The Choice Between
Private and Public Convertible Bond Offerings
Devrim
Yaman
Department
of Finance and Commercial Law
Haworth
College of Business
Western
Michigan University
Kalamazoo,
MI 49008-3811
e-mail:
devrim.yaman@wmich.edu
Abstract
In this paper we study
the impact of information asymmetry, agency conflicts, flotation costs, and
financial distress costs on the choice between private and public placements
of convertible bond issues. We find that the level of information asymmetry
between insiders and outside investors is the most important factor in the
placement decision. This finding is consistent with the argument that private
offering investors are better informed about the firm and certify firm
quality. Firms susceptible to high agency costs due to risk-shifting,
underinvestment, and free cash flow problems also prefer private placements
for convertible bonds. However the economic impact of agency costs on the
placement choice is lower than that of information asymmetry. We also find
that firms with smaller issues prefer private offerings, possibly due to the
lower fixed issue costs of these placements. Expected financial distress costs
do not affect the placement decision of convertible bond issuers.
1.
Introduction
Bonds
can be sold as a public issue to investors at large or placed privately with a
few investors. In a public offering, after the issue is approved by the board
of directors, the issuing firm prepares and files a registration statement
with the Securities and Exchange Commission (SEC). If accepted, the
registration statement becomes effective 20 days later, and the bond can be
sold to the public. In a private placement firms do not have to go through the
SEC registration process and can raise funds faster with lower flotation
costs. It is also easier to renegotiate a private placement since there are
fewer investors involved. However since private placements are not offered to
the general public, they are less suitable for very large offerings. These
placements are also less liquid since they do not trade in secondary markets.
There
have been several studies on the placement structure of bonds. For example
Blackwell and Kidwell (1988) show that utility firms choose the type of debt
placement that provides the lowest transaction costs.
Easterwood and Kadapakkam (1991) argue that information asymmetry is a
major determinant in firms’ choice between private and public placement of
debt. Krishnaswami, Spindt, and Subramaniam (1999) show that firms with small
issues and high moral hazard problems have higher proportions of private debt
in their capital structure. Denis and Mihov (2003) study the choice among bank
debt, non-bank private debt, and public debt and find that the credit quality
of the issuer is the main determinant of the debt source.
The focus of
prior studies has been debt in general and none of them analyzes the placement
structure of convertible debt separately. These studies either exclude
convertible debt from their samples or combine straight debt and convertible
debt in one sample. However the determinants of the placement choice of
convertible debt might be different since convertible debt has equity-like
characteristics as well as debt-like characteristics. For instance Myers and
Majluf (1984) show that information sensitive securities face higher adverse
selection costs. Hence the equity-like characteristics of convertible debt may
cause information asymmetry problems to be a more important determinant in the
private-public placement choice.
In this paper we analyze the impact of information asymmetry, agency
conflicts, flotation costs, and financial distress costs on the choice between
private and public placements of convertible debt issues. Our sample consists
of 78 private and 229 public placements of convertible bonds issued by
industrial firms between 1983 and 2002. The results show that the primary
determinant of the placement choice for convertible debt is firms’
information asymmetry problems. We find that firms with higher levels of
information asymmetry are more likely to choose private placements. This
result is consistent with the argument that investors in private placements
establish closer relations with the issuing firm and can therefore value the
firm more accurately. The result also indicates that the commitment of better
informed investors in private placements signal favorable information to the
market.
We
find that agency costs of risk-shifting, underinvestment, and free cash flow
problems also affect the placement choice. Firms with higher agency costs are
more likely to choose private placements for their convertible bond offerings.
This result shows that when investors have a close relationship with the debt
issuing firm the risk-shifting and underinvestment problems are mitigated.
This result also suggests that the better monitoring offered by private
investors will reduce the misuse of free cash flows. Our analysis on the
economic impact of each variable in the logistic regressions shows that,
although important, agency problems are less influential than information
asymmetry in the choice between private and public placements. This finding is
consistent with the argument that by design convertible bonds already have low
agency problems since they mitigate the risk-sifting and underinvestment
problems.
Prior studies
show that the fixed component of the flotation costs of public placements is
higher than those of private placements. Based on this argument we hypothesize
that firms with smaller convertible bond issues will prefer private issues in
order to lower the flotation costs. Our results provide strong support for
this hypothesis. We find that flotation costs have a statistically and
economically significant impact on the convertible debt placement decision. We
do not find support for our hypothesis that firms with high expected financial
distress costs prefer private placements for their convertible bond offerings
due to rescheduling, renegotiation, and reorganizational advantages offered by
this placement structure.
The rest of the
paper is organized as follows: Section 2 presents the theoretical arguments in
prior studies on the convertible debt placement decision and section 3 reviews
prior empirical evidence. Section
4 presents the hypotheses we develop based on theoretical arguments and
empirical findings in literature. Section 5 describes the sample
characteristics and section 6 discusses the empirical results. Section 7
concludes the paper.
2.
Arguments on the Private-Public Placement Choice
2.1.
Information Asymmetry
Managers and
outside investors have asymmetric information because managers know more about
the firm’s future prospects than a typical investor. Hence firms with
unfavorable prospects will sell securities in order to bring in new investors
to share their losses. On the other hand when the prospects of the firm are
very good, the firm should avoid selling these securities in order to avoid
sharing the profits with the new investors. Therefore security offerings
signal that the firm’s prospects are not bright and result in adverse
selection costs.
Myers and Majluf
(1984) argue that under information asymmetry adverse selection costs are
particularly high for securities that have equity characteristics such as
convertible bonds. When firms lack internal resources to fund projects, they
might need to sell undervalued securities. In these cases managers might
choose to forego valuable investment opportunities in order to protect the
interest of the current shareholders.
Private
placements might alleviate the information asymmetry problems associated with
security issuance. James (1987) shows that private placement purchasers are
better informed about the issuing firm than public placement investors. In
private placements investors are better able to assess the true value of the
firm which in turn prevents the need to abandon profitable investment
opportunities. Private placements of debt might also signal positive
information to the market. Szewczyk and Varma (1991) argue that in a private
placement investors gain access to detailed financial information and their
commitment serves as quality certification. Dierkens (1991), Korajczyk,
Lucas, and McDonald (1992), and Bayless and Chaplinsky (1996) show that when
firms have high information asymmetry the effect of the signal associated with
security issues is higher. Therefore, private placements are more advantageous
for firms with high information asymmetry since the effect of the positive
signal of the offerings will be higher for these firms.
2.2.
Agency Costs
An
agency relationship is formed when a person (principal) hires another person
(agent) to perform services on his behalf and delegates him decision-making
authority. Agents will take actions in their self-interest if they do not have
incentives to do otherwise or they are not constrained in their actions.
Agency problems also exist in financial markets and the market participants
can take actions such as monitoring the agent in order to reduce the costs
associated with these relations.
Agency
problems of risk shifting and underinvestment arise between stockholders and
bondholders when debt financing is combined with the limited liability of
shareholders. Risk-shifting occurs when the firm substitutes riskier assets
for less risky assets (Jensen and Meckling, 1976). Shareholders of levered
firms have incentives to undertake riskier projects since they have bounded
downside potential due to limited liability but unbounded upside potential.
Underinvestment is the problem of foregoing positive NPV projects when the
value of the project is less than the face value of debt plus the initial
investment[1].
Shareholders might choose to underinvest since they receive only the cash
flows that remain after the liabilities of the firm are paid. There are also
agency problems between managers and shareholders when managers do not own 100
percent of the firm. This type of agency problem is exacerbated when the firm
has high levels of financial slack (free cash flow). Jensen (1986) argues that
managers might use financial slack for their own benefits, thereby reducing
firm value.
Smith
and Warner (1979) argue that private placements of debt provide better
monitoring of the issuing firm since they have detailed restrictive covenants.
As a result, private placements should reduce the agency costs of
risk-shifting, underinvestment, and free cash flow problems of convertible
bond issues. Since agency costs are anticipated by bondholders at the time of
the issue shareholders ultimately bear these costs. Therefore firms with high
agency costs have incentives to reduce these costs through private placements.
2.3.
Financial Distress
Financial distress occurs when a firm’s operating cash flows are not
sufficient to cover current financial obligations. Firms in financial distress
incur both direct and indirect costs. Direct costs include lawyer fees as well
as administrative and accounting fees. Indirect costs are the costs of
impaired ability to perform normal business activities. Altman (1984) shows
that the total direct and indirect costs of financial distress is often
greater than 20 percent of firm value.
Prior
studies suggest that firms with a high probability of financial distress and
firms that incur high costs in case of financial distress would prefer private
debt compared to public debt. Denis and Mihov (2003) indicate that regulations
provide private lenders a higher flexibility of renegotiation compared to
public placements. Any major changes to bond indentures must be approved
unanimously by public bondholders which might cause holdout problems for
publicly traded bonds. Cantillo and Wright (2000) argue that in financial
distress, privately held debt allows for ‘less damaging intervention’.
Firms with a high risk of default will need the reorganizational skills of
private lenders and hence prefer to borrow from these sources. Similarly Sy
(1999) argues that when credit risk is high managers prefer private debt
placements because of the value gains and the private benefits they obtain
from renegotiating the debt restrictions that are less flexible compared to
public issues.
2.4.
Flotation Costs
Flotation costs include the direct and indirect expenses of security
issues. Direct expenses are the underwriter’s spread, filing fees, legal
fees, and taxes. Indirect expenses include management time on the new issue.
Flotation costs can range from less than 1 percent to 20 percent of gross
proceeds.
Prior
studies suggest that the flotation costs for private issues are lower than
those for public issues. These studies show that there is a fixed component in
issuance costs which are higher for public issues. For example Blackwell and
Kidwell (1988) argue that compensation of the investment banker is higher in
public placements. Since the number of investors is higher, the distribution
costs are also higher for public placements. Moreover the legal fees,
accountants’ fees and trustees’ fees are higher for public offerings since
registration, certified financial statements, and bond counsel’s opinion are
required by SEC.
3.
Prior Empirical Evidence on Private and Public Debt Offerings
Several
prior empirical studies analyze the placement choice of debt. These studies
show that information asymmetry is an important factor in the choice between
private and public offerings. One of the prior studies that investigate the
influence of information asymmetry is Easterwood and Kadapakkam (1991) who
examine the choice between private and public placement of debt for Fortune
500 firms. They find that medium size firms prefer private markets more than
larger firms. Since smaller firms are more prone to asymmetric information
problems this finding indicates that information asymmetry is a critical
determinant of the debt placement choice.
Easterwood and Kapadakkam also find that private placement is an
important source of funds for firms in general because private debt
constitutes 60% of the long-term debt in their sample.
Szewczyk
and Varma (1991) provide evidence on the signaling affect of private debt
offerings under information asymmetry. They analyze the stock price effects of
private offerings of utility firms and find that these offerings result in a
positive stock price reaction. This finding is consistent with the arguments
that information asymmetry problems are avoided in a private placement and
that the willingness of the better informed investors to participate in a
private placement serves as a quality certification of the issuing firm.
Szewczyk and Varma also find that the stock price reaction is more favorable
for larger issues which indicates that larger positions by well-informed
buyers signal more favorable information to the market participants.
Fields
and Mais (1991) study the stock price effects of private placements of
convertible bonds. They find that the average announcement returns of the
private sales of convertible bonds is significantly positive. This result is
consistent with the argument that private placements avoid the information
asymmetry problems by selling the securities to a small number of
well-informed investors. As in Szewczyk and Varma (1991), Fields and Mais find
a positive relation between issue size and the announcement returns which
suggests that management’s ability to place larger issues conveys more
favorable information to the market.
Information
asymmetry is also one of the factors that Krishnaswami et al. (1999) analyze
in their study on the cross sectional variation of firms’ privately placed
long-term debt. Their sample consists of large firms that have access to both
private and public markets. Krishnaswami et al. find that firms with greater
potential information asymmetries issue more private debt. They also find that
firms that have favorable information about their future value and earnings
and have high information asymmetry use more private debt.
Prior
studies find that agency costs are also important in the placement choice of
debt. For example Blackwell and Kidwell (1988) examine the cost differences of
private and public placements of debt issued by utility firms and find that
firms choose private placements due to the higher agency costs in public
markets. Private placements have higher agency costs because they have more
detailed covenants and they include renegotiation provisions more frequently.
Krishnaswami et al. (1999) also find that firms with high agency costs of debt
prefer to place the issue privately due to the greater monitoring associated
with these placements. Szewczyk and Varma (1991) argue that their finding of a
positive reaction to private placements of debt can be the result of closer
monitoring of the management that reduces agency costs.
The
impact of financial distress on the placement choice of debt has also been
tested. Easterwood and Kadapakkam (1991) hypothesize that private debt can
avoid formal default by allowing the firms to reschedule the debt payments.
However their empirical analysis does not indicate a relationship between the
proportion of private debt and financial distress variables.
Prior
studies find support for the argument that flotation costs affect the choice
between private and public debt. For example for firms that have access to
both public and private markets Blackwell and Kidwell (1988) find that smaller
and riskier firms choose private markets due to the higher flotation costs in
the public markets. Private placements also provide lower yields when interest
rates are volatile and the search costs are high. These firms would have paid
an average of 132 basis points more to sell debt publicly. Hence Blackwell and
Kidwell conclude that firms choose the placement method that results in the
lowest transaction costs. The cross sectional analysis of Easterwood and
Kadapakkam (1991) also shows that firms with low levels of debt rely on
private markets since these firms have high fixed costs of public issuance and
small gains from the liquidity of public markets. Similarly Krishnaswami et
al. (1999) find that larger firms and firms with larger issues have lower
proportion of private debt since they can take advantage of the economies of
scale in issuance costs of public offerings.
5.
Data Selection and Sample Characteristics
Frequency
Distribution of Offerings
The
sample consists of privately and publicly placed convertible debt offerings
completed during the period 1983-2002 by industrial companies.
We require that issuer firms have at least $100 million of assets and
that all issues have issue dates recorded in the Securities Data Corporation
database.
Year |
Private
Offerings |
Public
Offerings |
||
|
Offerings |
Firms |
Offerings |
Firms |
1983 |
2 |
2 |
25 |
25 |
1984 |
0 |
0 |
4 |
4 |
1985 |
7 |
7 |
11 |
11 |
1986 |
7 |
7 |
37 |
35 |
1987 |
6 |
6 |
38 |
36 |
1988 |
2 |
2 |
8 |
8 |
1989 |
4 |
4 |
19 |
19 |
1990 |
3 |
3 |
4 |
4 |
1991 |
4 |
4 |
9 |
9 |
1992 |
4 |
4 |
4 |
4 |
1993 |
3 |
3 |
16 |
15 |
1994 |
2 |
2 |
6 |
6 |
1995 |
1 |
1 |
6 |
5 |
1996 |
4 |
4 |
11 |
11 |
1997 |
5 |
5 |
8 |
8 |
1998 |
3 |
3 |
6 |
6 |
1999 |
2
|
2 |
1 |
1 |
2000 |
2
|
2 |
7 |
7 |
2001 |
11 |
11 |
9 |
9 |
2002 |
6 |
6 |
0 |
0 |
Total |
78 |
78 |
229 |
223 |
|
Total
number of unique firms |
74 |
|
191 |
The
sample consists of privately and publicly placed convertible debt offerings
completed during the period 1983-2002 by industrial companies.
We require that issuer firms have at least $100 million of assets and
that all issues have issue dates recorded in the Securities Data Corporation
database. The table presents the mean (median) values of the descriptive
measures. Total assets is the current assets plus net property, plant, and
equipment plus other noncurrent assets. Market value of equity is the closing
common stock price at the fiscal year-end multiplied by the number of common
shares outstanding. Debt ratio is the sum of long-term debt and debt in
current liabilities, divided by total assets. Coupon rate is the ratio of
coupon payment to the face value of the bond. Maturity is the number of years
until final maturity of the bond. Call protection period is the number of
years before the bond can be called back. Probability of conversion is the
risk-neutralized probability that the bond will be converted into equity.
Firm-specific variables are obtained from Compustat and measured at the fiscal
year-end prior to the issue. Issue-related variables are obtained from SDC
database. Stock price data is from CRSP.
Variable |
All Offerings
|
Private
Offerings |
Public
Offerings |
Total
Assets |
3631.80 (690.30) |
7462.10 (828.20)
|
2327.15 (681.12)
|
Market
value of equity |
2386.92 (622.80) |
3180.35 (635.47)
|
2120.07 (620.83) |
Debt
ratio (%) |
31.57 (28.97) |
33.37 (29.88)
|
30.96 (28.85)
|
Coupon
(%) |
7.32 (7.25) |
6.48 (6.25)
|
7.56 (7.56)
|
Maturity |
25.15 (20.28)
|
46.82 (11.54)
|
17.76 (20.29) |
Call
protection period |
21.15 (5.15) |
26.77 (4.55)
|
17.74 (8.15)
|
Probability
of conversion |
0.36 (0.31)
|
0.07 (0.08)
|
0.37 (0.31)
|
Table
2 also shows the average and median probability of conversion for the
convertible bond issues in the sample. Lewis, Rogalski and Seward (1999) argue
that a high probability of conversion indicates that the convertible bond is
more equity-like whereas low probability of conversion indicates a debt-like
bond. As in Lewis et al. we
measure probability of conversion by the risk-neutralized probability that the
bond will be converted into equity. In the option pricing equation N(d2)
represents the ex-ante
probability that the option will be exercised in a risk-neutral world. N(•)
is the cumulative probability under the standard normal distribution and
d2=[ln(P/X)+(r-div-s²/2)T]/s√T
where,
P is the price of the firm’s
stock on the convertible bond issue date, X
is the conversion price, r is the
risk-free rate calculated as the continuously compounded annual yield on
10-year T-bonds in the issue month, div
is the continuously compounded dividend yield during the fiscal year preceding
the issue date, s is the standard deviation of the continuously
compounded equity return estimated over 240 to 40 trading days prior to the
issue day, and T is the number of
years until maturity. The results in Table 2 show that private offerings are
more debt-like than public offerings. The median probability of conversion is
8 % for privately placed debt and 31 % for public placements. This result is
consistent with the argument that firms that choose private issues are the
ones that face high informational asymmetries. These firms might structure
their bond issues to be more debt-like in order to reduce the adverse
selection costs.
In Table 3 we examine the reasons firms stated for the offerings. This
table shows that the reason for convertible bond issuance is different for
private and public debt placements. For example half of the firms that issue
privately placed debt state the intended use of proceeds as “general
purpose” while only 35 % of public bond issuers state this usage. The
proportions of general purpose issues in private and public offerings are
significantly different from each other at 5% level. Also the funds from about
40 % of public issues is used to refinance existing debt while the funds from
only 33 % of private issues is used for this purpose. Similarly about 6 % of
public issues are used for acquisition financing compared to 2 % of private
offerings.
The
sample consists of privately and publicly placed convertible debt offerings
completed during the period 1983-2002 by industrial companies.
We require that issuer firms have at least $100 million of assets and
that all issues have issue dates recorded in the Securities Data Corporation
(SDC) database. The table presents the reasons private and public convertible
bond issuers stated for the offerings. The stated purpose of each issue is
obtained from SDC database.
The numbers in brackets are Chi-square statistics testing the null hypothesis
that the proportion of offerings with each stated reason in the private
offerings sub-sample is equal to the proportion in the public offerings
sub-sample. a, b, and c denote significant
level at 1, 5 and 10 percent respectively.
Purpose |
Private
Offerings |
Public
Offerings |
Refinancing |
32.81
% |
39.63
% (0.9119) |
General
Purpose |
50.00
% |
35.37
% (4.1264
b) |
Acquisition
Financing |
1.56
% |
6.10
% (2.0619) |
Multiple
Purposes |
15.63
% |
18.29
% (0.2265) |
Other
Purposes |
0.00
% |
0.61
% (0.3920) |
|
|
|
Total |
100.00
% |
100.00
% |
6.
Empirical Analysis
6.1.
Univariate Results
To
test our hypotheses on the determinants of the choice between private and
public placements of convertible bonds we initially perform univariate
analysis. In Table 4 we first test whether firms that choose private
placements have higher levels of asymmetric information compared to firms that
place their bonds publicly. As in Dierkens (1991) and Krishnaswami and
Subramaniam (1999) we measure information asymmetry by the dispersion
in the market-adjusted daily stock returns in the year preceding the issue.
Information asymmetry is higher when managers have a large amount of
firm-specific information that market participants do not have. To the extent
that the investors and managers are equally well-informed about the
economy-wide factors that affect firm value, the residual volatility in the
firm’s stock returns indicates the level of information asymmetry between
investors and the managers about firm-specific information. Hence, the
dispersion in market-adjusted returns captures the level of firm-specific
uncertainty that remains after the uncertainty common to managers and
investors is removed from the total uncertainty. We hypothesized that private
convertibles alleviate information asymmetry problems since the investors of
these offerings are better informed and are more likely to value the firm
accurately (James, 1987). Private placements also signal positive information
to the market since the commitment of the better informed investors sends a
positive signal to the market (Szewcyk and Varma, 1991). Hence firms with high
information asymmetry should prefer private convertibles. Consistent with this
hypothesis we find that the average dispersion in returns is about 0.61 for
private bond issuers compared to 0.46 for public issuers. The difference in
the average dispersion of returns is significant at 1%.
The
sample consists of privately and publicly placed convertible debt offerings
completed during the period 1983-2002 by industrial companies.
We require that issuer firms have at least $100 million of assets and
that all issues have issue dates recorded in the Securities Data Corporation
(SDC) database. The table presents the univariate analysis of the determinants
of the choice between private and public offerings of convertible bonds.
Dispersion in stock returns is measured as the standard deviation of the stock
returns of the issuing firm in days –240 to –40 relative to the issue date
of each debt issue. Issue size is the proceeds of the debt issue divided by
the total assets of the firm. Market-to-book ratio is defined as the closing
price of the fiscal year multiplied by the number of common shares outstanding
divided by the book value of common equity. Financial slack is the ratio of
(operating income before depreciation minus capital expenditures minus change
in net working capital minus net taxes minus change in deferred taxes) to
total assets. Altman’s Z is a combination of several financial ratios where
a high z-value indicates low probability of bankruptcy. High bankruptcy cost
industry is the percentage of firms in the machinery and equipment industry
(SIC code between 3400 and 3999). Firm-specific variables are obtained from
Compustat and measured at the fiscal year-end prior to the issue.
Issue-related variables are obtained from SDC database. Stock return data is
from CRSP. For each continuous variable the mean value is presented first
followed by the median value in parenthesis. The
numbers in Difference column represent p-values of t-tests (Wilcoxon tests)
for the differences in means (medians) for the continuous variables. For high
bankruptcy cost industry indicator variable the figure in the Difference
column is the Chi-square statistics testing the null hypothesis that the
proportion of private offerings in the machinery and equipment industry is
equal to the proportion for the public offerings. a, b,
and c denote significant level at 1, 5 and 10 percent respectively.
Variable |
Private
Offerings |
Public
Offerings |
Difference |
Dispersion
in stock returns |
0.6057 (0.4783) |
0.4635 (0.4237)
|
0.0033
a (0.0179
b)
|
Issue size |
0.1567 (0.1152) |
0.3097 (0.1594)
|
0.0005
a (0.0235
b)
|
Market-to-book
|
3.9001 (2.0660) |
2.3431 (1.9220)
|
0.0708
c (0.4814)
|
Financial
slack |
85.2847 (13.8730)
|
24.606 (1.9170)
|
0.4304 (0.2675)
|
Altman’s
Z |
3.8291 (3.1620)
|
3.6577 (3.1440)
|
0.6649 (0.7058) |
High
bankruptcy cost industry (%) |
24.36 |
29.69 |
0.8156 |
We also
hypothesized that private placements reduce the risk-shifting, underinvestment
and free cash flow problems through better monitoring associated with these
offerings. Since the risk-shifting and underinvestment problems are more
severe for firms with high growth options we expect these firms to
particularly prefer private placements. Barclay and Smith (1995) argue that
the firm’s balance sheet does not include intangible assets like growth
options and growth options increase firm’s market value in relation to its
book value. Hence we proxy the extent of the growth opportunities of the firm
with the market-to-book ratio defined as the stock price multiplied by the
company's common shares outstanding divided by common equity. Consistent with
this hypothesis we find that the mean and median market-to-book ratio is
higher for private offerings compared to public offerings, although only the
difference in the means is significant. In Table 4 we define financial slack
(free cash flow) as the ratio of (operating income before depreciation minus
capital expenditures minus change in net working capital minus net taxes minus
change in deferred taxes) to total assets. This measure shows the cash flow in
excess of that required to finance the projects and pay for the tax
obligations of the firm. We find that firms that place convertible bonds
privately have higher levels of financial slack compared to firms that use
public placements. However the difference in the mean and median levels of
financial slack is not significant. Hence we do not find support for our
hypothesis that private placements reduce free cash flow problems.
We also test
whether expected financial distress costs affect the decision to choose
between private and public placements of convertible bond issues. Prior
studies show that private offerings might allow firms to avoid formal default
(Easterwood and Kadapakkam, 1991); provide flexibility in renegotiation of
debt (Sy, 1999; Denis and Mihov, 2003); and offer better reorganization
following financial distress (Cantillo and Wright, 2000). Hence firms with a
high probability of financial distress and firms that incur high
distress-related costs in case of default should choose private placements for
their convertible bond offerings. We measure the likelihood of financial
distress with Altman’s Z-score. This variable is a function of several
firm-specific variables and is calibrated so that higher scores represent
lower probability of distress[6].
Altman (1968) finds that his model can correctly assign 95 % of the sample
firms in bankrupt and non-bankrupt groups. Table 4 shows that the average and
median Altman’s Z-scores are not significantly different for private and
public offerings. Titman (1984) and Titman and Wessels (1988) argue that the
liquidation costs and hence the expected bankruptcy costs are higher for firms
with unique products. The workers and suppliers of these firms have
job-specific skills and capital, and the customers might find it difficult to
find alternative service providers for the unique products. Therefore
following Titman and Wessels (1988) and Fisher, Heinkel, and Zechner (1989),
we measure expected bankruptcy costs with an indicator variable that takes the
value of one for firms in the machinery and equipment industry (SIC code
between 3400 and 3999) and zero otherwise. The results indicate that the
percentage of private bond issuers with higher expected bankruptcy costs is
24.36 % while the percentage for public issuers is 29.69 %. The difference in
the percentages is insignificant. These results show that financial distress
costs do not affect the placement choice of convertible debt.
We also
hypothesized that firms with smaller issues prefer private offerings since the
fixed component in the flotation costs of these issues are higher for public
issues (Blackwell and Kidwell, 1988). For example the investment banker’s
compensation is lower in private placements because the investment banker
serves only as a finder that brings together the issuers and investors. The
distribution costs are also lower in private placements since the number of
investors is smaller. We measure issue size by the proceeds raised from the
issue divided by the total assets of the firm. Consistent with our hypothesis
we find that the issue size is smaller for private convertible bond issues[7].
On average the proceeds of the new issues represent about 16 % of the assets
of private issues compared to 31 % for public issues. The difference in the
mean and median issue sizes are significant and consistent with the findings
of Blackwell and Kidwell (1988).
6.2.
Logit Regression Results
In Table 5 we
use logit regressions to study the determinants of the choice between private
and public convertible bonds. With this incremental approach we are able to
relate the placement decision of the company with the firm characteristics. In
the regressions in Table 5 the dependent variable is an indicator variable
that takes the value of one for private offerings and zero for public
offerings. The regressions confirm our earlier findings for the influence of
information asymmetry on the placement choice. The coefficient of the
dispersion in returns is positive and significant at one percent in all
regressions. Hence, even after we account for other variables that might
affect the choice between private and public convertibles, the results show
that firms with high information asymmetry are more likely to use private
markets for their convertible debt. Logit regressions also show that firms
that issue smaller offerings are more likely to choose private placements.
This result confirms the univariate analysis results and shows that flotation
costs are important considerations in debt placement. Firms with smaller
issues do not want to incur the higher fixed costs of public placements and
therefore prefer private issues[8].
Logit Analysis of the
Placement Choice
The sample consists
of privately and publicly placed convertible debt offerings completed during
the period 1983-2002 by industrial companies.
We require that issuer firms have at least $100 million of assets and
that all issues have issue dates recorded in the Securities Data Corporation
database. The table presents the logit analysis on the determinants of the
choice between private and public offerings of convertible bonds. The
dependent variable takes the value of 1 for private offerings and zero for
public offerings. Dispersion in stock returns is measured as the standard
deviation of the stock returns of the issuing firm in days –240 to –40
relative to the issue date of each debt issue. Issue size is the proceeds of
the debt issue divided by the total assets of the firm. Market-to-book ratio
is defined as the closing price of the fiscal year multiplied by the number of
common shares outstanding divided by the book value of common equity. The
interaction variable return dispersion is the information asymmetry indicator
variable that takes the value of one for dispersion of stock returns greater
than the sample median and zero otherwise. Financial slack is the ratio of
(operating income before depreciation minus capital expenditures minus change
in net working capital minus net taxes minus change in deferred taxes) to
total assets. Bankruptcy industry is an indicator variable that takes the
value of one for firms in the machinery and equipment industry (SIC code
between 3400 and 3999) and zero otherwise. Altman’s Z is a combination of
several financial ratios where a high Z-value indicates low probability of
bankruptcy. Chi-squared statistics are presented in parentheses.
a, b, c represent significance at the 1%, 5%, and 10% levels
respectively.
Regressions |
|||||
|
1 |
2 |
3 |
4 |
5 |
Intercept |
-2.0392
a (24.3446) |
-1.6448
a (21.4126) |
-2.3922
a (20.9282) |
-2.4690
a 20.6067) |
-2.2448
a (17.2378) |
|
|
|
|
|
|
Dispersion
in stock returns |
1.6258
a (8.2278) |
1.5344
a (8.6086) |
2.0704
a (8.1350) |
2.0928
a (8.2440) |
2.1797
a (8.3220) |
|
|
|
|
|
|
Issue
size |
-1.8047
b (4.3336) |
-1.8813
b (4.8729) |
-1.7729
b (3.5664) |
-1.8303
b (3.7135) |
-2.0072
b (4.1889) |
|
|
|
|
|
|
Market-to-book |
|
|
0.2157
c (3.2447) |
0.1919 (2.3416) |
0.2391
b (3.6779) |
|
|
|
|
|
|
Market-to-book
* Return dispersion |
|
|
-0.1710 (1.9613) |
-0.1567 (1.6411) |
-0.1780 (1.9320) |
|
|
|
|
|
|
Financial
slack |
|
|
0.000746
c (2.6903) |
0.000779
c (2.8462) |
0.000811
c (2.9111) |
|
|
|
|
|
|
Financial
slack * Return dispersion |
|
|
-0.00098 (1.4418) |
-0.00101 (1.5226) |
-0.00100 (1.4652) |
|
|
|
|
|
|
Bankruptcy
Industry |
|
-0.3533 (1.1169) |
|
|
-0.8021
c (3.3892) |
|
|
|
|
|
|
Altman’s
Z |
0.0727 (2.0276) |
|
|
0.0315 (0.2546) |
|
|
|
|
|
|
|
Year |
0.4171 (1.7390) |
0.4916 (2.6572) |
0.3732 (1.1064) |
0.3662 (1.0703) |
0.2933 (0.6550) |
|
|
|
|
|
|
Likelihood
Ratio |
22.5157
a |
26.5423
a |
26.3576
a |
26.6027
a |
30.0797
a |
Concordant
Responses |
66.4
% |
68.4
% |
69.5
% |
69.8
% |
71.6
% |
N |
264 |
281 |
222 |
222 |
222 |
The coefficients
of the market-to-book ratio and financial slack are overall positive in the
regressions in Table 5. This result is consistent with our hypothesis that the
better monitoring offered by the investors of private placements results in
lower agency costs. Hence firms with higher potential risk-shifting,
underinvestment, and free cash flow problems prefer private offerings over
public offerings. We also hypothesized that firms with high information
asymmetry problems are more likely to choose private placements when they have
better investment opportunities and/or little financial slack. In Table 5 we
test this hypothesis using interactive variables. The first interactive
variable is composed of the product of the market-to-book ratio of the firm
and an information asymmetry dummy variable. The information asymmetry dummy
takes the value of one for firms with dispersion in stock returns higher than
the sample median and zero otherwise. A positive coefficient for this
interactive variable indicates that firms with better investment opportunities
and high information asymmetry are more likely to choose private placements.
The second interactive variable is the product of the financial slack and the
information asymmetry dummy. The coefficient of neither of these variables is
significant. Hence we do not find support for the argument of Hertzel and
Smith (1993) that information asymmetry problems are more costly for firms
with high investment opportunities and firms with little financial slack.
In
Table 5 we also confirm our univariate results on the influence of the
expected bankruptcy costs on the choice between private and public convertible
debt placement. We find that the coefficients of the bankruptcy cost dummy and
Altman’s Z variables are overall insignificant. Hence we do not find support
for our hypothesis that firms with high expected bankruptcy costs are more
likely to issue private convertible bonds. This result is consistent with
Easterwood and Kadapakkam’s (1991) finding for debt in general.
Our results
might have been influenced by the changing market conditions during the sample
period. In order to account for the impact of the market conditions on the
convertible bond placement choice in each regression we also include a dummy
variable that takes the value of one for issues made in the second half of the
sample period and 0 for those made in the first half. The coefficient of this
variable is insignificant in all regressions. This result indicates that the
time period of issuance does not affect the placement choice of convertible
bonds.
In
Table 6 we measure the relative economic impact of each variable included in
Models 4 and 5 of Table 5. We use two methods to estimate the relative impact
of each variable on the convertible bond placement choice: marginal
probabilities and step-wise regressions. In Panel A the marginal probabilities
show the relative magnitude of the effect of each variable on the choice
between private and public placements of convertible bonds. Marginal
probability is calculated as the increase in the probability that the
dependent variable takes the value of 1 for a one quartile change in the value
of each independent variable, while holding all other independent variables
constant at their medians. In other words the economic impact of each variable
is measured by the increase in the probability of issuing private convertible
bonds for a change in the independent variable from its median to either the
25th or 75th percentile (whichever leads to an increase
in the probability), holding all other variables at their median. The results
show that information asymmetry is the most influential variable in the
convertible debt placement decision. In Model 4, a one quartile increase in
the dispersion of stock returns increases the probability of issuing private
convertibles from 22.26 % to 48.70 % (an increase of 26.44 %).
The sample consists
of privately and publicly placed convertible debt offerings completed during
the period 1983-2002 by industrial companies.
We require that issuer firms have at least $100 million of assets and
that all issues have issue dates recorded in the Securities Data Corporation
database. The table presents the relative importance of the variables in
Models 4 and 5 in Table 5. Panel A shows the marginal probabilities of each
variable where marginal probability is computed as the increase in the
probability of issuing convertible bonds for a change in the independent
variable from its median to either the 25th or 75th
percentile (whichever leads to an increase in the probability), holding all
other variables at their median. Dispersion
in stock returns is measured as the standard deviation of the stock returns of
the issuing firm in days –240 to –40 relative to the issue date of each
debt issue. Issue size is the proceeds of the debt issue divided by the total
assets of the firm. Market-to-book ratio is defined as the closing price of
the fiscal year multiplied by the number of common shares outstanding divided
by the book value of common equity. The interaction variable return dispersion
is the information asymmetry indicator variable that takes the value of one
for dispersion of stock returns greater than the sample median and zero
otherwise. Financial slack is the ratio of (operating income before
depreciation minus capital expenditures minus change in net working capital
minus net taxes minus change in deferred taxes) to total assets. Bankruptcy
industry is an indicator variable that takes the value of one for firms in the
machinery and equipment industry (SIC code between 3400 and 3999) and zero
otherwise. Altman’s Z is a combination of several financial ratios where a
high Z-value indicates low probability of bankruptcy. The overall probability
of issuing convertible bonds based on each regression model is also presented
below each regression, and is computed at the median for all the variables.
Panel B shows summary of the step-wise regression procedure.
Panel A: Marginal Probabilities |
|||
|
Marginal
Probability |
||
Assumed
Change in Variable |
Model
4 |
Model
5 |
|
Dispersion
in stock returns |
0.1352 |
26.4352 |
25.4491 |
Issue size |
-0.1052 |
17.1993 |
17.7151 |
Market-to-book |
1.0605 |
17.8179 |
21.4358 |
Market-to-book
* Return
dispersion |
-0.3250 |
4.1084 |
4.3106 |
Financial
slack |
78.9810 |
5.0487 |
4.9688 |
Financial
slack * Return
dispersion |
0.0000 |
0.4296 |
0.3942 |
Altman’s
Z |
1.7460 |
4.6907 |
|
Bankruptcy
industry |
0.0000 |
|
0.0000 |
|
|
|
|
Private
issue probability |
|
0.2226 |
0.2583 |
PANEL B: Summary of
Step-wise Selection
|
|||||
Model |
Effect |
Score Chi-Square |
Wald Chi-Square |
Pr
> Chi-Square |
|
Entered |
Removed |
||||
4 |
Dispersion
in stock returns |
|
9.7318 |
- |
0.0018 |
|
Issue size |
|
4.0536 |
- |
0.0441 |
|
|
|
|
|
|
5 |
Dispersion
in stock returns |
|
9.7318 |
- |
0.0018 |
|
Issue size |
|
4.0536 |
- |
0.0441 |
|
Bankruptcy
industry |
|
2.7533 |
|
0.0971 |
|
|
Bankruptcy
industry |
- |
2.6982 |
0.1005 |
The
results in Panel A of Table 6 show that agency problems due to risk-shifting,
underinvestment, and free cash-flow problems are less influential than the
information asymmetry problem in the convertible debt placement choice. For
example in Model 5 the change in the probability of private placements is
25.45 % for the dispersion in stock returns compared to 21.44 % for the
market-to-book ratio and 4.97 % for financial slack. This result is consistent
with Green (1984) and Mayers and Smith (1987) who argue that risk-sifting and
underinvestment are not as important problems for convertible bonds. In
Green’s model the conversion option allows the bondholders to take advantage
of the upside potential of the firm, thereby reducing the value of limited
liability. As a result the risk-shifting problem is reduced. In Mayers and
Smith managers are less likely to forego valuable investment projects with
convertibles since the equity characteristics of the bond reduce the fixed
component. These results are also consistent with Goh, Gombola, Lee, and Liu
(1999) who find that information asymmetry is more influential in private
placements compared to agency costs[9].
Panel
A of Table 6 also shows that flotation costs have a high economic impact. In
Model 5 a one quartile decrease in the issue size increases the probability of
issuing private convertibles from 25.83 % to 43.55 % (an increase of 17.72 %).
However, consistent with Table 5, the results in Panel A of Table 6 show that
bankruptcy costs do not have a major impact on the probability of issuing
private convertibles. For example the marginal probability of Altman’s Z is
4.69 in Model 4. The economic impacts of the interactive variables are also
low.
Panel
B of Table 6 shows the results of the step-wise regressions used for Models 4
and 5. In these regressions each variable is evaluated on the basis of its
significance level and the model is built by adding and deleting variables
sequentially. The results show that dispersion of stock returns, our proxy
variable for information asymmetry, has the highest significance level. This
finding confirms our conclusion in Panel A that information asymmetry is the
most influential factor in the choice between private and public placements of
convertible bond offerings. The results in Panel B of Table 6 also confirm our
finding that flotation costs, measured by the issue size, have a high
influence on the probability of issuing private convertible bonds.
Agency cost and financial distress variables are not included in the
model and therefore are less influential on the placement choice.
7.
Conclusion
In this paper we
study whether information asymmetry problems, agency costs, financial distress
costs and flotation costs affect firms’ decision to place convertible bonds
privately instead of going to public markets. We study 78 private and 229
public offerings of convertible bonds issued between 1983 and 2002 by
industrial firms. Our results show that information asymmetry problems are the
major determinants of the placement decision of convertible debt. We
hypothesize that since investors of private placements are better informed
about the firm, the issuer will be valued more accurately with this type of
offering. The commitment of private investors to the issue also acts as a
quality certification. Consistent with these arguments we find that firms with
high levels of information asymmetry are more likely to choose private
offerings compared to public offerings.
Agency
costs also affect the choice between private and public offerings of
convertible bonds. We find that firms that are more susceptible to agency
problems of risk-shifting, underinvestment and free cash flow choose private
placements. However our results show that the economic impact of agency costs
are not as important for convertible bond issuers as information asymmetry
problems. This finding is consistent with the argument that due to their
equity-like nature convertible bonds face high adverse selection costs but
they do not have high agency costs since the conversion option reduces the
risk-shifting and underinvestment problems. We also test the interaction of
agency costs and information asymmetry problems on the placement choice. Our
results do not support the argument that the opportunity cost of foregoing
valuable projects will be higher for firms with high information asymmetry
when these firms have good investment opportunities and little financial
slack.
Our
results indicate that expected flotation costs are also important in the
placement choice of convertible bonds. We hypothesize that in order to avoid
the higher fixed costs of public offerings firms with smaller issues will
prefer private placements. Consistent with this hypothesis we find that firms
with lower offering proceeds relative to firm size are more likely to issue
private convertibles. We fail to find support for our hypothesis that firms
with high expected financial distress costs choose private placements due to
the higher flexibility of renegotiation offered by these offerings.
Altman,
E.I., 1968, Financial ratios, discriminant analysis and the prediction of
corporate bankruptcy, Journal of Finance 23, 589-609.
Altman,
E.I., 1984, A further empirical investigation of the bankruptcy cost question,
Journal of Finance 39, 1067-1089.
Barclay,
M. and C. Smith, 1995, The maturity structure of corporate debt, Journal
of Finance 50, 609–631.
Bayless, M. and S.
Chaplinsky, 1996, Is there a window of opportunity for seasoned equity
issuance? Journal of Finance 51,
253–278.
Blackwell
D.W. and D.S. Kidwell, 1988, An investigation of cost differences between
public sales and private placements of debt, Journal of Financial Economics
22, 253-278.
Bodie,
Z., A. Kane, and A.J. Marcus, 2002. Investments ((McGraw-Hill Irwin, New
York).
Cantillo,
M. and J. Wright, 2000, How do firms choose their lenders? An empirical
investigation, Review of Financial Studies 13, 155-189.
Denis
D.J. and V.T. Mihov, 2003, The choice among bank debt, non-bank private debt,
and public debt: evidence from new corporate borrowings, Journal of
Financial Economics 70, 3-28.
Dierkens,
N., 1991, Information asymmetry and equity issues, Journal
of Financial and Quantitative Analysis 26, 181–199.
Easterwood
J.C. and P. Kadapakkam, 1991, The role of private and public debt in corporate
capital structures, Financial Management 20, 49-57.
Eckbo,
B.E., 1986, Valuation effects of corporate debt offerings, Journal
of Financial Economics 15, 119-151.
Fields,
L.P. and E.L. Mais, 1991, The valuation effects of private placements of
convertible debt, Journal of Finance 46, 1925-1932.
Fischer,
E.O., R. Heinkel, and J. Zechner, 1989, Dynamic capital structure choice:
Theory and tests, Journal of Finance
44, 19–40.
Froot,
K., D. Scharfstein, and J. Stein, 1993, Risk management: Coordinating
corporate investment and financing policies, Journal
of Finance 48, 1629-1658.
Goh,
J., M.J. Gombola, H.W. Lee, and F. Liu, 1999, Private placement of common
equity and earnings expectations, Financial Review 34, 19-32.
Green,
R., 1984, Investment incentives, debt, and warrants, Journal
of Financial Economics 13, 115–136.
Grice, J.S. Sr., 2000,
Bankruptcy Prediction Models and Going Concern Audit Opinions Before and After
SAS No. 59, B>Quest.
Hertzel M. and R.L.
Smith, 1993, Market discounts and shareholder gains for placing equity
privately, Journal of Finance 48, 459-485.
James,
C, 1987, Some evidence on the uniqueness of bank loans, Journal of
Financial Economics 19, 149-171.
Jensen
M., 1986, Agency costs of free cash flow, corporate finance, and takeovers, American
Economic Review 76, 323–329.
Jensen,
M. and W. Meckling, 1976, Theory of the firm: Managerial behavior, agency
costs, and ownership structure, Journal
of Financial Economics 3, 305–360.
Korajczyk,
R. A., D. J. Lucas and R. L. McDonald, 1992, Equity issues with time-varying
asymmetric information, Journal of
Financial and
Quantitative
Analysis 27, 397-417.
Krishnaswami,
S., P.A. Spindt, and V. Subramaniam, 1999, Information asymmetry, monitoring,
and the placement structure of corporate debt, Journal of Financial
Economics 51, 407-434.
Krishnaswami,
S. and V. Subramaniam, 1999, Information asymmetry, valuation, and the
corporate spin-off decision, Journal of
Financial Economics 53, 73–112.
Lewis,
C., R. Rogalski, and J. Seward, 1999, Is convertible debt a substitute for
straight debt or for common equity? Financial
Management 28, 5–27.
Mayers,
D. and C. Smith, 1987, Corporate insurance and the underinvestment problem, Journal
of Risk and Insurance 54, 45–54.
Myers,
S., 1977, Determinants of corporate borrowing, Journal
of Financial Economics 5, 147–175.
Myers,
S. and N.S. Majluf, 1984, Corporate financing and investment decisions when
firms have information that investors do not have, Journal
of Financial Economics 13, 187–221.
Paxton,
D. and D. Wood, 1998. Encyclopedic Dictionary of Finance (Blackwell
Publishers, Malden, Massachusetts).
Ross,
S.A., R.W. Westerfield, and J.F. Jaffe, 2005.
Corporate Finance (McGraw-Hill Irwin, New York).
Slovin,
M. B., M. E. Sushka, and J. A. Polonchek, 1991, The information content of
multiple seasoned common stock offerings by bank holding
companies,
Journal of Banking and Finance 15,
633-646.
Smith,
C.W. and J.B. Warner, 1979, On financial contracting: An analysis of bond
covenants, Journal of Financial Economics 15, 3-29.
Smith,
C. W. and R.L. Watts, 1992, The investment opportunity set and corporate
financing, dividend, and compensation policies, Journal
of Financial Economics 32, 263–292.
Szewczyk,
S.H. and R. Varma, 1991, Raising capital with private placements of debt, Journal
of Financial Research 14, 1-13.
Sy, A.N.R., 1999,
Managerial entrenchment and the choice of debt financing, IMF Working Paper.
Titman,
S., 1984, The effect of capital structure on a firm’s liquidation decision, Journal
of Financial Economics 13, 137–151.
Titman,
S. and R. Wessels, 1988, The determinants of capital structure choice, Journal
of Finance 43, 1–19.
[1]
Myers (1977) argues that having a close relationship with the debt issuer
mitigates the underinvestment problem.
[2]
SDC is a database provided by Thomson Financial Services. The public
offerings in this database include all issues underwritten in the U.S. and
are updated daily. The data for public offerings is gathered from SEC
filings, prospectuses, news sources, wires, and daily surveys of
underwriters and other corporate finance contacts of the company. Private
placements in the database include all issues placed privately with U.S.
investors by private placement agents and are updated semi-annually. Private
placement data is obtained through surveys of placement agents and financial
news sources.
[3]
Denis and Mihov (2003) argue that SDC data excludes smaller issues. However
since our sample includes larger firms (which tend to issue larger issues)
this is not a problem for our study.
[4]
20 of the total 245 firms in our sample
issued both private and public convertible bonds during the sample period.
However none of these firms made both private and public offerings in the
same year. Hence the firms might have changed the type offering as a result
of the change in their characteristics over the years.
[5]
To see the whether outliers affect the statistics on debt maturity we also
calculate the mean and medians after winsorizing 1% of each tail. The
deletion of these observations decreased the mean (median) maturity of the
sample to 16.33 (18.11) years. The mean (median) maturity of the private
convertibles also decreased to 8.35 (7.06) years while those of public
offerings remained the same.
[6]
As in Altman (1968) the Z-score is
calculated as Z = 0.012X1 + 0.014X2 + 0.033X3 + 0.006X4 + 0.999X5 where X1=
Working Capital/Total Assets, X2=Retained Earnings/Total Assets, X3=
Earnings Before Interest and Taxes/Total Assets, X4= Market Value of
Equity/Book Value of Total Debt, and X5= Sales/Total Assets. Grice (2000)
shows that the bankruptcy prediction models like Altman’s are better than
auditors at signaling the future prospects of companies.
[7]
Our finding on issue size is also
consistent with the argument that firms with smaller issues find private
offerings more convenient and avoid the hassles of public offering
procedures.
[8] We also use the natural logarithm of the proceeds of the issue to measure issue size. This alternative proxy also has a negative coefficient in the logit regressions and confirms our results.
[9]
The focus of Goh et al. (1999) study is equity issues.