Print Reference This Reddit This

Determinants Of The Choice Of Convertible Debt Design Finance Essay

Increasingly Liberal capital controls by the Indian Central Bank coupled with a growing appetite for emerging market investments among international investors has enabled Indian companies to raise capital in international capital markets. Smaller Indian companies until recently were unable to access bond markets, while bigger firms weren’t able to make large issuances due to less liquid and underdeveloped bond markets. As Indian firms make their operations global, borrowing internationally provides them with a natural hedge since receivables and borrowing occur in the foreign currency.

Foreign Currency Convertible Bond is the security of choice among Indian firms accessing international capital markets due to the flexibility it provides to the management of these companies in controlling the design aspects of the security which then determine the informational content of the issue.

This study analyses various aspects of the foreign currency convertible bond market viz. the type of issuers, prominent industry groups represented in the market, stated use of proceeds, size of the market and trend in coupon rates over the years. The next and more significant part of this study is the identification and characterization of issuers based on the announcement date conversion probability of the issue. Selection and measurement of financial variables that affect decisions regarding the design of issue viz. debt-like, equity-like or hedge-like. The study explores if there is a consistent framework which could be used to predict the design of convertible debt issue based on issuer characteristics. Finally, the study checks if there is a time trend in the choice of security design.

Stein (1992) argues that convertible securities make financial markets efficient by enabling different types of firms to invest in their future growth opportunities without bearing significant distress costs. In the absence of convertible debt, Stein conjectures that some firms wouldn’t be able to pursue their investment opportunities.

Surveys done by Brigham (1966) and Hoffmeister (1977) point out that firms choose convertible debt as a means to add equity into their capital structure in a delayed manner to prevent the problem of adverse selection caused due to informational asymmetry between management and investors.

According to Green (1984) convertible securities help get around the problem of asset substitution by constraining shareholders’ incentive to overinvest in risky assets. Kahan and Yermack (1996) state that by giving bondholders the right to convert their claims into equity, management is able to reduce agency costs of debt by allowing debt holders to participate in increases in shareholder value which results from an increase in the company’s overall appetite for risk. Nyborg (1994) states that benefits of convertible debt arise only when the conversion is voluntary instead of force calling the convertible debt when the share price exceeds the call price. This is because the investors perceive a call covenant negatively. Moreover, empirical research has shown that there is an adverse price impact during the force call announcement. On the other hand, Dann and Mikkelson (1984) have suggested that convertible debt offers convey unfavorable information about issuers, although the nature and specific details regarding the information is not identified.

2 Literature Review

2.1 Delayed Equity Hypothesis

Convertible debt alleviates the problem of informational asymmetry between investors and the firm. This helps the firm to avert adverse price impact to some extent which is common for an equity offering. Stein (1992) has developed a three-period stylized model to analyse capital raising decisions of three different types of firms. These firms have similar investment opportunities but different probabilities of a successful outcome to the investment opportunity. The probability of the investment being unsuccessful can be thought of as the risk borne by the firm. The firm which has the lowest risk profile would prefer raising capital through straight debt offering since an equity or convertible debt issue would cause adverse impact on stock price at the announcement date and will make the offering underpriced. The firm with the highest risk in the investment opportunity would prefer a pure equity offering instead of straight debt since if the investment opportunity doesn’t materialize, which is very probable for this type of firms, then it would be unable to repay its debt and would be under financial distress. Raising capital through convertible debt offering has a similar downside. As the investment opportunity slips away, its stock price would continue to remain low and the convertible debt would not be called, forcing the firm to bear a debt burden. For firms that have a relatively low risk profile, convertible debt is the best means to raise capital. These firms would avoid the straight debt route to prevent chances of financial distress. They are also reluctant for an equity offering as the market would view the announcement negatively since investors do not have the same private information as the managers causing an adverse impact on the share price. Moreover, since this type of firms know with some amount of certainty that the conversion value of convertible debt would be positive in the intermediate period, they would force call the convertible debt once the stock price reaches the call price. Thus, this type of firm is best suited to raise capital through convertible debt instruments. He argues that if the costs of financial distress are sufficiently high then a separating equilibrium exists in which no firm mimics any other firm.

Stein further states that convertible securities make financial markets efficient. In other words, all the different types of firms described above invest in their future opportunities and no expected distress costs are borne by any of the firms. In the absence of convertible securities, some of these firms wouldn’t be able to pursue their investment opportunities.

Surveys done in studies by Hoffmeister (1977) and Brigham (1966) confirm that the most firms issue convertible debt to add equity into their capital structure although in a delayed manner to prevent an adverse stock market reaction due to the informational asymmetry described above. Although in his current model, since information asymmetries get resolved in the intermediate period, it seems possible for the firms which have relatively low risk profile to raise capital through a short-term debt offering which matures in the intermediate period and then issue an equity offering since the asymmetries in information are resolved by then. Stein posits that this is an unrealistic assumption and suggests an iterative process where managers are always one step ahead of investors, so informational asymmetry exists at all times. Thus, the model need to be developed further.

Nyborg (1994) extends Stein’s work but focuses on the issue of forced versus voluntary conversion. He states that Stein’s stylized three period three firm model is not able to capture all the dynamics faced by firms in financial markets. In particular firms pursuing the strategy of issuing convertible debt as delayed equity face a two-phase fall in stock price, once when the convertible is issued and the next when it is called. So the net drop in share price may be higher for such a capital raising strategy. Stein’s hypothesis was supported by an empirical study done by Smith (1986) in which the cumulative abnormal returns of equity and convertible debt issues were compared at the announcement date. The study concluded that the cumulative abnormal return in equity near the announcement date was -3.14% while for convertible debt it was around -2.07%. In research performed by Mikkelson (1981) it was observed that forced conversion at maturity resulted in a cumulative abnormal return in stock price of -2.08%. Thus the total return in equity for a convertible issue is around -4.15% which is even greater than the average drop observed in equity offerings. Nyborg emphasises on the information content of call policies rather than on financial distress costs as a deterrent to low quality firms to issue convertible debt. Nyborg’s model comprises of four dates starting from zero to three in which an all-equity firm has an investment opportunity where the cash flow occurs only on the last date. If the manager issued convertible debt he can decide in the two intermediate periods whether to call the convertibles or not. If the convertible bond is called at a later date either forcefully or voluntarily, the equity of the firm suffers a dilution which also results in the dilution of the convertible bond. Nyborg proposes that there is a separating equilibrium in which manager force calls the issue if and only if he receives further bad information about the investment opportunity. Nyborg proposed that signalling through a call policy at intermediate dates may increase the variance in stock prices in those dates. The manager being risk averse might prefer an equity issue. In the absence of informational asymmetry the preferred financing option for the manager is equity. In a separating equilibrium, a manager who has proprietary information about positive prospects of the investment opportunity would issue straight debt despite the fact that it adversely exposes the firm to stock price volatility in later periods. Nyborg (1994) further claims that Callable convertible debt provides insurance against stock price depreciation. This can therefore be treated as a cost to prevent financial distress. Any call policy pursued by the firm sends a negative signal to investors so it might make the stock price more volatile than if such a call feature was not part of the bond indenture. The author further states that for firms where the probability of the investment opportunity materializing are high, a repurchase of debt financed through equity would cause a greater adverse impact on price than a callable debt since the repurchase cost in a callable convertible is implicit in its initial price. Nyborg argues that call protection may decrease the probability of conversion. If the cost of financial distress was high, managers wouldn’t offer call protection. One of the key propositions of Nyborg which differs from Stein is that the adverse effect of issuing equity is less as compared to issuing convertible debt if the conversion is forced. This is because the investors generally perceive the announcement of force call negatively and therefore the stock prices suffers a negative impact during such an announcement. The author’s main finding is that convertible debt cannot by itself treated as a mechanism to add equity in the capital structure rather it provides a means to delay the decision to add equity or to add debt.

As an aside, Harris and Raviv’s (1985) proposition on signalling through convertible debt policy emphasises that after tax gains for investors holding dividend paying stocks are greater than coupons of convertible debt on those stocks. Thus, investors might be better off holding stocks than convertible debt. This is not an issue for non dividend paying stocks. Brennan and Her (1993) showed that on average, the dilution suffered by convertible bonds was around 16.3%.

2.2 Asset Substitution Hypothesis

Brennan (1987) states that a firm's financing strategy could be considered as a signalling device to investors since they could evaluate these strategies. Financing per se should be costless if Modigliani Miller propositions hold. He further says that if the adverse selection due to information asymmetry between firm managers and investors can be overcome by financial policy, it will be efficient to do so. Adverse selection may prevent firms from pursuing profitable investments due to inability to raise adequate capital in financial markets. The firm’s existing capital structure may also play a role in defining an efficient financing strategy.

Green (1984) demonstrates in his paper that convertible securities may help the firm mitigate problems surrounding asset substitution by checking shareholders’ incentive to overinvest in risky assets. Jensen and Meckling (1976) state that external financing can distort the firm’s optimal operating strategy since it may happen that shareholders’ incentives don’t go hand in hand with the firm’s enterprise value. If the firm raises capital through debt financing bringing in substantial debt to its capital structure, shareholder’s incentive’s lie in the firm assuming more risks i.e. substituting less risky assets for more risky ones, since their claim on the firm’s assets is convex. Green (1984) demonstrates that such a “risk-shifting” cost can be eliminated by selecting an appropriate security design i.e. instead of raising capital through straight debt, the firm could use a convertible structure to reduce the conflict of interests between debt holders and shareholders by balancing the concave and convex regions of the contract to form a equity like security.

The extent to which such agency costs can be reduced depends upon the level of financial leverage. Jensen and Meckling (1976) have stated that cost related to risk-shifting increases with the level of debt in the capital structure.

2.3 Financing real options

Mayers (2004) states that convertible debt financing is the most cost effective way for high growth firms that have a sequence of investment opportunities. He proposes a stylized three period model to explain the dynamics. He argues that for such firms the cost of debt or equity financing would be quite high and that a strategy involving convertible debt would circumvent the problem of adverse selection. A firm that has potential investment opportunities in two distinct periods may raise capital by issuing a two period straight debt offer. If the first opportunity materializes, the firm can use the proceeds from it to fund the next opportunity if it becomes profitable. The problem with this approach is that if the second opportunity doesn’t become profitable, the firm would be left with excess cash. Since investors assume that the firm would most likely invest in negative NPV projects in such a scenario (i.e. the problem of overinvestment), giving rise to adverse selection. The same would be true for an all equity offering. On the other hand a convertible debt issue provides a perfect financing strategy. The firm issues a convertible debt and if the second project becomes profitable in the intermediate period, the issue would be voluntarily converted into equity as information asymmetries vanish. Thus the firm wouldn’t need to repay the face value on its debt and the funds would be used to for the next project preventing an adverse market reaction. So the problems of overinvestment and adverse selection can be eliminated.

3 Sample Selection and Data description

3.1 Sample Selection Procedure

The data on foreign currency convertible bonds (FCCB) denominated in US dollars and issued by Indian companies from 2003 up to 2010 was collected from Bloomberg. The dataset comprises of 197 securities and 155 companies.

Some securities had to be excluded from the study due to the following reasons:

Conversion premium was missing. This is required to analyse issues concerning security design which is central to this study.

Financial Accounting data for the issuing firm wasn’t available near the announce date of the issue.

The data on stated uses of proceeds from the issues over the period 2003-2010 was procured from Reserve Bank of India’s website.

3.2 Data description

This study explores some of the implications of recent work done in this area by Stein (1992), Nyborg (1994), Green (2003) and Seward (2003). They can be stated as:

Firms’ motivations for issuing foreign currency convertible bonds

Characteristics of firms issuing these bonds.

Growth opportunities these firms have.

Call policy on these securities.

How security design vary with firm characteristics and business cycle?

Table 1: Industry subcategories of Indian firms which issued foreign currency convertible bonds from 2003-2010.

INDUSTRY

No.

INDUSTRY

No.

Agricultural Biotech

2

Hotels & Motels

3

Agricultural Chemicals

1

Identification Sys/Dev

1

Agricultural Operations

2

Lighting Products & Sys

1

Airlines

1

Mach Tools & Related Products

1

Apparel Manufacturers

1

Machinery-Material

1

Appliances

2

Medical-Drugs

18

Applications Software

10

Medical-Hospitals

1

Auto/Truck Parts & Equip

5

Metal Processors & Fabrication

4

Auto-Cars/Light Trucks

3

Metal-Copper

1

Auto-Med & Heavy Duty Trucks

1

Miscellaneous Manufacturing

1

Beverages-Wine/Spirits

2

Motion Pictures & Services

3

Bldg Prod-Cement

1

Networking Products

1

Building & Construction-Misc

5

Oil Refining & Marketing

1

Building-Heavy Construct

6

Oil-Field Services

2

Chemicals-Diversified

4

Paper & Related Products

4

Chemicals-Specialty

3

Petrochemicals

1

Coal

1

Pharmacy Services

1

Communications Software

1

Power /Supply Equip

6

Computer Aided Design

2

Recreational Centers

1

Computer Services

2

Retail-Jewelry

2

Computer Software

2

Rubber/Plastic Products

2

Computers-Integrated Sys

3

Rubber-Tires

1

Computers-Peripheral Equip

2

Shipbuilding

1

Containers-Paper/Plastic

1

Steel Pipe & Tube

4

Data Processing/Mgmt

1

Steel-Producers

9

Distribution/Wholesale

1

Storage/Warehousing

1

Diversified Minerals

1

Sugar

4

Diversified Operations

8

Tea

1

Educational Software

1

Telecom & Fiber Optics

2

Electric-Integrated

2

Telecom Services

3

Electronic Instruments

1

Telecommunication Equip

2

Energy-Alternate Sources

2

Textile-Apparel

1

Engineering/R&D Services

3

Textile-Products

11

E-Services/Consulting

2

Theaters

3

Finance Loan/Banker

1

Tools-Hand Held

1

Food-Misc/Diversified

1

Transport-Marine

1

Gas-Distribution

1

Transport-Services

2

3.2.1 Industry Subgroups

This data was obtained by aggregating data for dollar denominated convertible debt securities issued by Indian firms abroad over the years 2003-2010. It can be observed that Application software firms and medical drugs firms have been most aggressive in issuing FCCBs. This goes hand in hand with the theoretical prediction that firms that have high growth opportunities but face substantial risk in their investment opportunities are the ones most active in convertible debt market. I further classified firms into broader categories by combining some of the industrial subsectors. This was done to limit the number of industry subgroups from 74 to 22. The top ten most active industries in the foreign currency market have been plotted in the graph below to get the distribution of firms across broad industrial sectors.

Table 2: Percentage of Firms by Industrial sector

3.2.2 Size of the Market

Next this study looks at the size of the FCCB market over the years since its inception.

Table 3: Size of FCCB Market

Year

Total amount issued per year

Number of issues

Average size

Median size

2003

$45,000,000.00

1

$45,000,000

$45,000,000

2004

$460,000,000.00

3

$153,333,333

$110,000,000

2005

$2,329,250,000.00

34

$68,507,352

$48,500,000

2006

$4,629,500,000.00

64

$72,335,937

$42,500,000

2007

$7,426,300,000.00

64

$116,035,937

$50,000,000

2008

$545,203,171.00

10

$54,520,317

$20,000,000

2009

$3,243,851,854.00

15

$216,256,790

$150,000,000

2010

$441,940,420.00

5

$88,388,084

$75,000,000

There was a steady increase in the number of firms issuing FCCB until 2007. Since then there was a substantial fall in the numbers and didn’t quite pick up to the pre-2007 levels. From 2005 till 2007 there was an increase in the average size of issues. The financial crisis of 2008 lowered both the number of issues and the size of issues. From mid 2009, both the average and median size of issues has considerable risen. This may be due to the fact that bigger firms are resorting to foreign currency convertible debt as a viable market to raise money.

3.2.3 Stated use of proceeds

Further we look at the stated use of proceeds. This data was collected by consolidating data from monthly tables for the seven year period (2003-2010) available through Reserve Bank of India’s website.

Table 4: Stated Use of Proceeds

Expansion of Activity

3.23%

Import of Capital Goods

12.90%

Modernisation

45.16%

New Project

9.68%

Overseas Acquisition

22.58%

Refinancing of old loans

6.45%

Modernisation has been the stated purpose for raising capital through convertible debt in over 45 percent of securities issued till date. Overseas acquisition is the next big purpose of raising capital in this market. In only 6.5% of the cases, refinancing of old loans is the stated use of proceeds. This implies that issuing convertible debt to repay loans is not a preferred strategy for firms. Capital obtained through convertible debt offerings is primarily used in financing those activities that have a huge potential to be profitable but there is some uncertainty involved. This also explains the reason why funding new projects using proceeds from convertible debt is more common than refinancing old loans.

3.2.4 Trend in coupon payments

This study also examined the trend in average and median coupons offered on FCCBs. There has been a sharp increase in the amount of coupon offered post-2008. In the years up to 2008 the coupon varied between 0.8% to 1%. Post-2008, average coupon rate jumped to over 4.5%. The median coupon rates during 2009 and 2010 were at 5%.

The reason for this sharp increase is the following. In the aftermath of the financial meltdown, buyers of FCCBs turned have cautious. Prior to the financial meltdown, FCCBs were mostly zero coupon bonds that did not involve intermediate interest payments, but were sold at a discount to the face value. For firms issuing FCCBs, this meant that they could get access to funds at cheaper rates, and the payments were only due at maturity.

Table 5: Average coupon on FCCBs

Year

Average Coupon

Median Coupon

2005

0.76

0.50

2006

0.81

0.00

2007

0.99

0.00

2008

0.80

0.00

2009

4.63

5.00

2010

4.65

5.00

4 Results

4.1 Identifying issuers based on security design characteristics

Recent theory in this area says that the likelihood of conversion of convertible debt offerings at announcement date can be used as a measure for identifying the security design characteristics. The risk neutral probability of conversion can be obtained from Black and Scholes (1973) option pricing formula. This could be interpreted as the likelihood of conversion assuming a risk-neutral world. It is represented as N(d2) in the option pricing formula where N is cumulative probability function of a standard normal random variable. The parameter d2 is defined as follows:

Firms issuing convertible debt can choose conversion premium, maturity period or call period in a way to make the offer more debt-like or equity-like. The study observed that the mode of the maturity period across observations in the foreign currency convertible bond market is 5 years. This restricts firms to design the security primarily based on conversion premium.

Stein mentions three kinds of firms that raise money in capital markets. In a separating equilibrium proposed by Stein, each of them raise money through one of equity, debt or hybrid securities. In Seward (2003), convertible securities are categorized into three classes based on announcement date conversion probabilities. The paper describes a convertible bond as ‘debt-like’ if the conversion probability is less than 40% and equity like if the conversion probability is greater than 60%. If the probability lies between these two values then it is termed as a hybrid security. The author justifies these cut-offs as reflecting the simple observation that a higher conversion probability is interpreted by the market as delayed equity. This has also been mentioned in Stein (1992) that convertible debt is a means of adding equity in the firms’ capital structure albeit in a delayed manner. Hybrid security design as described above can be considered hedge-like securities because of the fact that it is equally likely that the convertible debt is converted or remains as debt. Convertible securities designed as “debt like” instruments have a lower probability of conversion into equity but they still provide investors a signal that the manager’s incentive to invest in risky asset is constrained. Thus according to Green (1984) it could be used to alleviate problems related to asset substitution.

After screening the data by removing securities which did not have crucial data on security characteristics like conversion premium, only 70 securities were left. I used the following methodology to classify each security into one of the three categories listed above. I obtained information on conversion premium and 360-day volatility for all the above securities from Bloomberg. I used the annual interest rate in India as a proxy for risk-free rate since that is the rate one can realize on a one year investment in an Indian government debt which is almost riskless. The following distribution was obtained. 20% of the securities were equity like, 51.43% were debt like and the remaining 28.57% were hedge-like. I found that the average risk neutral probability of conversion was 44.7%. In a similar study done by Lewis (1999) on a sample of US convertible securities the average was 50.3%. While in a study done by Dutordoir and Gucht (2006) for Western European firms, the value was 32.44%. This implies that the conversion probabilities of convertible securities issued in India are somewhere between those issued by US and Western European countries where the former is more equity like and the latter is more debt like.

4.2 Selection and measurement of explanatory variables

The study now explores what explanatory variables can be used to identify security design characteristics.

Table 6: Selection and Measurement of Explanatory Variables

Explanatory Variable

Full Sample

Debt-like

Equity-like

Hedge-like

Change in Book Value (in percent)

73.82

58.15

67.06

106.31

(48.83)

(30.3)

(76.29)

(141.67)

Price-Earnings Ratio

31.05

47.92

15.87

11.37

(11.86)

(15.29)

(14.65)

(10.43)

(ROE/ROA)

3.39

3.34

4.26

3.42

(3.03)

(2.39)

(3.61)

(3.43)

Long-term debt/total assets

40.66

40.14

40.15

43.14

(43.38)

(40.27)

(39.66)

(46.49)

360 Day Volatility

57.96

61.30

46.86

59.24

(57.83)

(59.72)

(47.49)

(58.96)

firm-size (In Billion Rs)

143.924

64.969

438.421

74.763

(47.056)

(33.018)

(326.805)

(55.263)

Financial Leverage

3.80

3.74

4.08

3.64

(3.46)

(3.32)

(3.62)

(3.84)

Slack (Cash+Mkt Sec/Total Assets)

10.82

11.29

7.78

12.78

(7.44)

(5.85)

(6.13)

(14.03)

Market-to-Book ratio

1.94

0.97

2.40

1.23

(1.48)

(0.97)

(2.86)

(1.69)

Net Income/Total Assets (ROA)

6.81

6.20

5.26

6.81

(8.05)

(3.87)

(5.34)

(8.05)

4.2.1 Variables signifying investment opportunities

According to Green (1984) and Stein (1992), one of the most prevalent motives for convertible debt offerings is to fund the firm’s uncertain investment opportunities. Seward (2003) analyzed two attributes associated with the investment policies of firms issuing convertible debt:

Future investment growth rate and

Profitability of future investments

Future investment growth rate can be measured as the change in total assets during the year surrounding the announcement. Change in total assets was calculated as the difference between the book value of assets at fiscal year after announcement date and the book value of assets for the fiscal year prior to the issue announcement date, divided by the latter. This measures the rate of change in investment during the period around the convertible debt offer. Ceteris paribus, firms experiencing rapid growth in assets require greater amount of investment capital. Capital expenditure should have to be commensurate with the investment to prevent the possibility of overvaluation. In table 6 we observe that both equity and hedge like issuers have experienced high changes in assets around the convertible issue. Ideally firms which have an equity like offering are the ones that fall under Stein’s classification of firms seeking backdoor equity.

The profitability of future investments can be measured using the price-to-earnings ratio. The price-to-earnings ratio is computed as market price of shares divided by earnings per share (EPS). McConnell and Servaes (1995) suggest that firms with high market-to-book ratios and high price-to-earnings ratios have a greater likelihood of facing high financial distress costs since investors are vary about their ability to profit from future investment opportunities.

4.2.2 Effect of Financial Leverage

Stein (1992) states that managers of firms seeking backdoor equity force call the convertible issue once the likelihood of investment turning profitable increases. Nyborg (1994) raises a criticism to Stein’s model by saying that the problem of adverse selection is not eliminated by this strategy and the firm further experiences a drop in share price at the call date. His main criticism to Stein is that managers will only force-call their issues if the investment opportunity worsens. On the other hand, if the likelihood of the investment opportunity turning profitable increases, then the manager will never force call the issue, but rather wait for voluntary conversion as and when asymmetries vanish.

In the foreign currency convertible bonds issued by Indian firms from 2003 to 2010, only 8 out of 197 issues have call provisions in the bond indenture. This provides support to Nyborg’s hypothesis that firms less interested in force-calling their convertible debt. They rather wait for the asymmetries of information to vanish. On analysing the three categories of convertible issues defined in the previous section separately, following observations are made:

Firms that have higher leverage ratios have a more equity like offering to ensure that investors voluntarily convert into equity.

Firms that have lower leverage ratios have a more debt like offering. This could mean that such firms do not view their convertible debt as a form of delayed equity as proposed by Stein (1992). Alternatively, the managers of such firms are very optimistic about the firms’ investment opportunities and subsequent rises in the share price as they have private information about the likelihood of success of their future investment opportunities.

4.2.3 Optimal convertible debt design and explanatory variables

Seward(2003) follows an event study to measure the share price response to the announcement of convertible debt issues over a two-day period using market model as the pricing benchmark. The return on an equally weighted average of returns on the NYSE, AMEX, and NASDAQ market indices serves as a proxy for the market return. Excess returns are based on the parameter estimates of the market model over 280 trading days. This provides an understanding about investors perceptions to issue announcements. He further performed a multivariate logistic regression of firm-specific and industry characteristics of convertible debt issuers. Characteristics of these firms is compared with the industry median of each firm for each explanatory variable.

CD = intercept + market-to-book ratio + net income/total assets +

change in total assets + long term debt/total assets + firm size

+slack + volatility + pre issue run up in stock price + e

In this study, instead of performing a similar regression due to paucity of data, I examine and compare the mean and median values of various explanatory variables to examine the role of each of these in determining the optimal design of convertible debt so as to minimize the impact of investors’ negative reactions regarding the convertible offer.

Seward (2003) states that financial distress cost are high for firms with high market-to-book ratios and/or low earning-to-price ratios. Such firms avoid raising capital through straight debt issues. Thus such firms would resort to convertible debt as a means to raise capital. Moreover such firms would design the debt such that the issue date conversion probability is high. This is consistent with the delayed equity hypothesis of Stein (1992). The results obtained from Indian markets confirm this result since the mean value of market-to-book ratio is 2.4 for firms that have equity like convertible structure which is substantially higher than that of firms with a more debt like structure.

Further, Asset substitution costs are high for firms with low market-to-book ratio. Such firms avoid raising capital through debt straight offers since the shareholders incentives would no longer be aligned with that of the firm’s enterprise value and shareholders would invest in risky assets to earn higher returns. Such firms may issue convertible debt but with a more debt like offering. The results obtained from the data on Indian FCCBs verify this result.

Seward (2003) states that firms which have high leverage ratios and/or have volatile firm value are expected to face high debt related financing costs. Firms that have high net income to total assets i.e. high return on assets often do not need to resort to external financing. This is also true if the firm has a high slack ratio i.e. the ratio of cash and marketable securities to total assets. If such firms issue convertible debt, the market views this positively and therefore there isn’t any problem of adverse selection. There is no clear relationship between return on assets and the security design a firm may chose for its convertible debt.

An anomaly to Seward’s analysis was found in the data on Indian FCCBs. Seward (2003) conjectured that large firms having higher financial leverage prefer a more equity like convertible issue. On the contrary, I found in my data sample that the average size of firms that issued equity-like convertible bonds was considerably higher than the ones issuing debt-like convertible bonds. Further, these firms also had higher financial leverage ratios although these firms belong to disparate industries. In his paper, Stein (1992) suggested analysing firms’ motives for issuing convertible debt. From the stated use of proceeds it was evident that these firms are investing primarily in acquisitions abroad and/or modernization. Thus raising capital in foreign currency to fund foreign acquisitions provide a natural hedge to the investment. The second stated use of proceeds i.e. modernization, is an uncertain investment and the outcome of such corporate activity is perceived cautiously by investors. Thus instead of raising capital through equity, these firms are better off resorting to the convertible debt market to cause a less adverse stock market reaction on announcement date.

4.2.4 Time variation in security design decisions

Table 7: Security Characteristics by Year

Year

Equity-like

Debt-like

Hedge-like

2005

33%

56%

11%

2006

25%

55%

20%

2007

4%

70%

26%

2008

20%

30%

50%

2009

30%

10%

60%

2010

33%

33%

33%

Several studies in the past have documented that security design choices seem to vary with to business cycles. Choe (1993) states that adverse selection on seasoned equity announcements is much lower in expansionary periods. Ulrike Hoffmann-Burchardi (1997) outlines two effects which can lead to the formation of hot issue markets in the case of IPOs where there is incomplete information about the industry’s prospects. He says that if investors rely on the information acquired for the first IPO within the industry to access subsequent IPOs, the so called “free ride” problem then the investors “uninformed valuation” can become greater than the firm’s private value. Secondly, if the prospects of the industry become more uncertain then managers of both high and low quality firms would go for an IPO to prevent the revelation of poor industry prospects by some other firm.

There is considerable evidence of time variation in convertible security design in the data on FCCBs. Years 2006 and 2007 were “hot markets” where the total issue volume was over four times the issue volume in 2008 or 2009. The latter could be termed as “cold markets”. Hot markets are characterized by a debt-like convertible issues while cold markets have more equity-like or hedge-like convertible debt issues.

Conclusion

Foreign currency convertible bonds provide managers of Indian firms to design securities according to different debt and equity related financing costs. Moreover, for firms in certain industries convertible securities is the only viable means to raise capital as they face substantial costs of financial distress. Issue date conversion probability seems to be an important factor in explaining the relationships between security design choices and various firm characteristics measured by variables like change in total assets, financial leverage, market-to-book equity etc. Stein’s (1992) hypothesis that certain firms issue convertible debt to infuse equity in their capital structure in a delayed manner is confirmed. Nyborg’s (1994) hypothesis about voluntary conversion seems to hold in the foreign currency convertible bond market. Green’s (1984) arguments about agency conflicts and asset substitution is also one of the factors that prompt certain types of firms to enter this market. Lastly, business cycle seems to have an impact on the choice of security design.

Need help with your literature review?

Our qualified researchers are here to help. Click on the button below to find out more:

Literature Review Service

Related Content

In addition to the example literature review above we also have a range of free study materials to help you with your own dissertation: