 1
Definition 2
Investor purchases > 50% of shares issued
Investor purchases ! 5% of shares outstanding
OLS Regressions
Simultaneous Models
OLS Regressions
Simultaneous Models
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Intercept
-2.016
-2.229
-2.047
-2.244
-2.000
-1.971
-2.004
-1.924
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
(0.000)
Pre-placement Relationship Variables
DOM
-0.086
-0.094
0.018
0.014
(0.134)
(0.161)
(0.707)
(0.750)
MGT-DOM
0.218
(0.000)
0.200
(0.000)
0.290
(0.002)
0.317
(0.002)
KEY-DOM
0.128
(0.001)
0.180
(0.000)
0.142
(0.012)
0.134
(0.047)
GOV-DOM
0.093
(0.118)
0.076
(0.203)
0.028
(0.735)
0.030
(0.731)
Pre-existing directorship(s)
-0.114
(0.000)
-0.146
(0.000)
-0.115
(0.000)
-0.142
(0.000)
-0.141
(0.000)
-0.426
(0.000)
-0.093
(0.000)
-0.423
(0.000)
Pre-existing large investor(s)
0.096
(0.003)
0.109
(0.022)
0.100
(0.002)
0.119
(0.008)
0.081
(0.006)
0.081
(0.310)
0.089
(0.002)
0.080
(0.269)
0.166
(0.000)
0.151
(0.000)
0.174
(0.000)
0.111
(0.014)
—
0.120
(0.004)
—
0.004
(0.014)
0.003
(0.006)
0.003
(0.045)
0.003
(0.009)
0.002
(0.462)
0.003
(0.004)
0.002
(0.423)
-0.014
(0.214)
Proxies for Relationship-Specific Investment
New economy
0.151
(0.000)
-0.008
(0.495)
-0.016
(0.146)
-0.012
(0.315)
-0.018
(0.129)
-0.017
(0.358)
-0.018
(0.094)
-0.019
(0.278)
0.149
(0.000)
0.136
(0.001)
0.111
(0.020)
0.169
(0.000)
0.178
(0.004)
0.166
(0.000)
0.169
(0.037)
Patenty+3
-0.000
(0.222)
-0.001
(0.107)
-0.000
(0.850)
-0.000
(0.893)
-0.001
(0.038)
-0.001
(0.007)
-0.001
(0.189)
-0.001
(0.152)
Firm specific risk (%)
2.237
(0.000)
2.344
(0.000)
2.202
(0.000)
2.418
(0.000)
2.040
(0.000)
1.450
(0.021)
1.975
(0.000)
1.272
(0.059)
0.007
(0.000)
0.009
(0.000)
0.006
(0.000)
0.007
(0.000)
0.006
(0.003)
0.008
(0.003)
0.006
(0.000)
0.007
(0.000)
Placement Herfindahl index
0.110
(0.160)
0.104
(0.258)
0.161
(0.070)
0.117
(0.242)
0.041
(0.417)
0.129
(0.158)
0.037
(0.442)
0.118
(0.238)
Winsorized price discount 5% tail (%)
0.001
(0.059)
0.001
(0.022)
0.003
(0.042)
0.005
(0.013)
0.001
(0.045)
0.000
(0.823)
0.001
(0.098)
0.001
(0.989)
0.033
(0.304)
0.049
(0.152)
0.033
(0.307)
0.036
(0.326)
0.035
(0.281)
0.015
(0.727)
0.041
(0.186)
0.005
(0.910)
Post-placement New Relationship Variables
New non-directorship relationship(s)
0.145
(0.000)
Proxies for CEO Power and Entrenchment
CEO Share (%)
0.004
(0.002)
CEO Tenure
Placement Characteristics
Block size (%)
Control Variables
Dummy for multiple placements in
sample
Dummy for multiple dominant investors
in a single placement
Year fixed effect
No.
P-value
R-squared
0.158
(0.241)
0.142
(0.255)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1,190
(0.000)
0.161
971
(0.000)
0.196
1,170
(0.000)
0.161
955
(0.000)
0.198
1,152
(0.000)
0.134
506
(0.000)
0.155
1,137
(0.000)
0.137
500
(0.000)
0.150
Table 8. Regressions of Winsorized issue price discount on relationship, characteristics of the private placement, and the issuing firm.
The dependent variable is offer price relative to closing price one day prior to the private placement announcement date and is winsorized at 5% in either tail. DOM
equals one in placements with dominant investors and zero in placements without dominant investor. MGT-DOM equals one when a manager is a
dominant investor and zero for placements with other types of dominant investors. KEY-DOM equals one when a key business partner is the dominant
investor and zero for placements with other types of dominant investors. GOV-DOM equals one when the dominant investor is a director or 5% or
greater shareholder prior to the placement and zero otherwise. New large investor(s) equals one if a new 5% or greater blockholder is established as
part of the placement and zero otherwise. Added directorship(s) equals the number of director appointments over the three years following the
placement for investors with board seat rights. Results presented are OLS regressions and the three-stage least-squares regressions in which we first
estimate a model of added directorships, then we estimate a model of Winsorized price discount using predicted directorship rights from the first
model as one of the explanatory variables, and the final step is to simultaneously estimate CAR regression models (using predicted added directorships
and predicted Winsorized price discount as two of the explanatory variables) and re-estimate directorship and Winsorized pricing models using
seemingly unrelated regression (Zellner (1962)). P-values (in parentheses) are based on White-corrected standard errors. Coefficients with p-values
of .10 or lower are highlighted in bold face type.
Dependent variable: Winsorized price discount (%) relative to closing one day prior to
announcement
Definition 1
Definition 2
Investor purchases > 50% of shares issued
Investor purchases ! 5% of shares outstanding
OLS Regressions
Simultaneous Models
OLS Regressions
Simultaneous Models
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Intercept
18.035
15.099
19.927
21.011
19.849
29.416
21.851
29.084
(0.000)
(0.041)
(0.000)
(0.004)
(0.000)
(0.000)
(0.000)
(0.000)
Pre-placement Relationship Variables
DOM
0.730
3.277
2.763
2.557
(0.851)
(0.427)
(0.191)
(0.220)
MGT-DOM
-0.975
(0.756)
-1.921
(0.561)
9.485
(0.024)
10.045
(0.035)
KEY-DOM
-12.442
(0.000)
-13.331
(0.000)
-12.276
(0.000)
-11.915
(0.000)
GOV-DOM
0.677
(0.830)
0.385
(0.903)
-3.100
(0.342)
-2.723
(0.447)
Post-placement New Relationship Variables
New large investor(s)
2.304
(0.222)
2.956
(0.233)
1.741
(0.357)
1.756
(0.448)
—
—
—
—
-1.836
(0.120)
-1.408
(0.328)
2.878
(0.014)
5.780
(0.000)
-1.854
(0.161)
-1.858
(0.284)
2.917
(0.023)
-3.950
(0.017)
0.083
(0.324)
0.084
(0.388)
0.093
(0.206)
0.081
(0.348)
0.023
(0.781)
0.110
(0.338)
0.038
(0.612)
0.175
(0.123)
0.544
(0.435)
0.638
(0.418)
0.512
(0.441)
0.550
(0.460)
0.281
(0.696)
-0.269
(0.798)
0.318
(0.639)
0.101
(0.919)
Proxies for Relationship-Specific Investment
New economy
6.514
(0.008)
7.791
(0.005)
5.863
(0.016)
7.304
(0.009)
6.405
(0.009)
13.135
(0.000)
5.899
(0.016)
13.368
(0.000)
Added directorship(s)
Proxies for CEO Power and Entrenchment
CEO Share (%)
CEO Tenure
Patenty+3
-0.087
(0.001)
-0.099
(0.004)
-0.074
(0.024)
-0.083
(0.016)
-0.058
(0.085)
0.002
(0.951)
-0.045
(0.166)
-0.004
(0.921)
Firm specific risk (%)
0.909
(0.016)
1.247
(0.004)
0.989
(0.001)
1.367
(0.000)
0.808
(0.018)
1.657
(0.002)
0.886
(0.001)
1.650
(0.000)
-5.204
(0.000)
-4.816
(0.000)
-5.489
(0.000)
-5.222
(0.000)
-5.390
(0.000)
-6.500
(0.000)
-5.647
(0.000)
-6.035
(0.000)
Block size (%)
0.664
(0.000)
0.646
(0.000)
0.594
(0.000)
0.527
(0.000)
0.594
(0.000)
0.541
(0.000)
0.532
(0.000)
0.531
(0.000)
Placement Herfindahl index
-15.566
(0.002)
-5.681
(0.357)
-19.425
(0.000)
-9.491
(0.127)
-15.808
(0.000)
-9.956
(0.058)
-17.337
(0.000)
-11.439
(0.033)
1.085
(0.581)
3.638
(0.100)
0.721
(0.714)
2.791
(0.206)
0.431
(0.824)
3.820
(0.135)
0.065
(0.974)
3.804
(0.135)
Placement Characteristics
ln (issue) $ millions
Control Variables
Dummy for multiple placements in
sample
Dummy for multiple dominant investors
in a single placement
Year fixed effect
No.
P-value
R-squared
2.576
(0.635)
1.865
(0.792)
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
1,190
(0.000)
0.154
971
(0.000)
0.158
1,170
(0.000)
0.155
955
(0.000)
0.159
1,157
(0.000)
0.145
510
(0.000)
0.214
1,137
(0.000)
0.144
500
(0.000)
0.212