Over examined by bles and tr that issued 310 millior 1ndHiduaxl dlnated det flow and re rate, 0311 818° 9184111 r Off betweer gated. The ordinated C 908””er ulat1on "a: fox-man“ as “lated assx ABSTRACT CORPORATE EXPERIENCE WITH CONVERTIBLE DEBT BY John Hagerty Burns Overall corporate eXperience with convertible debt was examined by compiling aggregate data on all forms of converti— bles and their issuers over the period 1950-1969. U. 8. firms that issued convertible subordinated debentures larger than $10 million during the period were then selected to study individual firm performance. The effects of convertible subor- dinated debentures on earnings per share, incremental cash flow and rate of return were initial variables. Conversion rate, call policy, capital structure and conversion value were also examined to complete the experience. Finally, the trade- off between interest rates and conversion prices was investi- gated. The hypothesis was that financing with convertible sub- ordinated debentures, instead of subordinated debentures, should positively improve the company's subsequent performance. Sim- ulation was the technique employed in this study. Firm per- formance as reflected by the variables mentioned above was sim— ulated assuming the substitution of a pure debt issue for the actual convertible debt outstanding. Actual performance was then compared to simulated performance and tested using non- parametric statistical techniques. Analy tween actuall fluences. ri at a signifi the use of s equity. The earnings per the longer mental cash issue. The I probability return Were minimally. Capl issuers ten sion from d CORVertlble CEgree’ thq Joyed bett. a Convent Cur tWEQH int“ JOHN HAGERTY BURNS Analysis results showed significant differences be- tween actual and simulated performance due to non-random inp fluences. However, the data showed that convertible issuers were at a significant earnings per share disadvantage relative to the use of straight debt which was later replaced with common equity. The ratio of actual earnings per share to simulated earnings per share was consistently below one and deteriorated the longer an issue remained outstanding. Time adjusted incre- mental cash benefits were related to the length of time from issue. The longer an issue was outstanding, the greater the probability that net present value would be negative. Rates of return were positively influenced by convertible usage but only minimally. Combined results indicated rejection of hypothesis. Capital structure analysis indicated that convertible issuers tend to use debt more heavily than the average for all manufacturing companies. Examination of conversion values, how ever, showed that many convertible issuers eXpect early conver- sion from debt to equity. Conversion rate data revealed that convertibles tend to be converted quickly or remain outstanding for relatively long periods. These results explain, in some degree, the earnings per share data. Convertible issuers en— Joyed better relative earnings performance in the early life of a convertible and when it was converted quickly Curves were developed which reflected the tradeoff be- tween interest rates and conversion prices. Simple curvilinear regression was used with the data classified by time period, industry am issue date i The correla issue date results wit categories. JOHN HAGERTY BURNS industry and growth. A further test employed data taken at issue date and market data observed one year after issue. The correlation coefficients observed were generally low for issue date categories. The lagged market data produced better results with correlations of a magnitude of 0.8 for industrial categories. 1n CORPORATE EXPERIENCE WITH CONVERTIBLE DEBT JOHN RAGERTY BURNS A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Accounting and Financial Administration 1972 C9 Copyright by JOHN HAGEBTY BURNS 1972 To nary, Vh three livel To Mary, whose Joint venture with the author has resulted in three lively subsidiaries. 11 A thesi: pletion of h: are my commi Coleman and deserving p tYped and r ins the n ACKNOWLEDGMENTS A thesis writer relies on many people for aid in the com- pletion of his work. Among those I would like to acknowledge are my committee composed of Dr. Alan Grunewald, Dr. Bruce Coleman and especially my chairman, Dr. Myles Delano. Also deserving recognition are Mrs. Lillyan Fahy who cheerfully typed and retyped the rough draft and Miss Susan Boy for typ- ing the final copy. 111 LIS‘I‘ 0? TA: LIST OF PIG LIST OF APE Chapte r I. H 2?. *‘rf 9 £1: (0 II. TABLE OF CONTENTS Page LIST OF TABLES C O O O C O O O O 0 O O C O O O O O 0 O O O O O O O O O O O O O O O 0 V1 LIST OF FIGURES O O 0 O O O O O O O O O O O 0 I O O O O O O O O O O O O O O O O O O V111 LIST OF APPENDICES 00000000000.0000000000000000... 1X Chapter I. INTRODUCTION 0.000000000000000...0.0.0000... 1 Convertible Securities Convertible Bonds Historical Importance Problem Theoretical Model Factors Influencing Convertible Bond Usage Statement of the Problem Significance of the Study Absence of Comparable Studies Implications For Policy Hypothesis The Basis Hypothesis Assumptions Methodology The Theory Revisited Plan Of The Dissertation II. THE VOLUME OF CONVERTIBLE DEBT FINANCING.... 32 Introduction Data Sources Volume and Character of Convertible Debt 1950-1969 Overall Financing Activity Issue Characteristics Retirements Repeat Issuers Summary III. BACKGROUND AND METHOD........................ 63 Introduction Firm Selection Issue Characteristics Simulation Straight Debt Alternative iv Ll Var Sum IV. ELI-1P I d: Introd Actual Ear Bur Preser Actual Capite Det Cor Cor Conve: Summai V- TRADEC Int Pre Cu] Sun VI° REVIE; Int Cor Fir Int StL P : ~ IV. Variables Earnings Per Share Time Values Rates of Return Conversion Rates and Calls Capital Structure Conversion Value Summary EIVIPIHICAL ANALYSISooeooccoooooooeoo00090000000. 86 Introduction Actual Relative to Simulated Earnings Earnings Per Share Run Tests Present Value Cash Benefits Actual Relative to Simulated Rates of Return Capital Structure Debt Capacity Conversion Values at Issue Conversion Rates and Calls Conversion Versus Earnings Per Share Summary TRADEOFF CURVES................................117 ' Introduction Previous Research Current Study Purpose Proposal Method Results Summary REVIEW AND EVALUATION..........................13“ Introduction Convertible Debt Aggregate Financing Financial Performance Interest and Conversion Price Tradeoff Study Implications Summary BIBLIOGRAPHYCOOOCOOO00.0.0.0....0.0O.0.00.00.00.00000143 APPENDICESCOOOOOOOOOOOOOOOOOOOOOOOOOOO00.0.0000000000150 Table 1.1 1.2 2.1 2.2 2.3 2.1+ 2-5 rev New New Net I nc Net I-‘ioc {Lei Cor Si: Is: Yes Moc 3".0c KOc Cor Pre stt Act LIST OF TABLES Table Page 1.1 New Corporate Securities Offered For Cash In The United States, l950-l969............ 9 1.2 New Corporate Bonds Offered For Cash In The United States, l956—1969............... 10 2.1 New Total Public Convertible Debt, 1950-1969.. 35 2.2 New Public Convertible Debt Forms, 1950-1969.. 36 2.3 Industry Use of Convertible Debt.............. 37 2.4 New Public Convertible Debt By Industry, 1950- 1969000000.0000000000000OOOOOOOOOOOOOOOOOI. 38 2.5 Moody's Convertible Bond Survey Total Amounts Outstanding For Industrials Utilities and TransportationOOOOOOOOOOOOOOIOOO00.0.0.0... 1+0 2.6 New Convertible Bonds Offered For Cash In The United States.............................. #1 2.? Convertible Bonds Relative To Other Financing. #2 2.8 New Public Convertible Bond Series............ 46 2.9 Sizes Of Issues By Issue Year................. 48 2.10 Issue Size and Issuing Firm Size.............. 50 2.11 Years To Maturity............................. 51 2.12 Moody's Rating Versus Issue Year.............. 53 2.13 Moody's Ratings Versus Maturity............... 5“ 2.1“ Moody's Ratings And Coupon Interest........... 56 2.15 Convertible Bond Calls........................ 5? 2.16 Premiums At Which Calls Occur................. 59 2.1? Frequency Of Convertible Bond Usage........... 61 3.1 Study Group Statistics........................ 6? h.l Actual Relative To Simulated Earnings Per Share: Totals For Study PeriodCOOOOOOOOOOOOOOOO0.. 87 vi £52 11.3 at t.6 11.? “.8 “.9 l$.10 4.11 “.12 M3 wt Ave Ler L8! Nu: AV: FIN F1 Co We Pe u.3 the 4.5 4.6 4.7 4.8 4.9 4.10 4.11 4.12 4.13 c.1u 4.15 4.16 5.1 Average Actual Relative To Average Simulated Earnings Per Share: Each Year During StudyCOOOOOOOOOOOO0.000...00.000.00.00... 89 Number of Runs Above And Below 100 Per Cent. 91 Length Of Runs Above And Below 100 Per Cent. 92 Lengths 0f Runs Of Ascending And Descending OrderOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 92 Number Of Runs Above and Below 100 Per Cent By InduStryOOOOOOOOOOOOOOO00.00.000...... 9“ Summary Of Net Present Value Benefits During StudyOO.00...COOOOOOCOOOOOOOOOOOOOOOO.00. 96 Average Net Present Value Cash Benefits From DateOOOOOOOOOOOOOO0.0.0.0...0.0.0.0000... 97 Frequency Distribution Of Actual Relative To Simulated Rates Of Return................ 100 Firm Capital Structure At Issue Date........ 102 Conversion Value At Issue Expressed As Dis- count From Market Price.................. 105 Weighted Average Discount From Market Price. 106 Per Year Conversion Of Convertible Subordin— ated Debentures Into Common Stock........ 108 Cumulative Conversion Of Convertible Subordin- ated Debentures Into Common Stock........ 110 Call And Conversion Expiration Years From IssueOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. 111 Cumulative Conversion of Called Convertible Subordinated Debentures Up To Call Date And Conversion EXpirations............... 113 Correlation And Regression Measures For Trade— Off CurveSOOOOCOOOOOOOOOOOCOCO0000.00.00. 12“ vii Figure 1.1 1.2 5.1 5.2 "Y? o. '4 .L Tra Figure 1.1 1.2 5.1 5.2 LIST OF FIGURES Hypothetical Model of a Convertible Bond IssuBCOOOOOOOOOOOOO.00......0.00.00...OI Hypothetical Tradeoff Opportunities........ Tradeoff Curves: Tradeoff Curves: Market Dataoocooocccoocoo Lagged Industrials...,... viii Page 12 30 126 132 Appendix A B C 0.1 D.3 D.4 0.5 Stud List Stat ActL C Pres Issx Exec Indt LIST OF APPENDICES Appendix Page A Study Group IssueSOOOOOOOOOOOOOOOOO0000...... 150 B List Of SymbOJ-Scocoo-000000000000000000000000 155 C Stat18t1081 MethOdsogao00000000000000.0000... 157 D.l Actual Relative To Simulated Earnings Per Share, 1950-196900000000000000000009000... 159 0.3 Present Value Cash Benefits, 1950-1969....... 163 D.4 Issue Size And Straight Versus Convertible IntereStOOOOOOOOOOOOOOOOOOO.O..00.0.000... 16“ D.5 Excess Shares Required Under Actual Versus Sim- ulated SituationSOOOOOOOOO0.0.000.00...... 165 E Industrial Classifications................... 166 ix In th for: enjoy preen fleet deleg Part. ties of c: The “Pit. “1‘1 ety o 1” 01’ de Etc Ck 3 Co hue be” designed these c Cc CRAPTER I INTRODUCTION In the United States' economy, the corporate business form enjoys a predominant position. Dewing said: This book...1ooks upon the corporation as preeminently an ingenious creation of man's re- flective imagination to which our modern world has delegated a large part -- an increasingly large part, as time goes on -- of those economic activi- ties of society requiring a considerable amount of capital. The capital referred to is typically raised by using a wide variety of instruments which fall broadly into the categora iee of debt or equity. Familiar forms are bonds, preferred stock, common stock and retained earnings. These basic forms have been modified to produce a number of hybrid securities designed to meet the needs of the issuing corporation. Of these, convertible bonds appear to have found increasing favor in the corporate quest for additional capital..This study examined the financial effects on the issuing corporation that arose from the use of convertible debt in general and convertible subordinated debentures in particular. 1 Arthur Stone Dewing, A.Stud of Co oration Securities: Their Nature and Uses in Finance iNew TbrE: HonaId Press, 1% pn- " t1 Ii sent ed a blem the converti vertible tion fo: assumpt methodo the at‘ an app broad bond 0 holde] Share: eldeia h81d 1 In this chapter the nature of convertible bonds is pre- sented as a foundation for later development. Next, the pro- blem that was investigated is discussed in the context of convertible bond theory. This historical importance of con- vertibles and the study's significance are offered as motiva- tion for the research. Finally, the hypothesis and unverlying assumptions are presented along with a statement of basic methodology. Convertible Securities Convertible securities reflect an attempt to combine the attractions of debt and equity in one security form in an appeal to several classes of capital suppliers. In a broad sense, the convertible security has been defined as a bond or preferred stock, exchangeable at the cption of the holder, under specified terms or conditions for the common shares of the same corporation.2 This is not an all in- clusive definition. The cption to convert is sometimes with- held by the issuer. Also, the conversion security, in rara situations, can be senior to the convertible, or be securities representing more than one firm. These cases are deviations from the mainstream of convertible issuers which will be eigher 2Montgomery Rollins, "Convertible Bonds and Stock,” Annals of American Academy of Political and Social Science, XXXV, No. 3 (I910). 97. bonds or firm's co Convertit P issuer‘s red stoc many of debt org of a 00‘ underwr a consi In add} for oti nature secure Call f DRbllc are C: bonds or preferred stocks convertible into the issuing firm's common stock. Convertible Bonds This study focused on convertible bond usage from the issuer's viewpoint and did not consider convertible preferb red stock. Convertible debt can, and usually does, exhibit many of the characteristics of pure debt. The convertible debt ordinarily will be some form of bond or note. Issuing of a convertible can be publicly on an underwritten or non- underwritten basis, and rights can be employed. Alternatively, a considerable amount of convertible debt is placed privately. In addition, convertibles are sometimes offered in exchange for other securities. Convertible bonds can be of a serial nature but most are straight bonds. Debenture bonds and secured issues are used, as is the technique of subordination. Call features are also employed. As will be seen, most modern publicly offered convertible debt issues are subordinated and are callable. Making a bond convertible alters a fixed income security to one that also possesses equity characteristics. Unless otherwise noted, in this paper the term convertible means __5' Thomas R. Atkinson, Tgpnds in Corporate Bond gualit , National Bureau of Economic Research (New York: Columbia University Press,"1967), p. 104. exchang Thus, t and the that is The hyt may be terms u may be life. cent ’ time, t time b. the nu: Changl. every Option Conver verslo ohce t Share bOnd W 1nt 20 4 exchangeable for the common stock of the same corporation. Thus, the convertible security will be a bond or debenture and the conversion security will be common stock. A convertible bond is similar to a "pure" bond in that is pays a fixed amount of income on a given face value. The hybrid nature of the security is due to the fact that it may be converted into shares of common stock. Conversion terms may be either fixed or variable and the conversion period may be some time span other than that covered by the bonds life. A typical convertible bond would be a $1,000, a per cent, 25 year, subordinated debenture, callable at 10# at any time, and convertible into 20 shares of common stock at any time before call or maturity. A variation on this would have the number of shares of the conversion security obtainable, changing from 20 to 18 after 5 years and decreasing by 2 shares every 5 years up to 20 years at which time the conversion option would expire. This bond would be said to have a variable conversion ratio. The conversion ratio is the number of con— version security shares obtained when the bond is converted. Once the conversion ratio is stated, the effective price per share is determined. Thus, investing in a $1,000 convertible bond with a conversion ratio of 20 would be equivalent to buy- int 20 shares of the firm's common at $50 per share, ignoring tnnsacti 20 to 18 $50 to 35 It i acquire a Recent wo: securitiea The 1 found be c and the s: horizon. end Will (3 market for 18 success return in “111ml ‘ Primarily The E 'hich allo “bond, '7. “13“?er transaction costs. A change in the conversion ratio from 20 to 18 would increase the common's conversion price from $50 to $55.56 per share. It is helpful to maintain the investor viewpoint to acquire an understanding of the convertible's dual nature. Recent work examining the theoretical aspects of convertible securities divides the security's value into two parts.“ The first is the security's present value which can be found be discounting the fixed stream of interest payments and the security's expected value at the end of a given horizon. The security's expected value at the horizon's end will depend on the firm's profitability and on outside market forces which influence security values. If the firm is successful in generating profits and a favorable rate of return in a healthy economic climate, the convertible's terminal value may exceed its debt value and be determined primarily by equity considerations. The second part emphasizes the security's dual nature which allows the investor to treat it as either a stock or a bond. Thus the investor has both the insurance of a relatively stable debt value and a call or option on the u Otto H. Poensgen, "The Valuation of Convertible Bonds: Part I,” The Industrial Management Review, VII (Fall, 1965), 78-79; anMWalkiel, and Richard E. Quandt, ”The Valuation of Convertible Securities," Quarterly Journal of Economics, Lxxx (February, 1966), “8-59. {111's comm! hypothetica stock rose in conversi If the inve be determix is worth 8: other hand develop as realizable basic natu is determi hold the t has the 11 3-1118 the ‘ convert and inc W Com firm's common stock. Assume that an investor brought the hypothetical bond referred to earlier and the company's stock rose to sell at $60 per share. The band's value in conversion would be $1,200 (20 shares x $60 per share). If the investor cashed out atthat point, his final value would be determined by the band's stock equivalent. This cption is worth something to the investor in and of itself. On the other hand, if the market for the firm's stock does not develop as predicted this capital gain element will not be realizable. However, the investor is protected by the bond's basic nature and may dispose of his investment at a price that is determined by the bond value of the convertible or he may hold the bond and receive face value at maturity. Thus, he has the insurance of limiting his risk in the market vis- a—vis the purchaser of common stock. For an investor, then, a convertible can be viewed as two separate elements -- a fixed income portion and an equity portion. Hi sto ri cal Impo'rtance Convertible bonds are not a new financing tool. Pilcher cites a situation in which King Charles I of England was allowed to convert his stock of an early London Water Company into bonds during his reign of 1625-16h9.5 Taylor relates 5 C. James Pilcher, Raising Ca ital With Convertible Securities (Ann Arbor: B'u'r'ea'u' o!" fi'u'T's ne—s? 'He'Tea—F—Ufifrc , - W Michigan, 1955), p.2. that, "This corporation eighteenth . light that c manipulator: Conver‘ American fi] J” Gould e1 short of Br: 2. H. Harri: this centur; offerings 1., Period 1900. of Zero per 01‘ “0 Per (:1 that, "This practice is recorded in the literature of corporation finance as early as the first part of the eighteenth century."6 However, he adds a colorful side- light that convertibles were used mainly by financial manipulators in shady deals. Convertible bonds are mentioned in the history of American finance. Daniel Drew used them in the late 1850's. Jay Gould employed convertibles in covering himself when caught short of Erie.8 Later, convertible bonds were a tool of E. H. Harriman in building up the union Pacific system.9 In this century, Hickman reports a total of 1,989 convertible bond offerings representing 89,128.3 million of par value during the period 1900-1943.10 Convertibles have ranged from a low of zero per cent of all bond offerings in 1943 to a high of #0 per cent of all bond offerings in 1929, as shown 6 w. Bayard Taylor, Financial Policies of Business Enterb prise (New York: D. AppIeEoanentury, 1923), p. 292. Arthur Stone Dewing, Financial Policyof Corporations, Vol. I: Corpgrate Securities—(New Torre Ronald Press, 1950), p. 137. 8 Ibid. 9 Ibide’ p0 1380 10 W. Braddock Hickman, Statistical Measures of Corporate Bond Financing Since 1 00,'Nitiona1 Bureau of Economic Research [Pfificefbhk N. 3.: Pfghceton University Press, 1960), p. 210. by a study c In the increased it billion and One source 1' market. The 1.1. The us importance, 1 total time: non stock fl] Vhile total 1 half, in doll 01' this tweni The figl importance 01 'el‘tible bonc bottom of 3.: to a IeVel of years. It Sil differ rI‘Om 1I 11 by a study covering the years 1900-1965. In the 20 year period 1950-1969, American business increased its net working capital by approximately $13“ billion and spent over $861 billion on new plant and equipment.12 One source for financing this expansion was the new securities market. The importance of this sector can be seen in Table 1.1. The use of preferred stock has dropped steadily in importance, indeed preferreds comprised 10.9 per cent of total financing in the period 19u5.19u9. Both bond and com- mon stock financing increased slightly on a relative basis while total financing multiplied by a factor of two and a half, in dollar terms, between the first and last quarters of this twenty year period. The figures shown in Table 1.2 reflect the relative importance of convertible bonds in the years 1956-1969. Con- vertible bonds, as a per cent of total bonds, drop to a bottom of 3.3 per cent in 1963 before beginning a rapid rise to a level of approximately 20 per cent over the last 3 years. It should be noted that the figures in the SEC series differ from those of the Investment Dealers' Digest series of new ll Atkinson, op. cit., p. 78. 12 U. 8., Board of Governors, Federal Reserve System, Federal Reserve Bulletin, XLIII (December, 195?), 1396; cem er, , 3 LVI (March, 1970), Afl9. «mmu__oo ac mco____xc . ooo_u0mc~ mm.~<~.m DMH~ZL NIH Zn 7—th 10k memn—ko mmhhhszum Whtmczzou 3bz ~.— M._$¢F .noeazo xmxx .enumnnam neonomaneum x: .32: .coummeEou mmcosuxm one mo«u«u:oom ..m.: .mee .Aooon .umnouoov >4 .cfluo-sm o>uemom ammonmm .Eouwxm o>ummom “phenom .mhocuo>oo mo cameo ..m.p .mm .xnocn .auoncmnc xnnx .cwumfifiam m>ummom Heuopom .Emuaxm m>uomom doumuom .muocuo>oo no canon ..m.: .em .«mmon .auaacewe an .c«um-3m o>uomom ~mueocm .Eoummm o>uemom “phenom .muocum>om wo apnea ..m.: ”mousom e.oa Non.osN m.o~ nem.me e.e neo.nn mw~.mo~ ooo~-omon o.0m ooo.om o.mn mon.an m.m mom.m eem.ao~ ooo~.noo~ ~.om ~a~.we m.on moo.o e.n omo.~ ~o~.oo eco~-ooa_ n.0a -~.~e r.mn mom.o_ o.m ema.m ooe.mm omo~-mmon o.m~ mma.~m n.en ~n¢.m o.s mam.m omo.~e emo_-omon ammo and mumdfioa ucmu nod mama—co ucmu you oumflfioo mecca xooum COEEou xoODm pmuuwwmua mmwwwmmmu whom» wannaomm mo omww “nook «mama—co mo meowfiawzv ooo~u0m0~ .mmeehm omeaza mze 2H zmcu mom ommmmmo muaeampomm unexCamoo :mz ~.fi m4m¢h 10 TABLE 1.2 NEH CORPORATE BONDS OFFERED FOR CASH IN THE UNITED STATES, 1956-1969 (Millions of Dollars) Total Bonds Convertible Bonds Year Dollars Dollars Per Cent 1956 8,002 925 11.6 195? 9.957 1.06“ 11-7 1958 9.653 1.11;? 11.9 1959 7,190 628 8.7 1960 8,081 #62 5.7 1961 9,h20 710 7.5 1962 8,969 #45 5.0 1963 10 .856 357 3. 3 i96h 10,865 #25 3.9 1965 13.720 1 .26» 9.2 1966 15. 561 1.872 12.0 1967 21 095“ “.“75 20-“ 1968 170 383 3.281 18.9 1269 18 .3118 4.041 22.0 lleézé959 169.959 _gi.096 1;,t Note: BBC series on convertibles not available prior to 1956. Source: Total Bonds: Table 1.1 Convertible Bonds: U.S., Securities and Exchange Commission, Statistical Bulletin, XXIX (March, 1970) , 19. offerings- ths SEC 86 while the ‘ As a final bonds, 1&1 j New York 8 converti b1 The p! tible bond In Order t, “£901? on 1 10p“ by B: that; 01’ Poc t° ea1‘11 er A 813] 1.1 "here : \ 13 % ll} genturfggel Dim“: .Jc J 8 ed ‘ . Gallon s 11 offerings. The basic reason for this difference is that the SEC series covers all new issues offered for cash while the Digest reports primarily on underwritten issues. As a final indication of recent activity in convertible bonds, #1 per cent of the number of bonds quoted on the New York Stock Exchange for Friday, March 13, 1970, were 13 convertible. Problem Theoretical Model The purpose of this study was to investigate converb tible bond usage from the issuing corporation's viewpoint. In order to get a clearer picture of the convertible's effect on the firm it is helpful to examine a model deve- loped by Brigham.1u The line of development is similar to that of Poensgen and Baumol, Malkiel and Quandt referred to earlier. A graphic presentation of the model is given in Figure 1.1 where IP=issue price of bond, Mcmaturity, CP=call price 13 The Wall Street Journal, March 13, 1970, p. 23. in Eugene F. Brigham, ”An Analysis of Convertible De- bentures: Theory and Some Empirical Evidence,” Journal of Finance, XXI (March, 1966), 35-59. While Brigham first puBIIsEed the model, it was presented in lecture by Myron J. Gordon as early as 1963. 12 Time Figure 1.1 Hypothetical Model of a Convertible Bond Issue. The shaded area shows the premium above theoretical value. Adapted from Brigham. of bond, and of bond, and initial mark is equal to bond invests a lower coup debt issues. approaches ; ‘0 Par as me 20 is assum 1? since the Will ordinal the common 5 equal to th1 "I‘Sion rati Company's e: ket Prise, first be in 18 h18her t Value as 8r lore Valuab| i that for a I that the I attitude to The h) 13 of bond, and Dv=pure debt value of bond, C’=conversion value of bond, and Mv=market value of bond. As pictured, the initial market value is the same as the issue price which is equal to par value. Without the conversion feature, the bond investment value is DV since convertibles typically carry a lower coupon interest rate than that of corresponding pure debt issues. As maturity approaches, the DV value approaches par. Also, the call price typically decreases to par as maturity nears. If a constant conversion rate of 20 is assumed, the initial conversion value will be below IP since the conversion price, assumed at $50 per share, will ordinarily be above the prevailing market price of the common at time of issue. The conversion value being equal to the common's market price multiplied by the con, version rate at any point in time. Assuming growth in the company's earnings and therefore growth in the common's mar- ket price, CV will increase as shown. The market value will first be influenced by the debt only value since initially it is higher than the conversion value and then by the conversion value as growth in the common's price makes the convertible more valuable as an equity instrument than as a bond. Note that for a convertible bond to react as shown, it is assumed that the market rate of interest is constant as is the market's attitude toward risk and that earnings are growing. The bond's market value will reflect a premium over theoretical pony is suc value of ti Second, in duce his ri mon associs to 360, the ket drops 5 83161 a come convertible lished by t Vestor won] YeI‘tibles t Seneml Ina: level. Thi from holdh Order to a( bond. 0 PC lement the for °°nver1 There the 18811e 1 either the \ 15 w. lb theoretical value for several reasons. First, if the com- pany is successful and the stock price rises, the positive value of the conversion option will result in a premium. Second, in buying convertibles, the investor is able to re- duce his risk exposure. For example, if the price of the com- mon associated with the convertible referred to above rises to $60, the CV is $1,200 (360 X 20). Then, if the stock mar- ket drops sharply and the common falls to 825, the CV is $500 and a common stock investor suffers a bad loss. However, the convertible's price will, at a minimum, hit the floor estab- lished by the band's debt only value. A convertible bond in- vestor would encounter a relatively smaller loss and so conp vertibles tend to sell at a premium. Naturally, changes in general market interest rates will also influence the floor level. Third, it has been argued that institutions restricted from holding pure equities bid up the price on convertibles in order to acquire the equity characteristics of convertible bonds . Fourth, margin requirements on convertibles are more lenient than on common stock and this may also increase demand for convertible bonds.15 There are three reasons why the MV approaches the CV. First, the issue is callable and if called, the investor will receive either the conversion value or the call price both of which are 15 Ibid., pp. 36, 53. generally 1" Second, due the convert or DV. Th1 crease to 8 Factors Inf An exs tor action convertible First. common sto! Btook mark 90881b1e a Iarket Prl “flatworm: not to f] b°nd3 are ”W's an ment can vergion l: Natal-allI \ 16 lb 1'7 AI sleep t 1 o 15 generally below the market price for a ”successful” issue. Second, due to the loss potential for an investor buying the convertible at a price that is high relative to the CV or DV. Third, the stock's current dividend income may in» 16 crease to a point above the band's current interest income. Factors Influencing Convertible Bond Usage An examination of the model and consideration of inves- tor action in the market place reveals several reasons why convertible bonds have become a popular financing instrument. First, convertibles allow the firm to conditionally "sell” common stock at a price higher than that prevailing on the stock market at the time financing is undertaken. This is possible since the conversion price is ordinarily set above the market price of the common at the time of issue.17 Thus, if management feels the company's stock is undervalued and doesn't want to float pure debt for one reason or another, convertible bonds are a convenient way to market equity. Then, if the com- pany's and not the market's expectations are borne out, manage- ment can force conversion by calling the bond or encourage con- version by raising the dividend rate on the common stock. Naturally, if the common's price does not rise as anticipated 16 Ibid., 9 p0 370 1? ATT's post World War II convertible bonds are an exception. by the 188\ the companl also be 81‘! rise in the pure equity the convert ponable fir other forms Second a debt issu leverage rs Iould be to “I be emu Third is IOMer th 9 term“ Dur income in o "rubles to amount, of 1 Can hold on a 1°" inter lent has a I stock p no this 18 th 10! intere "m b°usht 16 by the issuer, call or conversion may not be feasible and the company would be saddled with unwanted debt. It can also be argued that if the common's price is expected to rise in the near future, management should wait and sell pure equity at the higher prices instead of that fixed in the convertible contract. However, in the case of unpost- ponable financing, convertibles present a solution when other forms cannot be marketed. Second, the conversion option provides a sweetener for a debt issue. If the firm desires to sell debt but its leverage ratio is already comparatively high or plain debt would be too expensive, the added enticement of conversion may be enough to attract investors. Third, convertible bonds carry an interest rate that is lower than the rate on equivalent pure debt issues. Po- tential purchasers are willing to give up current interest income in order to acquire the conversion option. Con- vertibles mean a relatively low interest cost with a given amount of leverage. Management seeking debt money, then, can hold out the carrot of convertibility in order to float a low interest issue. This can be done even though manage- ment has a very low level of expectation that the common stock price will rise appreciably in the near future. If this is the case, management will have achieved a desired low interest cost debt issue to the detriment of those who bought the convertible hoping for capital gains. Naturalll bably in of convel latively starting ing littl sent pays take plac. standing. Fouri bility to Scales dos “Tl. conv Nudtt wit the equity sun1’18 Wis Fifth 1? Naturally, in a situation like this, the company is pro- bably in the doldrums and not prospering. The attraction' of convertible bonds in this sense, though, is that re- latively cheap financing is available when a firm is first starting a project and that project is likely to be generat- ing little or no return. Later, assuming the firm's invest- ment pays off, the stock price will rise, conversion will take place, and the more ”expensive” equity will become out- standing. Fourth, convertible bonds introduce a dimension of flexi- bility to the capital structure since conversion automatically scales down long term debt turning it into equity. To the firm, conversion can represent the redemption of a senior se- curitt without the disbursement of cash. At the same time the equity base is built up preparatory to new borrowing, as- suming wise investment by the firm. Fifth, related to the convertible's marketability, is an appeal to a broad spectrum of investors. Dewing said that “broadly speaking, the demand for any security comes from three groups of human beings —- and the groups shade imper- ceptibly into one another."18 There is the person who invests for security and assured income. There is the large scale in- vestor looking for income and enhancement of principal. Last, 18 Arthur Stone Dewing, Financial Polio of Co orations, Vbl. I (5th ed.; New York: HBnaId Press, I953), p. 259. there is U looking in can appeal three grou] deund for restricted vertibles, Sixth, all terms c is another ments, the to absorb d best soluti Seven dilution to bonds. Ea by dmcn net income Senerate a tranfilalfied dilution 1; 1mmellehta in the ear t° “no o 8hare due 18 there is the speculator looking for capital gains. By looking in two directions at the same time, the convertible can appeal to all three classes. The individuals in these three groups bid against each other and may increase the demand for convertible bonds. Large institutions, legally restricted from investing in equities but eligible for con— vertibles, may also increase demand. Sixth, the possibility of the firm receiving better over- all terms on a convertible than on alternative security issues is another attractions. Depending on current margin require- ments, the risk posture of the market and the market's ability to absorb debt or equity issues, convertible bonds may be the best solution for a firm in need of funds. Seventh, avoidance or postponement of earnings per share dilution may have been an incentive to employ convertible bonds. Earnings per share have traditionally been calculated by dividing the number of common shares outstanding into the net income figure. A firm using convertibles was able to generate a return from the convertible's proceeds which, when translated to per share figures, did not reflect the potential dilution built into the hybrid security. This, because the incremental shares to be issued on conversion did not appear in the earnings per share calculation. If management wished to avoid or postpone showing any possible drop in earnings per share due to an increase in the number of shares, it could issue convertibles and the earnings per share would not reflect the secur abuse has recent so on this m: Essential. ing conve: lents musi the numbei tible, as: period in as common calculatic conversior called fill centive of “”1138 es Should hay The diluti Dressmn c into a 105 whetr WnVertib] \ 19 0 1 Allie: 1862;021:140 19 the security's equity portion. However, this alleged abuse has been attacked by the accounting profession in recent action. The Accounting Principles Board first moved on this matter in December,1966 and again in May, 1969.19 Essentially, these pronouncements mean that a firm employ- ing convertible bonds that qualify as common stock equiva- lents must reflect in the primary earnings per share figure, the number of shares that would be issued due to the conver- tible, assuming conversion took place at the start of the period in question. If the convertible bonds do not qualify as common stock equivalents they should not enter into the calculations of the primary earnings per share until actual conversion takes place but should appear in a companion figure called fully diluted earnings per share. Therefore, the in- centive of using convertible bonds for the purpose of de- ferring earnings per share dilution and confusing investors should have been eliminated by the Accounting Principles Board. The dilution referred to here is the initial short-run de- pression of earnings per share which may or may not translate into a loss of stockholder wealth. Whether these rules have discouraged any firms issuing convertibles cannot be answered here. However, consideration 19 American Institute of Certified Public Accountants, Opinion No. 9 Reporting the Results of Operations, December, 1966; and OpinI8H“N6T‘I3‘EEEfiIfiEE‘PE?‘SHEF§T‘ME§T 1969. of the Pa in Table reasons I The A prior-1 the eme< prelilse, by the u: more fuli 20 of the pace of convertible bond financing, as reflected in Table 1.2, indicates that there must be other compelling reasons for employing convertible bonds. Stgtement of the Problem The basic problem was twofold: 1- to investigate overall corporate ex- perience with convertible debt, and 2- to determine the financial impact on firms issuing convertible subordinated debentures. A priori convertible financing should not be employed unless the expected results benefit the issuing firm. Given this premise, was the issuing firm and its residual owners benefited by the use of convertible bonds? The problem is developed more fully in later sections. Significance Of The Study Absence of Comparable Studies Research indicated that scholarly work in the convertible bond area tends to fall into two categories. The first cate- gory can be termed theoretical, or model building since the writers were primarily concerned with the predicted valuation of the security in the market place. Recent works here are those by Brigham, Poensgen, and Baumol, Malkiel and Quandt mentioned earlier. The second category leans more to the empirical ering and Hiclman's earlier. as part c ancing. Pilcher s several 0 appear to presents analyzes Studies 5 the resul issued to BeSe pact of c Further, bond fine 21 empirical side and has been concerned mainly with fact gath- ering and analysis. The first modern study of this type was Hickman's20 which was updated by the Atkinson work mentioned earlier. Both Hickman and Atkinson include convertible bonds as part of a larger study encompassing all corporate bond fin- ancing. Also in this category are the previously cited Pilcher study and those by Broman and McKenzie.21 There are several other studies relating to convertible bonds but these appear to be the most outstanding. Each of these studies presents figures documenting convertible bond financing and analyzes the statistics to some degree. Both categories of studies shade into one another. For example, Brigham gives the results of a survey of large corporations as to why they issued convertible bonds. Research did not reveal any studies assessing the im- pact of convertible bonds on the issuing firm's performance. Further, those studies tracing the volume of convertible bond financing tend to restrict themselves to aggregate figures 20 w. Braddock Hickman, Statistical Measures of Co crate Bond Financi , National Bureau of Economic Research (grlnce- Eon, N. J.: Princeton University Press, 1960), pp. 210—215. 21 Keith L. Broman, ”The Use of Convertible Subordinated Debentures by Industrial Firms 1989-1959," uarterl Review of Economics and Business, III (Spring, 1963), 65-75; and RSBert 5. ficKenzie, lConvertible Securities, 1956-1965," uarterl Review of Economics and Business, VI (Winter, 196%), 51-35. on volume ' stances. factors an: to the i 331 In licatim In adc should have should: (1 structure 1 evaluate pe A recs 1 any hr the cc ferenc 0f tw: mum sq Vealt} firm, ricer lead 1 5" undeI‘Stgl ble bODd “It ASSume 3 Pure deb1 or (2) 8 cc 1‘“ and e. ~~..‘___N 22 I cheI Lern . some“, 3‘ 22 on volume with little consideration to attendant circum- stances. This study attempted a deeper analysis of these factors and sought to determine the financial consequences to the issuing firm. Implications For Policy In addition to academic merit a study of this nature should have practical significance. Generally, this study should: (1) aid management decision-making regarding capital structure formulation and (2) help the investment community evaluate performance. A recent paper discussing the finance function states: Although the firm can and often does hold any number of preferences simultaneously...under the conditions of economic rationality these pre- ferences are assumed to be organized under one of two metavalues: maximum current profit or maxi- mum shareholder wealth. If maximum shareholder wealth is, in fact, the major preference of the firm, then the principal task of the finance of - floor is to build an implement models which best lead to such a goal. 2 An understanding of the consequences resulting from converti- ble bond financing would help management achieve this goal. Assume a firm has the alternatives of financing with (i) a pure debt issue carrying a face interest rate of 9 per cent, or (2) a convertible debt issue carrying a 5 per cent interest rate and exchangeable for no shares of common stock. The 22 Joseph S. hoag, Willard T. Carleton and Ewgene M. Lerner, ”Defining The Finance Function: A Model Systems Approach,” Journal of Finance, XXII (December, 1967). Sh9. rational I greatest as correct or Since man: tible bonc' deliberati covering I corporate specific 1 Inve: Corporate be aware : Well as W] Selves. bond Will tors. Th luating a W The 1 subOrtn na‘ 3h0u1d PO: Therefore ference 1] The 1 effects 01 23 rational firm will choose the issue that generates the greatest amount of wealth for a given amount of risk. The correct choice depends upon probabilistic expectations. Since management seems to be opting more and more for conver- tible bonds they should have a set of guidelines to aid in the deliberations pertaining to the financing decision. This study, covering a relatively long period of time, delved into overall corporate experience with convertibles and then concentrated on specific financial results for a selected group of issuing firms. Investor experience is the opposite side of the coin from corporate experience. Present and prospective investors should be aware of the convertible's effects on residual equity as well as what might be expected by convertible bondholders them- selves. A company's relative success in floating a convertible bond will have an important impact on the wealth of its inves- tors. The results of this study should aid investors in eva- luating a firm using convertible bonds. gzpothesis The Basicglpothesis The hypothesis was that financing with convertible subordinated debentures, instead of subordinated debentures, should positively improve the company's subsequent performance. Therefore, the null hypothesis would be that there was no dif- ference in subsequent performance. The intent of the research was to determine the financial effects on the firm from using convertible bonds. This goal was accomplls investiga in analys sideratlo: The debenture. subordina‘ lars 81011 as those c 141 and 1" minimum of Years 19 5c (1) earn” ci8!" flow of COHVerg Study srou IGthodg ell 2H accomplished in two stages. The initial, macro phase of the investigation was concerned with overall corporate experience. An analysis of general data provided a perspective for con- sideration of the study's second stage. The study then concentrated on convertible subordinated debentures. The study group consisted of all new convertible subordinated debentures of size greater than $10 million dol- lars excluding foreign and financial corporation issues such as those of banks and credit companies. These issues numbered 1&1 and were floated by 90 companies. In order to insure a minimum of 5 years Operating experience, bonds issued in the years 1950-1965 were studied. The variables examined were: (1) earnings per share, (2) rate of return on equity, (3) cash flow costs and benefits, (h) capital structure, (5) rate of conversion, (6) tradeoff terms. A detailed statement of study group selection, variables, and the specific analytical methods employed is presented in Chapter Three. Assumptions Several assumptions were made in this study. First, it was assumed that managers are wealth maximizers and that firms issuing convertible bonds have done so on a rational basis. Maximizing behavior may not have been the motivation for firms seeking to defer the dilution of reported earnings per share. To the extent that this is a factor, bias would be introduced in the study results. only 310 m the U restr (2) 1 issue trodu able ' 25 Second, in designing the study it was decided to include only subordinated convertible debenture issues larger than 310 million floated by domestic non-financial firms and sold in the United States. This limitation was introduced to: (1) restrict the study group size to manageable proportions, and (2) insure the availability of data. Stratification of the issue population ran the risk that directional bias was in- troduced that would affect the study results. Third, it was assumed that the financing options avail- able to the firm were: (1) subordinated convertible debentures, or (2) subordinated debentures whose quality and maturity were comparable to the subordinated convertible debentures. This implies that either of these forms could have been mar- keted with equal facility. The straight debt would naturally bear a correspondingly higher face interest rate. To the ex- tent that this assumption did not hold, the study results would not be as general. Fourth, it was assumed that a study rooted in the past would have relevance for future decision making. Methodology By providing a design within which to perform the study, scope is limited and it is more probable that the research will bear fruit. This empirical study followed positive theory. Weston states: "Positive theory seeks to describe and underb stand how the object of the study or description behaves. To d: not I impli descr alyze the t sever stitu‘ outst; timum were 1 in a I “nan. and t1 Vestit 313me With , (Engle 2L 26 23 To describe is to help understand." This study's purpose was not merely to describe. Weston's use of the word "understand" implies an additional, more productive benefit. He also says: The important point about explanation is not explanation itself. It is control. The reason we cannot accept the purist position thatprediction is the end of science is because man chooses to alter his environment. He seeks explanation for control or policy purposes so he can influence outcomes and not as an end in itself.2“ Thus, the present study, while positive in nature, first, described overall convertible debt financing and second, an- alyzed its effect for a selected issue group. Simulation was the technique employed in this study. Firm performance for several variables was statistically analyzed assuming the sub- stitution of a pure debt issue for the actual convertible debt outstanding. The general questions of optimum form size, op- timum cost of capital and efficient asset portfolio structure were not considered per-gs. The study was not intended to result in a general theory. What did come under examination was the financial decision to employ convertible subordinated debentures and the subsequent performance of the firm. This approach, in— vestigating individual firm performance of the firm. This approach, investigating individual firm performance combined with a consideration of the convertible debt market over time 23 J. Fred Weston, The Scope and Methodolo of Finance (Emglewood Cliffs, N.J.: Prentice-Hall, I556), p. 53. 24 Ibid., p. #1. should 8 bonds. The Theo 27 should allow a Judgement as to the wisdom of using convertible bonds. The Theorygaevisited An empirical study of this nature is based on convertible bond theory and the predictions that can be derived from the theory. Keynes states: Even when we are engaged in the mere collection and registration of events, it is often advantageous, as Jevons pointed out in the case of the physical observer, that our attention should be guided by theoretical anticipations. Industrial phenomena are ex- ceedingly complex, and unless we know what special facts to look for, it is quite possible that some of the most vital circumstances may fail to attract our notice.2 With this and the study's purpose in mind, convertible bond theory can be examined. First, compared to issuing pure equity, the use of convera tible bonds will result in fewer shares of common stock ulti- mately outstanding. Even.if a bond issue is fully converted, the fact that the conversion price is set higher than the mar- ket price at time of issue will mean a savings in the number of shares issued. Assuming that $1 million is to be raised and that the common is currently selling at 380 per share, 12,500 shares will be required if they can be sold at the market. Using a convertible bond with the conversion price set at 8100 means that 10,000 shares will eventually be needed. This argument 25 ' John Neville Keynes, The Scope and Method of Political Econo (London: Macmillan & Company, Ltd., I390, reprinted :Po 3. pu th th su re; th. to: 1201 81c the Ohc de; 01‘ Off 28 should be combined with the evidence that many convertible bond issues are never fully converted and so even more shares are saved. Therefore, whether earnings per share are calculated in the traditional manner or following APB No. 15 they should be higher for the firm issuing convertibles as opposed to common stock. This ignores the inescapable conclusion that even fewer shares may have been necessary if the financing were postponed and equity sold at a time when the market for the common stock was above $100 per share. Second, risk should be considered when evaluating alter- native financing forms. Comparing a convertible issue to a pure debt issue the degree of financial risk as measured by the leverage ratio will be the same under either alternative at the time of flotation for issues of equal size. If it is as- sumed that the proceeds are to be used for the same investment regardless of their source then business risk is fixed. However, the fixed charge inserted in the stream of earnings will be less for the convertible issue than for the pure bond. Therefore, the total earnings available for common and the total return on re- sidual equity should be greater for the firm using convertibles than for the simulated pure debt issue since the firm actively chose the convertible. The effect on earnings per share will depend on whether they are calculated in the traditional manner or following APB No. 15 as well as on the rate of conversion. Finally, Brigham and others have pointed out that a trade- off exists between the level of interest on a convertible and the con: is dete: equity 1 return 1 the bonc drawn, 1 should c plotting respondi to the s horizont can be d I means 00 terest 8: Yield 6-- p0“ (P011 corresDO: gains 001 Logicall; with lnv 1 F nVestOr \ 2 29 26 the conversion price. If the return on a convertible is determined by the bond's fixed income aspect and by its equity nature, management can influence themakeup of this return when it sets the coupon rate and conversion price in the bond indenture. Theoretically, iso-yield curves can be drawn, for given growth rates and terminal values, which should describe tradeoff opportunities. This is done by plotting the ratio of straight debt interest (is) to the cor- responding convertible coupon rate of conversion price (GP) to the stock's initial market price (MP0) is plotted on the horizontal axis. In this manner hypothetical tradeoff curves can be developed as shown in Figure 1.2. If the iso-yield curves are indeed the tradeoff curves, it means convertible bond investors are indifferent between in- terest and capital gains. Investors buying a bond priced to yield 6-1/2 per cent would be indifferent to one with a low cou- pon (point A) and one with a high coupon (point B) relative to correSponding straight debt interest. The pfospective capital gains component would be high for point A and low for point B. Logically, this means that investors would be equally happy with investments of the same return but unequal risk. Assuming investors would not be satisfied with this situation, it appears 26 Brigham, op. cit. pp. “2-48. >9“)? 1.6 1.5 1.14 1.3 1.2 1.1 F181!!! is 1.6 1.5 1.“ . ,F 1.0 \_ \ F Figure 1.2 Hypothetical Tradeoff Opportunities adapted from Brigham, 92. 959. p. M. l w that the yield cu: flatter 1 tradeoff rough, p] The the effec In t convertib tigated h. °°1'Porate enllbhasizix Chapter I applied tc 0f the Stt chamter v. remlts a! 31 that the tradeoff curve is something other than the iso- yield curve. In particular, the tradeoff curve should be flatter than the iso-yield curve to maintain a balance in the tradeoff of interest versus capital gains. Brigham's rather rough, preliminary data indicates flatness. The ideas discussed above provided the basis for analyzing the effects of convertible bonds on firm performance. Plan Of The Dissertation In this Chapter, introductory material was presented on convertible bond theory and on the specific problem inves- tigated here. Chapter II contains an analysis of overall corporate financing with convertible subordinated debentures emphasizing their relation to total corporate financing. Chapter III comprises a detailed discussion of the method applied to the study group. Chapter IV is a presentation of the study results. Tradeoff curves are developed in Chapter V. Finally, Chapter VI offers a discussion of the results and suggestions for future research. buslm volume the re vertib popula: evaluat America some he convert: Mined 1 V01ume a Vertible financmé securi ty. can Prov “dean: Prim Utilities. ‘ CHAPTER II THE VOLUME 0F CONVERTIBLE DEBT FINANCING Introduction The use of convertible debt in financing American business has varied considerable over time. Both the volume of debt and its characteristics have changed as the requirements for funds sources have changed. Con— vertible debt has been enjoying a period of relative popularity over the last four to five years. In order to evaluate the effect convertible debt has had on the American corporation it is necessary to first answer some basic questions. In this chapter the importance of convertible debt in American corporate financing is ex- amined for the period 1950-1969. first, convertible debt volume and its characteristics are reviewed. Next, con» vertibles are placed in the perspective of total corporate financing and also related to the firms using this hybrid security. Finally, the retirement of convertibles via the call provision is considered for underwritten and non- underwritten calls. Data Sources Primary data sources used were various Moody's publica- tions. Specifically, Moody's Industrial Manual, Moody's Utilities Manual, and Moody's Transportation Manual were 32 CI 86 St. fro: Conu E fin: Mom tic] ing abli 33 examined for each year of the study. The blue page inp serts contain a list of convertible debt securities out- standing at the publication date. These lists were comp pared year by year to establish a population of new public issues for domestic nonfinancial firms. The characteris- tics of these issues and the firms involved were then com- piled from Moody's. Data obtained in this way was verified and supplemented (where necessary) by consulting Mood s Convertible Bond Surggy and the semi-annual Corporate Fi- nancing Directory of the Investment Dealer's Digest. Other sources reviewed from time to time were prospectuses, list- ing statements and annual reports. Data relating to calls and retirements was also taken from Moody's. Overall corporate financing activity was compiled from several sources. Both the Securities and Exchange Commission in its Statistical Bulletin and the Investment Dealer's Digest publish a continuing series covering macro financing data. These were compared with figures from Mggdyfs Convertible Bond Survey and with the author's data. It should be noted that unless specific source citap tions are required for clarity, the sources for the follow- ing tables compiled by the author are the publicly avail- able records referred to above. 3h Vblume and Character of Convertible Debt m Overall Financing Activity Moody's data.--Concentrating first of all on con- vertible debt issues reported by Moody's Investors Service it is evident that on an absolute basis a considerable amount of convertible debt financing has taken place in the period 1950-1969. The recent popularity of conver- tible debt is revealed by an examination of new total public convertible debt (Table 2.1) which shows #0 per cent of the 1,350 issues floated came in the years 1966-1969. In dollar volume 5“ per cent of the approximately $21 billion total was brought to the market during the same four years. The average issue size fluctuated widely during the fifties, reached a low level of 34-37 million in the first half of the sixties and then Jumped to an average of 819 million per year in the last half of the sixties. This data is limited to new, domestically offered issues of American firms and excludes finance corporation issues such as those of banks and credit compani es . The debt instruments employed varied but the most pOpular was the subordinated convertible debenture (Table 2.2). The ”Other” classification includes all forms not covered by the specific categories and is made up primarily of secured issues. Subordinated convertible debentures emerged as practically the exclusive form of public conver- tible debt in the later years. TABLE 2.1 NEH TOTAL PUBLIC CONVERTIBLE DEBT, 1950-1969 (Millions of Dollars) Number Mean lumber Mean Year Dollars of Issue Year Dollars of Issue Issues Size Issues Size 1950 125.8 7 18.0 1960 31#.6 80 3.9 1951 2““.5 1n 17.5 1961 6H2.6 100 6.“ 1952 926.5 27 3M3 1962 331.9 62 5.1: 1953 950.3 23 “1. 3 1963 263. 8 97 5. 6 19511 139.9 21 6.7 196': 2711.0 39 7.0 1955 1 e 373. 2 58 23.7 1965 919. 3 61 15. 1 1956 809A 61 13.3 1966 1,650.0 83 19.9 1957 9'67.“ 61+ 1M8 1967 $799.8 209 17.9 19 58 1,055.8 1&8 22.0 1968 2. 793.8 1&3 19. 5 1959 53?.“ 8h 6.1+ 1969 30 2‘13. 3 119 27. 2 To 1 0 1 8 Some: Mocdy' s hauls. 36 TABLE 2.2 NEW PUBLIC CONVERTIBLE DEBT roams. 1950-1969 (Number of Issues) Subordinated Convertible Subordinated Industry Total Convertible Debentures Conv. Notes Others Debentures and Conv. Notes Indus- 1,198 989 11b 39 6 trial Utility 94 30 57 6 1 Trans- Portation 108 91 ll 1 5 Totals 1,350 1,110 182 “6 12 Source: Moody's Manuals. Comparing the data in Table 2.2 with that compiled by Pilcher in his study covering 1933-1952 reveals a shift in the use of convertibles among industry groups (Table 2.3).1 It should be noted that Pilcher's data includes a small number of bank issues. The shift in usage is rather pro- nounced in the decreased popularity of the conversion op- tion for utility issues. Industrial firms continued to be the heaviest employers of conversion and increased their proportion over the earlier period. Comparing the two periods, the difference in the industrial category is rein- forced by the fact that as of 1952 airlines were transferred 1 Pilcher,‘gp. cit.. p. 19. 37 TABLE 2.3 INDUSTRY USE OF CONVERTIBLE DEBT 1933-1952 1950-1969 Percentage Percentage of Total of Total Offerings Offerings Industrials 73.7 85.1 Utilities 22.0 6.9 Railroads 3.8 8.0 Banks 0 . 5 '- 100.0 mow Source: 1933-1952: Pilcher 1950-1969: Table 2.2 from industrials to transportation. The rails classification was changed to transportation at the time. A more detailed breakdown of convertible debt usage by industry is shown in Table 2.4. It is interesting to note the relatively high dollar volume in the transporta- tion category in the last five years. This reflects the considerable amount of financing being pursued by the air- lines for Jet aircraft. It is also illuminating to examine the effect of the American Telephone and Telegraph Company on this data. Removing ATT's 5 convertible debenture issues lowers the mean size of all public utility issues from 899.8 million to $17.9 million. On a total basis the mean of all public convertible debt issues drops from $15.8 million 38 .mfimocox m.mnoox "wouSOm n.- mos ~.H_sq~ w.se so l.mo~4¢ w.~s memes s.eso.ss ”sauce 2.2m o o.oms o.m o o.sm n.o~ sol ~.~na.~ cool ~.o~ cm c.n~n n.em e ~.mn~ o.o~ mwfi m.~nn.~ moo~ n.6s N2 s.oom «.mn a m.em~ B.ns oos ~.mmo.~ soo_ o.an NH m.mme _.e~ e m.om o.n~ n e.wnl.~ boom o.e~ e m.~qs o.m~ m o.m- m.o~ Nm m.oem meal _.sa a n.na m.~ a m.~ m.o an o.qoa soo— m.o a m.oo m.m N o.~l o.e mm m.om~ mood ~.~ e ~.m m. a m. ».m am o.-m woos o.o~ m e.em ~.m 6 m.me 2.6 om 6.6mm Leon 0.0 m c.me m.o m «.56 ~.m 0n n.-~ coo. m.~a n o.~m ~.aa e s.mo~ L.m ms e.lmn once o.m m o.ss o.e- o N.oes m.a 6m s.lo~ mass m. H m. ~.m~ as o.ob~ ~.n~ wm n.0mo anal L.m~ o o.ma o.- n m.~s~ «.ms om o.mse omen o.e N N.o L.~o a m.ono o.e~ as ~.emo «was N. l N. n.- m m.~o~ ~.m ml e.wm smog q.q m ~.ml m.r- m o.mmo 6.6“ ms r.mo~ mmoa q.~ n _.q ~.mm~ q c.0em _.o~ ON r.~wm Nmofl ~.s m n.~ 6.ooa N w.~o~ o.e om c.0e anal s.an a s.sm o.« s o.« m.o~ m ~.ew omos mn—m mmvmmw mmam mosmmu uwww mmsmmw mammu we mammu mo 03mmu mo Emmi bun-:32 mhm~ HOD cwmz hmDEDZ mhmd HOD cam—a. thEnz mhm~ ~00 Ln 0% coauouaoamcoas no.u«~_u: m~maaumsosu «measles no mcosssszs ooos-omol .sxamaazs an ammo mamsazu>zou oedema 2mz Q.~ mqmtk 39 to $13.9 million. In order to obtain a picture of the amount of con- vertible debt outstanding at a given time, Moody's Convertible Bond Surggy was consulted. Beginning with September, 1955, Moody's has published a monthly sup- plement to its Bond Survey which tabulates widely traded convertible bond issues. The total amount of converti- ble debt outstanding in the January issue of each year was compiled to arrive at the figures in Table 2.5. These figures include only industrial, utility and transporta- tion issues, excluding issues of foreign corporations, foreign offered issues of American firms and finance cor- porations issues. Further, this data should not be inter- preted to be the absolute total outstanding since the Convertible Bond Survey does not include many small issues or those in which there is not a relatively active market. Thus, a convertible given in an exchange offer and for which a market does not develop will not be included in the list until active trading develops. In the same sense, a "hung” convertible as well as those that have been substantially converted may be removed from the list. Therefore, the Convertible Bond Survey cannot be considered a universe‘of issues or a completely accurate gage of new issues in the same way as the previous Moody's data. However, since it does include the bulk of the maJor issues it re- presents the convertible bond market to a reasonable degree. ll 40 TABLE 2.5 HOODY'S CONVERTIBLE BOND SURVEY TOTAL AMOUNTS OUTSTANDING FOR INDUSTRIALS, UTILITIES AND TRANSPORTATION (Millions of Dollars) Number T18 Number Year Dollars of Year Dollars of Issues Issues 1955 1,717.0 52 i 1963 2,668.5 211 1956 1,572.4 53 1964 2,740.5 213 1957 1,419.6 56 I 1965 2,671.3 197 1958 2,142.2 108 ' 1966 3,142.7 202 1959 2,295.6 112 g 1967 4,160.4 209 1960 2,443.5 187 ' 1968 6,400.7 233 1961 2,338.4 206 Q 1969 8,291.5 324 1962 2,600.6 206 _h 1970 10,749.1 401 Source: Mood '8 Convertible Bond Surve , January issues. Nets: SerIesBEgan,‘Séptember, I9 . Securities and Exchange Commission data.-- Data on absolute convertible debt financing was presented in the previous section. The SEC compiles and publishes continu- ing series on corporate financing for bonds, preferred stock, and common stock. Since 1956, the series includes data on convertible bond financing (Table 2.6). A dis- cussion of the issues included in the SEC data will be pre- sented in a later section. Basically, these are all issues sold for cash in the United States and include private placements. 41 TABLE 2.6 NEW CONVERTIBLE BONDS OFFERED FOR CASH IN THE UNITED STATES (Millions of Dollars) Total Publicly Offered Privately Placed Year Dollars Dollars Percentage Dollars Percentage 1956 925 763 82.6 163 17.4 1957 1.064 995 93.5 69 6.5 1958 1.147 1.071 93.6 77 6.4 1959 628 536 85.3 92 14.7 1960 426 356 77.2 105 22.8 1961 710 625 88.0 84 12.0 1962 445 346 75.5 99 24.5 1963 357 234 65.6 122 34.4 1964 42 366 86.2 59 13.8 1965 1,264 1,181 93.4 83 6.4 1966 1,872 1,764 94.3 109 5.7 1967 4,475 4,108 91.8 367 8.2 1968 3,281 2,663 81.2 619 18.8 1 6 4,941 3,099 76.7 942 23.3 'TEIEIE“21,096 18,107 85.8 2i990_ 1&12 Source: U. 8., Securities and Exchange Commission, Statistical Bulletin, XXIX (March, 1970), 19. Comparing the volume of convertible bonds to other financing forms shows the relative popularity of converti- bles. Stated differently, Table 2.7 analyzes on a relative dollar basis in what magnitude management actively chose to employ the conversion option. This data includes both pub- lic and private issues with the exception of those issues Specifically labeled as public. It can be stated that both absolutely and relatively the .mocom odn0uue>cou mucemouaou .m .0 "00oz .mmammH msomum> .saumflfinm 000000.000m .sommmueeou owsonoxm one 00.000900m ..m .2 "oopsom 42 0.00 00m400 0.00 000.00 0.00 0000000 0.0 000.000 000000 0.00 000.00 0.00 000.0 0.00 000.02 0.00 000400 0000 0.00 000.00 0.00 000.0 0.00 000.00 0.00 000.00 0000 0.00 000.00 0.000 000.0 0.00 000.00 0.00 000.00 0000 .- 000.0 0.00 000.. 0.00 000.00 0.00 000.00 0000 000.0 0.00 000.0 0.0 000.00 000.00 000_ 000.0 0.00 000.0 0.0 000.00 000.00 0000 000.0 0.00 000.0 0.0 000.00 ~00.00 0000 000.0 0.00 000._ 0.0 000.0 000.00 0000 000.0 0.00 000.0 0.0 000.0 000.00 0000 000.0 0.00 000.0 0.0 000.0 000.00 0000 000.0 0.00 000.0 0.0 000.0 000.0 0000 000.0 0.00 000.0 0.00 000.0 000.00 000_ 000.0 0.00 000.0 0.00 000.0 000.00 000. 000.0 0.00 000.0 0.00 000.0 000.00 0000 mucom 0-0~3m mtcom xuOum xooum mvc0m mncom mcaucmc—h manoeucrm mo N n ma 0‘1“?“ GOEEOU mo COEEOU Mayo“. mo ku08 7.508 No .0. man-OH hQU> .m.u 0033.0 00000 N mm .m .u ~38. .0 mm .md mm .m .u «mama—co we acoufi~02v wzHuzHHon 0.N m0m¢b 43 popularity of convertible bonds has increased markedly in the last half of the sixties. Beginning in 1956, convertible bonds relative to total financing and to total bonds dipped to a low point in 1963 and then rose to his- torically high levels at the end of the period. The re- lationship of convertible bonds to common stock has been somewhat more erratic. However, it can be assumed that in the later years management has perceived the convertible to be an increasingly desirable alternative to the issuance of common stock. A possible explanation of this is that the convertible is viewed as a first step in the evolution of what will eventually be equity financing. Table 2.7 also includes the relationship of public con- vertibles to total public bonds. The pattern is quite similar to that of total bonds with the percentages some- what higher in the public bond category. Private conver- tibles relative to total private bonds reveals that converti- bles play a rather minor role in private placements. In al- most all years private convertibles are approximately two per cent of all privately placed bonds. However, in the years 1967, 1968, and 1969 the relationship is 5 per cent, 9 per cent, and 17 per cent, respectively. It may be that private capital suppliers are recognizing the advantage of the conver- sion option or it may merely reflect the overall popularity of convertibles. gagparison of data sources.--In this section a compari- son is made of the four maJor data sources examined in this 44 study. First, Mgggy's Industrial, Utilitygand Transportation Manuals are considered. The data includes all domestic non-financial public convertible debt reported in the manuals. Private issues are not included since it ap- peared that this source does not present a complete picture of private financing. The second source, Maggy's Converti- ble Bond Suryey, is related to the first source. The data from the Survey is essentially similar to that from the Manuals except that it is almost exclusively composed of widely traded subordinated convertible debentures. Further, some issues appearing in the Survey for the first time may not be new issues as discussed in a previous section. The third source is the Securities and Exchange Com- mission's Statistical Bulletin. This is a monthly publi- cation which reports substantially all new issues of securi- ties offered for cash sale in the United States in excess of $100,000 and with terms to maturity greater than one year. Fourth, the Investment Dealer's Digest series includes all underwritten security issues in the United States and foreign issues of United States firms. The Digest also re- ports any non-underwritten securities that come to its atten- tion. It is felt that this includes practically all such issues. These four sources are compared in Table 2.8 for public issues. It is evident that the two Moody's series will not agree I’lj 95 with each other nor will they agree with the Investment Dealer's D_ig_e_s£ or SEC. This is due primarily to the ex- clusion of foreign and financial issues in the author's compilation. The Investment Dealer's Digest data is more of a ”universe” of offerings in that it also includes is- sues of U. 8. foreign subsidiaries, exchange offers and SEC registered secondaries. On the other hand, the SEC data does not include registered secondaries, offers of closed end investment companies, SBIC's and other differences, but the SEC does make periodic revision of proceeds as reported by a company. An attempt to reconcile the Investment Dealer's Digest data with the SEC data for selected years was successful. Another insight of the degree to which management, over all industry groups, chose to use a convertible rather than a non-convertible is found in theIgnvestment Dealer's Digest data. Over the period 1959-1969, for which data was available, 32.6 per cent of the total number of public bond offerings were convertible. This compares to Pilcher's findings of 9.3 per cent for the period 1933-1952 using essentially the same source.2 In terms of dollar volume reported by the Digest, 18.1 per cent of public bond offerings were convertible during 1959-1969. These figures can be compared 2 Pilcher, 22. 011., p. 21. .mouooou 0000~0o>0 00000000 5000 nosuso 00 00000500 ”000200 ‘46 000.000.00 00040 0.000.00 0.00 000 0.000.00 0.00 000 0.000.00 000000 00040 0.00 000 0.000.0 0.00 000 0.00040 0.00 00" 0.00040 0000 000.0 0.00 000 0.000.0 0.00 000 0.000.0 0.00 000 0.000.0 000_ 000.0 0.00 000 0.000.0 0.00 000 0.000.0 0.00 000 0.000.0 0000 000.0 0.00 00 0.000.0 0.00 00 0.000.0 0.00 00 0.000.s 0000 000.0 0.00 00 0.000.0 0.00 00 0.000.0 0.00 00 0.000 0000 000 0.0 00 0.000 0.00 00 0.000 0.0 00 0.000 0000 000 0.0 00 0.000 0.00 00 0.000 0.0 00 0.000 0000 000 0.0 00 0.000 0.00 00 0.000 0.0 00 0.000 0000 000 0.0 00 0.000 0.00 00 0.000 0.0 00_ 0.000 0000 000 0.0 00 0.000 0.0 00 0.000 0.0 00 0.000 0000 mmsmmu mwamuu mmammn maofigoo seen we muoLLoo :00: mo mue-oo see: mo managed anew uoBEsz 009532 wonE:z umommo ANu>usm poem 0mm 0.000000 usoeumo>su o~nmuao>soo 0.xoooz Lesson 0.00ooz «mumugoa wo msoudguzv mmuzum nzom mamahmm>zOu Uuqnzm 3mz m.~ mqm0 00000030 6000 nonuam 00 00000560 "mousom o.” 00 0.00 0- 0.00 000 m.m~ 00m 0.00 000 0.0 00 000.0 mdouoa 0.0 0 0.00 000 0.00 000 0.00 000 0.0 00 0.0 00 000 00-0000 0.0 0 0.0 00 0.00 00 0.00 000 0.00 00 0.0 00 mun 00.0000 0.0 0 0.00 00 0.00 no 0.0m no" 0.0_ No 0.0 00 000 00-0000 0.0 0 0.0 0 0.00 0H 0.00 an 0.0n mm 0.0 0 00 00-0000 N noses: N 000552 MI‘ 000522 N nonesz AM 000602 M1 000632 +00 00-00 00:00 00-00 00:00 0-0 mosmmH mo 0000200: on mums» nonfinz manor wyumph0 00000930 5000 005020 an 000005ou 0000som 000 mmn 000 N0 000.0 0000o0 0mm 000 000 mm 000 00002 0oz 0.000 Own 0.000 000 0.000 ~0~ 0.000 00 0.000 000 0000a 00090 0.0 0 0.~ 0 0.0 N n- u- 0.0 0 no 0 00o 0.00 000 ~.00 00 0.o~ 00 0.00 0 0.00 00m 0 0.~0 ~00 ~.00 00 0.00 000 0.00 0N 0.00 000 mm 0.0 mm ~.o0 00 0.00 00 0.0m 00 0.00 00 man 0.~ 00 0.0 m 0.0 00 0.0 0 0.0 On ¢ E. E. 0.0 0 0.0 0 n.0 0 N.0 00 0< ml 00 J0 00 .0. 00 0. 00 .01 0 0000 - 0000 0000 - 0000 0000 - 0000 0000 - 0000 0000 - 0000 0000000 m mummu many—“.5 0250.50 mtg—ODE ~0.N uqm<fi 51+ .0000000 0000000>0 00000000 5000 005000 00 00000500 0000000 00 EN 000 000 000 00 00 000.0 0030.0 0.0 m 0.0 m0 0.0N 0N0 N.mn 0mno.mN m00 N.m 00 0.0 00 000 00000 0oz 0.00 0 nu un 0.0N N 0.0N N 0.0N N un nn nu un 0 mu m 000 0.0 0 0.00 00 0.00 000 0.00 09 0.00 00 0.0 0 un un 00N m 0.0 0 0.00 0N0 0.00 000 0.0N 00 0.0 00 0.0 N nn nu 000 0m N.0 0 0.00 00 N.00 N0 0.0 0 N.0 0 nn nu un nn 00 00m 0.00 0 0.00 0 0.00 0 0.00 00 0.0 N nu nu nu nn 00 m 0.0N N 0.00 0 nn nn 0.00 0 0.00 0 nu nu nn un 00 00 .0. 00 0. 000 .0. 000 0.. 000 0. 00 .0 000 .0 00 003.00 .00 00 00 00 00 00 - 00 00 00 0 0 0 o smug 05000 00000002 00 0000» whuxnh mmzHH0 00000000 6000 005000 00 00000600 0000000 56 0.0 0 w.N mm w.mN 000 0.00 N00 0.0N 000 0.0 00 0.0 0 000.0 000000 0.0 0 0.0 00 N.00 00m 0.00 000 0.0 00 0.0 0 0.0 0 000 000mm 002 nu uu uu uu 0.N0 m N.00 0 nu nu uu nu nn nu 0 00 0 000 nu nu 0.0 N 0.00 00 0.N0 000 0.N0 00 0.0 0 u. nu 000 0 u uu 0.0 0 0.0 0N 0.00 000 0.00 000 0.0 00 un uu 000 00 -u u- u- -u 0.0 0 0.00 00 0.00 00 0.00 00 uu u- 00 000 nu nu nu nu uu uu 0.0 0 0.00 00 0.00 00 nu un 00 0 -u uu u- u- u- u- u. uu 0.00 0 0.00 0 0.00 0 00 00 x x m 0 x w s 000000 +0.0 0.0 n 0.0 0.0 n 0.0 0.0 n 0.0 0.0 n 0.0 0.0 n 0.0 0.0 30000 00 000000 qu 000602 0000 00000000 coaaou 00000020 200000 020 0020000 0.00002 q—.m w0m¢H 57 TABLE 2.15 CONVERTIBLE BOND CALLS (Millions of Dollars) Number Dollar Standard & Year of Amount Poor's Index Issues Called (1941 - A3 - 10) 1°56 2 38.3 A6.62 1°57 l 8.0 64.38 1958 11 126.8 h6.2& 1959 1 0.7 57.38 1960 12 186.3 55.85 1961 16 105.6 66.27 1962 2 29.5 62.38 l963 lO 8A.8 69,87 I964 15 87.5 81.37 1965 12 216.8 88.17 1966 12 100.1 85.26 1967 28 662.6 01.93 1968 24 £20,0 98.70 1969 12 ' 229.3 97.84 Totals 158 2,296.1 Source: Compiled by author from publicly available records. Stock market activity could be expected to have a major influence on calls. If the purpose of calling is to force conversion, periods of rising stock prices would be the most conducive to successful calls. The probability of conversion value being safely above the call price would be greater given rising prices as Opposed to falling prices. Management must make a decision as to how wide a differential they desire between the stock's market price and the conver- sion price before they risk a call, If market developments 58 push the price of the common stock back toward the conversion point it may become monetarily disadvantageous to con- vert and more attractive to accept the call price. Table 2.16 illustrates actual corporate practice regard- ing calls of convertible bonds for which data was available. Based on conversion value almost a half of all calls took place in the range 111-160 per cent. Brigham reported that almost a quarter of the companies he surveyed (22) force conversion as soon as conversion value exceeds call price by 20 per cent.5 As stated above, the company is risking mar- ket price deterioration between the time of announcing the call and the actual redemption date. In order to mini- mize this risk the technique of underwriting the call has evolved. In a survey of 29 underwritten industrial con- vertible bond calls, Williams and Letwat found that 19 calls went out when conversion value was in the range of 120-160 per cent. In the present study, approximately 75 calls were underwritten. Regarding calls below one hundred, possible explanations are the removal of a ”hung” converti- ble or the cleaning up of the capital structure preparatory to a merger. Brigham, 22. cit., p. 52. 6 B. S. Williams and Marvin Letwat, 'Underwritten Calls of Industrial Convertible Securities", Quarterly Review of Economics and Business, III (Spring, 1963), 76. filllulll u as, - as... l .._ 7 P" 59 TABLE 2.16 PREMIUMS.AT WHICH CALLS OCCUR Premium Based on Premium Based on Premiums Bond conversion Value Bond Market Price Above (Number of Bonds) (Number of Bonds) Gall Price 100 and Below 17 3 101 - 110 8 23 111 - 120 18 20 121 - 130 18 17 131 - 1&0 12 10 1H1 - 150 1a 15 151 - 160 9 11 161 - 170 6 6 171 - 180 5 5 181 - 190 6 4 191 - 200 i u 201 - 210 6 2 211 - 220 5 h 221 — 230 3 h 231 - 2&0 2 8 ahi - 250 6 O 251 - 260 2 h 261 - 270 1 1 271 - 280 1 0 281 - 290 1 1 291 - 300 O 0 301 and Above 5 5 Conversion Egpiggg _;_fi - Totals “in? 155: Source: Egggz's Convertible Bond Survey, January 1956 - December r9 0 6O haturities and defaults.--Convertible bonds have a tendency not to mature since it is often management's goal to use convertibles as a means to effect equity financing. Most issues will end up in either voluntary or forced conversion. If market conditions favorable to conversion do not develop the security will often be called in order to pave the way for future financing. During the study period, 53 convertible bonds matured as compared to 1,350 new issues. The maturing face value was $30.1 million. Only 7 issues had amounts greater than $1 million outstanding and all were below $5 million. Default experience with convertible bonds has been re- latively rare. During the study period 8 issues were in default. There were also 10 firms in reorganization of which 2 had defaulted on convertible securities. Repeat Issuers A thesis sometimes seen in the literature is that con- vertible bond financing is used only rarely in the firm's life.7 This is when the firm is relatively small but growing and a unique set of external and internal conditions pro- vide a setting favorable to floating a convertible. In order to examine this idea, a tabulation was made of the number of convertible issues and their respective issuing firms (Table 2.17). 7 Eli Schwartz, Corporation Finance (N.!.: St. Martin‘s Press, 1962), p. 17?. \.Il“llll|ll,'r| { Jurll. 1111' II 61 TABLE 2.17 FREQUENCY OF CONVERTIBLE BOND USAGE 1950-1969 Number of Total Number Issuing ~Issues of Firms Per Firm Issues 802 1 802 145 2 290 N6 3 138 19 h 76 6 5 30 l 6 6 l 8 8 Totals 1,020 1,350 Source: Compiled by author from publicly available records. It can be seen that 80 per cent of the issuing firms floated one convertible during the study period. However, those firms that employed convertibles more than once ac- counted for approximately #1 per cent of the total number of issues. Evidently, there is a group of companies that have enjoyed a favorable enough response to use them more than once. Summary This Chapter presented data pertaining to convertible 62 debt financing by domestic corporations for the years 1950-1969. Convertible debt enJoyed its greatest pop- ularity on both an absolute basis and relative to other financing forms during the latter years of this period. Additionally, mean issue size was rising at the study's end after hitting a low in 1960. Empirical data supported theory in many areas. It was found that, measured by assets, large companies tended to float large issues as compared to small companies putting out small issues. In a similar vein, stronger firms, as guaged by bond ratings, employed the largest maturities for convertible debt instruments. The most popular maturi- ties fell in the 15 to 20 year range with longerblived bonds becoming more pepular at the study's end. Additionally, high bond ratings generally meant low face interest rates and vice-versa. 1t was also found that more large firms have been issuing convertible debt which may account for the rise in mean issue size and the lengthening of maturities. Contrary to this idea is the fact that the general level of bond ratings for convertible debt has been dropping over the last 20 years. Data relating to calls was also examined and it was found that 50 per cent of all convertibles were called while conversion premiums ranged from 111 to 160. Also, 50 per cent of all calls studied were underwritten. Finally, several firms have issued convertibles more than once. CHAPTEB.III BACKGROUND AND METHOD Introduction In this Chapter the study's scope is narrowed in order to examine financial effects on the firm from con- vertible bond usage. First, the procedure followed for selecting firms that were included in the study is pre- sented. Next, the convertible issues floated by the selected firms is delineated. The method employed in the study group examination is discussed as an illustration of the experimental design. This discussion includes a presentation of the variables adcpted in the study and implements the theory presented in Chapter 1. Thus, this Chapter provides the details of the study issues and the method used for determining their effects on the respective firms. The results will be presented in Chapter IV. Firm Selection The primary criterion followed in firm selection was to obtain as complete a picture of convertible debt as possible. Therefore, it was determined to examine a relatively long time series as opposed to adopting cross section analysis. In this way, the effects of time as reflected by changing 63 6“ market conditions and individual firm fortunes were encom- passed. Chosen for investigation were issues floated during the period 1950 through 1965. Stopping at 1965 insured that actual operating data would be available for a minimum of five years. Referring to Chapter II, it can be seen that a large num- ber of convertible debt issues of various categories were floated during the period in question. In order to hold the study to manageable proportions, it was decided to select do- mestic nonfinancial corporations that issued subordinated con- vertible debentures in sizes greater than 310 million.1 This form of convertible debt was and continues to be the most fre- quently employed. Limiting the study to issues over $10 million eliminated a large number of relatively small issues and eased the burden of analysis. Further, and more importantly, it gave the assurance that data would be available from public sources. Convertible debt issues outstanding are listed yearly in the blue page inserts of Eggdyjs Manuals. These lists were compared on a year by year basis for the Moody's Industrial, Utility and Transportation Manuals. In this way new issues were detected. Other sources such as Moody's Bond Survey and the Investment Dealer's Digest were scanned to insure that no new issues were missed. The size of each new issue and its specific form was determined from Moody's individual company 1 For the remainder of the paper, the term convertible will mean subordinated convertible debenture. 65 data. Those issues meeting the stated criteria were put in the study group. The issues selected are listed in Appendix A. Once a firm was admitted to the study group, the pro- blem arose of subsequent subordinated convertible debenture issues, particularly those following the 1965 cutoff date. All subsequent convertible subordinated dubentures were put in the study group, including the post 1965 issues, pro- vided they met the initial size criterion. The study firms floated five issues during the 20 year study period that did not meet the size requirement. Another consideration was that some firms offered sub- ordinated convertible debentures in exchange transactions -— either for another firm's assets or for its own securities. The trouble here was that in the case of five issues sub- sequent operating data was not available and so these issues were dropped from the study. Mergers and liquidations caused a further problem. A num- ber of companies in the original group disappeared during the study period due to mergers and one, Merritt Chapman Scott, was in the process of liquidation. These firms were examined individually and left in the study as long as seemed relevant. The prime consideration here was whether the convertible continued to exert influence on the sure viving company's performance. This was determined based on the relative sizes of the merging companies as measured by 66 assets. Warren Petroleum was absorbed by Gulf Oil in 1956 and Warren's convertible was retired in the process. The company was dropped from.the study. However, Sinclair Oil, Richfield, and Atlantic Refining were all in the original study group and all merged to become Arco. Arco and its component issues was carried through to the study's conclusion. There were 90 companies in the study group and these firms issued 1&1 convertible subordinated deben- tures during the study period. . Issue Characteristics Table 3.1 presents background data on the study group issues. Issues ranged in size from 310 million to $200 million with approximately half the total number falling between $10 million and $30 million. Industrial firms were the heaviest users of convertibles. Utilities have virtually ignored convertibles and this is even more significant when it is considered that two of these issues were put out by General Telephone and Electronics Company, which performs a certain amount of manufacturing. Convertibles have be- come popular with airlines. In the transportation category, 17 issues were floated by airlines since 1960. It should be noted that maturities on convertibles tend to be similar to those of straight debt. The question might be raised as to why such long maturities are used if re- latively rapid conversion is expected. In a similar vein, 67 TABLE 3.1 STUDY GROUP STATISTICS Dollars Number of Issues {Millions} Mayor Per Cent A. Sise of Issues 10 - 19.9 38 26.9 20 - 29.9 32 22.7 30 - 39.9 18 12.8 ”0 - “9.9 9 6.4 50 - 59.9 10 7.1 60 - 69.9 6 u.) 70 " 79e9 ‘l’ 208 80 - 99.9 3 2.1 100 -12#.9 10 7.1 125 -1“9.9 3 2.1 150 -17u.9 6 4.3 175 -199.9 2 1.9 B. Industrial Cstogories 222823.! Rumba; Per Cent Industrial 11? 83.0 Transportation 21 ill. 9 Utility 3 1.5 C. hturity Years to Maturity number Per Cent 10 1 0.7 13 1 0.7 15 8 5.7 18 1 .7 20 #2 29.8 21 1 .7 22 3 2.1 25 7“ 52.5 26 2 1.1. 30 8 5.7 D. G. I. TABLE 3. 1- Continued thlity Ratings Coupon lute Offerings Conversion Price Sinking Fund Issues per Company 68 Moody aetigg Number Per Cent A u 2.8 Baa 37 26.2 Ba 89 59.6 B 19 9.9 Not rated 2 1.9 Per Cent Number Per Cent 0 - 3 2 1.4 3 1/8 - 3. 11 7.8 3 5/8 - 4 25 17.7 a 1/8 - hi 52 36.9 h 5/8 - 5 29 20.6 5 1/8 - 9} 16 11.3 5 5/8 - 6 6 u.3 Number Per Cent Public Offerings 75 53.2 lughte Offerings 61 113.3 Exchange Offerirgs 5 3.5 Number Per Cent Fixed 10? 75.9 Variable 3“ 24.1 W Sinking nuns 133 99.3 No Sinking Fund 8 5.7 Number of Number of .____..qu°8 92W 1 59 60.0 2 23 25.6 3 11 12.2 b 2 2.2 69 over 9b per cent of the issues included sinking funds. With a convertible, however, most sinking funds did not commence until several years after issue-typically, around 10 years. The sinking fund would not be a maJor factor then, if early conversion occurs. The funds were designed to retire from 50 to 90 per cent of the issue amount. Bond ratings, as shown, tend to fall on the lower end of the rating scale. About 70 per cent were Ba or lower. Baa is the lower limit for investment grade and this may have implications for regulated institutional investors. However, it should be remembered that convertibles tend to be downgraded due to their equity aspects and thus con. vertible ratings are not an absolute measure of firm strength. About half the issues were offered publicly and there were relatively few direct exchange offers. Approximately 75 per cent of the issues employed a fixed conversion price. The variable conversion price is sometimes used to encourage conversion before the converb sion ratio drops. While not shown here, a number of firms provided for the conversion privilege to expire before maturity. Alternatively, a few companies delayed the con- version privilege's effectiveness until some time after the issue date. This is done if the conversion price is set below the common market price at issue. Thus, immediate conversion is precluded but eventual conversion virtually 7O assured assuming no drastic turn of events. Also not shown here but related to conversion is the fact that all the issues in the study group were callable. Finally, it can be seen from Table 3.1 that #0 per cent of the study firms issued more than one convertible. The two companies with four issues a piece were both airb lines. Simulation Procedures followed for selecting the study firms and their respective issues were presented in the preceding section. The study group consisted of 1&1 convertible sub- ordinated debentures issued over a 20 year span. In this section, the analytical techniques employed in assessing the convertible's impact on the firm are detailed. As will be recalled from Chapter I, the study hypothesis was that the use of convertible bonds should positively improve company performance. Support for the hypothesis would be shown if convertibles had a favorable effect on such variables as earnings per share, rate of return and others. These are developed in the following material. The fundamental approach was to build a model that would simulate study group firm performance under an al- ternative financing plan. Simulated performance was then compared to actual performance in order to measure the COD! vertible's impact. 71 Straight Debt Alternative It was assumed that, at the time the convertible was issued, management could have sold straight debt as opposed to convertible debt. The straight debt would carry a core respondingly higher face interest rate than the convertible. There is a precedent for such an assumption since Brigham found that, "All but two respondents indicated that straight- debt could have been sold."2 There were 22 respondents. The problem of the interest rate to be assigned the hypo- thetical straight debt was solved by adopting the rates found in Moody's Bond Surv_y. Moody's regularly evaluates converti- bles based on prevailing yields of nonpconvertible bonds of comparable quality and maturity. These estimates, taken at the issue date, formed the basis for comparison with the actual rates. To the extent that this is not a viable assumption the study results would not be general. Another operational assumption incorporated in the model was that portions of the straight debt issues would be called from time to time and replacedby an equivalent amount of com- mon equity. The purpose here was twofold. First, the actual convertibles were sometimes voluntarily converted and sometimes called in entirety. Retirement of the hypothetical issues was necessary to maintain comparable capital structures. Second, it has been suggested by several of the writers cited in Chapter I that the main purpose for using convertibles is that they are 2 Brigham, op, cit., p. 51. 72 an expedient way of selling common stock. Brigham found that 73 per cent of his respondents were primarily interested in obtaining equity.3 Further evidence relating to this point will be found later in this paper. Therefore, a decision rule was adopted that whenever a cumulative 25 per cent of the actual issue was converted or called to force conversion that a corresponding amount of the hypothetical issue would be re- placed by common equity. The 25 per cent figure was used to eliminate the need for replacing minor amounts of debt and to avoid hypothetically constant trips to the capital market. There are two acceptable methods to account for the con- version from debt to equity.“ The face amount converted can be transferred to equity at book value, or the market value of the stock surrendered on conversion can be transferred. The latter method involves recognizing a gain or loss on the income statement due to revaluing the bond before retirement. Investigation revealed that most firms follow the book value method. This is probably to avoid the loss recognition at- tendant in the market value method for a ”successful” converb sion. A loss, under this method, would arise whenever a company exchanges its convertible for stock whose market value exceeds the debt's book value. The book value method was used in this analysis since it was the most commonly encountered in practice. 3 Ibid. 11 Eldon S. Hendriksen, AccountinggThecrz (Homewood: Richard D. Irwin, 1970), pp. 522—523: 73 Thus, the simulation model replaced the actual con- vertible with an equivalent straight debt issue which was gradually supplanted by common equity as the actual issue was converted. If the convertible was called without forcing con- version, the straight debt was also merely called without issu- ing common stock. Non-conversion calls were determined by com- paring conversion values with call values in Moody's Converti- ble Bond Survey. Additional aspects of the simulation model are considered in the following discussion of financial vari- ables. Variables Earnings Per Share Support for the study hypothesis that convertible usage should improve operating performance would be shown if earnings per share under the convertible are greater than under the pro- posed alternative straight debt. It was noted that, ceteris paribus, the difference in interest rates should generate a difference in net earnings under the two alternatives. The ex- pected increment in earnings after taxes for a given time period was expressed as: AF. = EA - ES Where: AB 8 change in net earnings E8= simulated net earnings under straight debt EA? actual net earnings under convertibles A complete list of symbols used in this paper can be found in Appendix B. 71+ Letting: EBT 8 earnings before taxes t = time tr’= tax rate = nth issue of a given firm dollars of interest P H :3 11 a face interest rate P = principal outstanding EBIT 8 earnings before interest and taxes. Then the analysis proceeded as follows using the subscript A to denote actual situation terms and the subscript S to denote simulated situation terms. .AE = EAn- E8 =- (l-tr) (EBITA-IA) - (l-t) (EBITS-Is) After some rearranging and the fact that EBITA=EBITS, AB = (l-tr) (IS-IA) For operational purposes, use was made of the fact that IA - iAPA and IS - 1393 so that, AE =(1-tr) (131’s - iAPA). Since is was greater than iA it was expected thatnoE would be positive. Once the earnings increment was calculated, simulated earnings were found by substitution in Es 8 EA - AB. The model shown above illustrates the calculation for one time period and one convertible issue outstanding. This situation was generalized by the addition of a time subscript, t, and an issue subscript, n. Thus, PA35 would denote the actual principal 75 outstanding, for the third convertible issued since 1950, in the fifth year since issuance. The model would then become: AEt = (but) “1319316 * isnPSnt) - (111P11t+---+ input} This method of calculation fits reality since it allowed the actual interest paid on the convertible to decrease over time as conversion took place. Further, interest paid on the alter- native straight debt droppod only at those points where the firm replaced the debt with equity. The next step was to calculate earnings per share. Actual earnings per share, EPSA, were available from published data. Simulated earnings per share were calculated by dividing the previously determined E3 by the number of simulated common shares, 83. In order to arrive at 88 it was necessary to re- move the number of shares issued under the actual convertible and replace them with any shares issued under the simulated straight debt for each year in question. If the convertible, PA’ was being voluntarily converted or called to force conversion, the number of common shares out- standing, SA’ had to be reduced accordingly by the number of shares issued on conversion. This was to get the number of shares back to where they would have been if straight debt had been issued originally. The number of shares that were removed, ant’ was found by dividing the face amount converted by the conversion price, GP, for each year in which conversion shares were outstanding. The calculation was made according to the 76 formula: m T Sent.“ E Z APAnt: UP nsl t=1 nt where T a maximum number of years the firm was in the study and m I the total number of issues outstanding at a given time by the firm in question. These cumulative figures were used to reduce the actual shares year by year. However, as soon as a cumulative 25 per cent of the actual issue was converted, it was assumed that the same amount of P8 was called and equity issued in the same dollar amount. The number of shares that were hypothetically issued were found by dividing the face amount issued by the common's market price, HP, at the time of the simulated issuing. Again, the cumulative amount issued up to any time was calculated by: m SDnt ' E APSnt . n81 t=l iat The total number of shares outstanding under the simulation alternative, SS, was determined for each year by subtracting the number issued under actual conversion up to that time from the actual shares outstanding and then adding back the number issued under simulation up to that time. m T m sst = SAT - E E SCnt + Z ism“ ' t: nal t=1 n=1 Or: Then the calculation for simulated earnings per share was made by: E EP8 s St . St SSt 1 in." m 77 Before continuing with the development, it is necessary to digress briefly. APB Number 15. -- It will be recalled from Chapter I that the Accounting Principles Board has established criteria regarding earnings per share computations for firms with con, vertible debt. Under the Opinion, companies whose converti- bles meet the test as common stock equivalents have their earnings per share calculated as though the convertible debt had been converted. This results in a primary earnings per share and a fully diluted earnings per share neither of which resemble the traditional earnings per share measure. The pro- cedures put forth under APB number 15 were not followed in this study since it was felt that the resulting measures were not relevant for either managerial decision making or invest- ment analysis. . There are several maJor criticisms of APB Number 15—- Hawkins lists 18, and some others are presented here.6 First, convertibles are Judged as to whether they are common stock equivalents gt'igggg and since conversion will never take place for some issues, the diluted earnings figures would not be relevant. Second, there is evidence that investors have accounted for possible dilution in establishing market prices.7 6 David P. Hawkins, Corporate Financial Reporting (Hemewood: Irwin, 1971), pp. 243-ZU6. 7 Eugene M. Lerner and Rolf Auster, "Does the Market Dis- count Potential Dilution?”,p§inancial Analysts Journal, XXV (July-August, 1969), 118-121. 78 Third, there was the inconsistency of relating diluted earnings to a capital structure in which no part of the proceeds received from floating the convertible are con. sidered equity until conversion. Fourth, most pertinent to this study was the operational problem of second quessing management on issues which came out prior to APB Number 1 . The question was whether or not the debenture would have been issued as a common stock equivalent. If the cash yield on.the convertible is less than two-thirds the prime rate at issue the debenture is a common stock equivalent. Further emphasis was added to this point by applying the two-thirds rule to the study issues. Only 2 of the lhl issues would have been classified as common stock equivalents. The cash yield measure was based on offering price for all issues since Opinion Number 15 does not provide any guidelines for handl- ing debentures offered on a rights subscription basis. Frank and Weygrandt performed the same test for all categories of convertibles issued in 1965. The only debenture they found qualifying as a common stock equivalent had shown no converb sion at the study's end in 1968.8 Since it was felt that neither management nor investors would benefit from a study based on a measure which is misleading at best, the tradi- tional method for earnings per share calculation was adopted for this analysis. 8 Werner*G. Frank and Jerry J. Weygrandt, ”Convertible Debt and Earnings Per Share”, Accounting Review, XLV (April, 1970), 280-289. 79 Comparative measures. -- Returning to the research design, the last step under the earnings simulation was to compare the actual and simulated earnings per share figures. As a means to affect this comparison, the ratio of actual earnings per share to simulated earnings per share was calculated. This ratio, EPSA/EPSS, was generated for each company for each year the company appeared in the study. The comparative measure was also run on the total earnings per share figures spanning the entire study time for each company. Support for the hypothesis would be shown if EPSA was greater than EPSS. Various comparisons and statis- tical tests were run on this data as presented in Chapter IV. Time Values One way to examine the cash flow costs and benefits associated with the alternative financing plans was to ad- Just these flows for time by finding their present values. The differential in interest rates between the simulated and actual debentures created an incremental benefit, AE. How- ever, a future cost under wither of these plans was the dividend payment attached to the shares that resulted from actual conversion, SC, or from selling simulated equity, SD. Therefore, the incremental benefit,‘AEh could be compared to the incremental dividends arising under the two plans. It was expected that interest costs on the simulated straight debt would be greater than those on the actual con- vertible and a positive AE would result as shown in the pre- vious section. However, it was more difficult to make a priori- 80 statements about dividends since the number of shares varied with circumstances. The number of shares issued under the convertible were fixed by the conversion price, CP. The number of shares sold under the simulated debt depended upon the common's market price at the time of its hypothetical issuing. Then, the incremental shares, at a given time, could be found from the expression developed in the previous section: m T m SSt = SAt - a. i, SCnt + a :4 SDnt n: ts n: t: rearranging, T s _ s = i g, s - s =- AS At St n= t. Cnt n; is But It could be rather tentatively expected thatlcs would be positive since the common's market price generally would be greater than the conversion price if conversion was taking place. The procedure followed was to, first, multiply the divi- dend rate by4cS to find incremental dividends paid. It was assumed that the dividend rate would be the same under both plans as the rate that prevailed in the actual situation. Next, the incremental dividends were netted against AE and the net present value benefit was calculated using this figure. 0r: NPV 8 Present Value (AE - ADIV) Fowllowing Bowlin's argument, the discount rate used was the 81 after tax net yield on the subordinated convertible deben- ture.9 His analysis, for a bond refunding, concluded that the proper discount rate is the after tax net yield on the refunding bonds. Simply stated, this is because of the certainty of the interest savings due to refunding at a lower interest rate. Similarly, financing with a suborb dinated convertible debenture as opposed to a subordinated debenture creates an opportunity savings which gives rise to incremental earnings. Notice that the certain interest savings does not mean the discount rate equates to the pure interest rate since earnings could be spent by the firm be- fore the lender is paid. If more than one issue was out- standing in a given year, a weighted average rate was used based on the principal amounts outstanding in that year. Call premiums, and issue premiums or discounts were ignored here since they are relatively minor'factorsfor'convertibles. Underwriting expenses were also ignored. In summary, the discount rate employed was the after tax net yield for the subordinated convertible debenture in ques- tion since this was the rate that gave rise to the incremental earnings. The net present value for each firm for each year of the study and in total was calculated. The results are given in Chapter IV. Oswald D. Bowlin, "The Refunding Decision", Journal of Finance, XXI (March, 1966), 63-64. 82 Rates of Return Support for the study hypothesis would be shown if the rates of return on equity were greater for firms using convertibles as opposed to their using straight debt. In a manner similar to that followed under the earnings per share analysis, it was determined to compare rates of return according to the ratio: Actual Rate of Return Simulated Rate of Return This ratio could be calculated from: Actual Net Income / Simulated Net Income Actual Stockhamer s ui y Simulafiam However, two factors simplified the analysis considerably. First, since book values were used and the model forced the simulated equity to be the same as actual equity, the ratio became dependent upon the difference in earnings. Second, since it was expected that AE would be positive due to the interest rate relationships, it was expected that BRA/RRS would be greater than one in almost all cases. However, the motive for actually making the calculation was to observe the magnitude of the difference in the rates of return. The re- sults are presented in Chapter IV. Conversion Rates and 03115 Actual conversion rates and convertible subordinated de- benture calls for the study group were detailed in order to explain study results in the earnings and rate of return areas. This procedure also was intended to reveal the historical 83 pattern of conversions and calls over the study period. The method followed was to determine the principal amount outstanding at each year's end. Any changes from year to year were considered conversions and this amount expressed as a per cent of original issue amount. The yearly per cents were also cumulated to show the pattern of conversion from debt to equity. The results are presented in Chapter IV. ggpital Structure An examination of capital structures measured the impact of convertible usage on the debt-equity mix for the study firms. Capital structure was defined as the difference be- tween total assets and current liabilities. The ratio, long term debt to capital structure was then calculated for each firm at issue date. These ratios were classified by firm asset size to make than comparable to government figures. The Federal Trade Commission and the Securities and Exchange Com- mission regularly publish capital structure data for manufac- turing companies. Next, it was assumed that the convertible was entirely converted at issue date and the long term debt to capital structure ratio calculated again. The amount of equity issued under the "if converted“ status was determined by multiplying the debenture's conversion value (conversion ratio x HP) times the issue size at issue date. These figures were then compared to the original capital structure ratios and to the government figures. The years 1956 to 1965 were used so as to make maxi- mum use of the data available. Data collection for the new 84 issue population terminated in 1965 and the number of issues prior to 1956 were considered too few for com- parison. The data was also adjusted to include preferred stock and minority interest, first as long term debt and then as equity. In summary, this analysis revealed capital structures for firms using convertibles-—.both as convertibles and as- suming equity had been issued in place of the debentures to show the change in structure due to the convertible. Govern- ment figures were used as standard. These calculations also provided a test of the assumption that straight debt could have been issued under the simulation model. The results are given in Chapter IV. Conversion Value An aspect of convertibles incorporated in the previous section was the issue's conversion value at offering. Firms may have used the convertible as temporary financing with the expectation of early conversion to equity. If so, con, version value at issue would be established so as to encourage early conversion. This idea was tested by expressing the conversion value as a discount from the debenture's face value at issue date. Again, conversion value was determined by multiplying each issue's conversion ratio times its associated common stock market price. Conversion value was then related to debenture face value and the results tabu- lated for both public and rights offerings for the original 85 study group issues. It was anticipated that conversion value would be close to face value if quick conversion was desired by management. The results are given in Chapter IV. SUMMARY In this Chapter, the study's research design was dis- cussed. The selection of 1&1 issues representing 90 firms was outlined along with the pertinent issue characteristics. As seen, the basic technique applied in the study was simu- lation of firm performance. Simulation was adopted in order to compare actual operating results with performance expected under alternative straight debt financing. The discussion then centered on the method followed for analyzing earnings per share, rates of return and time values. Finally, several other aspects of convertibles were examined with the purpose of explaining convertible subordinated debenture influences on the firm. Chapter IV will present the results of this analysis. CHAPTER IV EMPIRICAL ANALYSIS Introduction Research results are presented in this Chapter. The results were generated by applying the method outlined in the previous Chapter. Discussed first are the various as- pects of earnings per share. Next, time adjustments for cash flows are presented. Third, the rate of return analysis is given in the same context as earnings per share. Finally, the study's investigation into convertible effects on capital structure is discussed. The emphasis here is on conversion rates and their relation to earnings per share. Actual Relative to Simulated Earnings Convertible bond theory, as discussed in Chapter I, re- vealed several factors pertinent to this phase of the analysis. It was expected that the number of common shares ultimately outstanding due to a convertible issue would be less than for a comparable equity financing. If straight debt, gradually supplanted by common stock, was the alternative to a converb tible it would be more difficult to make statements about share numbers. The determining factor would become the price at which future equity could be sold. The theory also indicated that total earnings should be higher under the convertible than under straight debt-equity due to the difference in face in- 86 8? terest paid. Study group earnings per share hinged on these factors. Support for the hypothesis required that actual earnings performance, based on convertible financing, be greater than simulated performance based on the straight debt alternative. Total earnings per share and earnings per share over time are examined here. Earnings Per Share Total for studygperiod. -- Over the study period there was a significant difference between actual and simulated earnings per share. See Table 4.1. TABLE 4.1 ACTUAL RELATIVE TO SIMULATED EARNINGS PER SHARE TOTALS FOR S'IUDY PERIOD Actual Earnings Probability Actual to per Share Relative of as Many Simulated to Simulated Observations (Per Cent) EPSA/EPSS Below Simulated Equal or Above 23 Negligible High 104.72 Below 62 Low 71.13 Total 90 Average95.51 Note: See Appendix C for'a brief description of statistical tests used in this paper. The probability here was based on the binomial with P=.5. The average was obtained by summing, over all the observations, the total actual earnings per share divided by the total simulated earnings per share and dividing by the num- ber of observations. 88 This data provides evidence for rejection of the hypothesis. The results do not confirm the hypothesis because of the test result direction. During the study period, 1950-1969, simulated earnings were generally greater than actual earnings. A frequency distribution of actual relative to simulated earnings per share displays the dif- ferences that prevailed under the alternative financing forms. See Appendix D.1. §3£_year during study_period. -- On a per year basis, average actual relative to average simulated earnings per share never exceeded 100 per cent. See Table 4.2. This table illustrates the average impact on earnings from conver- tibles relative to straight debt-equity as measured from the issue date. Throughout the study period, average simulated earnings per share were always greater than average actual earnings per share. This was true regardless of the number of years from the time a firm first issued a convertible. The relationship hovered around 99 per cent during the first four years but thereafter dropped to a point where actual earnings per share were approximately 5 per cent below simulated earn- ings per share. Subsequently, the ratio deteriorated even further to the study's end. See Appendix D.2 for a complete distribution of the ratio. Therefore, it was concluded that the use of convertible subordinated debentures may slightly decrease earnings per share below what might have been ex- perienced under straight debt-equity. 59 TABLE 4. 2 AVERAGE ACTUAL RELATIVE TO AVERAGE SIMULATED EARNINGS PER SHARE: EACH YEAR DURING STUDY Average Number of EPSA Number Years From fist— of Issue Date 3 Companies (Per Cent) 1 98.80 76 2 99.00 87 3 99.08 36 4 99.30 81 5 96.46 78 6 95.95 71 7 97.55 69 8 98.54 66 9 96.80 37 10 94.75 46 11 95.35 41 12 93.1? 38 13 93.79 33 1a 94.08 19 15 94.99 11 16 93.06 4 17 86.89 3 18 69.95 3 90 Run Tests One approach to time series analysis is the use of distribution free, nonpparametric statistics. The parti- cular tool employed here is the runs test. A run is a succession of items of the same class. By examining the total number of runs in a series or the length of the long- est run it is possible to determine the randomness of a series. The study series were suited to runs test analysis since the earnings relatives divided naturally into classi— fications above and below 100 per cent. If the ratios were independent, it would be expected that the data would be normally distributed around the 100 per cent level. Firm runs. -- Three versions of runs tests were applied to the individual firm actual relative to simulated earnings per share series. These tests were meant to examine the supposition that there was a difference in earnings perfora mance under the two financing forms. The proposition was that there was no clustering. Rejection of this proposition and acceptance of the alternative would mean that there were too few runs for the observations to be independent. The first test was involved with the number of runs above and below the 100 per cent level for the ratio. Table 4.3 shows the number of firms whose earnings series were found to show no difference and those whose actual versus simulated earnings per share were significantly different. A lower tail test was appropriate since too few runs would indicate rejection. 91 TABLE 4.3 NUMBER OF RUNS ABOVE AND BELOW 100 PER CENT Firms Whose Earnings Series Were Level ~ of IndependentINot Independent Significance 1m { in .05 Note: Firms in study less than 5 years were not included. For the firms whose ratios were not independent it can be said that the difference between actual and simulated earnings per share tended to be high for several years and then low for several years. For these firms, to estimate the next years earnings relationship, the latest year's ratio is likely to be closer than the series average. Another test of independence was an examination run lengths on either side of the 100 per cent level. It was possible that while the number of runs would indicate independence, the probability of encountering a run of given length or greater would show nonrandom influences. Table 4.4 shows the number of firms whose earnings series were found to be independent under this test. Thus, 63 firms were found to exhibit runs too long to be considered random. A third test concerned with independence involved the length of runs of ascending or descending order. This test examined consecutively increasing or decreasing runs regard- less of their position relative to the 100 per cent level. 92 TABLE 4.4 LENGTH OF RUNS ABOVE AND BELOW 100 PER CENT Firms Whose Earnings Series Were Level of Independent Not Independent Significance 22 63 .05 It was possible that directions of movement, once begun, tended to continue. This might be the case for a firm when conversion starts. Table 4.5 shows the results of the test. TABLE 4.5 LENGTHS OF RUNS OF ASCENDING AND DESCENDING ORDER Firms Whose Earnings Series Were Level of Independent Not Independent Significance 72 13 .05 Industrial runs. -- An attempt was made to reduce the runs tests results from a macro level to the point where they would be more useful for decision-making. The first approach was to classify runs with conversion rates. However, no sys- tematic pattern was detected. Another approach was to segre- gate the data by industry. Study group firms were placed in industrial classifications by Standard Industrial Classification 93 (SIC) numbers.1 It was found that most study group firms fell into a relatively small number of industries. See Appendix E for a complete industrial classification. Results of the test for the number of runs above and below 100 per cent were then classified by industry. These results are shown in Table 4.6. Most industries contained approximately the same number of firms whose earnings ratios were independent versus those whose earnings were not independent. However, two industries, petroleum refining and transportation equipment, were pre- dominantly independent and not independent, respectively. It should be noted that all firms in both these industries had average actual earnings per share relative to simulated earn- ings per share below 100 per cent. The results indicate the experience that firms issuing convertibles in these industries might expect. Runs based on averages. -- It was noted earlier that the average earnings ratio never exceeded 100 per cent on a per year basis. See Table 4.2. An apparently steady decrease in the ratio was also observed. This pattern was analyzed by using runs tests. A test for ascending and descending order showed that the direction of movement tended to persist going from relatively higher to lower levels as the number of years from issue date increased. In other words, the probability of so few runs up and down, if there were no real persistance of movement in the same direction, would be .031 based on the l Dun and Bradstreet, Million Dollar Directory, 1971, (New York: Dun and Bradstreet. 105“). 94 TABLE 4.6 [HUBER OF RUNS ABOVE AND am 100 PER CENT BY INDIBTRY Egg; Whose Ea: shim Series More Industry Indeandeni Air Transport 4 Chemical 5 Conglomerate 5 Distribution 6 Electrical Equipment 3 Food Processing 2 hehinery 6 Office hchinery 4 Paper 1 Petroleum Refining 2 Steel 1 Textile 3 Transportation Equipment 8 t Not Inde ndent 2 a 5 5 3 2 3 3 2 7 2 a l 95 Table 4.2 data. It could be condluded that, based on averages, a firm issuing convertibles would experience steadily decreasing earnings relative to what could be expected from employing straight debt-equity. In summary, the runs test analysis shows that a large number of firms experienced significant differences in the relationship of actual to simulated earnings per share due to non-random influences. This was true not only for in- dividual firms but to some extent on an industry basis as well. Placed in the context of the earnings levels the analysis shows that convertible issuers were at a signifi- cant earnings disadvantage relative to the use of straight debt-equity. This can be seen by comparing the number of firms that experienced ratios above 100 per cent to those below that level. Further, the disadvantage becomes worse as the number of years increases from date of issue. This point will be considered again under the conversion rate topic. Present Value Cash Benefits Cash benefits were defined as the net of the incremental earnings generated and the incremental dividends paid under the alternative financing forms. Yearly net cash benefits were discounted by the convertible's after tax interest cost to arrive at a net present value cash benefit for each firm. Table 4.? summarizes the results. A complete frequency dis- tribution of the net present values is given in Appendix D.3. 96 TABLE 4.? SUMMARY OF NET PRESENT VALUE BENEFITS DURING STUDY Present Value Probability of as Benefits Many Positive (Number of Firms) Observations Positive 60 Negative 29 Negligible Total 90 Note: Probabilities based on binomial with P=.5. A distribution of the weighted average net present value cash benefits by year of issue is presented in Table 4.8. This table shows the average net present value cash benefits experienced by study group firms ranked according to when they first issued a convertible. For example, firms issuing convertibles for the first time during the study in 1963 exp perienced a positive net present value benefit of $750,000 to the study's end. The actual averages should be viewed cautiously for the years represented by a small number of issues since extreme values may have affected the average. As can be seen from Table 4.8, firms issuing convertibles in the study's early years experienced a loss in terms of net present value. The results from the study's latter years were predominantly positive. These figures are supported by the number of firms that experienced positive versus negative net 97 TABLE 4.8 AVERAGE NET PRESENT VALUE CASH BENEFITS FROM ISSUE DATE Average Net Present Number of Firms Whose Value For Given Net Present Value Was Issue Year Issues all -1 Year (Millions of Dollars) Positive Negative 1951 Zero 1 0 1952 - 2.500 1 2 1953 -12.500 0 l 1954 -12.500 0 1 1955 - 1.238 3 5 1956 - .511 6 3 1958 - 1.375 3 3 1959 1.433 6 3 1960 1.583 2 1 1961 1.654 11 1 1962 1.610 5 0 1963 .750 2 1 1964 .516 3 0 1965 1.691 10 l Note: Time zero for firms issuing more than one convertible was taken as the first issue's flotation date. 98 present value. A cyclical pattern also appeared and was confirmed by a runs test for runs of ascending and descending order. Thus, the net present values have tended to move con- secutively in the same direction more often than would be the case for independent observations. However, a second runs test for overall trend did not show significance and so there has been no general shifting from a lower to higher level of net present value cash benefits -- just the cyclical pattern. The shift from negative to positive net present values over the study period can be partially explained by the inp crease in interest rates (and interest rate differentials) over the study period and by the trend to larger size issues. Larger issue sizes and greater differences between convera tible and straight debt interest rates means larger incremen- tal earnings. See Appendix D.4. However, it is felt that another important factor is the conversion rate. This can also be seen in Table 4.8, since issues from the early study years have had a greater chance at conversion than those of the latter years. Once conversion begins, more dividends tend to be paid under the convertible as opposed to the straight debt alternative. Thus an increase in dividends plus a de- crease in incremental earnings would tend to create negative net present value as time passes. One way for firms to avoid this problem might be to replace a convertible with another convertible in order to maintain the incremental earnings advantage. Further information on this point is presented under the conversion rate topic. 99 Actual Relative to Simulated Rates of Return The rate of return is a measure of the owner's equity productivity. As defined by the model, actual rate of re- turn relative to simulated rate of return was expected to exceed 100 per cent. Of interest here was the magnitude of the effect convertible issues would have on rates of return. A frequency distribution showing the observations calculated for all firms over all the study years is given in Table 4.9. It can be seen that 840 out of 898 data points fell in the range 100-105 per cent. Within this range the large bulk of the observations were below 102 per cent. There were isolated instances of ratios below 100 per cent due to Moody's straight interest estimates. The re- turn on the scale's higher end were also isolated. No one firm or any time period enjoyed consistently high actual re- lative to simulated rates of return. From this it was con- cluded that further analysis related to this ratio would be non-productive. In summary, the use of convertible subordinated deben- tures raised rates of return to only a slight degree com- pared to the straight debt—equity alternative. Combining this result with the findings related to earnings per share provided only minimal support for the study hypothesis. Therefore, the study hypothesis that financing with conver- tible subordinated debentures, as apposed to straight debt- equity, would positively improve financial performance was 100 TABLE 4.9 FREQUENCY DISTRIBUTION OF ACTUAL RELATIVE TO SIMULATED RATES OF RETURN Actual Relative to Number Simulated Rates of of Return Firms (Per Cent) Less than 100 30 100 - 105 840 106 - 110 9 111 - 115 10 116 - 120 6 121 - 125 2 Total: 898 rejected. Further consideration is given to these results in Chapter VI. Capital Structure Convertible subordinated debentures have a direct effect on capital structure. Capital structure as defined in Chapter III is total assets minus current liabilities and thus encom- passes all long term financing sources. Convertible debt's place in the structure has generated a certain amount of con, troversy. This is reflected in the measurement of earnings per share as discussed earlier. The question is when to 101 change the convertible's classification from debt to equity. Figure 1.1 shows that as time passes, assuming rising stock prices, the convertible's value and nature are determined, first by its debt characteristics and then by equity influences. J~t has been suggested that conver- tibles provide a means for automatic debt retirement with- out a corresponding dollar outlay. These concepts are put into perspective in this section by examining the study group firm's capital structures over the study period. Debt Capacity One capital structure measure is the ratio long term debt to total capital structure. Table 4.10 presents the re- sults from applying this ratio to the study group firms at issuance. The asset size categories were dictated by the Federal Trade Commission figures. These figures reflect the impact on capital structure due to the issuance of converti- bles by the study firms. The actual long term debt to capi- tal structure is compared to the same ratio assuming full con- version at issue and to the ratio for all manufacturing com- panies. Results show that long term debt as a percentage of capital structure was much higher for the study group firms compared to the macro data for manufacturing firms. The difference was greatest for the smallest and largest size categories. This evidence weakly supports the theory that when smaller firms engage in convertible financing they do so in high volume since 102 on oouuo>coo messy -m new mounmau mqugoo ham macaw: mwumuuomo .coammweeoo mmcozoxm one mmguapaomm a sommmweeou mocha dsumommo .mmwmucoouma .mummx unswuu> .mmwsomeouimswuauoouscm: m ~.n~ s.wn o.es + ooo.~ mm o.- m.qm w.ee o.ooo . 0mm mm o.- o.- ~.mm o.oq~ a 00— mH o.w~ ~.n~ q.we o.oo u 0n moo“ u owed m o.ma n.0m o.qm + ooo.~ mm o.n~ m.mn o.oq o.ooo u OmN ea e.- a.a~ a.ee o.oe~ - cos s 0.0m m.~a ~.mn o.oo - on moo_ o ~0o~ m o.e~ m.m~ e.mw + ooo.~ n~ m.H~ ~.nn o.~e o.ooo u Own «N eflaw ~.n~ non. o.oq~ .. on: m ~.sa o._n e.om o.oo - cm ooo~ n ono~ ooaoaum Aucmu ammo Aucmu ummv aucou uomv «maowaduxv msuwm ammusmfim umuumso mammH um commuo>cou use» momma we saw aumm we saw .emucmaeou -=m monasmm< ouauonuum um muauosuum fimuwamu somm< umseaz mamasuomwacmz -< “muwamu cu upon Esme moon cu anon anew mcoq o~.« wqmdh wh¢o mummu H< mmthDMHm 4m so Ewes—H ~.m museum 127 indicates that investors were buying both convertible categories for yield rather than appreciation. As the in- terest rate ratio approaches unity, the convertible tends to sell on a straight yield basis since the interest rate spread approaches zero. This was probably true due to the state of the stock market and the U. S. economy's uncertain future at that point. Investor‘s were buying security and speculating on a distant conversion possibility. These regression curves represent one point in time and their predictive ability would be limited to markets re- flecting the same conditions. In addition, they ignore other differences in firms, such as credit strengths, stepped up conversion prices and the broad definition of growth versus non-growth categories. other’growth curves. -- Another approach to the problem of differing growth rates was to define growth more narrowly and derive curves from the study group data. The method fol lowed here was to use ex post data from called convertibles. In this way, it was possible to calculate growth using Brig- ham's model as presented earlier in this chapter. Since only called debentures were included, terminal values and the con- vertibles' lives were known with certainty. The procedure was to scan the study group issue list and determine the number of years each called bond was out- standing as well as terminal value. Terminal value was taken as conversion value if conversion was being forced, 128 call price if the common's market price was below the conversion price. This was done for both public and rights offerings. Prices paid for the convertibles depended upon the type of offering. Growth was then calculated using the model. The resulting growth rates range from -4.7 per cent to +67.6 per cent. Converti- bles floated from 1952 to 1965 with lives from 1 to 12 years were represented. There were 27 public offer- ings and 24 rights offering. In order to accumulate a sufficient number of points for meaningful curves, the growth rates were characterized as low, medium, or high and the data clasified accordingly. Low growth was defined as ranging from 0 per cent to 6.5 per cent, medium from 6.6 per cent to 15.5 per cent and high growth as above 15.5 per cent. Scatter diagrams were then plotted for publicly issued convertibles that were called, for rights issues that were called and for combina- tions of public and rights issues-—.all in the proper growth categories. Inspection revealed that only the low and me- dium growth categories, for public issues, exhibited the expected shapes and positions. The correlation coefficients for these categories, as shown in Table 5.1, are disappointingly low. Both curves have the proper shape but the medium growth curve inter- sects the low growth curve from above and stays below it for price ratio values greater than 1.10. 129 Since the scatter diagrams were relatively nonp descriptive, the data for the same two categories were run again as straight lines. The results are labeled Low Growth-Straight and Medium Growth-Straight in Table 5.1. The improvement in correlation coefficients is minor under this configuration and the standard errors are larger. These curves also have the expected negative slope but they too intersect. Further discussion of these curves will be deferred until Chapter VI. New issues. -- As a means of solving the time series problem yearly new issues were analyzed. This is similar to the Moody's and Value Line curves except that they re- flect market determined parameters. Here, the interest rate and stock price ratios were determined at issue. Scatter diagrams showed 195? issues to come closest to the expected plot. However, the correlation coefficient was only .067 and the curve sloped positively upward. Asset sizes. -- Another approach was to categorize the issuing firms according to asset size. The size categories were the same as those employed for the capital structure analysis reported in Chapter IV. or the four size cate- gories, the best results were obtained for firms whose sizes ranged from $250 million to $1,000 million. This was the second largest size category. However, the correlation 130 coefficient is a low .319 for issues from the study group. Thus, the analysis based on parameters taken at the issue date for growth, new issue and asset size tradeoff curves reveals only minor evidence to support the hypothesis. Further discussion will be found in Chapter VI. 8 Industrial curves. -- Industrial categories were examined as a final attempt to develop curves management might use for guidance in establishing convertible interest rates and conversion prices. The industries analyzed, petroleum and air transport, were selected since they are rather narrowly defined, provided a sufficient number of issues during the study period, and because of the particular flotation dates. In the petroleum sample 11 of the 14 issues were offered prior to 1959 and only 1 was offered subsequent to 1963. For the air transports, all of which were regularly scheduled airlines, 13 of the 19 issues were offered after 1964 and 14 of the 19 were floated by 4 firms. Further, these 4 firms were all in the same asset size category throughout the study. The issue date bunching minimized the time span pro- blem. Regressions were run on the data from both industrial samples as of the issue date and lagged one year after issue. Again, issue date convertible interest rates and conversion prices were those set by management, and the straight interest rates were Moody's estimates. The lagged interest rates 131 were market determined yields for the convertibles and Moody's estimates for comparable straight debt at that date. Correlation data is presented in Table 5.1. The correlation coefficients for the industrial issue date curves are relatively good compared to the other issue date category curves. However, both petroleum and air transport curves have the wrong shape being positively shoped. On the other hand, curves based on the lagged data exhibit the best correlations encountered being .84 for petroleum and .83 for air transport. In addition, they both have the proper shape and position. Standard errors are the greatest encountered though not excessive for the range of values. These curves are shown in Figure 5.2 Based on the issue date industrial data, if the hypo- thesized tradeoff curves do exist, either this analysis has failed to uncover them or management has chosen to ignore their dictates in setting parameters. However, once trading ensues market forces tend to produce data that reflects the expected relationship. The usefulness of these curves de- pends on their stability over time. Summary The hypothesis that there are tradeoffs between converti- ble interest rates and conversion prices was examined in this chapter. Tradeoff curves were developed for a variety of 132 a: Cu o.~ m.“ o; . so a \T uwodmsoua um< x oououodouuxm G Esofiouumm . m.~ o.~ m.~ m4naive to detect the relationship. Another approach would be to query converti- ble issuers as to the possible interest-conversion price com- binations they considered in designing the convertible issue. This would reveal whether the postulated tradeoff exists. However, two factors discouraged this approach. First, it must be assumed the various managements did consider more than one choice of interest-conversion price. Second, management must be willing to reveal this data. Brigham attempted some- thing of this nature but usable reSponses were received from only 5 firms out of 42 sampled.1 This paper's approach could be improved and a more sophis- ticated analysis utilized. Multiple curvilinear regression and correlation could be attempted including such variables as 1 Brigham, op. cit., p. 45. 141 expected growth, firm size, issue year and others. This appears to be the most promising avenue for future research. Growth curves were developed in this paper with mediocre re- sults. Different growth measures should be incorporated since the present definition has inherent limitations. The results presented here are encouraging and should lead to future research. Study Implications As a result of this research several items were deter- mined that should be important to management. Firms prob- ably should not issue convertible subordinated debentures and eXpect superior financial performance vis a vis the al- ternative of temporary straight debt-equity. Study results showed that the hypothesized advantage did not really develop for either earnings per share or rate of return and net pre- sent value cash benefits depend on conversion rates among other factors. However, if a convertible is issued the best management policy is to call the debenture as soon as convert sion becomes attractive. Earnings per share would be opti- mized under this policy. Research data analysis found that convertibles are often a thinly veiled disguise for selling common equity. One theory supporting this practice is that the number of common shares ultimately issued under the convertible will be fewer 142 than if common were sold directly. However, if the choice is between financing with a convertible now versus straight debt now that will be replaced by common to be sold later -- it is better to take the latter course. Under the straight debt- equity alternative fewer shares will be needed since the com- mon can be sold at a higher price than under the convertible. Earnings per share would be directly affected. Two convertible parameters that can be set by management are interest rates and conversion prices. Theory and this paper's empirical evidence show that tradeoffs exist between these two parameters as a means of balancing risk versus re- turn. The crude curves derived here offer encouragement that more sophisticated curves can be developed. These curves could be used by management in designing convertible issues. Summagy In this chapter the entire study was reviewed and its implications discussed. First, the highlights of aggregate convertible financing over the last 20 years were presented. Next, the research results of the study group's financial performance were discussed in the context of convertible debt management. The research into tradeoff curve develop- ment was reviewed and suggestions proposed for future re- search in this area. Finally, the study's results were put into perspective by presenting the research's policy implica- tions. B IBLIOGRAPHY BIBLIOGRAPHY Books Atkinson, Thomas B. Tiends In Corporate Bond Quality. National Bureau of Economic Research. New York: Columbia University Press, . 1967. Ashly. Techniques For Investing In Convertible Bonds. New York: Salomon Brothers and Hutzler, 1966} Bladen, Dewing, Arthur Stone. A Study of Corporation Securities -- Their Nature and Uses in Finance. New York: Ronald Press, 1934. . Financial Policy of Corporations. Vol. 1: Corporate Sec- urities. New York: Ronald Press, 1920. ‘ . Financial Polipy of Corporations. 5th ed. New York: Ronald Press, 1953. Dun and Bradstreet. Million Dollar Directory, 1271. New York: Dun and Bradstreet, 1970. Ezekiel, Mordecai, and Fox, Karl A. Methods of Correlation and Re- gression Anaiysis -- Linear and Curvilinear. 3rd ed. New York: John Wiley and sons, Inc., 1959. Hawkins, David F. Corporate Financial Policy. Homewood, Illinois: Richard D. Irwin, 1971:, Hendrickson, Eldon 8. Accounting Theopy. Rev. Ed. Homewood, Illinois Richard D. Irwin, 1970. Hickman, Walter Braddock. Statistical Measures of Corporate Bond Fi- nancing. National Bureau of Economic Research. Princeton, New Jersey: Princeton University Press, 1960. Keynes, John Neville. The Scope and Method of Political Economy. London: Macmillan and Co. Ltd., 1890 reprinted, 1930. Moodyjs Bank and Finance Manual. New York: Moody's Investors Service, Inc., 1949-1970. 143 144' Moody's Industrial Manual. New York: Moody's Investors Service, Inc., Moody's Transportation Manual. New York: Moody's Investors Service, Inc. 1949-1970. Moody's Utility Manual. New York: Moody's Investors Service, Inc. Pilcher, C. James. Raising Capital With Convertible Securities. Ann Arbor, Michigan: Bureau of Business Research, university of Michigan, 1955. Richmond,Samue1 B. Statistical Analysis. 2nd ed. New York: Ronald Press, 1964. Siegel, Sidney. Nonparametric Statistics for the Behavioral Sciences. New York: MoGraw-Hill, 1956. Taylor, W. Bayard. Financial Policies of Business Enterprise. New York: D. Appleton Century, 1942. Wallis, W. Allen, and Roberts, Harry V. Statistics -- A New Approach. New York: Free Press, Collier Macmillan, 1956} Weston, J. Fred. The Scope and Methodology of Finance. Englewood Cliffs, New Jersey: Prentice Hall, 1966; Articles Baumol, William J.‘ Malkiel, Burton, 3 and Quandt, Richard E. "The Valuation of Convertible Securities." gnarterly Journal of Economics, LXXX (February, 1966), 48-59. Bowlin, Oswald D. "The Refunding Decision: Another Special Case in Capdggl Budgeting." Journal of Finance, XXI (March, 1966). 55- . Brigham, Eugene F. "An Analysis of Convertible Debentures: Theory and Some Empirical Evidence." Journal of Finance, XXI (March, 1966) 9 35-51}. Broman, Keith L. "The USe of Convertible Subordinated Debentures by Industrial Firms 1949-1959." gnarterly Review of Economics and Business, III (Spring, 1963), 5-7 . Frank, Werner G., and Weygandt, Jerry J. "Convertible Debt and Earnings Per Share." Accounting Review, XLV (April, 1970), 280-289. 145 Lerner, Eugene M., and Auster, Rolf. "Does the Market Discount Potential Dilution?" Financial Analysts Journal, xxv (July-August, 1969). 118-121. McKenzie, Robert R. "Convertible Securities, 1956-1965." §parterly 1" 90 9 Review of Economics and Business, VI (Winter, 19 Moag, Joseph S.: Carleton, Willard T.; and Lerner, Eugene M. "Defining The Finance Function: A Model Systems Approach." Journal of Finance, XXII (December, 1967), 543-556. Poensgen, Otto H. "The Valuation of Convertible Bonds." Industrial ggfagement Review, VI (Fall, 1965), 77-92; and VII (Spring, 1966). 498, Rollins, Montgomery. "Convertible Bonds and Stocks." Annals of American Academy of Political and Social Science, XXXV (November 3, 1910, 97. Williams, B. 3., and Letwot, Marvin. "Underwritten Calls of Industrial Convertible Securities, 1950-1961." gparterly Review of Be- onomics and Business, (No. 3, 1910) . 71-77. Miscellaneous American Institute of Certified Public Accountants. Opinion Number 9: Reporting the Results of Operations. New York: American Institute of Certified Public Accountants, December 1966. . Opinion Number 15: Earnings Per Share. New York: American Institute of Certified Public Accountants, May, 1969. Annual Financial Statements. Various Firms. 1950-1969. Investment Dealer's Digest, Corporate FinancipgDirectony, (Semiannual, 1950-1971). Moody's Bond Survey. (Weekly, 1949-1971). Moody's Convertible Bonds. (Monthly, 1955-1971). New York Stock Exchange Listing_Statements. Various Issues. 1950-1959. "The 100 Hundred Stocks With the Greatest 3-5 Year Appreciation Potential." Value Line Investment Survey, (January, 1971). 1H6» U. 3., Board of Governors. Federal Reserve System. Federal Reserve Bulletin. Washington, D. 0.: Division of Administrative Services, Board of Governors, January, 1955 - March, 1970, various issues. U. S. Federal Trade Commission - Securities and Exchange Commission. Quarterly Financial Report for Manufacturing Companies. Washington, D. C. : Government Printing Office, 1951-1970, various issues. U. S. Securities and Exchange Commission. Statistical Bulletin. Washington D.C.: Government Printing Office, XXIX (March, 1970). Wall Street Journal. New York Stock Exchange Bonds, March 13, 1970. General References Books Archer, Stephen B., and D'Ambrosio, Charles A. Business Finance: Theory and Management. New York: Macmillan, 1956. Bellemore, Douglas H., and Ritchie, John 0., Jr. Investments -- Principles, Practices Anal sis. 3rd ed. Cincinnati: Southwestern Publish- ing Company, 1909. Bierman, Harold, Jr. Financial Policy Decisions. New York: Macmillan, 1970. Bowker, Albert H., and Lieberman, Gerald J. Engineering Statistics. Englewood Cliffs, New Jersey: Prentice Hall, 1959. Cohan, Avery B. Yields On Corporate Debt Directly Placed. National Bureau of Economic Research. New York: Columbia University Press ' 1967. Cohen, Jerome B., and Zinbarg, Edward D. Investment Analysis and Port- folio Management. Hemewood, Illinois: Richard D. Irwin, 1967. Cootner, Paul H., ed. The Random Character of Stock Market Prices. Cambridge, Massachusetts: The MIT Press, l96h. Dewing, Arthur Stone. Corporation Finance -- A Textbook For Colleges and Schools of Business Administration. New York: Ronald Press, 1922. Duncan, Acheson J. Quality Control and Industrial Statistics. Home- wood, Illinois: Richard D. Irwin, 1955. 147 Gerstenberg, Charles W. Materials of Corporation Finance. 5th ed. New York: Prentice Hall, 192R. Graham, Benjamin, D.: Dodd, David L.: Cottle, Sidney; and Tatham, Charles. Security Analysis -- Principles and Analysis. New York: McGraw Hill, 1962. Grunewald, Adolph B., and Nemmers, Erwin Esser. Basic Managerial Finance. New York: Holt, Rinehart and Winston, 1970. Hickman, Walter Braddock. Corporate Bond QualityLand Investor Ex- perience. National Bureau of Economic Research. Princeton, New Jersey: Princeton University Press, 1951. Johnson, Robert W. Financial Management. 3rd ed. Boston: Allyn and Bacon, 1966. Home , H. L. Corporate Finance. New York: Henry Holt and Co, 191+9. Lindsay, J. Robert, and Sametz, Arnold W. Financial Management -- An Analytical Approach. Homewood, Illinois: Richard D. Irwin, 1967. Meigs, Walter B.; Johnston, Charles E.; Keller, Thomas F.; and Mbsich, A.N. Intermediate Accounting. 2nd ed. IVew York: McGraw Hill, 1968. Mumey, Glen A. Theory of Financial Structure. New York: Holt, Rine- hart and Winston, 1969. Skelly ,William S. Convertible Bonds - A tudy of Their Suitability for Commercial Bank Portfolios. New York: Salomon Brothers and Hutzler, 1959. Weston, J. Fred, and Brigham, Eugene F. Managerial Finance. 3rd ed. New York: Ho1t Rinehart and Winston, 1969. Articles Bacon, Peter W., and Winn, Edward L., Jr. "The Impact of Forced Con- version On Stock Prices." Journal of Finance, XXIV (December, 1969) 1 871-87'4’0 Berle, A. A., Jr. "Convertible Bonds and Stock Purchase Warrants." Yale Law Journal, XXXVI (March, 1927), 6h6-666. Brigham, Eugene F., and Pappas, James L. "Rates of Return On Common Stock." Journal of Business, XLII (July, 1969), 302-316. 148 Briloff, Abraham J. "The Funney-Money Game." Financial Analysts JOurnal, XXV (May-June. 1969). 73-79. Conrad, Gordon R., and Plotkin, Irving H. "Risk/Return: U. S. Industry Pattern." Harvard Business Review, XLVI (March-April, 1968), 90-99 plus supplement. Gathman, Harry G. "Measuring the Dilution Effects of Convertible Securities." Journal of Business, IX (January, 1938), nu-5o. Hayes, Samuel L. III. "New Interest in Incentive Financing." Harvard Business Review, XLIV (July-August 1966), 99-112. Imdieke, Leroy F., and Weygandt, Jerry J. "Classification of Convertible Debt." AccountingyReview, XLIV (October, 1969), 798-805. Johnson, Keith. "Stock Splits and Price Change." Journal of Finance, XXI (December, 1966), 675-686. Johnson, R. W. "Subordinated Debentures: Debt That Serves As Equity." Journal of Finance, X (March, 1955), 1-16. Mautz, R. K., and Skousen, K. Fred. "Some Problems in Empirical Research in Accounting." Accounting Review, XLIV (July, 1969), ##7-456. Miller, Alexander B., "HOw To Call Your Convertibles." Harvard Business Review, XLIX (May-June, 1971), 66-70. Pogue, Thomas F., and Soldofsky, Robert M. "What's In a Bond Rating." Journal of Financial and Quantitative Analysis, IV (June, 1969), 201-228. Soldofsky, Robert M. "Performance of Convertibles." Financial Analysts Journal, XXVII (March-April, 1971), 61-66. Vinson, Charles E. "Rates of Return on Convertibles." Financial Analysts Journal, XXVI (July-August, 1970), 110-115. Weil, Roman L., Jr.: Segall, Joel E.; and Green, David, Jr. "Pre- miums On Convertible Bonds." Journal of Finance, XXIII (June, 1968), 445-h6h. Williams, Howard A., and Williams, Charles M. "Incentive Financing -- A New Opportunity." Harvard Business Review, XXXVIII (March- April, 1960), 123-13#. Miscellaneous American Accounting Association Committee on Accounting Concepts and Standards. Accounting and Reporting Standards for Corporate Financial Statements and Preceding Statements and Supplements. Columbus, Ohio: American Accounting Association, 1957. n+9 American Institute of Certified Public Accountants. Opinion Number 10: Omnibus 0 inion - 1966. New York: Amera ican’InStitufe of CerEITIed Public Accountants, Dec- ember, 1966. . Opinion Number 12: Omnibus 0 inion - 1967. New York: American Institute of Certified PuSlic Accountants, December, 1967. . O inion Number 14: Accounting for Convertible Debt and DeBt Issued with Stock Purchase warrants. *Néw YSrk: ‘American Institute of CerEITIed PuEIIc Account- ants, March 1969. Investors Bond values Table. Boston: Financial Publishing 0’ 110., 19620 Pinches, George E. ”Financing Corporate Mergers and Acquisi- tions with Convertible Preferred Stock." Unpublished Ph.D. dissertation, Michigan State University, 1968. APP EN DI CES APPENDIX A STUDY GROUP ISSUES This appendix contains a listing of the study group issues. The list is arranged chronologically according to the year a firm first qualified for inclusion in the study. Companies are listed alphabetically within years along with face rates, maturity dates, and principal at flotation in millions of dollars for the attendant issues. 1951 Warren Petroleum 333, '66; 15.0 1952 Dow Chemical 33, '82; 100.0 International Minerals & Chemical 3.65s, '77; 20.0 and 43, '91; 50.0 Union Oil 3 1/83, '72; 35.0 and 33, '75; 60.0 and his, '91; 60.0 1953 Sinclair 313, '83; 101.8 and a 3/83, '86: 167.2 195“ Columbia Gas System 3&3, '64; 50.0 1955 Burlington his, '75; 30.0 and 53, '91; 40.0 Continental Baking 3 5/83, '80; 13.0 and 4 3/83, '83; 13.2 Freuhauf 3 3/hs, '75: 15.0 and 43, '76; 37.5 and 5&3, '9#; 60.0 150 151 Grace 333, '75; 30.0 and his, '90; 83.“ Lockheed 3 3/hs,’80: 30.0 and his, '92; 125.0 Merritt Chapman Scott hés, '75; 25.0 National Tea 333, '80: 15.0 Radio gorporation of America 3&3, '80; 100.0 and hfis, '92; 1 0.0 1956 Allegheny Ludlam Industries hs, '81: 16.h American Machine and Foundry hzs, '81, 10.8 and 5s, '77; 12.7 and hzs, '81; 39.9 Burroughs his, '81: 30.2 and 3 3/h3, '93; 75.0 and h 5/83, '9h; 100.0 Capital Airlines his, '76; 12.0 Combustion Engineering 3 3/8 3, '81; 15.0 General American Transportation hs, '81; 23.8 Lowenstein h 3/8, '81: h0.0 National Cash Register h/és, '81: 28.2 and his, '92; 88.6 Textron 5s, '71; 20.0 1957 Atlantic Refining hés, '87; 100.0 Carrier a 1/8, '82; 18.0 Chance Vbught 5:3, '77; 12.5 Commonwealth Oil 63, '72: 20.0 and his, '92; 20.0 Douglas Aircraft he, '77; 27.9 and h 3/h, '91; 75.0 Dresser Industries h 1/83, '77; 20.0 Fairbanks Morse 5 3/h, '72; 15.0 General Portland Cement 5s, '77; 15.0 Macy 53, '77; 12.3 and his, '90: 21.8 and 5s, '92; 23.0 152 McDermott 5s, '72; 20.2 National Cylinder Gas 5 1/83, '77; 17.5 Olin 5%, '82: 60.0 and 553, '83: ho.o Phillips Petroleum his, '87; 171.7 Shamrock Oil 523, '82; 17.5 Thompson Products h 7/83, '82; 19.7 1958 Boeing his, '80: 30.6 and 5&3, '91: 129.9 Case 553, '83; 20.1 Continental Airlines 5 3/hs, '73; 12.5 and 33$, '92: 35.0 ITT h 7/83, '83; 28.7 Richfield 011 h 3/83, '83; 50.0 Sylvania and General Telephone & Electronics his, '83; 18.0 and hs, '90; 100.0 and 5s, '92; 150.0 1959 Avco 5s, '79; 15.0 Cerro 538. '79: 26.3 Champion Paper his, '8h: 20.0 Food Fair hs, '79: 21.2 General American Oil of Texas h 3/hs, '8h; 20.0 Hooker Chemical 53, '8h; 2h.h Pan American World Airways h 7/83, '79; h7.0 and his, '8h; 60.h and his, '86: 175.0 and 5:3, '89: 175.0 Philco his, '8h; 22.0 Spiegel 53, '8h; 15.0 1960 Collins Radio h 3/hs, '80; 12.0 and h 3/h, '83: 12.5 and 4 7/88’ .87; 1‘000 153 Kayser Roth 5&3, '80; 16.1 United Airlines h 7/8, '85; 25.0 and hs, '90; 66.5 and 5s, '91; ’103.h and his, '92; 130.8 1961 Allied Stores his, '81; 27.0 and his, '92; 50.0 Armour his, '83; 32.6 Automatic Canteen h 3/hs, '81; 22.6 Brunswick his, '81; 25.6 City Products 53, '82; 15.0 Crowell Collier his, '81; 12.0 and hs, '92; 25.0 FMC 3 1/83, '81; 30.0 and his, '92; 100.0 Hunt h 3/83, '86; 38.7 Keystone Steel his, '81; 20.0 Ling Temco VOught 533, '76; 55.2 and h 3/hs, '76; 23.3 United States Freight 5s, '81: 15.h Xerox his, '81; 15.0 and hs, '8h; 51.2 1962 Air Reduction 3 7/83, '87; h5.0 Control Data h 3/hs, '77; 35.0 and 3 3/hs, '89; 15.0 Litton 333, '87; 50.? Rapid American 5 3/hs, '77; l3.h Stokley Van Camp his, '82; 15.0 1963 Ashland Oil 3 7/83, '93; 35.0 Trans World Airlines 5 3/hs, '83; 38.5 and hs, '92; 100.0 and 53, '9h: 150.0 154 United Aircraft hts, '88; h2.8 and 5 3/83, '91; 58.9 and hés, '92; 119.8 196h Cluett Peabody has, '8h; 12.5 Granite City Steel h 5/83, '9h; 30.0 MSL his, '8h; 12.0 1965 American Airlines hs, '90; 53.1 and 533, '91; 81.7 and his, '92; 167.h . Beaumit his, '90; 25.0 Celanese hs, '90; 78.8 Eastern Airlines 5 3/8s, '83; 22.0 and 5s, '92; 75.0 and 1* 3/“89 .93; 5000 General Instrument his, '85; 12.0 and 5s, '92; 50.0 Grant hs, '90; 35.0 Insilco hs, '85; 12.0 and 5s, '93; 35.? Standard Packaging 5&3, '90; 20.0 Stevens hs, '90; 30.0 Twentieth Century Fox his, '90; 18.3 and 5 3/hs, '92; 28.1 United Merchants and Manufacturers hs, '90; h0.0 EBIT EBT tr SDnt = CPnt EpsA EPSS as NPV ADIV APPENDIX B LIST OF SYMBOLS actual net earnings under convertible simulated net earnings under straight debt change in net earnings earnings before interest and taxes earnings before taxes time tax rate nth issue of a given firm, a subscript dollars of interest interest rate outstanding principal for nth convertible issue subscript to denote actual convertible subordinated debenture subscript to denote simulated subordinated debenture actual number of shares outstanding at time t simulated number of shares outstanding at time t. number of shares issued due to actual conversion of nth convertible in year t. number of simulated shares arising from nth issue of straight debt in year t. conversion price of nth issue at time t. market price of common stock at time t. actual earnings per share simulated earnings per share So - SD net present value incremental dividends 155 1:! ass TA LTD scp CL CAPT ss IP DV CV MV TV 156 net interest cost to firm after taxes actual rate of return on equity simulated rate of return on equity total assets long term debt subordinated convertible debenture current liabilities capitalization = TA - CL stockholder's equity total number of issues of a given firm number of years a firm was in the study SCD's issue price maturity value pure debt value SCD conversion value of SCD market value of SCD number of years the bond was held growth rate internal rate of return on bond terminal value APPENDIX C STATISTICAL METHODS The binomial test is a test for goodness of fit. Its purpose is to determine the belief that the pr0portions obser- ved in a sample could have come from a population having a specified proportion. In this study, a P of 0.5 was adopted to check the hypothesis of no difference related to earnings per share and net present value benefits.l Runs tests are tests to determine the randomness of a series. A run is a succession of items of the same class. Data can be classified in a number of ways and the resulting runs examined to determine if non-random influences are pre- sent. Classifications can segregate the data according to lengths of runs, number of runs, runs of ascending order and runs of decending order as well as runs above and below the mean. The tests then determine randomness. It is also possible to check for cyclical patterns and for trend.2 Regression and correlation analysis is a statistical method used to measure the relationship between two or more variables. In this study simple curvilinear techniques were 1 Sidney Siegel, Nonparametric Statistics For The Behav- ioral Sciences (New York: *McGraw Hill, I956), pp. 36Lh2. 2 W. Allen Wallis and Harry V. Roberts, Statistics: A New Approach (New York: Free Press, CollieF'MEEMITIEH7'I956), 157 158 employed. First, to determine the average relationship be- tween two variables as described by the regression equation. Second, to determine the closeness of the relationship as re- flected by the correlation coefficients and standard errors.3 3 Mordecai Ezekiel and Karl A. Fox, Methods of Cor- relation and Regression Analysis - Linear and CurviIInear 13rd ed.; New York: *Wiley,‘19597} APPENDIX D.1 ACTUAL RELATIVE TO SIMULATED EARNINGS PER SHARE 1950 - 1969 Actual Earnings Per Share Relative to Simulated Earnings Frequency Per Share (Per Cent) below 80 3 80 - 8h 3 85-89 6 90 - 9“ 15 95 - 99 36 100 -10h 2h 105 -109 Total firms 9g 159 APPENDIX 0.2 ACTUAL RELATIVE TO SIMULATED EARNINGS PER SHARE BY COMPANY Year Company EPSA / EPSS (per cent) 1951 Warren Petroleum 100.00 1952 Dow Chemical 102.56 International Mineral & Chem- 1081 90.02 Union Oil 90.25 1953 Sinclair Oil 96.72 195h Columbia Gas System 98.51 1955 Burlington 9h.96 Continental Baking 95.8 Fruehauf 93.65 Grace 9 .91 makheed 97072 Merritt Chapman Scott 99.9h National Tea 100.21 RCA 92.89 1956 Alleghany Ludlam 96.h0 American Machine & Foundry 91.06 Burroughs 100.28 Capital Airlines 10h.55 Combustion Engineering 97.77 General American Transportation 98.22 Lowenstein 101.10 National Cash Register 99.19 Textron 96.26 1957 Atlantic 97.01 Carrier 91.h5 Chance vought 101.12 Commonwealth 011 75.12 Douglas Aircraft 99.31 Dresser Industries 92.00 Fairbanks Morse 91.65 General Portland Cement 100.86 Macy 87.66 Mc Derrmott 100.26 National Cylinder Gas 96.05 Olin 99.h3 ll'll'llIlII'lil’ll l 1958 1959 1960 1961 1962 1963 196h 161 Phillips Petroleum Shamrock Oil Thompson Products Boeing Case Continental Airline ITT Richfield Oil Sylvania - GTE Avco Cerro Champion Paper Food Fair General American 011 Texas Hooker Chemical Pan American World Airways Philco Speigel Collins Radio Kayser Roth United Airlines Allied Stores Armour Automatic Canteen Brunswick City Products Crowell Collier FMC Hunt Keystone Steel & Wire Ling Temco Vought U.S. Freight Xerox Air Reduction Control Data Litton Rapid American Stokely Van Camp Ashland Oil Trans World Airlines United Aircraft Cluett Peabody Granite City Steel MSL 99.06 97.h7 91.39 98.08 72.27 93.38 92.hh 99.h3 98.70 99.59 96.20 100.00 99.03 100.82 100.h9 103.06 92.61 100.20 98.6? 92.80 96.93 96.91 99.51 105.38 95.37 99.32 101.31 99.87 95.37 103.30 125.26 100.0h 100.18 101.39 96.92 90.82 101.91 87.72 9h.20 92.71 97.75 97.35 104.30 101.81 1965 162 American Airlines Beaunit Celanese Eastern Airlines General Instrument Grant Insilco Standard Packaging Stevens Twentieth Century Fox 100.hh 93.hh 100.19 99.31 98.h6 9h.86 105.07 100.80 98.36 United Merchants & Manufacturerlel.27 APPENDIX D.3 PRESENT VALUE CASH BENEFITS 1950 - 1969 Net Present Value Cash Benefits Frequency (Millions of Dollars) -11 thru -15 2 — 6 thru -10 Z - 1 thru - 5 1 0 31 + 1 thru + 5 35 + 6 thru +10 3 Note: The zero category includes all observations ranging from minus $500,000 to plus $500,000. 163 APPENDIX D.h ISSUE SIZE AND STRAIGHT VERSUS CONVERTIBLE INTEREST Year Average Issue Number Equivalent Straight Size of Interest Versus Face (Millions of Dollars) Firms Interest - Averages 1s / 1A 1951 15.0 1 1.07 1953 01.8 1 1.03 195h 50.0 1 0.93 1955 35.3 9 1.07 1956 36.5 11 1.07 1957 35.0 16 1.06 1958 28.8 7 1.10 1959 23.5 9 1.1h 1960 17.7 3 1.22 1961 27.2 12 1.20 1962 31.h 1.2? 1963 30.h 5 1.27 196h 33.5 6 1.23 1965 no.3 17 1.2a APPENDIX D.5 EXCESS SHARES REQUIRED UNDER ACTUAL VERSUS SIMULATED SITUATIONS Number of Shares Frequency (in Millions) -0.2 1 -0.1 0 0 32 0.1 9 0.2 8 0.3 7 O.h 7 0.5 3 0.6 7 0.7 2 0.8 3 0.9 0 1.0 1 1.1 1 1.2 2 1.3 1 1.h 1 1.5 1 over 1.5 4 Illllliillilllllllllilll‘l'l. APPENDIX E INDUSTRIAL CLASSIFICATIONS The following industrial classifications were established from the study group companies. Eligibility for industrial classification was determined by reference to Standard Indus- trial Classification (SIC) numbers for all but two cases. The Office Machinery and Conglomerate Categories were determined . by inspection. Air Transport American Airlines Capital Airlines Continental Airlines Eastern Airlines PanAmerican Airlines Trans World Airlines United Airlines Chemicals Air Reduction Chemetron Diamond Shamrock Dow FMC Hooker International Mineral and Chemical Olin w. R. Grace Conglomerates AMF Brunswick FMC International Telephone and Telegraph Litton LTV Olin Textron 166 167 TRW W. R. Grace Distributors Allied Stores Automatic Canteen City Products Dresser Food Fair Grant Lowenstein Macy MSL National Tea Spiegel Electric Equipment Manufacturers Collins Radio Fairbanks Morse General Instrument International Telephone and Telegraph Philco RCA Sylvania Food Processors Armour Continental Baking Hunt Foods Stokley Van Camp Machinery Manufacturers AMF Burroughs Carrier Combustion Engineering Control Data Dresser J. I. Case Litton NCR 168 Office Machinery Brunswick Burroughs Control Data figgeral Instrument Xerox Paper Champion Paper Crowell Collier Standard Packaging Petroleum Ashland Oil Atlantic Refining Commonwealth 011 General American Oil of Texas Mc Derrmott Phillips Petroleum Richfield Sinclair Union Oil of California Warren Petroleum Steel Alleghany Ludlam Granite City Steel Keystone Steel and Wire Textile Mills Beaunit Burlington Celanese Cluett Peabody Kayser Roth Stevens Textron United Merchants and Manufacturers 169 Transportation Equipment Avco Boeing Chance Vought Douglas Fruehauf General American Transportation Lockheed Merritt Chapman Scott Thompson Products United Aircraft "7:111:11?leflifltflfflm‘flfll’flififlmfi'es