Balancing operational- and financial
risk in the dry bulk shipping market
ABSTRACT: This thesis examines the relation between operational risk, defined as the spot market exposure a shipping company has, and financial risk on leverage. Spot market contracts are considered riskier than time charter contracts due to risk of unemployment, relocation and bunker costs.
Moreover, employing vessels under time charter contracts results in more favorable financing terms on bank loans due to higher predictability of cash flows. Using a sample of twelve firms operating in the dry bulk market between 2007 and 2012, it is hypothesized there exists a tradeoff between operational and financial risk. The results found, however, show a positive relation between spot market exposure and leverage. This might be attributable to the shipping cycle, leading ship owners to employ their vessels under spot charters in expectation of a market upswing, keeping leverage constant. At the same time, a ship owner might employ different methods to reduce operational risk, for instance by only chartering his most operationally flexible vessels under spot contracts or by hedging freight rates by using derivatives. Furthermore, it is concluded fleet employment reporting must be standardized in order to ensure proper comparison can be established, which will lead to a deeper understanding of the link between operational risk and financial risk in the (dry bulk) shipping industry.
By Laurens van Brandenburg Student number: 10013806 Supervisor: M.A. Dijkstra May 30, 2014
Table of contents
1 INTRODUCTION ... 1
2 LITERATURE REVIEW ... 4
2.1 Theory ... 4
2.2 Empirics... 12
3 DATA DESCRIPTION & METHODOLOGY ... 14
3.1 Methodology ... 14
3.2 Variables ... 15
3.3 Sample ... 17
3.4 Descriptive Statistics ... 18
4 RESULTS ... 21
4.1 Standard leverage regressions ... 21
4.2 Robustness checks ... 23 5 DISCUSSION ... 24 6 CONCLUSION ... 26 APPENDIX ... 27 REFERENCES ... 28
1 INTRODUCTION
Dry bulk shipping is defined as transport by sea on a one cargo/one ship basis (Stopford, 2009),
shipping commodities in unpackaged form. Dry bulk vessels can be contracted by spot contracts and
time charter contracts. These contracts are settled by brokers on the freight market (Stopford, 2009).
Spot charter contracts cover a single voyage or trip and allocate bunker costs (fuel costs, port taxes
etc.) to the ship owner. Time charter contracts are long-term contracts and allocate bunker costs to the
charterer. Freight rates determine the price a ship owner is able to demand for chartering a vessel.
Freight rates reflect the equilibrium between supply and demand for dry bulk vessels and are reported
in the Baltic Dry Index (BDI). The Baltic Dry Index is published by the Baltic Exchange in London
and is based on US dollars. It is measured by taking the 23 most important trading routes on a time
charter basis to provide an assessment of the price of moving the major raw materials by sea (Baltic
Index, while time charter rates are believed to be based on the market’s expectations of future spot
rates (Alizadeh & Nomikos, 2010). Marsoft, the world’s biggest shipping forecasting consulting firm
(Lorange, 2005), provides strategic vessel employment advice to ship owners based on the state of the
Baltic Dry Index (upswing or downswing). Vessel employment pertains to the decision of chartering
vessels under a spot or time charter contract.
Spot charter rates expose a firm to the risks of unemployment, relocation costs and fuel cost
fluctuations (Zannetos, 1966, Eriksen & Norman, 1981, Kavussanos, 2006). Furthermore, spot
contracts are deemed riskier (Kavussanos, 2006), because under a time charter contract the charterer
pays a fixed amount per day for the time of the contract and time charter rates are less volatile than
spot freight rates. Hence, cash flow volatility will increase when a ship owner employs more of his
ships under spot charters. As a result, Zannetos (1966) and Eriksen and Norman (1981) argue there is
a positive risk-premium for spot charter rates to compensate for these risks, which will be referred to
as operational risk. At the same time, the shipping industry is highly levered, exhibiting a mean book
leverage ratio of 41% (book value of short- and long-term debt over book value of assets) as opposed
to 25% for firms across all industries in the G7 countries covered in Compustat Global (Drobetz et al.,
2013)1. The leverage ratio in this sample is as high as 48% (for book leverage). Opler & Titman
(1994) find that highly levered firms perform worse by 26,3% than less levered companies during a
recession based on sales growth, constituting financial risk. The exposure to risk on the spot market
and the risk pertaining to leverage through increased costs of financial distress is balanced by
managers according to the Shipping Corporate Risk Tradeoff Hypothesis formulated by Merikas et al.
(2011). The research question of this thesis therefore is: Does there exist a relation between the
amount of ships a dry bulk firm operates in the spot market and the leverage ratio of the firm? Based
on a sample of twelve dry bulk firms in the period from 2007-2012, it is tested whether there exists a
relation between the spot market exposure and the leverage ratio of a dry bulk company by using a
fixed-effects regression. The control variables tangibility, profitability, market to book value, size and
dividend-payer are included to isolate the effect of spot market exposure. The results show a positive
1
G7 firms are firms from the Group of 7, seven advanced economies consisting of Canada, France, Germany, Italy, Japan, the United Kingdom and the United States of America.
but insignificant relation between spot market exposure and leverage. This relation can be explained
by the period under examination. The dry bulk market experienced a downfall in 2008/9 causing the
Baltic Dry Index to arrive at an all-time low (Financial Times, 2008). Marsoft developed a Shipping
Market Cycle model in 2003 to provide advice on vessel employment management (Lorange, 2005).
According to this cycle, a dry bulk firm needs to employ vessels in the spot market when the Baltic
Dry Index is low to anticipate a market upswing (Lorange, 2005). This might have led to an increase
in spot market employment. At the same time, shipping companies tend to deleverage significantly
slower during recessions (a 9,3% lower speed of adjustment for market leverage, Drobetz et al., 2013),
contributing to a positive relation between spot market exposure and leverage2.
The remainder of this thesis is organized as follows, section 2 provides the theoretical
framework, section 3 contains the data and describes the methodology, section 4 displays the results,
section 5 is a discussion on the results and finally section 6 offers concluding remarks.
2 LITERATURE REVIEW
This section defines operational risk associated with spot market exposure and discusses other factors
driving vessel employment decisions in the dry bulk industry, such as vessel size and the shipping
cycle. In addition, the rationale behind standard leverage-related decisions in the dry bulk shipping
industry is explained to ensure the spot exposure effect on leverage can be properly identified. To
assess which capital structure factors explain financing decisions and how these factors affect the
leverage ratio, theories on leverage will be discussed.
2.1 Theory
Operational risk
The dry bulk shipping freight market has time charter contracts (contracts for a period of time), spot
charters (contracts for a particular voyage, based on tonnage) and trip time charters (TTC, essentially a
charter contract for one voyage but charged by day instead of based on tonnage). Moreover, less
common contracts such as contracts of affreightment (CoA’s, contracts for the carriage of a fixed
2
volume of cargo) and bareboat charters (charter contracts without crew or management ships) exist
(Gratsos et al., 2012). A main difference between time charters and spot charters lies in the carriage of
the bunker costs. Bunker costs consist of fuel costs, port charges, canal dues and cargo handling costs
(Kavussanos, 2002). Under time charter contracts the charterer pays for these costs, while under a spot
charter contract the ship owner pays. Figure 1 displays the distribution between the use of time
charters, spot charters and trip time charters.
Figure 1: Fixture counts per contract type per month (in percentage share), 2001-2009. The y-axis shows the percentages, while the x-axis shows the time period.
Source: Gratsos et al. (2012).
Different types of contracts represent different levels of risk. According to Kavussanos (2010)
a one-year time charter and the sum of a series of spot rates over the same period plus a risk-premium
should theoretically equal each other on average by assuming the freight market is efficient. Ship
owners should not be able to make a profit by choosing either contract type. This is called The Term
Structure relationship and it is based on the expectations hypothesis, which was originally developed
for the bond market (Mankiw & Miron, 1986). Empirical testing, however, has rejected this
expectations hypothesis because the decision to employ ships in the spot- or time charter market is
found to vary with the volatility of the excess returns between the spot and time charter rates (Hale &
Vanags, 1989, Veenstra, 1999 and Kavussanos, 2002). Apparently, ship owners are willing to take a
market is perceived as riskier than the time charter market is threefold. First, spot rates fluctuate more
than time charter rates because time charter rates are smoothened by aggregated expectations over the
period of the contract. This results in higher levels of volatility for spot charters compared to time
charters (figure 2). Secondly, shipping companies decide to employ their ships on the time charter
market to reduce the risk of vessel unemployment or the risk of having to relocate a vessel to a
different port. Finally, the bunker costs constitute a part of the risk. Since 20-30% of bunker costs are
fuel costs, spot charter contracts expose dry bulk companies to fluctuations of fuel costs (Stopford,
2009). Deciding on which market to employ its ships is thus an important commercial decision for
ship owning companies, since risk and return vary between spot and time charter contracts. The higher
perceived risk on spot charters is defined as operational risk.
Figure 2: Panamax sector: Spot vs time charter rate volatilities in returns (US$) (SD’s), 1980-1996. Volatilities are shown on the y-axis, while time is shown on the x-axis.
Source: Kavussanos (2010)
Operational risk varies with vessel size. The dry bulk market is divided into three major
segments based on the capacity of a ship, as measured in dead weight tons (dwt). In addition, their
names reveal the routes they are able to sail. Cape-size bulk carriers can carry between 80.000 and
180.000 dwt and cannot sail through the Panama Canal or the Suez Canal because of their size.
Panamax carriers have a capacity between 55.000 and 80.000 dwt and are able to sail through the
to carry between 25.000 and 35.000 tons and sail through both canals (Lorange & Fjeldstad, 2012).
The volatility in daily freight rates increases with vessel size (see figure 3). Vessel size might
influence the decision on whether ships are employed in the spot market or under time-charter
agreements. Alizadeh and Nomikos (2011) find that returns for Handy-size vessels are less volatile
(average standard deviation of 0,22) compared to the other sizes (0,50 and 0,43, for Panamax and
Capesize respectively) in the spot market, based on 817 weekly observations from 1992 to 2007.
Figure 2 shows similar results based on freight rate volatilities between 1980 and 1996, where the
volatility for Capesize vessels is highest (avg. std. dev. of 0,21) followed by Panamax and Handy-size
vessels (0,15 and 0,09 respectively) (Kavussanos, 2010). Smaller sized vessels are more flexible in
terms of trading routes, access to ports and serve more, and more varied commodities (Kavussanos,
1996). Increased operational flexibility makes smaller vessels easier to employ by serving more trades
and trade routes and thereby decreasing volatility in returns. Contrarily, Capesize vessels are restricted
to approaching specific ports and fewer canals are accessible to them. This contrast results in a
reduced number of trading routes for Capesizers, affecting demand for these vessels. The quantity of
trading routes a vessel is able to navigate determines the size of the market it operates in, which in turn
implies dry bulk market can be regarded as a disaggregated market (Alizadeh & Nomikos, 2010).
Figure 3: Spot freight rate volatilities by vessel size; dry bulk sector, 1980-1996. Volatilities are shown on the y-axis, while the time is shown on the x-axis.
Alizadeh and Nomikos (2010) and Strandenes (2012) argue, however, that substitutions
between the size segments can occur when the demand for ships is relatively higher in one market than
in the other and the possibility of making profitable journeys by accepting part cargoes is present.
Secondly, a company might prefer smaller vessels because it uses just-in-time inventory management,
meaning they ship multiple relatively small cargoes instead of one large cargo to press inventory costs.
The reason why these differences in volatilities are so important is because they offer investors an
insight to the risk/return strategies of dry bulk companies and help diversify their portfolio. At the
same time, volatility differences across vessel size offer the ship owner a way to minimize the total
risk exposure of their fleet (Kavussanos, 2010).
Decisions on fleet employment are subject to shipping cycles (Stopford, 2009). Stopford
(2009) examined the shipping market from 1872 to 2008 and identified 15 cycles with an average
length of 7,7 years and a standard deviation of 2,6 years. The freight rates in the shipping market are
determined by supply and demand for transportation, rising when demand exceeds supply and falling
when supply exceeds demand. The shipping cycles are driven by the world economy through
declining industrial activity and decreased demand for raw materials in times of recession, leading to a
lower demand for transport, while in times of economic prosperity demand is high (Stopford, 2009).
On the supply side, shipbuilding contributes to the cycle because of the time lag between ordering and
delivery of a vessel, which can take up to three years and during which vessel demand may have
changed (Stopford, 2009). Deciding on which market to employ its vessels is to some extent
dependent on the shipping cycle (Lorange, 2005). Marsoft, a consulting company in the shipping
business, advises companies how they should employ their vessels based on which part of the cycle
the shipping industry is in (Lorange, 2005). The key is to employ vessels under spot contracts when
one expects a market upswing, and fixing the vessels on longer-term charters when the market is at its
peak (Lorange, 2005). In practice, not including an entire cycle might bias results because fleet
employment could be affected by the stage of the cycle that is overweighed in the period under
In all, ship owners must take vessel size and the shipping cycle into consideration in the
decision-making process on contract type. Because the spot market entails more risk than the time
charter market, fleet employment decisions determine the operational risk for a dry bulk firm, which is
part of its total risk strategy.
Financial risk
In addition to operational risk, a ship owner faces financial risk in the form of costs of financial
distress. Costs of financial distress are an indirect result of leverage (Opler & Titman, 1994). Opler &
Titman (1994) find that high leverage (deciles 8-10, 10 being the most highly levered companies)
results in a 26,3% lower sales growth performance compared to less levered companies in an industry
(deciles 1-7) during periods of negative sales growth and declining stock performance (a recession),
indicating that financial distress and thus leverage is costly. It also forces companies to fulfil their
obligation to pay interest on the loans. The shipping industry is highly levered, exhibiting a mean book
leverage ratio of 41% (book value of short- and long-term debt over book value of assets) as opposed
to 25% for firms across all industries in the G7 countries covered in Compustat Global (Drobetz et al.,
2013). In order to understand the high leverage ratios and the consequent financial risk in the shipping
industry, three theories on leverage and their relevance to the dry bulk market will be described: the
trade-off theory (Kraus & Litzenberger, 1973), the agency theory (Jensen & Meckling, 1976) and the
pecking order hypothesis (Myers & Majluf, 1984). These theories might explain the leverage-related
decisions in the shipping industry.
The tradeoff theory considers the optimal debt ratio of a firm as being determined by the gains
of borrowing, the tax-shield, and the costs of adjustment and financial distress caused by the
compulsory interest payments. For the dry bulk shipping industry, however, the potential benefits of a
tax shield are negligible (PwC, 2009), since shipping companies enjoy special tax incentives. Of the
twelve companies in this sample, ten are registered on the Marshall Islands, one company is registered
in Panama and one company is registered in Liberia. In Panama, income from shipping activities is
are managed in Panama. As a result, there is a tax free environment for shipping companies registered
in Panama. In Liberia and the Marshall Islands, a shipping company registered in either one of these
countries, can register its vessels there as well and only tonnage tax (port taxes) and register fees are
due. No corporate tax has to be paid (PwC, 2009).
Secondly, the agency theory regards debt as disciplinary tool for managers in order to mitigate opportunistic behavior. When a firms’ free cash flow is high, this might result in adverse selection and
moral hazard. Furthermore, shareholders can use debt overhang to compromise debt holders by
making suboptimal investment decisions (Myers, 1977). The role of debt is thus primarily to motivate
managers, across all industries (Jensen, 1986). Since the dry bulk industry is so highly levered, it is to
be expected the disciplinary aspect of debt will ensure managers to act in line with a long-term
perspective.
Thirdly, the financing pecking order constitutes that firms will always prefer internal financing
for their investment opportunities. 75% of external ship financing is traditionally covered by bank
loans (ABN Amro, 2011). Theoretically, the pecking order arises because issuing equity is a signal of
bad expectations to shareholders. In shipping, an increased appetite for debt and a risky environment
for investors might distort the effects of the pecking order. Firstly, the secured debt hypothesis might
explain the appetite for debt issuance. Scott (1977) formulated this secured debt hypothesis, stating
firms can borrow funds at lower interest rates if their debt is secured with tangible assets. The results
of Bradley et al. (1984), who conducted research on 821 firms across 25 industries between 1962 and
1981, support this hypothesis by suggesting that firms investing in tangible assets generate high levels
of depreciation and tend to exhibit higher leverage. Since the shipping industry displays significantly
higher tangibility ratios, 63% as opposed to mean tangibility ratios of 34% and 28,9% across all
industries of US and G7 firms (Drobetz et al. 2013), it is likely the lower costs of debt for dry bulk
companies contribute to the high degree of debt financing. Secondly, the cyclical nature of the
shipping industry, where higher ship earnings spark shipping investment and new-building prices, but
predict low future returns due to excess supply of transportation (Greenwood & Hanson, 2013) does
shipping IPO’s from 1984-2007 and find that investors buying immediately after listing and hold the
stock for three years will make a loss of 15,72%.
Finally, the market timing theory suggests that when stock prices have risen, firms are more
likely to issue stock than when stock prices have declined (Myers, 1984). Baker and Wurgler (2002) argue firms generally don’t care which method of financing they employ, but rather employ the most
valuable method as valued by financial markets. However, the average shipping company does not
have easy access to the capital market. Drobetz et al. (2013) report only 22% of the companies in their
sample are unconstrained in their debt capacity, which is defined as companies that have a higher
rating than the median G71 firm based on 115 shipping companies during the period between 1992 and
2010. Denis and Mihov (2003) state that 75% of the companies that issue public debt have an
investment-grade rating (rated BBB/Baa or higher by S&P and Moody’s respectively), while bank
borrowers such as dry bulk shipping companies only have an investment-grade rating in 20% of the
cases. They base their findings on 2,338 new debt financings in 1995 and 1996. Supply-side
constraints make it harder for dry bulk companies to time the market efficiently.
Even though the tax benefits of debt are negligible for dry bulk companies, high leverage
ratios persist. 75% of external financing is covered by bank loans. Because of the proportion of bank
debt, the high leverage ratios and the poor returns on IPO’s in shipping, the theoretical pecking order
is affected. This makes it harder to identify the real driver behind the capital structure-related decisions
in the dry bulk market from a pecking order perspective. In addition, supply-side constraints make it
difficult for dry bulk companies to time the market based on stock prices or bond prices. In all, the
tradeoff theory and the pecking order hypothesis best explain the rationale behind leverage-related
decisions in the shipping industry. Using these capital structure theories, it is possible to determine the
proper signs of the capital structure determinants influencing the leverage, because increasing leverage
means increasing financial risk, constituting the second part of the total risk strategy of a dry bulk
2.2 Empirics
This paragraph will identify the capital structure determinants most relevant to the dry bulk shipping
industry in order to isolate the effect of spot market exposure on leverage in the market. This will lead
to a hypothesis on the relation between operational risk and leverage.
Kayo and Kimura (2010) look at time-, industry- , country-, and firm-level determinants of
capital structure and define a hierarchical relationship between these levels, where
industry/country-level determinants are ranked higher because they may affect lower industry/country-levels (firm). Their sample
comprises 127.340 observations across 40 countries from 1997 to 2007. The firm-level determinants
include growth opportunities, profitability, size and tangibility. On the industry level, the influence of
industry dynamism, munificence and industry concentration are measured and this level accounts for
11,6% of the variance in firm leverage.
Country-level determinants are hard to interpret for the dry bulk market because the countries
they reside in offer tax benefits for shipping companies. Moreover, the country effect Kayo and
Kimura (2010) report is relatively low, accounting for only 3,3% of the variance in firm leverage,
while the variance in leverage over time accounts for 35,6% and intrinsic firm characteristics account
for 42,5% of total variance in firm leverage.
On firm level, tangibility is measured by the ratio of property, plant and equipment to total
assets. Kayo and Kimura (2010), and Rajan and Zingales (1995), who conducted research on all
consolidated firms in the G7 countries from 1987-1991, report a positive influence of tangibility on
leverage. Drobetz et al. (2013) underscore this result for the shipping industry in particular, and
observe tangibility to be the most important driver of capital structure in the shipping industry by
examining the elasticity coefficients. They find that a 1% increase in the proportion of fixed to total
assets results in a rise of 0,46% in leverage, the highest value for all determinants they examine.
Ultimately, tangibility boils down to market tangibility, which means that the balance sheet value of a
fixed asset doesn’t necessarily match the commercial value in liquidation. The price creditors wish to
receive for a vessel in liquidation might not be attainable in the second-hand market, since a
Size, as measured by taking the natural logarithm of total assets, is a determinant of capital
structure. The direction of the sign of this determinant is ambiguous. The pecking-order hypothesis
predicts a negative sign, since the degree of information asymmetry between management and
investors is deemed smaller in large firms because of increasing transparency (Myers & Majluf, 1984).
Large firms experience less negative consequences from an equity issue than do relatively smaller
firms, resulting in lower leverage for large firms. The trade-off theory, on the other hand, predicts the
sign to be positive. Large firms tend to be more diversified and will have a lower probability of default
of 13,2% for long-term book leverage, based on a sample of 469 firms from 1974 to 1982 (Titman &
Wessels, 1988). Driven by decreasing bankruptcy probability, large firms are able to take on higher
levels of debt. Rajan and Zingales (1995) document a significant positive relation in the United States,
supporting the trade-off theory.
Profitability, measured by the ratio of operating income before depreciation to total assets, is
proven to influence firm leverage by giving a firm the opportunity to finance investments with
retained earnings. The pecking order states firms will use retained earnings before issuing external
equity, resulting in a negative relation between profitability and leverage. A negative relation is
reported by Rajan and Zingales (1995), who use EBITDA divided by assets, and by Kayo and Kimura
(2010), who use operating income to total assets. Volatility in freight rates and freight risk of the
industry, which results in income volatility for the ship owner, affects profitability in the shipping
industry (Kavussanos, 2010). These freight rates are dependent on the world economy through
demand for transportation. This results in a risky environment for managers of shipping companies as
well as investor (Albertijn et al., 2011). Profitability is identified by Drobetz et al. (2013) as being an
important capital structure driver in the shipping industry, with a reported elasticity coefficient of
0,13%.
In addition to tangibility, size and profitability, two other control variables, dividend-paying
status and the market to book ratio, are included. These two determinants have been extensively
investigated in capital structure literature and have been identified as influential in the shipping
Zingales, 1995). Together, the control variables will be used to investigate the effect of spot exposure
on leverage.
Since leverage is accompanied by costs of financial distress and interest obligations, this poses
financial risk to the ship owner (Merikas et al., 2011). The financial risk a company is willing to take
depends in part on the level of debt it wishes to attain and at what financing terms they are able to
receive loans. Kavussanos and Alizadeh (2002) point out these financing terms are more favorable for
ship owners who employ their ships under time charter contracts, due to the perceived lower
probability of loan default by banks. Furthermore, the higher volatility of spot rates increases the
volatility of returns for ship owners, possibly affecting their ability to fulfil their financial obligations
(Merikas et al., 2011). Less favorable financing terms on bank loans in combination with more volatile
cash flows, the risk of unemployment, relocation costs and bunker costs associated with spot exposure
lead to believe high operational risk will be balanced with lower financial risk. Operational risk is
operationalized as the percentage of ships a company deploys on the spot market divided by the
number of ships it operates during a calendar year. Due to the aforementioned risks associated with the
spot market and based on the findings of Merikas et al. (2011) and Drobetz et al. (2013), it is
hypothesized that there exists a tradeoff between spot market exposure and firm leverage in the dry
bulk shipping industry, resulting in a negative relation.
3 DATA DESCRIPTION & METHODOLOGY
3.1 Methodology
In order to establish the relation between the degree of operational risk a dry bulk shipping company is
willing to take considering its debt level, a model specification is tested. This model is a fixed effects
regression:
Where Lit is leverage for firm I at time t, α is an intercept, the β’s represent the coefficients
corresponding to a variable: X1 is the charter coverage ratio, X2 is asset tangibility, X3 is profitability,
X4 is asset size and X5 is market to book ratio. Finally, X6 is a dummy-variable reflecting whether a
firm pays dividends in a certain year. The last two effects represent firm fixed effects (Cf) and year
effects (Ct) and εit is the error term. The reason for these fixed effects arises from research by Lemmon
et al. (2008), who find that leverage is not fully explained by the capital structure determinants of
Rajan and Zingales (1995), Frank and Goyal (2009) and Faulkender and Petersen (2006) and draw the
conclusion there must be unobserved firm-specific factors at play, resulting in heterogeneous
intercepts. This thesis follows Drobetz et al. (2013), who base their methodology on Petersen (2009)
and so standard errors are clustered at the firm level to account for heteroscedasticity using robust
standard errors and autocorrelation of the error terms. The year fixed effects control for time events
affecting all firms in a given year, removing cross-sectional correlation (Petersen, 2009).
3.2 Variables
Financial risk arises when a firm is levered because of costs of financial distress and compulsory
interest payments. The dependent variable in this thesis is the debt to assets ratio, commonly referred
to as leverage. Furthermore, a distinction between the book and market values of assets is made. Frank
and Goyal (2009), who study a sample of non-financial US firms from 1950 to 2003, define leverage
as the ratio of short- and long term debt to assets, based on book and market values. In this thesis, the
definition by Frank and Goyal (2009) will be followed because Drobetz et al., (2013) use the same
measures of book and market leverage in a broader investigation of the shipping market.
The measure of operational risk is obtained by dividing the number of vessels being employed
on the spot market by the total number of ships a company has in its fleet. This measure reflects a company’s strategic course and to what extent a company is exposed to the risk of unemployment or
Vessel employment reporting by the companies in this sample is taken from their annual
filings by the SEC3. The spot exposure serves as a proxy for operational risk. It is hypothesized the
charter coverage will have a negative impact on the debt to assets ratio.
To isolate the effect of the charter coverage ratio, control variables are also included to control
for capital structure decisions not driven by the spot charter exposure. Tangibility is the ratio of
property, plants and equipment to total assets following Drobetz et al. (2013), Frank and Goyal (2009)
and Rajan and Zingales (1995) and is predicted to have a positive influence on debt. Profitability is
measured by dividing operating income before depreciation by total assets and is predicted to
negatively influence debt; firm size is measured by the natural logarithm of total assets and is expected
to positively influence the level of debt (Frank & Goyal, 2009, Rajan & Zingales, 1995). A commonly
used proxy for growth opportunities is the market to book ratio. This variable is included and
measured as the ratio of the market value of assets divided by the book value of assets; its sign is
predicted to be negative (Frank & Goyal, 2009, Rajan & Zingales, 1995). The definitions of the
variables measuring profitability, firm size and market to book ratio are also calculated based on the
papers by Frank and Goyal (2009) and Drobetz et al. (2013). Finally, a dummy-variable will be used
to distinguish between dividend-paying firms and firms that do not pay dividends. This dummy is set
equal to one if the firm pays dividends in a given year, and zero otherwise. Dividend-paying firms
tend to exhibit lower levels of leverage due to this obligation, resulting in a negative coefficient for
this variable (Frank & Goyal, 2009)
3
Although this is the most accurate information on charter coverage possible, the distribution of vessel employment in these annual filings might not always represent the accurate distribution during the entire year. For example, suppose Dryships Inc. states it operates 12,5% of itsvessels on the spot market in its SEC filing, as filed on the 3rd of March 2012. This does not mean it will hold on to this particular distribution of employment all year long.
3.3 Sample
Table 1: Average spot market exposure and average book- and market leverage per firm, ordered alphabetically. Company Firm-year
observations
Average chart cov ratio Average book leverage Average mkt leverage Baltic Trading Ltd. 2010-2012 (3) 1 0.27 0.46 Danaos Corp. 2008-2012 (5) 0 0.76 0.77
Diana Shipping Inc. 2008-2012 (5) 0 0.24 0.29 Dryships Inc. 2007-2012 (6) 0.22 0.49 0.57 Euroseas Ltd. 2008-2012 (5) 0.02 0.23 0.41 Excel Maritime Carriers 2011-2012 (2) 0.38 0.38 0.73 Freeseas Inc. 2009-2012 (4) 1 0.60 0.80 Genco Shipping & Trading 2008-2012 (5) 0.46 0.56 0.71 Global Ship Lease Inc. 2008-2012 (5) 0 0.53 0.66 Globus Maritime Ltd. 2008-2012 (5) 0.46 0.48 0.70 Safe Bulkers 2007-2012 (6) 0.14 0.73 0.51 Star Bulk Carriers 2008-2012 (5) 0.20 0.39 0.66
Total 56
The sample consists of 12 dry bulk shipping companies listed on the NYSE or on the Nasdaq, who are
all covered in the Compustat North America database. These firms employ dry bulk vessels only,
which makes the companies comparable. The period in which the companies are under examination is
between 2007 and 2012. The data are retrieved on an annual basis and since all companies are listed
on a U.S. stock exchange, the data is in US dollars. Data regarding debt levels, fixed assets, total
assets, operating income before depreciation, shareholders’ equity and other financials were taken
from the Compustat North America database. The information on the spot charter coverage was
retrieved manually by examining annual filings by the SEC. Outliers were removed from the data
manually by winsorizing the upper and lower one percentile, excluding two data points. The data is
unbalanced, meaning not all firms have the same number of observations, due to firms going public
later than 2007. In addition, some firm-year observations were lost because of inadequate reporting on
fleet employment, meaning a firm did not release exact data on the distribution of time charter or spot
market employment of its vessels. This leads to a sample of 56 firm-year observations for the dry bulk
shipping industry (table 1). Table 1 gives an overview of the average spot charter coverage ratios for
their vessels on the time charter market, and two firms employ all their vessels on the spot market.
Two firms, Baltic Trading Ltd. and Genco Shipping & Trading, have hybrid contracts which they call ‘spot market-related time charters’. Their vessels are employed under medium term contracts linked to
the Baltic Indices (BI), instead of an agreed upon price per day for the entire period of the contract.
This results in returns moving up and down with the Baltic Dry Index and for that reason it is
considered a spot market oriented firm. A potential reason for these contracts is that they can take the
place of Forward Freight Agreements, which can be used to hedge freight risk by settling a contract on
the difference between the contracted price and the average price of the Baltic Index (Kavussanos &
Visvikis, 2006). By linking a time charter contract to the Baltic Index directly, the need for Forward
Freight Agreements might diminish. These hybrid contracts, however, make it difficult to differentiate
between spot- and time charter agreements and in a robustness test they will be regarded as time
charter contracts.
3.4 Descriptive Statistics
Table 2: Descriptive statistics
Variable Mean SD Median Min Max
Charter coverage 0.278 (0.35) 0.110 0 1 Fixed/Total assets 0.838 (0.11) 0.871 0.395 0.975 Profitability 0.112 (0.08) 0.089 -0.086 0.363 Log (Assets) 6.868 (1.14) 6.826 4.739 9.091 Market to Book 0.827 (0.30) 0.765 0.395 1.829 Dividends 0.518 (0.50) 1 0 1 Book leverage 0.483 (0.19) 0.493 0.164 0.971 Market leverage 0.595 (0.18) 0.626 0.188 0.866 Obs. 56
The descriptive statistics show the number of observations (Obs.), the mean, the standard deviation (SD, between brackets), the median as well as the minimum (Min) and the maximum (Max) based on the reported financials in the annual reports (in US$). The sample consists of 12 listed shipping companies during the period from 2007 to 2013. Data are annual and obtained from the Compustat North America database.
Table 2 provides descriptive statistics on the variables that are used. The number of ships a dry bulk
firm employs on the spot market differs between the mean (0.278) and the median (0.110), because
coverage ratio of zero (table 1). This could pertain to commercial decisions made by senior
management to avoid risk when it is deemed unnecessary, since long-term contracts were made for all
vessels under employment. The downside to this risk-averse strategy is that they fail to capture the
upside potential the spot market has to offer (Kavussanos, 2010).
The results are compared to Frank and Goyal (2009) and Drobetz et al. (2013). The difference
between the shipping industry and other industries is best described by tangibility. In this sample the
mean is 0.84 as opposed to 0.34 in the US sample. This difference is explained by the substantially
higher level of fixed assets due to the vessels dry bulk firms operate. In terms of profitability, the dry
bulk shipping sector seems to be slightly higher than the US firms, 11.02% against 9.71%. Drobetz et
al. (2013) report a profitability coefficient of 10,4%. The same goes for shipping companies in terms
of size. In the sample of US firms an average size of 4.58 is found, while this sample reports a size
average of 6.87. Drobetz et al. (2013), report a 6.48 size average. In terms of the ratio of market to
book value of assets, these results show similarity with the findings by Drobetz et al. (2013), reporting
substantially lower market to book values (1.00) than the US sample (1.79). Concerning the market to
book ratio, the mean of 0.82 is lower than the value Drobetz et al. (2013) report (1.165) and
substantially lower than across the US firms, reporting a value of 1.76 (Frank & Goyal, 2009). In
terms of dividends paid, there seems to be a difference between the findings in this sample (0.518) and
the sample Drobetz et al. (2013) use, who report an average of 0.778. This could be due to the period
under examination in this sample, a period of economic downturn, which has had its effect on the
ability for dry bulk companies to pay dividends (Campello et al., 2010). Campello et al. (2010)
examined the effects of the crisis on 1050 non-financial firms from the US, Europe and Asia and
found dividend-payments to have decreased 14,2% for constrained (non-investment grade) firms.
Finally, these statistics show a higher book leverage ratio than do average US industrial firms, 48.3%
for dry bulk shipping firms as opposed to 29%. Market leverage is found to be 59,5%, against 28% for
the average US firm. This underscores the principle that fixed assets can be used as collateral and
gives firms the opportunity to attain higher levels of leverage. In figure 6, book leverage and charter
Figure 4: Book leverage and spot charter coverage by company over 2007-2012. The y-axis shows the leverage ratios and charter coverage ratios in percentage shares, while the x-axis shows the time.
4 RESULTS
4.1 Standard leverage regressions
Table 3A: Standard leverage regressions with book leverage as the dependent variable.
1 2 3 4 5 6 7 8
Dependent variable: Book leverage Charter coverage 0.002
(0.074) 0.042 (0.091) 0.062 (0.094) 0.041 (0.081) 0.000 (0.010) 0.152** (0.076) 0.084 (0.075) 0.108 (0.079) Fixed over Total
Assets -0.141 (0.152) 0.186 (0.247) -0.129 (0.150) 0.089 (0.133) 0.126 (0.221) 0.124 (0.132) Return on Assets 0.155 (0.210) 0.402 (0.452) 0.032 (0.295) 0.046 (0.173) -0.084 (0.407) 0.052 (0.233) Market to Book 0.043 (0.063) 0.394** (0.090) 0.027 (0.06) Log (Assets) -0.232*** (0.054) 0.001 (0.026) -0.278*** (0.067) Dividend-payer -0.054 (0.038) -0.075 (0.055) -0.077** (0.038) Constant 0.481*** 0.224** 0.335* 0.413* 0.452** 1.430*** 0.131 1.632*** (0.033) (0.105) (0.184) (0.277) (0.192) (0.332) (0.283) (0.036) Firm fixed effects No Yes Yes No Yes Yes No Yes Year fixed effects No No No Yes Yes No Yes Yes
Observations 56 56 56 56 56 56 56 56
R² 0.00 0.81 0.82 0.09 0.85 0.90 0.42 0.93 Adjusted R² -0.02 0.76 0.76 -0.06 0.78 0.86 0.27 0.87
Table 3B: Standard leverage regressions with market leverage as the dependent variable.
This table shows the results of standard leverage regressions using a sample of twelve listed shipping companies active in the dry bulk sector during the period from 2007 to 2012. Firm and year fixed effects indicate whether firm firm-level and calendar year effects are included in the specification. Standard errors are in parentheses.
*** Statistical significance at 1% level. ** Statistical significance at 5% level. * Statistical significance at 10% level.
1 2 3 4 5 6 7 8
Dependent variable: Market leverage Charter coverage 0.176**
(0.067) 0.327*** (0.099) 0.244** (0.105) 0.086 (0.070) 0.070 (0.088) 0.196** (0.080) 0.103 (0.077) 0.105 (0.083) Fixed over Total
Assets -0.094 (0.233) 0.450* (0.255) -0.070 (0.178) 0.181 (0.215) 0.454* (0.271) 0.053 (0.204) Return on Assets -0.558** (0.240) -0.644* (0.387) -0.236 (0.234) 0.024 (0.217) -0.669 (0.455) 0.000 (0.234) Market to Book -0.512*** (0.090) 0.064 (0.104) -0.259** (0.115) Log (Assets) -0.305*** (0.077) 0.010 (0.026) -0.160* (0.086) Dividend-payer -0.023 (0.041) -0.060 (0.059) 0.029 (0.046) Constant 0.548*** (0.029) 0.129* (0.113) 0.321* (0.228) 0.088 (0.265) 0.159 (0.178) 2.238*** (0.450) -0.038 (0.352) 1.130** (0.54) Firm fixed effects No Yes Yes No Yes Yes No Yes Year fixed effects No No No Yes Yes No Yes Yes
Observations 53 53 53 53 53 53 53 53
R² 0.12 0.78 0.81 0.34 0.91 0.90 0.39 0.93 Adjusted R² 0.11 0.72 0.75 0.22 0.86 0.86 0.22 0.88
The results are shown in table 3A and 3B. Table 3A displays the results when book leverage, defined
as debt over the book value of assets, is the dependent variable. Table 3B shows the results for market
leverage, defined as debt over the market value of assets. In the single regression, the charter coverage
coefficient does not significantly differ from zero, implying no effect on book leverage. The effect on
market leverage is positive and significant, especially when firm fixed effects are included (column 2).
This is contrary to the predicted negative sign of this operational risk measure. Furthermore, a positive
effect of spot market exposure on financial risk is counterintuitive because it increases the costs of
financial distress by increasing the volatility of cash flows (Kavussanos, 2010). However, because the
shipping industry went into a recession in 2009 and most data points are from 2009 and onwards,
results are potentially biased upwards due to a preference for spot contracts from a shipping cycle
management perspective. In addition, the charter coverage might not proxy well enough for
operational risk due to a changing fleet employment during the year. The positive effect persists when
more control variables are introduced and heterogeneous intercepts are allowed, as can be seen in
column 6 (for both book and market leverage). The difference between including firm fixed effects in
the specification and leaving them out results in a substantial decrease in explanatory power, which is
reflected by the adjusted R², differing from 0.87 and 0.88 as opposed to 0.27 and 0.22 in columns 8
and 7 respectively. This reveals the importance of controlling for unobserved variation at firm level, as
shown by Lemmon et al. (2008); implying capital structure is (to a certain extent) driven by a
time-invariant component. On the other hand, coefficients such as charter coverage, market to book and
size decrease in significance compared to column 6 (for market leverage). The latter effect is also
described by Drobetz et al. (2013), who report tangibility, size and other variables decrease in
magnitude and significance due to the inclusion of firm fixed effects.
Tangibility has a positive influence on book leverage in four out of the six regressions it is
included in, although not significant. In columns 4 and 7 it is significant at the 10% level when market
leverage is the dependent variable. This was expected, seeing as how fixed assets can provide
collateral for loans, which increases debt capacity. Statistical insignificance when firm fixed effects
profitability varies between book and market leverage. In the case of book leverage it is almost always
positive, although not significant, while in the case of market leverage, it is negatively related to debt
and significant in columns 3 and 4. A negative profitability supports the pecking order theory,
assuming managers prefer to use internal funds over external funds (Myers & Majluf, 1984). A
positive profitability would conjecture the tradeoff hypothesis, because low profitability may increase
the risk of bankruptcy, which would lead to lower leverage to reduce that risk (Fama & French, 2002).
Drobetz et al. (2013) report a negative profitability coefficient for the shipping industry, which is also
more significant for market leverage. The market to book variable behaves similarly, from being
positive and significant for book leverage (column 7) to being negative and significant with respect to
market leverage (columns 6 and 8). A negative relationship is predicted by the pecking order
hypothesis, because firms with high market to book ratios will have higher costs of financial distress
due to lower asset values in liquidation (Myers, 1984, Rajan & Zingales 1995). Since the firms in this
sample exhibited a relatively low mean value of the market to book ratio (0.82, paragraph 3.3), a
negative relation would point toward lower costs of financial distress, which is likely due to the high
level of tangibles. The size of dry bulk firms is negatively related to their debt levels and is significant
at the 1% level when firm fixed effects are included (for both book and market leverage). In the case
of dry bulk shipping, it might be easier to understand by taking the supply side of equity or public debt
into account. As mentioned before, Drobetz et al. (2013) reported only 22% of their sample could be
classified as unconstrained, implying only the large companies are able to attract public equity or debt.
The negative coefficient reported in table 3 supports this, because larger companies apparently issue
less debt. Finally, dividend-paying firms tend to exhibit lower levels of leverage, although the
coefficient is significant in just one regression (column 8, book leverage). This is in line with the
results from Frank and Goyal (2009) and Drobetz et al. (2013).
4.2 Robustness checks
This paragraph is included to provide robustness checks for the variables charter coverage and
spot charters are tabulated in the appendix as table A1. The results are similar to the regressions above.
There is no evidence these contracts affect the sign of the charter coverage variable.
5 DISCUSSION
The positive relation between spot market exposure and leverage can be attributable to the shipping
cycle. The industry went into a recession in 2009 (Drobetz et al, 2013), which made it relatively less
attractive to secure vessels under time charter contracts at very low freight rates, since ship owners
might expect an upswing. The trough of the cycle is overweighed in the data and because shipping
companies tend to deleverage 9,3% slower during recessions (Drobetz et al., 2013), which might lead
to a positive relation. However, the positive effect of spot market exposure on leverage contradicts the
hypothesized negative relation. Three factors possibly influencing the results are identified. Firstly,
the difference between outcomes in Merikas et al. (2011), reporting a negative correlation between
operational risk and financial risk, and this thesis, potentially depends on how the operationalization of
the operational risk measure. In contrast with Merikas et al. (2011), this thesis regards the employment
of a portfolio of vessels a dry bulk shipping company has annually. The difference potentially arises
from new contracts made during 2012 and 2013, which were not taken into account by Merikas et al.
(2011). For example, reported time charter coverage for 2013 may have been only 20% in the third
quarterly filing of 2010, but by securing new contracts in 2012, the actual coverage level for 2013 is
most likely higher than 20%. This influences the correlation coefficient. Furthermore, dry bulk
companies expand their fleet on an annual basis, also distorting the percentage. These two reasons
combined lead to an inaccurate measure, as their reported charter coverage levels do not represent
actual coverage levels.
The data on spot market exposure in this thesis is also manually accumulated. There is no
reporting standard for dry bulk companies as to their fleet employment. To exemplify, Star Bulk
Carriers Inc. has fourteen ships under employment as of March 18, 2013. In its 2012 annual filing,
considered to be employed in the spot market, due to the short duration of their charter agreements.
Genco Shipping & Trading ltd., on the other hand, does not make such a statement in its 2012 annual
filing, even though their charters are in some cases of the same length. Because a reporting standard on
fleet employment is absent, it is thus difficult to determine the actual operational risk. The
approximation used to calculate spot market exposure is not waterproof.
Finally, there might be other ways dry bulk companies try to reduce the volatility of their
operational cash flows. Firstly, by employing their most flexible vessels on the spot market.
Kavussanos (2010) has shown volatility of returns is higher for larger size vessels due to lower
operational flexibility, meaning in practice that larger vessels, such as Capesize vessels, can sail only a
few routes and are able to access less ports than smaller-sized ships. To examine if this affected the
decision on which market a ship was employed, the number of ships employed in the spot market was
investigated. For all firm-year observations across all twelve companies, a total number of 195 vessels
were employed in the spot market. 32% of these vessels fall in the Handysize/Handymax segment
(smaller-sized), 56% falls in the Panamax segment (middle-sized), while only 12% of the vessels
employed under spot charters are Capesize (large-sized) vessels. Four companies did not operate
Handysize vessels, meaning Panamax vessels were the most flexible ships they operated. This shows
that firms try to avoid employing their most volatile ships on the more volatile spot market, thereby
decreasing operational risk. Secondly, freight derivatives are used to hedge freight risk, which pertains
to the volatility in freight risk. Both spot charter and time charter contracts can be hedged and there
exist OTC markets as well as futures markets. Forward Freight Agreements can be sold to compensate
for the loss in the freight income through a gain in the forward position. In addition, options are used
to hedge downside freight rate risk. For instance, a ship owner who wants to protect his income
against a decline in freight rates can buy put options; these will compensate the income loss in the
physical market by expiring in-the-money. Together, these three factors could explain why the
distribution of time charter and spot charter contracts does not have the predicted effect on leverage.
When assessing the external validity of this thesis, a caveat of the dry bulk shipping market is
and a large number of small firms. The difference between listed firms and family-owned firms is
described by Slack and Frémont (2009), who state private companies are more flexible in changing
their strategic course and in addition are able to carry out big decisions more swiftly than listed
companies with dispersed ownership. Furthermore, Brav (2009) found that private firms have a 10%
higher (32,7% vs 22,7%) debt ratio than listed firms do across all industries in a large UK sample.
However, this was not investigated for the shipping industry, making it difficult to apply the results of
this thesis to private dry bulk firms.
6 CONCLUSION
This thesis examines the effect of operational risk, as measured by the spot market exposure a dry bulk
company exhibits, on leverage. The risks associated with spot charters lead to believe a company will
balance additional operational risk by decreasing its leverage ratio. The results found, however, are not
in line with the hypothesized outcomes, not suggesting a tradeoff between the distribution of a company’s vessels under spot and time charter contracts and leverage. Ship owners’ expectations on
market recovery might have induced them to employ more vessels under spot contracts from a cycle
management perspective. In addition, propositions were made to explain the absence of the tradeoff
between spot exposure and leverage. For one, companies employing ships on the spot market mostly
do so with their most operationally flexible ships. This flexibility makes it easier to find a charterer
and volatility in returns is lower for these ships, thereby smoothening income from operations. In
addition, the use of financial derivatives such as futures, forward contracts and options help to
stabilize a firms’ cash flow. These issues are both related to operational risk. One thing which became
clear is that reporting on fleet employment must be standardized, thereby making it easier to compare
commercial strategies and the degree of risk such strategies entail. Only then will a deeper
understanding between operational risk and financial risk in the dry bulk shipping industry be
possible, providing opportunities for future research in this area and thereby making the cyclical
APPENDIX
Table A1: Standard leverage regressions, regarding hybrid charter contracts as spot contracts.
1 2 3 4 5 6 7 8
Dependent variable: Book leverage Charter coverage 0.083
(0.088) 0.078 ( 0.100) 0.083 (0.101) 0.149 (0.095) 0.054 (0.101) 0.142 (0.085) 0.182** (0.778) 0.032 (0.078) Fixed over Total
Assets -0.137 (0.151) 0.293 (0.252) -0.123 (0.150) -0.016 (0.218) 0.363 (0.225) -0.170 (0.192) Return on Assets 0.138 (0.204) 0.491 (0.442) 0.030 (0.295) 0.074 (0.222) -0.717* (0.379) 0.053 (0.221) Market to Book 0.031 (0.093) 0.582*** (0.087) 0.272** (0.111) Log (Assets) -0.256*** (0.080) 0.027 (0.023) -0.117 (0.082) Dividend-payer -0.047 (0.042) -0.021 (0.050) (0.006) (0.045) Constant 0.466 0.244 0.371 0.300 0.439 1.847 -0.818 0.567 (0.031) (0.059) (0.160) (0.281) (0.171) (0.461) (0.311) (0.526) Firm fixed effects No Yes Yes No Yes Yes No Yes Year fixed effects No No No Yes Yes No Yes Yes
Observations 56 56 56 516 56 56 56 56
R² 0.02 0.82 0.82 0.13 0.86 0.91 0.60 0.94 Adjusted R² 0.00 0.76 0.76 -0.01 0.78 0.86 0.50 0.90 This table shows the results of standard leverage regressions using a sample of twelve listed shipping companies active in the dry bulk sector during the period from 2007 to 2012. Firm and year fixed effects indicate whether firm firm-level and calendar year effects are included in the specification.
*** Statistical significance at 1% level. ** Statistical significance at 5% level. * Statistical significance at 10% level.
REFERENCES
Adland, R., & Jia, H. (2008). Charter market default risk: A conceptual approach. Transporation Research Part E: Logistics and Transportation Review 44(1), 152-163.
Albertijn, S., Bessler, W., Drobetz, & W. (2011). Financing shipping companies and shipping operations: a risk-management perspective. Journal of Applied Corporate Finance, 70-82. Alizadeh, A., & Nomikos, N. (2010). An overview of the dry bulk shipping industry. In C. T.
Grammenos, The Handbook of Maritime Economics and Business (pp. 319-353). London: Lloyd's List.
Alizadeh, A., & Nomikos, N. (2011). Dynamics of the term structure and volatility of shipping rates. Journal of Transport Economics and Policty (JTEP), 45(1), 105-128.
ABN Amro. (2011). Shipping Finance and Investment: Current Trends in Ship Finance, Istanbul: 3rd Mare Forum in Ship Finance.
Baker, M., & Wurgler, J. (2002). Market timing and capital structure. Journal of Finance 57, 1-32.
Baltic Dry Index, Methodology: How Baltic market information is pronounced. Retrieved April 28, 2014, from http://www.balticexchange.com/market-information/methodology.shtml
Bradley, M., Jarrell, G., & Han Kim, E. (1983). On the Existence of an Optimal Capital Structure: Theory and Evidence. The Journal of Finance, 857-878.
Brav, O. (2009). Access to capital, capital structure, and the funding of the firm. Journal of Finance, 263-308.
Brown, J., Fazzari, S., & Petersen, B. (2009). Financing innovation and growth: Cash flow, external equity and the 1990s R&D boom. The Journal of Finance, 151-185.
Campello, M., Graham, J., & Harvey, C. (2010). The real effects of financial constraints: Evidence from a financial crisis. Journalof Financial Economics, 97(3), 470-487.
Denis, D., & Mihov, V. (2003). The choice among bank debt, non-bank private debt, and public debt finance from new corporate borrowings. Journal of Financial Economics, 3-28.
Drobetz, W., Gounopoulos, D., Merikas, A., & Schröder, H. (2013). Capital structure decisions of globally-listed shipping companies. Transportation Research Part E (52), 49-76.
Fama, E., & French, K. (2002). Testing trade-off and pecking order predictions about dividends and debt. The Review of Financial Studies(15), 1-33.
Faulkender, M., & Petersen, M. (2006). Does the source of capital affect capital structure? Review of Financial Studies, 45-79.
Ferri, M., & Jones, W. (1979). Determinants of financial structure: A new methodological approach. The Journal of Finance, 631-644.
Fischer, E., Heinkel, R., & Zechner, J. (1989). Dynamic Capital Structure Choice: Theory and Tests. The Journal of Finance, 19-40.
Frank, M., & Goyal, V. (2009). Capital structure decisions: which factors are reliably important? Financial Management, 1-37.
Gratsos, G., Thanopoulou, H., & Veenstra, A. (2012). Dry Bulk Shipping. In W. Talley, The Blackwell Companion to Maritime Economics (pp. 187-204). West Sussex: Blackwell Publishing Ltd. Greenwood, R., & Hanson, S. (2013). Waves in Ship Prices and Investment. Not published, Harvard
Business School.
Hale, C., & Vanags, A. (1989). Spot and period rates in the dry bulk market: Some tests for the period 1980-1986. Journal of Transport Economics and Policy (JTEP), 281-291.
Jensen, M. (1986). Agency costs of free cash flow, corporate finance and takeovers. American Economic Review 76, 323-329.
Jensen, M., & Meckling, W. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 305-360.
Kavussanos, M. (1996). Comparisons of volatility in the dry-cargo ship sector: Spot versus time charters, and smaller versus larger vessels. Journal of Transport Economics and Policy, 67-82. Kavussanos, M. (2010). Business risk measurement and management in the cargo carrying sector of
the shipping industry - an update. In C. T. Grammenos, The Handbook of Maritime Economics and Business (pp. 709-743). London: Lloyd's List.
Kavussanos, M., & Alizadeh, A. (2002). The expectations hypothesis of the term structure and risk premiums in dry bulk shipping freight markets. Journal of Transport Economics and Policy, 267-304.
Kavussanos, M., & Visvikis, I. (2006). Shipping freight derivatives: a survey of recent evidence. Maritime Policy and Management, 33(3), 233-255.
Kayo, E., & Kimura, H. (2011). Hierarchical determinants of capital structure. Journal of Banking & Finance (35), 358-371.
Kraus, A., & Litzenberger, R. (1973). A State-Preference Model of Optimal Financial Leverage. The Journal of Finance, 28 (4), 911-922.
Lorange, P. (2005). Shipping Company Strategies: Global Management Under Turbulent Conditions. Oxford, UK: Elsevier.
Lorange, P., & Fjeldstad, O. (2012). New Business Modelsand Strategies in Shipping. In W. Talley, The Blackwell Companion to Maritime Economics (pp. 263-280). West Sussex: Blackwell Publishing Ltd.
Mankiw, N., & Miron, J. (1986). The changing behavior of the term structure of interest rates. The Quarterly Journal of Economics, 211-228.
Merikas, A., Gounopoulos, D., & Nounis, C. (2009). Global shipping IPOs performance. Maritime Policy & Management, 481-505.
Merikas, A., Sigalas, C., & Drobetz, W. (2011). The Shipping Corporate Risk Trade-Off Hypothesis. Marine Money.
Miller, M. (1977). Debt and Taxes. Journal of Finance, 261-275.
Modigliani, F., & Miller, H. (1958). The cost of capital, corporation finance and the theory of investment. American Economic Review, 271-297.
Myers, S. (1984). The Capital Structure Puzzle. The Journal of Finance, 575-585.
Myers, S. C., & Majluf, N. (1984). Corporate Financing and Investment Decisions, when firms have information investors do not have. Journal of Financial Economics, 187-221.
Opler, T., & Titman, S. (1994). Financial distress and corporate performance. Journal of Finance 49(3), 1015-1040.
Petersen, M. (2009). Estimating standard errors in fnancial panel data sets: comparing approaches. Review of Financial Studies, 435-480.
PricewaterhouseCoopers. (2009). Choosing a Profitable Course Around the Globe: Corporate Taxation of the Global Shipping Industry.
<http://www.pwc.com/en_GX/gx/transpor-tation-logistics/assets/choosing-profitable-course.pdf>.
Rajan, R., & Zingales, L. (1995). What Do We Know about Capital Structure? Some Evidence From International Data. The Journal of Finance, 1421-1460.
Scott, J. J. (1977). Bankruptcy, Secured Debt, and Optimal Capital Structure. The Journal of Finance, 1-19.
Slack, B., & Frémont, A. (2009). Fifty years of organisational change in container shipping: regional shift and the role of family firms. Geojournal, 23-34.
Stopford, M. (2009). Maritime Economics, third edition. New York: Taylor & Francis.
Strandenes, S. (2012). Maritime Freight Markets. In W. Talley, The Blackwell Companion to Maritime Economics (pp. 107-120). West Sussex: Blackwell Publishing Ltd.
Titman, S., & Wessels, R. (1988). The determinants of capital structure choice. The Journal of Finance, 1-19.
Veenstra, A. (1999). The term structure of ocean freight rates. Maritime Policy & Management, 26(3), 279-293.
Wright, R. (2008, November 27). Dry bulk shipping rates approach all-time low. Financial Times, accessed May 2, 2014.
