Exchangeable debt gives the purchaser the option to exchange the debt for stock of a second company, referred to throughout this paper as the "convert" firm. For example, in March of 1985, Petrie Stores issued $150 million of exchangeable callable debt, due in 2010. The exchange feature enabled the purchaser of the debt to exchange each $1000 face value of debt for just over 27 shares of Toys "R" Us common stock. Petrie Stores owned a minority interest in Toys "R" Us and deposited a sufficient number of Toys "R" Us common with an escrow agent to guarantee the exchange option. Exchangeable debt has been offered by firms since the early 1970s and accounted for ten percent of equity-linked debt offered by corporations in 1981 while averaging almost six percent of this market from 1981 through 1987 (these estimates exclude six exchangeable debentures offered as Eurobonds).
The major conclusion of this research is that exchangeable debt constitutes a divestment strategy which has no unique tax advantages over other divestment strategies such as block sales or secondary distributions. Despite the fact that investment houses marketed exchangeable debt as a means by which corporations could capitalize on specific features of the tax code, none of the tax explanations considered provides a compelling motivation for issuing exchangeable debt. The most compelling justification for issuing exchangeable debt is non-tax-related and has previously been ignored -- namely, the underwriting fees and valuation effects associated with exchangeable debentures are substantially less than those associated with secondary distributions where the investor is left holding the convert's stock.
The first conclusion of this paper is that firms issue exchangeable debt conditional on having made a decision to divest of an intercorporate holding. Firms, on average, divest of a block of stock when there is negative information regarding the future prospects of the firm. The documented negative price reactions at the announcement of secondary distributions (Mikkelson and Partch |5~), block sales (Holthausen, Leftwich and Mayers |3~), and equity issues (Asquith and Mullins |1~) are all consistent with this view. Exchangeable debt is an alternative method of divesting of a block of stock which likewise is used when the convert firm's future prospects are not good. For example, IBM disposed of its Intel stock through an exchangeable debt issue because of Intel's "gloomy prospects."(1) Similarly, Intermark disposed of a portion of its Anthem Electronics common through an exchangeable debt issue after several years of poor earnings performance by Anthem. These facts are consistent with the view that exchangeable debt is essentially a divestment strategy.
The average abnormal price response of the convert firm's stock on the announcement of an exchangeable debt offering is -1% over a two-day event window and reliably nonzero with an associated t-statistic of -2.30. This price response is less than that documented by Mikkelson and Partch |5~ of -1.96% for nonregistered secondary offerings and -2.87% for registered secondary offerings. I argue the negative price response in the convert's stock is a result of information brought to the market regarding the convert's stock by the announcement of the exchangeable debt issue and is smaller in magnitude than the negative abnormal price response observed on the announcement of the sale of a block of equity by an informed party because of the repurchase guarantee implicit in exchangeable debt. Issuing firms offer a repurchase guarantee to investors in the exchangeable debt issue by guaranteeing to retain the convert firm's stock should that stock not appreciate subsequent to the issue. This is a much different outcome compared to that of an investor purchasing stock in a secondary distribution where the investor is left holding the stock.
The second question this research addresses is why exchangeable debt is chosen over alternative divestment strategies. Unfortunately, the answer here is not obvious. In this paper, I analyze two tax benefits posed by Jones and Mason |4~ and frequently cited in the financial press as motivating exchangeable debt issues. First, exchangeable debt allows the issuing firm to defer the realization of the capital gain on the convert's stock, if any, until the conversion feature is exercised. Second, a kind of tax arbitrage is available to issuers of exchangeable debt. The coupon payments to the bondholders are deductible as interest expense while the dividends received on the convert's stock receive a corporate dividend tax exclusion. Jones and Mason |4~ argue that this differing tax treatment effectively lowers the coupon paid by the issuing firm on exchangeable debt issues. Neither of these tax hypotheses receive support from the data. Though particular issues of exchangeable debt were arguably motivated by tax considerations, many of the offerings clearly were not motivated by an attempt to capture dividend income on the convert stock (eight issues involved convert firms which were non-dividend-paying) or to defer the realization of capital gains (three issues involved convert stock on which the issuing firm would have realized a capital loss). Most importantly, though, there is no evidence that the tax considerations are a potential source of value for the issuing firms. Issuing firms, on average, experience no abnormal price response on the announcement of exchangeable debt issues.
Finally, I document that the underwriting costs of exchangeable offerings are less than those of secondary offerings making it a lower cost means of disposing of a block of stock. Though this appears to be a reasonable justification for using exchangeable debt as a divestment alternative, the paucity of exchangeable debt issues suggests this fact is either not well known or that there are other unidentified costs (for example, the negative information revealed by the issue) associated with the issuance of exchangeable debt.
The remainder of this paper is organized as follows. Section I contains a summary of the data used. Section II analyzes the price response of the convert and issuing firms. Section III describes possible motivations for issuing exchangeable debt. Concluding remarks are made in Section IV.
TABULAR DATA OMITTED
I. The Sample
I identified exchangeable debt offerings by reading footnotes to convertible bonds identified in Moody's Bond Record. The Capital Changes Reporter, Corporate Finance Sourcebook, and Registration and Offerings Statistics (ROS) database (compiled by the Securities and Exchange Commission) were used to identify the issue date of the debt securities. To identify the announcement dates of the debt offerings, I obtained the filing dates of the offerings. This procedure resulted in the identification of 37 exchangeable debt offerings over the period 1970 through 1987. Exhibit 1 lists the 37 exchangeable debt offerings and the percentage of the convert firm owned by the issuing firm, which ranges from no stake in the case of a subsidiary issuing debt exchangeable into the common stock of a parent to wholly owned in the case of a parent issuing debt exchangeable into the common stock of a subsidiary. The maturity of the debt ranges from ten to 25 years, with the majority having 25 years to maturity when issued.
TABULAR DATA OMITTED
The 37 exchangeable debt offerings constitute 32 unique issuer/convert pairs. Exhibit 2 summarizes the major means of acquisition for the 32 unique issuer/convert pairs. The means by which the issuing firms acquired the convert firms are quite varied. Appendix B contains detailed descriptions of the means of acquisition for each of the issuer/convert pairs. The most common relation between the issuing and convert firms is that the convert firm was the subject of an attempted takeover by the issuing firm. This indicates that, at some point, the issuing firm viewed the convert firm as undervalued (possibly conditional on a change in management). The second most common relation between the issuing and convert firms is one in which the convert firm was the subject of an equity carve-out(2) by the issuing firm.
I obtained stock returns for the issuing and convert firms from the daily stock files compiled by the Center for Research in Security Prices (CRSP). Of the 30 issuing firms (representing 37 observations), returns data are available for 25 firms (representing 31 observations). Six issuing firm observations are not used in this analysis: three due to the lack of publicly traded common stock, one due to poor documentation of the announcement date, and two due to the absence of returns data around the announcement date. Of the 32 convert firms (representing 37 observations), returns data are available for 27 firms (representing 30 observations). Seven convert firm observations are not used in this analysis: one due to poor documentation of the announcement date and six due to the absence of returns data around the announcement date.
Of the 37 exchangeable debt offerings, 16 were outstanding as of December 31, 1988;(3) 21 had been called or otherwise retired. Six of the exchangeable debentures were offered as Eurobonds.(4)
Descriptive statistics for various characteristics of the 37 exchangeable debentures are provided in Exhibit 3. On average, the percentage of the convert's common stock owned by the issuing firm is 22% and represents 18% of the market value of the issuing firm. There is no clear relation between the size of the issuing firm and the size of the convert firm, and, while the majority of the conversion premiums(5) lie in the 20 to 30% range, they range as high as 130% and as low as five percent.
II. The Valuation Effects of Issuing Exchangeable Debt
In this section, I consider the valuation effects of issuing exchangeable debt to the convert and issuing firm. In the TABULAR DATA OMITTED next section, I analyze how the valuation effects documented here relate to the various hypotheses posed to explain the existence of exchangeable debt. The major conclusions of this section are:
(i) The convert firms experience, on average, a -1.0% abnormal price response on the announcement of an exchangeable debt issue.
(ii) The average abnormal price response of the issuing firm's common stock is not reliably different from zero at the announcement of the issue.
(iii) The price response of the issuing firm at the announcement of an exchangeable debt issue is positively correlated with the price response of the convert firm's price response at announcement of the issue.
A. Event Study Methodology
The market model(6) is used to estimate abnormal returns (A|R.sub.it~) and cumulative abnormal returns (CA|R.sub.it~). The announcement day is defined as the earlier of the SEC filing date or the day preceding the Wall Street Journal publication date for each exchangeable debt issue. The parameter estimation period begins 51 days after the announcement day and ends 250 days after. The event window comprises the 101 days centered around the announcement day. Test statistics are estimated using the cross-sectional standard deviation of the abnormal return for each day.
Exchangeable debt is a divestment strategy with a potentially informed investor (the issuer) disposing of a block of stock (the convert firm). For example, IBM was a major purchaser of Intel's products at the time of IBM's exchangeable debt issue, three of the issuers were parent companies and several of the issuers were prospective bidders in acquisition attempts. Accordingly, the abnormal price response of the convert firm at the announcement of the exchangeable debt issue should be negative and similar to other documented abnormal price responses at the announcement of the sale of a block of stock by an informed party. The financial press is replete with anecdotal evidence concerning the negative implications of an exchangeable debt issue about the future prospects of the convert firm.
Data constraints result in 27 convert firms providing 30 observations and 25 issuing firms providing 31 observations. Exhibit 4 contains the average abnormal return for event day t, |Mathematical Expression Omitted~, and the cumulative average abnormal return within the event window through event day |Tau~, CA|R.sub.|Tau~~, the associated t-statistics, and the percentage of abnormal returns greater than zero around the announcement day. I discuss the results for the convert firms and issuing firms in turn.
Around the announcement day, the cross-sectional average abnormal return for the convert firms, |Mathematical Expression Omitted~, is TABULAR DATA OMITTED negative from event day t = -2 through day t = 6 and is reliably negative on day O. Moreover, the average abnormal return summed over the announcement event days t = (0,1) yields a negative abnormal return of -1.10% (t = -2.30) with 73% of the sample firms experiencing negative abnormal returns over the two-day window.(7)
The most plausible explanation of the negative two-day abnormal return for the convert firms on the announcement of an exchangeable debt issue is that information is revealed regarding the relation between the issuing and convert firms. The most transparent example is the case of a prospective merger. Issuing exchangeable debt is essentially a divestment strategy and thus would signal that there is at least a decreased likelihood of the existence of a bidder-target relation between the issuing and convert firms. This reduced likelihood in essence reveals not so much that the convert firm is overvalued, but, that the convert firm is no longer considered undervalued by the issuing firm and is thus a less likely target candidate. This revelation could account for the observed negative abnormal price response in the convert's stock.
This explanation is consistent with the fact that a significant portion of the convert firms were acquired by the issuing firms in attempted acquisitions. Anecdotal evidence can also be found in the case of CIGNA issuing exchangeable debt on its holding in Paine Webber. The Wall Street Journal (November 17, 1982) reported:
The |Paine Webber~ stock has fallen for the past two days |following the announcement of the exchangeable debt issue~ after running up sharply on rumors of a merger with CIGNA.
Such coverage of an exchangeable debt issue is common in the financial press.
The abnormal price response of the convert firm's common stock at the announcement of an exchangeable debt offering is -1.0%. Significantly, this price response is less than the -1.96% abnormal price response for nonregistered secondary offerings and -2.87% abnormal price response for registered secondary offerings documented by Mikkelson and Partch |5~. The less pronounced negative price response on the announcement of an exchangeable debt offering is consistent with the notion that exchangeable debt offers an investor a repurchase guarantee limiting the losses of investors to the price of the implicit call option. Unlike secondary offerings or equity issues, exchangeable debt offers an investor the guarantee of a "floor" should the stock price of the convert firm fall subsequent to issue. The guarantee of the floor on losses to investors appears to mitigate the information effects of the announcement relative to the information effects of secondary distributions. A second plausible explanation of the smaller price response associated with exchangeable debt issues relative to secondary distributions is that an issuer of exchangeable debt, although well-informed, may, on average, not be as well-informed as the issuers of secondary distributions.(8)
Around the announcement day, the cross-sectional average abnormal returns for the issuing firms, |Mathematical Expression Omitted~, are negative from event day t = -4 through event day t = 1, and the average abnormal return on event day t = -1 is significant. However, the average abnormal return summed over announcement event days t = (0, 1) are not reliably different from zero (-0.17%, t = -0.32) with 45% of the issuing firms experiencing two-day abnormal returns less than zero.(9)
An investigation of the cross-sectional variation in the issuing firms abnormal returns on announcement suggests that much of the variation in these abnormal returns is a result of the revaluation of the convert firm's stock. The two-day abnormal returns for the issuing firms, |Mathematical Expression Omitted~, are regressed on the two-day abnormal returns for the convert firms, |Mathematical Expression Omitted~, weighted by the ratio of the market value of the convert firm's stock held by the issuer to the market value of the issuing firm's equity, |S.sub.i~/|E.sub.i~. Absent any other valuation effects, the negative price response in the convert firm's stock, on average, would lead to a negative price response in the issuing firm's stock. The results of the regression estimation were (t-statistics in parentheses):
|Mathematical Expression Omitted~
The |R.sup.2~ of the regression was 24.5%. The intercept of the regression is not significantly different from zero. The slope coefficient of the regression is significantly positive, indicating much of the cross-sectional variation in the issuing firms' price response is attributable to the revaluation of the convert firms' equity.
III. Motivations for Issuing Exchangeable Debt
In this section, I describe several hypotheses put forth to explain the existence of exchangeable debt. First, I consider two tax hypotheses, often mentioned in the financial press, to explain the existence of exchangeable debt, the deferral of capital gains income and the capturing of dividend income. Neither of the tax hypotheses is strongly supported by the data presented here nor do either of the tax hypotheses require the issuance of exchangeable debt. That is, there are alternatives to issuing exchangeable debt which would capture the same tax advantages. Second, I consider the non-tax motivations for issuing exchangeable debt -- specifically, the underwriting costs and valuation effects associated with an exchangeable debt issue. Ultimately, I conclude that exchangeable debt is not sufficiently attractive relative to the alternatives (secondary offerings, block sales, etc.) to make it a more common divestment strategy. Exchangeable debt is a classic example of a "neutral mutation" -- a financial innovation likely born from tax considerations without real economic benefits.
A. Tax Motivations for Issuing Exchangeable Debt
The financial press cites two tax features of exchangeable debt as motivating its issuance. First, the issuing firm collects the tax-preferred dividend income on the convert firm's stock (if any). Second, the issuing firm does not realize any capital gains on the convert stock until the conversion feature of the exchangeable debt is exercised. If there is a large accrued capital gain, the deferral of it can result in significant tax savings to the issuing firm. Descriptive evidence presented here suggests that these two features of exchangeable debt do not provide a significant motivation for issuing exchangeable debt. I describe each of the two hypotheses and discuss the empirical evidence regarding each of the hypotheses.
1. Dividend Tax Arbitrage
Jones and Mason |4~ argue that exchangeable debt provides an opportunity to capitalize on the differing tax treatment of interest payments and dividends received. Corporations are allowed to deduct interest expense from taxable income while receiving an 85% corporate dividend tax exclusion as long as the convert's stock was acquired prior to October 1984. After this date, the tax laws changed and prohibited the use of the 85% corporate dividend tax exclusion to be used on securities held as collateral for a possible future transaction, such as exchangeable debt.(10) I refer to this motivation for issuing exchangeable debt as the "dividend stripping" hypothesis, as corporations are attempting to divest a block of stock while retaining the rights to dividend income.
In at least two particular issues of exchangeable debt, there is specific evidence that the dividend yield of the convert's stock was a motivating factor in the issue. In the case of Prudential Overseas issuing debt convertible into the stock of AT&T, Prudential entered the secondary market to purchase the AT&T shares (this is the only documented example of a firm purchasing shares at the time of an exchangeable debt issue). According to the Institutional Investor, February 1984:
What makes the |exchangeable debt~ deal work, from Pru's standpoint, is that the company will collect the AT&T dividends for as long as it owns the stock .... |T~he fact that the equity feature lowered Pru's interest cost on the deal by some 125 basis points, made the deal decidedly attractive.
In the case of General Cinema's issue convertible into R.J. Reynolds, General Cinema justified the issue in their annual reports because of the "... substantial dividends paid by R.J. Reynolds."
However, the "dividend stripping" hypothesis does not hold up to closer scrutiny. The hypothesis does not justify the bundling of debt with a call option on the stock. Corporations could accomplish the same stripping strategy by issuing debt to purchase stock without bundling the debt and a call option on the stock.
Furthermore, the empirical evidence does not support the dividend stripping hypothesis. For example,
* The dividend yield on the convert firms is not substantially different from that of the median NYSE/AMEX firm. In fact, eight of the convert firms were non-dividend-paying.
* The valuation effects to issuing firms on announcement of the exchangeable debt issue is unrelated to estimates of the tax-preferred dividend income issuing firms would receive on the convert firm.(11)
2. Deferring the Realization of Capital Gains
The second central tax feature of exchangeable debt is that it enables the issuer to effectively sell the stock (recall exchangeable debt is essentially a divestment strategy) without realizing the capital gains from the sale until the conversion feature of the debt is exercised -- the "capital gains deferral" hypothesis. This method of deferral might dominate holding the stock for later sale in that it allows the issuer to procure a portion of the proceeds through the sale of the call option implicit in exchangeable debt. The financial press cited the deferral of accrued capital gains as a specific motivation for issuing exchangeable debt in the General Cinema issue (exchangeable for Columbia Pictures), and the Prudential, Internorth and Newmont Mining issues.
The benefit of this feature of exchangeable debt to the issuing firm is the difference between the value of the capital gains tax which would have been realized if the convert's stock had been sold at the issue of the exchangeable debt less the present value of the capital gains tax which will be realized on the convert's stock. When the conversion feature of exchangeable debt is exercised, the issuing firm realizes a capital gain (loss) on the difference between the book value of the stock (generally the acquisition cost or, in the case of a subsidiary, the initial investment of the parent plus accrued equity) and the conversion price of the stock. The market value of the stock is irrelevant to the issuing firm for tax purposes. Since exchangeable debt almost always offers a conversion feature which is out-of-the-money, the deferred capital gains are greater than the capital gains which would be realized had the stock been sold at the time of the exchangeable debt issue. The capital gains deferral hypothesis has two shortcomings. First, the deferral of capital gains can be accomplished by holding the stock for later sale. Second, the call option implicit in the exchangeable debt offering does not yield the same dollar value as it would had the stock been sold.
Furthermore, the empirical evidence is also inconsistent with the capital gains tax hypothesis. For example,
* In at least three cases, exchangeable debt was issued by a firm with a capital loss to be realized on the convert firm.
* Four issuers had positive tax loss carry-forwards making any tax consideration for issuance implausible.
* The valuation effects to issuing firms on announcement of the exchangeable debt issue are unrelated to estimates of the capital gains deferred (see footnote 11).
Thus, though the financial press has often cited tax motivations for the issuance of exchangeable debt, these motivations do not hold up well to empirical and analytical investigation. In the next section, I consider non-tax motivations for issuing exchangeable debt.
B. Non-Tax Motivations for Issuing Exchangeable Debt
Issuing exchangeable debt is equivalent to selling the convert's stock with a repurchase guarantee. That is, if the price of the convert's stock does not appreciate sufficiently to warrant conversion, the holder of an exchangeable bond is left with the rights to interest and principal payments on the bond while the issuer is left holding the convert's stock. This strategy would make sense if the issuing firm is concerned about the potential price effects of selling the block on the secondary market. As documented earlier, the price response of the convert firm's stock at issue is negative, but smaller in magnitude than the price response associated with secondary distributions, the likely alternative method of disposal. In addition, the cost of issuing exchangeable debt is less than the cost of a secondary distribution. Mikkelson and Partch |5~ document the median underwriting costs of a registered secondary offering as 4.7% of the offering value in a sample of 146 registered offerings. These costs do not include the price effect of -2.87% which they document at announcement of the offering. The median underwriting costs of the 31 registered exchangeable debt offerings are 1.56%, significantly less than those associated with a secondary offering.(12) This lower cost, in combination with the lower price response of the convert firm at issue, makes exchangeable debt less costly than a secondary distribution for the disposal of a block of stock. This is the best rationale uncovered for using exchangeable debt as a means of divesting of an intercorporate holding in lieu of the probable alternative -- a secondary offering. The fact that only 37 exchangeable debt offerings have been made suggests that there are either unidentified costs which are important or that the apparent cost advantage documented here is not well known.
The evidence cited in this research indicates that firms issue exchangeable debt conditional on having decided to divest of an intercorporate holding. The anecdotal evidence in the financial press and the observed abnormal price response of -1.0% in the convert firm's stock on announcement of an exchangeable debt issue are both consistent with this interpretation. The price response of the convert firm is less pronounced than the negative price response associated with secondary distributions or block sales. I argue this is a result of the repurchase guarantee implicit in the exchangeable debt offering. The issuing firm guarantees it will keep the convert firm's stock should its value fall below the value of the straight bond component of the exchangeable offering.
Exchangeable debt was probably originally conceived to capitalize on specific features of the tax code. However, these tax motivations do not appear to be potential sources of value for firms issuing exchangeable debt. First, there is no abnormal price response to the issuing firm on the announcement of an exchangeable debt issue. Second, there is no evidence that the price response of the issuing firms is cross-sectionally related to an estimate of the size of the tax benefits associated with issuing exchangeable debt.
A second criticism of the tax hypotheses posed is that they do not uniquely motivate exchangeable debt. Accrued capital gains taxation can be avoided by delaying the sale of the stock. Tax-preferred dividend income can be captured by issuing debt to purchase stock. Thus, the tax hypotheses posed do not provide a compelling rationale for the issuance of exchangeable debt.
I document that the median underwriting cost of issuing exchangeable debt (1.56%) is significantly less than the median underwriting cost of a secondary offering (4.7%). This is the best rationale for issuing exchangeable debt -- to divest of a large intercorporate holding in lieu of a registered secondary offering -- but it begs the question of why there have been only 37 exchangeable debt offerings. Nonetheless, corporations might be prudent to consider exchangeable debt as a divestment strategy on this evidence alone. Aside from this consideration, I conclude that exchangeable debentures are a neutral mutation of previously existing divestment strategies and represent an unconvincing attempt to capitalize on specific characteristics of the tax code.
1 Wall Street Journal, February 12, 1986. See Appendix A for a summary of quotations from the financial press which are indicative of motivations which firms cite when issuing exchangeable debt.
2 An equity carve-out occurs when a firm offers a portion of a wholly owned subsidiary's common stock for sale to the public.
3 The Tenax debenture exchangeable into the common stock of Timken was subsequently called in July 1989.
4 Both Newmont issues, the General Cinema issue convertible into Cadbury Schweppes, the latest Dart issue, and the IBM and Prudential offerings were all Eurobonds.
5 Defined as price per share specified in the exchange option divided by the price of the convert stock at the announcement date. For example, the Petrie Stores issue provided investors with the opportunity to exchange each bond for 27 shares of Toys "R" Us common -- a conversion price of $36. The price of Toys "R" Us common at announcement of the issue was $30 1/8). Thus, the conversion premium at announcement was 19.5%.
6 For the NASDAQ stocks, the value-weighted NASDAQ market return is used. For NYSE/AMEX stocks, the value-weighted NYSE/AMEX market return is used.
7 The results cited around the announcement of an exchangeable debt offering proved insensitive to the choice of estimation period, choice of market index (equal vs. value-weighted), and choice of beta estimate (OLS vs. Scholes-Williams). A search of the Wall Street Journal Index revealed no significant corporate events which might contaminate these results. These results are similar to those reported in research independently undertaken by Ghosh, Varma, and Woolridge |2~.
8 Mikkelson and Partch |5~ document that 37% of the 146 registered secondary offerings which they analyze were offered by directors and officers of the corporation who are arguably better informed than an outside corporation, while 63% of the registered secondary offerings were made by individuals, corporations, financial service companies, or trusts.
9 Schipper and Smith |7~ document positive abnormal gains to parent firms at the announcement of equity carve-outs. The three firms (Tridex, Bergen Brunswig, and Kaufman & Broad), which simultaneously issued exchangeable debt and performed an equity carve-out of the convert firm, were eliminated from the sample to circumvent the possibility of a contaminating event. The average two-day abnormal returns around the announcement date for the remaining 28 issuing firms are also insignificantly different from zero at conventional significance levels.
10 Dividends received by American corporations on foreign stock holdings are taxed as foreign source income on which a foreign tax credit could be claimed.
11 Previous drafts of this paper presented evidence from cross-sectional regressions of the issuing firm's abnormal returns on the convert firm's abnormal returns, estimates of the tax benefits of the dividend tax exclusion and capital gains deferral. While the convert firm's abnormal returns remain important in explaining the issuing firm's abnormal returns, neither of the coefficient estimates on the tax variables were reliably different from zero at conventional significance levels.
12 It is important to consider the difference in the size of the secondary distributions studied by Mikkelson and Partch |5~ (mean dollar value of $31.6 million) and the exchangeable debt offerings studied here (mean dollar value of $96.4 million, Exhibit 3). For the exchangeable debt offerings, 53% of the variation in underwriting fees can be explained by offering size (measured as the log of the dollar value of the offering). Using a simple OLS regression model to project the expected underwriting fees of a $31.6 million dollar exchangeable debt offering suggests that underwriting fees for an offering of this size would be approximately 2.75% of the offering price. Though higher than the reported average of 1.56%, this is still considerably less than the median cost of a secondary distribution as reported by Mikkelson and Partch.
1. P. Asquith and D.W. Mullins, "Equity Issues and Offering Dilution," Journal of Financial Economics (January/February 1986), pp. 61-89.
2. C. Ghosh, R. Varma, and J.R. Woolridge, "An Analysis of Exchangeable Debt Offers," Journal of Financial Economics (November/December 1990), pp. 251-263.
3. R. Holthausen, R.W. Leftwich, and D. Mayers, "The Effect of Large Block Transactions on Security Prices: A Cross-Sectional Analysis," Journal of Financial Economics (December 1987), pp. 237-268.
4. E.P. Jones and S. Mason, "Equity-Linked Debt," Midland Corporate Finance Journal (Winter 1986), pp. 47-58.
5. W. Mikkelson and M. Partch, "Stock Price Effects and Costs of Secondary Distributions," Journal of Financial Economics (June 1985), pp. 164-194.
6. M. Miller, "Financial Innovation: The Last Twenty Years and the Next," Journal of Financial and Quantitative Analysis (December 1986), pp. 459-471.
7. K. Schipper and A. Smith, "A Comparison of Equity Carve-Outs and Seasoned Equity Offerings: Share Price Effects and Corporate Restructuring," Journal of Financial Economics (January/February 1986), pp. 153-185.
Brad M. Barber is an Assistant Professor of Finance at the Graduate School of Management, University of California - Davis, Davis, California.
Appendix A. Financial Press Quotations
The following are excerpts from articles regarding the issuance of exchangeable debt. The quotes are presented to illustrate the motives for issuing exchangeable debt as covered by the financial press.
Regarding overvaluation of convert stock:
... the plan to back out of the Paine Webber stake in no way ... indicate|s~ any dissatisfaction with Paine Webber's performance. |Wall Street Journal, November 17, 1982~
The |Paine Webber~ stock has fallen for the past two days |after the announcement of the exchangeable debt issue~ after running up sharply on rumors of a merger possibly with CIGNA. |Wall Street Journal, November 17, 1982~
The |exchangeable debt offering~ enables CIGNA to effectively convert its holdings in Paine Webber into cash immediately. |Wall Street Journal, November 16, 1982~
The announcement |of the exchangeable debt offering~ triggered a wave of speculation about the state of the alliance between IBM and Intel... IBM stressed that the two companies continue to have close productive relationships in many aspects of their respective businesses. |Wall Street Journal, January 28, 1986~
... as some money managers see it, the |exchangeable debt offering~ is a red flag. An insider who knows much about Intel -- IBM -- is effectively dumping the stock
... They say IBM must be gloomy about the company's prospects to make some of its Intel stake available, even though many analysts predict big gains in 1986 for semiconductor stocks. |Wall Street Journal, February 12, 1986~
International Paper Co. said its previously announced $60 million of debentures convertible into the C.R. Bard, Inc., stock held by International Paper doesn't indicate any lack of confidence in Bard's management or prospects. |Wall Street Journal, October 12, 1971~
While no equity-related investment is without risk, exchangeables have generally offered the right to exchange for the stock of established companies. |Wall Street Journal, June 9, 1986~
Regarding tax motivations for issuing exchangeable debt:
In September, |General Cinema~ sold ... subordinated |exchangeable~ debentures. As a result of this financing and assuming that the debentures are exchanged prior to maturity, General Cinema has sold its Columbia shares at a premium, received the proceeds, but postponed the capital gain tax on the approximate $7.5 million presumed long-term gain. The net cost of the financing, after giving effect to the Columbia dividends, is less than 4% per year. |GCN Annual Report 1980~
What makes the deal work from Pru's standpoint is that the company will collect the AT&T dividends for as long as it owns the stock ... If the dividend is around 8% and the Pru is shielded from paying taxes on 85% of that, then it will earn about 6.8 % in dividend income. |Institutional Investor, February 1984~
Take the case of Internorth, whose 10 1/2s of 2008 are exchangeable for Mobil common stock. Internorth bought its Mobil shares long ago. It doesn't want to sell and get socked with capital gains taxes, but it does want to leverage the asset. With these debentures, Internorth gets a lower cost of funds than on straight debt and still collects the 85% tax-excludable dividends on its Mobil common. |Forbes, June 3, 1985~
... the transaction creates certain tax advantages for the insurance company, as its gains on the Paine Webber stock won't be taxed until the debentures are exchanged. CIGNA can also deduct its interest expense on the debentures. |Wall Street Journal, November 17, 1982~
Issuing the new |exchangeable~ debentures also permits Newmont to raise cash without selling DuPont shares which would increase its tax liability. |Wall Street Journal, July 22, 1986~
TABULAR DATA OMITTED
Appendix C. Pricing Exchangeable Debt
In this appendix, I provide an arbitrage proof for pricing exchangeable debt. The derived pricing formulae prove useful in discussing the motivations for issuing exchangeable debt.
Consider the case of a firm, with value today V, which issues exchangeable debt with face value X which is exchangeable into the stock of the convert firm, with value today S. This debt pays no coupon and the exchange feature is only exercisable at maturity. The bond has no call provision, is the senior claim on the firm's assets, and comes due in one period of length t. The goal of this section is to determine the price of the exchangeable debt today, D.
To price exchangeable debt, I assume the value of the issuing firm has two mutually exclusive components -- the value of the block of the convert's stock into which the debt is exchangeable, S, and the value of the remaining assets of the issuing firm, A:
V = S + A. (C1)
This final assumption is realistic since the issuing firm in general places the convert's stock in an escrow account at the time of the issuance of the exchangeable debt. This escrow prevents the value of the issuing firm from falling below the value of the convert's stock and thus prevents default on the conversion feature of the bond. Default on the principal payment is, however, still possible.
Given these simplifying assumptions, there are three relevant states at maturity of the exchangeable debenture. The first relevant state is when the issuing firm is in default on the face value of the bond. In this state, the conversion feature is unattractive to bondholders and the value of the issuing firm at maturity is less than the face value of the bond, S* |is less than~ V* |is less than~ X. The issuing firm, which is in default, must deliver the value of the firm at maturity, V*, to the bondholders. The second relevant state finds the issuing firm with a value at maturity in excess of the face value of the bond, but the conversion feature remains unattractive, S* |is less than~ X |is less than~ V*. In this state, bondholders receive the face value of the bond. The third and final relevant state is one in which the conversion feature has value since this feature exceeds the face value of the bond, X |is less than~ S* |is less than~ V*. In this state, bondholders surrender the face value of the bond in exchange for the convert's stock.
Given these three relevant states, the payoffs at maturity of the exchangeable debt are easily reconstructed by two different methods. The following three actions would duplicate the payoffs of an exchangeable bond at maturity:
* Purchase the block of convert's stock into which the exchangeable bond is convertible, S.
* Purchase a European put option which allows the buyer to sell the block of the convert's stock at a price equal to the face value of the exchangeable bond, P(S, X).
* Sell a European put option which allows the buyer to sell the value of the issuing firm at a price equal to the face value of the exchangeable bond, P(V, X).
Thus, the price of the exchangeable debt must be equal to the summed prices of the three individual securities today:
D = S + P(S, X) - P(V, X). (C2)
In this context, the firm which issues exchangeable debt can be viewed as selling the convert's stock with two attached features -- one which is value-enhancing to debtholders (the convert's stock can be put back to the firm for a price X) and one which is value-decreasing (the issuing firm has liability limited to the firm's value at maturity).
The following three actions would duplicate the payoffs of an exchangeable bond at maturity:
* Purchase the issuing firm, V.
* Purchase a European call option which allows the buyer to purchase the block of the convert's stock at price X, C(S, X).
* Sell a European call option which allows the buyer to purchase the issuing firm at price X, C(V, X).
Again, the price of exchangeable debt must be equal to the summed prices of the three individual securities:
D = V + C(S, X) - C(V, X). (C3)
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Security Design Special Issue; includes appendices|
|Author:||Barber, Brad M.|
|Date:||Jun 22, 1993|
|Previous Article:||Interpreting SIGNs.|
|Next Article:||Financial innovations and excesses revisited: the case of auction rate preferred stock.|