The Impact of Capital Structure of Electricity Genera- tion ... UETCL report further indicates that since 2003, the number of private electricity generation facilities selling power

Introduction1 Unreliable and inadequate electricity supply is one of the major challenges affecting Uganda’s economic growth and develop-ment. “Frequent power outages cause sig-nificant losses, equivalent to 6 percent of turnover on average for firms in the formal sector, and as much as 16 percent of turno-ver for informal sector enterprises” (Eber-hard & Shkaratan, 2011). Electricity in Uganda is highly unreliable because of the dilapidated distribution and transmission network infrastructure. Eberhard & Shkara-tan (2011), about 15 percent of installed ca- 1 V Mutyaba, N Marwa, 2015

pacity in Sub-Saharan Africa is not opera-tional because of inadequate maintenance of the existing facilities. According to the World Bank (2000:224) Ugandan firms lose an average of 91 days a year because of power outages. The impli-cation is that Uganda and Africa must not only address the challenge of electricity reli-ability but also ensure that the price of elec-tricity is affordable to attract investment into the country. According to Mawejje (2013), business executives perceive the cost of electricity as the biggest hindrance to doing business in Uganda. In order to address the challenges of unreliable electricity supply and high tariffs, in 2003, the Government of

The Impact of Capital Structure of Electricity Genera-tion Projects on Electricity Tariffs in Uganda

Vianney Mutyaba University of Stellenbosch Business School [email protected] Nyankomo Marwa University of Stellenbosch Business School [email protected]

Abstract The recent transformation in the Ugandan energy sector has led to a significant surge in private electricity generation companies in the country. These companies have a heterogeneous capital structure and they tend to charge different tariff rates for the electricity generated. While the capital structure might have an important role to play in differential tariff setting, it is not clear to what ex-tent it influences the tariff structure of electricity generation projects. Thus, the objective of this study was to examine the effect of capital structure on the tariff of electricity generation projects in Uganda after controlling for other factors such as operation and maintenance costs, technology used for generation, project development costs, and installed capacity of generation plants on the generation tariffs. Using cross-sectional data from 29 electricity generation companies as at Sep-tember 2014, a bootstrap linear regression analysis was used for estimation. The results of the study indicated that the higher the debt portion in the capital structure, the lower the generation tar-iff. However, the impact of debt in the capital structure was not statistically significant. What stood out is that renewable technologies have a much lower generating tariff than non-renewable tech-nologies. Key words: Capital Structure, Uganda, Electricity Generation Tariffs

Journal of Green Economy and Development, Volume 1, Issue 1, 2015

26

Uganda made a strategic decision to liberal-ise and unbundle the Electricity Supply In-dustry (ESI). Before 2003 electricity genera-tion was the sole responsibility of govern-ment but liberalisation encouraged private sector participation, increased investment, reduced inefficiency and reduced depend-ence on government financing. In addition, liberalisation also changed the operation of the electricity sector in Uganda. Since the liberalisation of the electricity sup-ply Industry, there are more operators with diverse financing structures for generation projects (Electricity Regulatory Authority (ERA), 2013) and use of debt financing by utility firms has been an increasing trend. According to the Uganda Electricity Trans-mission Company Limited (UETCL) (2013), in 2012 about 55 percent of the energy was generated by Independent Power Producers (IPPs) compared to five percent in 2005. The UETCL report further indicates that since 2003, the number of private electricity generation facilities selling power to the Na-tional Grid increased from four to twelve in 2012. This inflow of private sector capital in electricity generation created competition, diverse capital structures and therefore dif-ferent generation electricity tariffs for the generation plants (ERA, 2013). The pro-spective developers submitted proposals to ERA under a competitive framework with optimal cost structures in order to be issued a license. The World Bank (2000) stated that electricity tariffs in Africa are high compared to other regions of the world, mainly because of cost and policy differences. According to Umeme Limited (2011), the Uganda Electricity gen-eration segment has the largest contribution to the retail tariff, contributing 68 percent of the total electricity sector costs. Cornille and Dossche (2006) stated that the major de-

terminants of prices are quantity demanded, quality supplied and cost of production. Since investment in generation companies requires significant upfront cost, we hypoth-esise that the capital structure, for electricity generation companies may be one the ma-jor drivers of the generation tariff. It is ex-pected that if the cost of capital directly in-fluences the costs of production, it will be transmitted to the prices of commodities. This study will look at the generation tariff and how capital structures of generation plants influence the generation tariff in Uganda.

Overview of the Ugandan Electricity Supply Industry and the Capital Structure of Operators Uganda’s electricity supply industry started with the commissioning of the Owen Falls dam in 1954 and establishment of the Uganda Electricity Board (UEB). At the time of Uganda’s independence in 1962, it was clear that the energy future of the country was very bright. The national demand was low and generation was high. Uganda was generating adequate electricity to meet the national demand, and the rate of industriali-sation and economic growth was stimulated by the existence of a generation facility that supplied adequate electricity. The Uganda Electricity Board was a vertical-ly integrated fully state-owned enterprise that was responsible for generation, trans-mission and distribution of electricity in Uganda. Like in other countries in Sub-Saharan Africa (SSA) at the time, Uganda Electricity Supply Industries experienced challenges of inadequate investment amidst competing government priorities, poor maintenance, unreliable and expensive


27

electricity. According to Eberhard & Shkara-tan (2011), SSA is experiencing a power cri-sis with unreliable, inadequate, expensive electricity infrastructure and limited access to electricity with less than three in every ten people living in SSA having access. Eber-hard & Shkaratan (2011) further stated that power supply in SSA is unreliable, with 15 percent of the installed capacity not opera-tional mainly as a result of aging plant and inadequate/irregular maintenance with aver-age electricity tariffs in SSA twice the tariffs in other parts of the developing world. Addi-tionally, the World Bank (2011) states that SSA is experiencing an infrastructure financ-ing gap of US$ 31 billion per year resulting from the required additional investment for new infrastructure and addressing ineffi-ciencies in the existing infrastructure. SSA governments do not have the ability to fi-nance the required investment amidst a constrained resource envelope, thus there is a need to encourage partnership with the private investors to bridge the financing gap. In 2004, the Government of Uganda decided to encourage participation of the private sec-tor in the ESI. The purpose of encouraging private sector participation was to increase investment in the sector, increase genera-tion capacity, reduce power losses, and im-prove collection rates and efficiency. Since liberalisation of the ESI in 2004, there has been increased generation capacity installed in Uganda. The electricity generation ca-pacity has increased from 425MW in 2004 to 868 MW in 2013 (ERA, 2013). The num-ber of companies generating electricity and selling to the national grid increased from four in 2003 to twelve in 2012 (UETCL,

2013). According to UECTL (2013), electrici-ty generation increased from 1,887GWh in 2005, when the Government of the Uganda was the main generator of electricity, to 2,857GWh in 2012 after encouragement of private sector participation. As private sector participation increased, there was a need to ensure a cost recovery mechanism (includ-ing capital costs, financing costs and operat-ing costs), accordingly the weighted average generation tariff increased from US$ 0.02/KWh in 2005 to US$ 0.13/KWh in 2011. The participation of the private sector led to the existence of a diversity of capital structures for investors in the ESI, especially in the Generation segment where there is more private sector participation (ERA, 2013). The capital structures of companies operating in the ESI include debt composi-tion ranging from 100 percent: 0 percent, 80 percent: 20 percent, 70 percent : 30 percent, 65 percent : 35 percent, to mention but a few (ERA, 2013).

Review of Relevant Literature and Theoretical Framework

From the theoretical framework it should be noted that apart from the capital structure, there are other factors affecting the genera-tion tariff including;- the operations & maintenance costs, the technology used, the installed capacity of the generation facili-ty among others. The theoretical framework can be conceptually summarised as in Fig-ure 1. Accordingly, Figure 1 can be ex-plained by moving from the right-hand to left. That is, at macro-level, the generation tariff is affected by two major factors: the volume of electricity generated and the cost of generation.


28

The volume of electricity generated depends largely on the type of technology employed or fuel used such as hydropower or thermal-power deployed; the installed capacity of the plant; and other factors (for example climatic factors such as hydrology, solar intensity). In this study, the full cost pricing theory is considered appropriate in describing the generation tariff determination process. Ac-cording to the microeconomic theory, the typical approach of determining the price (P) under this method is to estimate the average

variable cost (AVC) and to add a fair margin of profit (Dwivedi, 2003). That is:

𝑃 = 𝐴𝑉𝐶 + 𝐴𝑉𝐶(𝑚) (1) …(1) Where P is the price (which in our case is the generation tariff), AVC is average varia-ble cost, and m is mark-up percentage of profit that is fixed to cover average fixed cost (AFC) and a net profit margin (NPM). Thus: 𝑃 = 𝐴𝑉𝐶 + 𝐴𝐹𝐶 + 𝑁 (2)

From the theory of costs (Koutsoyiannis, 1978), the average cost (AC) is an aggrega-tion of AVC and AFC; and AC is the ratio of total costs (TC) to total output (Q). That is:

𝐴𝐶 = 𝐴𝑉𝐶 + 𝐴𝐹𝐶 (3) …(3)

𝐴𝐶 = !"!

(4) …(4)

Figure 1: Conceptual framework of effect of debt-equity structure on electricity tariffs of generation project

Generation tariff

Volume of eletricity

Generation technology

Installed capacity

Other factors e.g climate

Total cost of production

Debt-equity structure abd Finance costs

O&M costs

Return on owner equity

Other factors


29

Substituting Equation 4 into Equation 3 im-plies that the electricity generation tariff (PE) is a function of total cost of production, out-put produced and the net profit margin for owners or equity providers for generation plants. This is presented in Equation 3.5 as:

𝑃! =!"!+ 𝑁𝑃𝑀 (5) …(5)

In the electricity industry literature, the right-hand side of Equation 3.5 is referred to as the revenue requirement, which is com-posed of the operating and maintenance costs, invested capital and profit margin. According to Bauer (2004:4), the capital structure is determined by the size of the firm, profitability of the firm, growth pro-spects / opportunities, existing tax shield and the industry in which the firm operates. From a project finance perspective, Gatti (2013:138) stated that project sponsors pre-fer to commit as limited equity as possible. Gatti (2013: 138) further stated that the ba-sis for the debt to equity ratio is: economic soundness of the project, the acceptance level of risk for the lenders, and precedent in the financial markets. Harris and Raviv (1991:334) stated that many studies have been conducted to pro-vide guidance on how specific characteris-tics of firms and industries determine their capital structure. Hatfield, Cheng & Da-vidson (1994:2) noted that firms operating in the same industry will have generally similar financing structures. However, other re-searchers such as Boquist and Moore (1984) found weak evidence to support this theory. Electricity generation firms in Ugan-da have diversified financing structures de-spite operating in the same industry (ERA, 2013). Bauer (2004:4) agreed with other research-ers that determinants of capital structure in

clude the size of the firm, profitability of the firm, growth prospects / opportunities of the firm, existing tax shield, tangibility of the firm and the industry of operation. Bauer (2004) stated that more profitable firms are ex-pected to have higher debt in their capital structure because they have income to shield from taxes. Bauer (2004) further ar-gued that companies with more tangibility have more access to credit/debt because of their ability to use their assets as collateral to obtain debt financing. Mark and Clifford (1992), however, found no evidence to sup-port the finding of Bauer (2004) in respect of factors that influence the capital structure of firms. Some researchers, such as Huang and Song (2006), Rajan and Zingales (1995) and Friend and Lang (1988) concluded that there is a positive relationship between size and capital structure. Song (2005) agreed with Bauer (2004) that the determinants of the capital structure include tangibility (asset structure), non-debt tax shield, profitability, size, expected growth, uniqueness and in-come variability. Titman and Wessels (1988), argue that firm’s choice of the capital structure is influenced by the costs and benefits of using debt or equity. Researchers such as Kester (1986), Kim and Sorensen (1986) and Titman and Wes-sels (1988) however disagree and conclude that there is a negative relationship between size and the capital structure. According to Oolderink (2013), the pecking order theory of capital structure states that in circum-stances where external financing is needed, firms prefer debt financing to equity financ-ing. Johnson (1998) stated that there is no predetermined rule for determination of capi-tal structure as every company experiences unique risks, threats and opportunities. He


30

further noted that although the traditional view is that any company has an optimal capital structure the determination of this optimal structure is subjective and complex. Scannella (2012: 85), like Johnson (1998), noted that there is no universally acceptable rule for choosing the capital structure of a firm. Scannella (2012: 85) noted that typical projects have 70–75 percent debt and 25–30 percent equity. The National Renewable Energy Laboratory (2008: 4-1) stated that between 35 and 45 percent equity is stand-ard for commercial projects. Berger (2011:6) stated that electric power projects in Europe have typical debt to total capital range of 60–70 percent. Clarke, Wilson, Daines and Nadauld (1988:167) noted that a study by Scott and Johnson concluded that there is a target capital structure for very large American companies that guides decisions regarding financing. He further noted that while this target capital structure varies between 26 and 40 percent there is no consensus on the exact percentage and every company uses its own. The capital structure is key as it determines the cost of capital. Johnson (1998) noted that the value of the firm/company increased as the weighted average cost of capital re-duced. Firms’ costs reduce as the weighted average cost of capital reduces up to the optimal capital structure. As the portion of debt in the capital structure increases, the weighted average cost of capital reduces, increasing the value of the firm. As the por-tion of debt further increases, the cost of both debt and equity increase to cater for increased financial risk. When the debt por-tion increases beyond the optimal capital structure, the increase in cost of equity off-sets the benefit of lower cost of debt. Leland and Toft (1996) noted that there is a need to

balance the tax advantages of debt with the risks of bankruptcy and high agency costs in the determination of the optimal capital structure. Puwanenthiren (2011:1) stated that the capital structure is the most signifi-cant discipline in the operations of any com-pany, and concluded that there is a negative relationship between the capital structure and the financial performance of any com-pany. Like Puwanenthiren (2011:1), Dhan-kar and Boora (1996: 29) stated that the capital structure is the most important deci-sion for a firm. The capital structure affects the profitability of the company, the cost of capital, earnings per share, dividends and liquidity position of the company. Duca (2010:523) concurs with Chowdhury and Chowdhury (2010), and Mark & Clifford (1992) that the capital structure decision is one of the most fundamental premises of the financial framework of a corporate entity. According to Chowdhury (2010:112), the decision in respect of how much debt and equity to employ is aimed at ensuring an op-timal financing structure that maximises shareholder returns. Many other researchers agree that the fi-nancing structure has an impact on firm’s costs and profitability. Mohammed Omran (2001) studied 69 Egyptian firms privatised between 1994 and 1998 that started using debt financing. The study concluded that firms that were privatised and used debt fi-nancing increased in profitability, operating efficiency, capital expenditure and divi-dends. Pirashanthini and Nimalathasan (2013) studied the capital structure of manufactur-ing companies in Sri Lanka and concluded that the debt to equity ratio is positively and strongly linked to the profitability ratios of gross profit margin and net profit margin. According to Russo, Weatherspoon, Peter-


31

son and Sabbatini (2000: 27), excessive leverage leads to financial distress, increas-es the firm’s costs in terms of high transac-tion costs and forgone profit opportunities and therefore influences prices. Russo et al. (2001:34) stated that there is a positive cor relation between the cost of financing and the portion of equity in the capital structure. When interest rates increase, the cost of debt financing increases and makes equity financing more attractive than debt. Financ-ing more assets using equity reduces the risk brought about by operating leverage and increases profitability. The National Re newable Energy Laboratory (NREL) (2012) states that firms whose financial structure is fully debt-financed generally yield a lower levelised cost of energy (LCOE) than those that rely purely on equity capital. Empirical results from a study by Dhankar and Boora (1996) on the capital structure of companies in India showed that there is no definite relationship between change in the capital structure and the value of a firm but 81 percent of the companies showed a neg-ative relationship between the capital struc-ture and the cost of capital. Cost of capital reduces as debt levels in the capital struc-ture increase because cost of debt is cheaper than cost of equity as interest rate payments are allowable deductions for tax purposes. Chowdhury and Chowdhury (2010:112) argued that increasing debt fi-nancing within the capital structure would increase the firm’s value up to a point where any further increase in debt financing in-creases the overall cost of capital. The Brattle Group (2005: 5) stated that the amount of debt in the capital structure influ-ences the cost of equity. Debt financing re-sults in financial risk which equity share-holders incorporate when pricing the cost of

equity. Scannella (2012:85), and Chowdhury and Chowdhury (2010:111) agreed that firms dedicate operating cash flows to debt service before paying returns / dividends to shareholders. Debt financing ensures that the company benefits from interest expens-es, which is an allowable expense for tax purposes, but too much debt makes debt issuers become nervous regarding the com-pany’s ability to pay. On the other hand, Modigliani and Miller’s findings in 1958 con-cluded that the decision in respect of the company’s capital structure is not important because of the existence of frictionless mar-kets and homogeneous expectations. Modi-gliani and Miller (1958) argued that all firms experience the same risk, their expectations are similar, there is absence of growth, agency costs and bankruptcy costs and the risk-free rate can be applied for lending and borrowing by individuals. Puwanenthiren (2011:2) explained that Mo-digliani and Miller’s (1958) relevance theory reveals that no capital structure is better than another, the benefits of increasing debt financing is offset by the higher risk in-curred. Some researchers agree with Modi-gliani and Miller’s (1958) findings that the decision in respect of the company’s capital structure is not important. Brigham and Gapenski (1990) in their study of the effects of capital structure on utilities’ cost of capital and revenue requirement concluded that decisions regarding capital structure have a negligible effect on firms’ revenue require-ments.

Data and Methodology

Data type and sources

The data used in this study is secondary da-ta, obtained from ERA, Uganda’s ESI regu-lator. The data was obtained for companies


32

that have obtained or are about to obtain licenses and permits from ERA to develop, own and operate electricity-generating facili-ties in Uganda.

Econometric model

Based on the theoretical framework pre-sented in section three, the empirical model to estimate the relationship between the capital structure and the tariff of electricity generation projects in Uganda can be stated in a general functional notation as: T! = β! + β!DE+ β!FC+ β!ROE+ β!OM+β!GT+ β!PC+ ε (6) …(6) According to Gujarati (2003), Equation 6 can be written as an econometric equation as follows: T! = β! + β!DE+ β!FC+ β!OM+ β!ROE+β!GT+ β!PC+ εWhere 𝛽!……𝛽! ( 7) Where are β!… . β! 𝑎𝑟𝑒 unknown parame-ters to be estimated and ε is the error term representing other unknown variables that may affect the generation tariff.

Estimation Procedures and Model Diagnostics

Data used in this study is from various com-panies and therefore cross-sectional in na-ture. According to Gujarati (2003), in order to get efficient estimates from cross-sectional data, the data of the dependent variable should be normally distributed and the explanatory variables should not be highly linearly correlated (multicollinear). In case the dependent variable is not normally distributed, Gujarati (2003) suggested that such data can be transformed such as into natural logarithms, squared, squared root, etc., so as to make the data normally dis-tributed. Ordinary least squares (OLS) was used to estimate regression parameters. Since our sample was limited we used ro-

bust standard error with bootstrap option in order to get reliable test statistics.

To improve the skewness and relationship patterns among the variables, the following variables were log transformed: DE (portion of debt in the capital structure), OM (opera-tion and maintenance costs), PC (installed capacity), DC (Development costs). In the analysis, the transformed explanatory varia-bles were used in the econometric analysis. Bivariate linear regression models were es-timated with each dependent variable. Then, a forward and backward selection approach was used manually to estimate the final full model. Cook’s Distance was used to detect the influential outliers using a cut-off point of 4/n which is 0.134 in our case (n=29). Three observations were identified as influential and were left out during the analysis. Based on the correlation matrix there was significant association between develop-ment cost and installed capacity (r=0.98). To avoid multicollinearity only one of the varia-ble installed capacities was included in the model. After the selection of the final model, we used robust standard error bootstrap option with 1000 iterations and seed of 1. Boot-strap was used in order to achieve correct inferences given a small sample size (Guan, 2003). Only the results from the final model with bootstrap option will be used for inter-pretation and discussion purpose. The analysis was done using STATA SE 12 Sta-tistical Program.

Estimation of the Model and Relia-bility of the Estimates

To identify the relationship between the cap-ital structure of electricity generation pro-jects and electricity tariffs in Uganda, Equa-tion 5 was estimated using econometric mul-tiple linear regression analysis. Similar stud-

33

ies that have used OLS include ECME Con-sortium (2010) to estimate the functioning of retail electricity. The econometric analysis involves running one OLS regression; in which the projects will be the observations, the Generation Tariff will be the dependent variable, and debt to equity ratio, return on equity, debt financing costs, operation and maintenance costs, technology employed, development costs and installed capacity will be the explanatory variables. As mentioned earlier, given that the sample for this study was relatively small and there are more than five variables, the analysis can lead to inefficient estimates (Deaton, 2003). To enhance efficiency and reliability of the estimates involving a relatively small sample, Efron (1979) introduced the boot-strapping technique. The bootstrap is a sta-tistical technique that artificially increases the sample size through replication (produc-ing a large number of “copies”) of a sample statistic, computed from bootstrap samples at five percent confidence interval (Boos & Munahan, 1986). In this study, the bootstrap technique is used to improve the efficiency of the estimates.

Results and Discussions Descriptive Statistics Table 1 describes the data used in the anal-ysis. In Table 1, Tg is the generation tariff which is the dependent variable measured in US$ per KWh. The figures in the table in-dicate that the average generation tariff is US$ 0.105/KWh and ranges from US$ 0.053/KWh to US$ 0.243/KWh. The type of technology was coded 1 if the technology was renewable and 0 if it was non-renewable. From Table 1, it can be observed that 89.7 percent of the generation technology is re-newable energy that includes hydro, bio-mass and bagasse co-generation. Operation and maintenance costs average US$ 0.031/KWh. On average, debt contrib-utes 70.7 percent to the capital structure of electricity generation projects, and the aver-age cost of debt is 7.9 percent per annum. The average return on equity is 16.8 per-cent. Project development costs range from US$ 4 million to US$ 755 million and aver-age US$ 67.7 million, and the average in-stalled capacity is 25.7MW with a maximum installed capacity of 250MW.

Table 1: Description of the data Variable Description of the variable Mean Std. Dev. Min Max

Tg1 Generation Tariff with only re-newable technology (without outliers)

0.905 0.135 0.053 0.107

Tg Generation Tariff with both re-newable and non-renewable technology (outliers included)

0.105 0.046 0.053 0.243

GT Type of Technology 0.897 0.310 0.000 1.000 Om Operation & maintenance costs 0.031 0.029 0.003 0.100

De Debt portion in the capital struc-ture 0.707 0.128 0.263 1.000

Fc Debt financing costs 0.079 0.026 0.012 0.120 Roe Return on equity 0.168 0.054 0.000 0.250 dc_100 Development costs 0.677 1.403 0.040 7.550 pc_100 Installed Capacity 0.257 0.473 0.010 2.500


34

Testing for reliability of the data Testing for reliability of the data was under-taken using the approaches below. Testing for multicollinearity Table 2 below shows the pairwise correla-tion of the variables to check for any possi-ble multicollinearity. When there is a perfect collinearity, the result is 1, when there is no

relationship between the variable the result is zero. According to Gugarati (2003), if the variables are perfectly collinear, then one of these should be dropped out of the analysis. Table 2 shows that there is almost perfect collinearity between installed capacity and development cost. Development cost was omitted from the regression analysis to avoid the problem of multi-collinearity.

Table 2: Pairwise correlation of the variables

TG GT Om de Fc roe dc_100 pc_100

TG 1 GT -0.9596 1

Om 0.7388 -0.7187 1 De 0.2443 -0.2523 -0.075 1

Fc -0.2221 0.2618 -0.1169 -0.2459 1 Roe -0.3582 0.3057 -0.243 -0.1923 0.2707 1

dc_100 0.0062 -0.0166 -0.1883 0.3681 -0.2675 -0.0483 1 pc_100 0.1527 -0.1774 -0.0722 0.4325 -0.2815 -0.0723 0.9789 1

Regression Results

Table 3 below shows the results of the re-gression using the OLS method with the bootstrapped standard error technique to improve efficiency and robustness. The bootstrap included 948 replications. The re-sults of the regression are robust given that the Wald chi2 statistic (70.18) is statistically significant at less than 1 percent level. The adjusted R_Square is 79.2 percent, implying that the model describes up to 80 percent of the relationship between the generation tariff and explanatory variables.

From Table 3, two variables, that is GT (Generation Technology) and installed ca-pacity (log_pc_100), are found to have a significant effect on the generation tariff (gt), while the portion of debt in the capital structure (log_de) is found to have a positive

but not statistically significant effect on the generation tariff (gt). Details of the analysis are provided below.

Capital structures of generation projects

Debt-equity (DE): Table 5.4 illustrates that there is a positive relationship between the debt portion of the capital structure and the generation tariff of electricity generation pro-jects. The results in Table 5.4 indicate that a one percent increase in the portion of debt in the capital structure results in a 15.4 per-cent increase in the electricity generation tariff. When the outliers are excluded, a one percent increases in the portion of debt in the capital structure results in a 24.7 percent increase in the electricity generation tariff. The result is however not statistically signifi-cant. Studies such as Russo et al. (2001) also concluded that there is a positive


35

relationship between the portion of debt in capital structure and prices, explaining that a high proportion of debt in the capital

structure leads to financial distress and high costs therefore increasing prices/tariffs.

Table 3: Multiple Regression Results

Results when both renewable and non-renewable technology are included

(outliers included)

Results for renewable technolo-gy only

(outliers excluded)

Variable Observed Coefficient

Boot-strap Std. Err. Z P>z

Ob-served Coeffi-cient

Z P>z

GT -1.024*** 0.167 -6.150 0 - - - log_om 0.034 0.051 0.660 0.509 0.223 2.550 0.011 log_de 0.154 0.342 0.450 0.653 0.247 0.540 0.592 Fc 0.701 2.021 0.350 0.729 0.108 0.040 0.968 Roe -0.657 0.893 -0.740 0.462 -1.346 0.900 0.369 log_dc_100 0.122 0.104 1.170 0.240 -0.066 -0.380 0.702 log_pc_100 -0.168* 0.091 -1.850 0.064 0.117 0.690 0.487 _cons -1.369 0.224 -6.120 0 -0.968 -1.930 0.054 N 29 29.000 R_Square 0.844 0.538 Adj R-squared 0.792 0.4115 Wald chi2(7) 70.180*** 9.580 Root MSE 0.155 0.260 Replications 948 1000

In order to further understand further why the relationship between the portion of debt in the capital structure and the generation tariffs was positive but not statistically significant, a polyno-mial graph of the relationship is plotted in Figure 1. Figure 1: Illustration of relationship between debt portion in the capital structure and generation tariffs

.04.06

.08.1

.12Ge

nerat

ion Ta

riff

0.20 0.40 0.60 0.80 1.00de (Portion of debt in the capital structure)

gt(Generation Tariff) predicted gt (Generation Tariff)


36

Figure 1 shows a slight inverse relationship when the portion of debt in the capital struc-ture is low and later a positive relationship as the portion of debt in the capital structure increases. Because of this mixed relation-ship, it is indeed clear that it cannot be sta-tistically significant. However other studies, such as that of Pirashanthini and Nimala-thasan (2013), found that high debt compo-sition in the capital structure leads to a lower tariff but nonetheless not significant. Technology of generation projects Technology (GT): Table 3 shows that there is a negative and statistically significant (p<0.01) relationship between the genera-tion technology and the electricity genera-tion tariff. To further understand why this may be the case, a comparison was done in respect of the average generation tariff for renewable and non-renewable generation technologies in Uganda. The generation tariff for renewable generation technology is about US cents 10.7 per KWh compared to about US cents 23 per KWh for the genera-tion tariff for non-renewable technology. The likely reason why the generation tariff for non-renewable generation is high in Uganda is due to the high costs of fuel utilised in non-renewable electricity generation. Installed Capacity Installed Capacity (PC): Table 3 shows that installed capacity is a statistically signif-icant variable in the determination of gen-eration tariffs at 6.4 percent. The table fur-ther shows that there is a negative relation-ship between the installed capacity of the generation plant and the generation tariff of the respective plant. A 16.8 percent in-crease in installed capacity results in a 1 percent reduction in the generation tariff. The results are consistent with Dwivedi (2003) who concluded that a large installed

capacity implies that costs are recovered over more units therefore reducing the gen-eration tariff. When the installed capacity is small, the generation tariffs are high and in-crease as the installed capacity increases up to a certain level, beyond which an in-crease in the installed capacity leads to re-duction in the generation tariff. Development costs of generation projects Development costs (DC): Table 3 shows that there is a positive relationship between the development costs and the generation tariff and that the development costs varia-ble is significant. Table 5.4 shows that for a one percent increase in development costs, the tariff increases by 12.2 percent. This finding is supported by Ragwitz et al. (2012) who concluded that there is a positive rela-tionship between development costs and the tariff (i.e. an increase in development costs leads to increase in the generation tariff). Operation and Maintenance costs (OM Table 3 shows that there is a positive rela-tionship between the operation and mainte-nance costs and the generation tariff. For a 3.4 percent increase in the operation and maintenance costs, the generation tariff in-creases by one percent. The variable is sig-nificant because it is dependent on the units generated as it is measured in US$/KWh. Literature on the relationship between the operation and maintenance costs and tar-iffs/prices indicates that the effect of the op-eration and maintenance costs on prices of commodities depends on the proportion of operation costs to the total costs. A high portion of operation and maintenance costs in the total cost structure will have a higher effect on the tariff / prices.


37

Costs of debt Debt financing costs (FC): The results in Table 3 indicate that there is a positive rela-tionship between the costs of debt and the generation tariffs, although the table further shows that debt financing costs are not sta-tistically significant in the determination of the generation tariff. An increase in the debt financing costs by one percent results in a 70 percent increase in the generation tariff. The results are in line with findings by Rus-so et al. (2001) that a high interest rate in-creases the generation tariff.

Return on equity Return on equity (ROE): The results in Ta-ble 5.4 indicate that there is a negative rela-tionship between return on equity and generation tariffs. When the return on equity reduces, the generation tariff increases. An increase in return on equity by one percent leads to a reduction in generation tariff by 65.7 percent. This result is not consistent with findings of the Brattle Group (2005), which concluded that a high return on equity has a positive effect on the generation tariff. Conclusion and Recommenda-tions

The objective of this study was to examine the relationship between the debt-equity capital structure and the electricity genera-tion tariffs in Uganda. The study found a positive but not significant relationship be-tween the debt portion in the capital struc-ture and the generation tariff of electricity generation projects. There is a negative and statistically signifi-cant relationship between the generation technology and the electricity generation tar-iff. The generation tariff is high for non-renewable technology and low for renewble

technology. Installed capacity is a statistical-ly significant variable in the determination of generation tariffs and there is a negative re-lationship between the installed capacity of the generation plant and the generation tariff of the respective plant. The study has identi-fied that generation tariffs are low for elec-tricity generation plants using renewable technology and high for electricity genera-tion plants using non-renewable technology. Government of Uganda and Electricity Reg-ulatory Authority should prioritise licensing and development of renewable technology generation projects in order to keep the generation tariffs low. It is instructive to note that generation pro-jects with small /low installed capacity tend to have higher generation tariffs. The Gov-ernment of Uganda and Electricity Regulato-ry Authority should prioritise development of bigger projects and ensure fully optimisation of the project capacities in order to ensure lower generation tariffs. Also the level of de-velopment costs is a major determinant of electricity generation tariffs in Uganda. Measures should therefore be implemented to minimise development costs in order to ensure lower electricity generation tariffs. References Bailey, P. (2003). The business and finan-

cial structure of the water industry in Eng-land and Wales, The University of Bath

Bauer, P. (2004). Determinants of Capital Structure: Empirical Evidence from the Czech Republic. Czech Journal of Eco-nomics and Finance, 54(1-2), 2-21.

Berger, R. (2011). The structuring and fi-nancing of energy infrastructure projects, financing gaps and recommendations re-garding the new TEN-E financial in-strument. Berlin/Brussels, European


38

Commission Directorate-General for Energy.

Boos, D.D. & Munahan, J.F. (1986). Boot-strap methods using prior information. Bi-ometrika. 73, 77-83.

Boquist, J.A. & Moore, W.T. (1984). Inter-Industry Leverage Differences and the DeAnge-lo-Masulis Tax Shield Hypothesis, Financial Management, 2: 5-9.

Brattle Group, The. (2005). The effect of debt on the cost of equity. Washington, D.C Edison Electric Institute

Brigham, E.F., and Gapenski, L.C. (1990). In-termediate Financial Management. 3rd edn. The Dryden Press, Chicago.

Chowdhury, A and, Chowdhury S.P. (2010): Impact of Capital structure on Firm’s Val-ue: Evidence from Bangladesh”. Business and Economic Horizon (BEH), 3(3), 111- 122.

Clarke, R.G., Wilson, B.D., Daines, R.H. & Nadauld, S.D. (1988). Strategic Financial Man-agement. Homewood: Irwin.

Cornille, D. and Dossche, M. (2006). The Pat-terns and Determinants of Price Setting in the Belgian Manufacturing Industry. Euro-pean Central Bank, Working Paper 618.

Deaton, A. (2003). How to Monitor Poverty for Millennium Development Goals. Journal of Human Development, 4(3), 353-378.

Dhankar, R. & Boora, A. (1996). Cost of Capi-tal, Optimal Capital Structure, and Value of Firm: An Empirical Study of Indian Companies, 21(3), 29-36.

Duca, F. (2010). What determines the capital structure of listed firms in Romania? The Bucharest University of Economic Stud-ies, Bucureşti, România. CES Working Papers, 3a,523-531, www.ceeol.com

Dwivedi, D.N. (2003). Microeconomics - Pricing Theories, Microeconomics: Theory and

Applications. Delhi, Pearson Education (Singa-pore) Pte.

Eberhard & Shkaratan (2011). Powering Africa: Meeting the financing and reform chal-lenges. Energy policy, 42, 9.8, University of Cape Town, Elsevier Limited.

Efron, B. 1979. Bootstrap methods: Another look at the Jackknife. The Annals of Sta-tistics, 7(1), 1-26.

Electricity Act, 1999, Chapter 145.

Electricity Regulatory Authority. (2006). Tariff setting guidelines. Kampala, Electricity Regulatory Authority.

Electricity Regulatory Authority. (2013). Elec-tricity Sector performance report. Kampa-la, Electricity Regulatory Authority.

European Consortium of Management of Con-sultants (ECME). (2010). The functioning of retail electricity markets for consumers in the European Union. European Union, Director General for Health and consum-ers.

Finscope Uganda (2009).Final Report: Results of a National Survey on demand, Usage and Access to Financial services in Uganda

Friend, I. & Lang, L. (1988). An Empirical Test of the Impact of Managerial Self-Interest on Corporate Capital Structure, Journal of Finance, 43,(2), 271-281.

Gatti, S. (2013). Project Finance in Theory and Practice.Amsterdam: Elsevier Inc.

Guan, W. (2003). From the help desk: Boot-strapped standard errors. The Stata Jour-nal, 3(1), 71-80.

Gujarati, D.N. (2003). Basic Econometrics. 4th edition. New York: McGraw-Hill.

Hall, R.L. & Hitch, C.J. (1939). Price Theory and Business Behaviour, Oxford Econom-ic Papers No.2 May 1939


39

Harris, M. & Raviv, A. (1991). The theory of capital structure. Journal of Finance, 46, 297-356.

Hatfield, G.B., Cheng, L.T.W. and Davidson, III W.N. (1994). The Determination of Optimal Capital Structure: The Effect of Firm and Industry Debt Ratios on Market Value. Journal of Financial And Strategic Decisions, 7(3) [Online]. Available from: http://www.studyfinance.com/jfsd/pdffiles/v7n3/hatfield.pdf (Accessed: 24 July 2014).

Huang, G. & Song, F.M.(2006). The Determi-

nants of Capital Structure: Evidence from China. China Economic Review, 17(1), 14-36.

Johnson, A.S. (1998). The effect of bank debt on optimal capital structure, Financial Management, 27, 47-57.

Kester, C.W. (1986). Capital and Ownership Structure: a Comparison of United States and Japanese Manufacturing Corpora-tions. Financial Management, 15, 5-16.

Kim, W.S. & Sorensen, E.H. (1986). Evidence on the Impact of the Agency Costs of Debt in Corporate Debt Policy. Journal of Financial and Quantitative Analysis, 21, 131-144.

Koutsoyiannis, A. (1978). The theory of eco-nomics. 2nd edition. United States of America, Rowman & Littlefield Publishers.

Lee, F.S. (1996). Pricing, the pricing model and post-Keynesian price theory. Review of Political Economy, 8(1), 87-99.

Leland, H.E. & Toft, K.B. (1996). Optimal capi-tal structure, endogenous bankruptcy, and the term structure of credit spreads. Jour-nal of Finance, 51(3), 987-1019.

Macmillan. S, Antonyuk. A, Schwind. H (2013), Gas to Coal Competition in the U.S. Pow-er Sector. International Energy Agency, insight series 2013.

Manohar, S. & Nejadmalayeri, A. (2003). Inter-nationalization, capital structure, and cost of

capital: evidence from French corporations. Journal of multinational financial manage-ment, 14(2), 153-169

Mark, S. & Clifford, F. (1992). Determinants of capital structure, Review of Financial Economics, 43(1.2), 1-19

Mawejje, J .(2013). Falling consumer demand, competition stifle Uganda Business Climate, Economic Policy Research Cen-tre, The Uganda Business Climate Index, 4 Means, G.C. 1962. Pricing Power and the Public Interest. New York: Harper and Brothers.

Mendelsohn, M., Kreycik, C., Bird, L., Schwabe, P. & Cory, K. (2012). The Im-pact of Financial Structure on the Cost of Solar Energy. Technical Report NREL/TP-6A20-53086. Colorado, Nation-al Laboratory of the U.S. Department of Energy, Office of Energy Efficiency & Re-newable Energy.

Ministry of Minerals and Energy South Africa. (2008). Electricity Pricing Policy, Policy No. 1398. Department of Minerals and Energy.

Modigliani, F and Miller, M.H. (1958). The cost of capital corporation finance and the the-ory of investment. The American Econom-ic Review, 3, XLVIII

National Planning Authority. (2013). Uganda Vision 2040.Kampala, National Planning Au-thority.

National Renewable Energy Laboratory. (2008). Renewable Energy Price stability benefits in Utility Green Power programs. Technical Report NREL/TP-670-43532. Colorado, National Laboratory of the U.S. Department of Energy, Office of Energy Efficiency & Renewable Energy.

Omran, M. (2001). The Performance of State- Owned Enterprises and Newly Privatized Firms: Empirical Evidence from Egypt. Working paper, The Arab Academy for Science and Technology, College of


40

Management and Technology, Alexan-dria, Egypt

Oolderink, P. (2013). Determinants of capital structure: Static trade-off theory vs. Peck-ing-order theory, Evidence from Dutch listed firms. University of Twente, Nether-lands

Pirashanthini, S. & Nimalathasan, B. (2013). Capital structure and firm value:

An empirical study of listed manufacturing firms in Sri Lanka. Merit Research Journal of Art, Social Science and Humanities, 1(6), 86-91.

Puwanenthiren, P. (2011). Capital structure and financial performance: evidence from selected business companies in Colombo stock exchange sri Lanka. International refereed research journal, 2(2), 171-183.

Ragwitz, M., Winkler, J., Klessmann, C., Gephart, M. & Resch, G. (2012). Recent developments of feed-in systems in the European Union. A report commissioned by the Ministry for the Environment, Na-ture Conservation and Nuclear Safety (BMU). European Union.

Rajan, R.G. & Zingales, L. (1995). What do we know about capital structure? Some evi-dence from international data. The Jour-nal of Finance, 50(5), 1421-1460.

Russo, C., Weatherspoon, D., Peterson, C. & Sabbatini, M. (2000). Effects of managers’ power on capital structure: a study of Ital-ian agricultural cooperatives. The Interna-tional Food and Agribusiness Manage-ment Review, 3(1), 27-39.

Scannella, E. (2012). Capital Regulation and Italian Banking System: Theory and Em-pirical Evidence. International Journal of Economics and Finance, 4(2), 31-43.

Song, H-S. (2005). Capital Structure Determi-nants: An Empirical Study of Swedish Companies. Stockholm. The Royal Institute of

Technology Centre of Excellence for Sci-ence and Innovation Studies.

Thoenes, S. (2011). Understanding the deter-minants of electricity prices and impact of the German nuclear moratorium in 2011. EWI Working Paper No 11/06. Institute of Energy Economics, University of Cologne. Cologne

Titman. S. & Wessels, R. (1988). the determi-nants of the capital structure choice. The Journal of Finance, 43(1), 1-19.

Uganda Bureau of Statistics (UBOS). (2013). 2013 Uganda Bureau of Statistics statisti-cal abstract, Kampala. Uganda Bureau of Statistics (UBOS)

Uganda Electricity Transmission Company Limited (UETCL). (2013). Annual report 2012. Kampala, Uganda Electricity Transmission Company Limited.

Umeme Limited. (2011). Annual Report and fi-nancial statements for the year ended December 2011. Kampala, Umeme Limited.

United Nations Development Programme (2013), United Nations Development Pro-gramme Human Development Index for 2013. New York: The United Nations De-velopment Programme

United Nations Population Fund Uganda (2013). The State of Uganda Population report 2013, Uganda, Kampala, The Unit-ed Nations Population Fund Uganda.

World Bank. (2000). Can Africa Claim the 21st Century? Washington DC: The World Bank.

World Bank. (2011). Towards better infrastruc-ture, conditions, constraints, and opportu-nities in Financing Public-Private Partner-ships in Selected African Countries. Washington DC: The World Bank.

World Economic Forum (2012), The Global Competitiveness Report 2012-2013, Ge-neva: World Economic Forum.


Documents

The Impact of Capital Structure of Electricity Genera- tion ... UETCL report further indicates that since 2003, the number of private electricity generation facilities selling power