Unbundling Eskom – a silver or lead bullet?

Lauralyn Kaziboni and Michèlè Capazario


1.1 Introduction

The primary mandate of the state electricity provider, Eskom Holdings SOC Ltd (Eskom), is to meet South Africa’s electricity demand. In recent years, the demand for electricity has far exceeded Eskom’s ability to supply, with bouts of blackouts since the mid-2000s leading up to early this year. The constrained supply led to deferment of maintenance, which contributed to the inability to meet South Africa’s growing energy demand. Despite the surge in alternative energy production through the Renewable Energy Independent Power Producer (REIPP) Procurement Programme, only a fraction of South Africa’s energy needs are met by IPP-based production, rendering Eskom the principal source of South Africa’s electricity.

The recent debt relief support offered by the National Treasury, amounting to R69 billion (not the requested R100 billion) over three years, came with conditions to improve financial operations and management at the utility (Omarjee, 2019). Together with the National Energy Regulator of South Africa’s (NERSA’s) approved electricity tariff increases, this debt relief can slightly improve Eskom’s outlook amidst unsustainably high operating costs. More crucial, the government has put forward a plan to restructure the power utility by means of unbundling its business processes into three distinct parts: Generation, Distribution, and Transmission.

Unbundling implies that the three subsidiaries would operate autonomously, with each subsidiary operating its own assets, licences, board and personnel, and ultimately publishing separate financial reports (Mboweni, 2019).

The unbundling strategy dates back to the 1998 White Paper on Energy Policy. Progress on this, however, only started in February 2018 when the President announced the unbundling of Eskom as a way to address its unprecedented dire financial situation (Ramaphosa, 2019; Mboweni, 2019). This policy directive was closely followed by another proclamation to establish Eskom’s Transmission subsidiary first, with the board set to be established by mid-2019 (Omarjee, 2019). The State of the National Address (SONA) delivered on 20 June 2019 makes no mention of the unbundling, yet the President noted that a significant proportion of the R230 billion support required over the next ten years, will be front loaded to aid the ailing utility.

This article discusses (i) the lessons from international experiences on unbundling; (ii) the accumulation of Eskom’s debt over time, including its impact on electricity tariffs, and (iii) how the unbundling will influence the debt.

1.2 Lessons on international experiences of unbundling

Very few studies have investigated the relationship between unbundling and public enterprise debt. A comprehensive study authored by Vagliasindi and Besant-Jones (2013) explored this empirical relationship by drawing on a panel of 22 emerging market economies. In this study, the authors illustrate that there is a possibility to decrease an electrical utility’s debt-to-equity ratio by between 45% and 60% if the utility unbundles and privatises(Vagliasindi & Besant-Jones, 2013).


Unfortunately, the Power Market Structure Database from which Vagliasindi and Besant-Jones draw their data from is not publicly available. As such, inference on utility-based debt and equity is less detailed than we would like. From the way the data is presented, we suspect that there are two possible forces that explain the decrease in the debt-equity ratios:


  1. Privatisation of the utility increases the equity – the issuance of shares props up the equity base, ultimately decreasing the debt equity; and/or
  2. The state absorbs a large portion of the utility’s debt, also reducing the debt-equity ratio.


Beyond changes to debt and equity, international literature points to other possible outcomes post unbundling. In general, unbundling of power utilities was met by increased tariffs, increased government (not utility) debt-burden and decreased employment in the sector, on the one hand; and on the other increased capacity generation and electricity security of supply (Vagliasindi & Besant-Jones, 2013).The literature discussed above is not necessarily causal given that there are a host of confounding factors which could influence all of the above findings, aside from the unbundling. The study though does indicate that unbundling comes with probable pros and cons.

From the South African perspective, the unbundling of Eskom could likely result in some or all of the effects discussed above. We argue that, the combination of Eskom’s rapidly growing debt as well as the sheer scale of production inefficiencies could result in various outcomes. Most probablistically, the effects of unbundling could initially be felt in a drop in the employment levels of Eskom, increasing debt levels, continued mismanagement and higher electricity tariffs. The unbundling could also encourage electricity generaton by IPPs, which would increase energy supply. Nonetheless, these outcomes, either negative or positive, are far from certain and will be better understood post-unbundling. 

1.3 Debt accumulation over time

Arguably the biggest barrier to a successful unbundling of Eskom is its level of debt – which more than quintupled from R80 billion in 2009 to approximately R420 billion in 2019 (Merten, 2019). This persistent upward trend over time is shown below (Figure1) The accumulation of Eskom’s debt stock can be explained by several factors including capital expansions and governance issues.

The main explanatory factor of debt accumulation is the poorly designed and planned construction of the Medupi and Kusile coal-fired power stations, two of the biggest coal-fired power stations in the world. The costs of constructing these power stations have reached approximately R145 billion and R161 billion for Medupi and Kusile respectively, just over R300 billion in total (Donnelly, 2019). This construction of has been severely delayed, with construction costs going over the initial expenditure. What is more concerning is that the Medupi and Kusile power station units that are complete are operating below capacity (Gosling, 2019), and are not providing the extra generation to the grid, which would translate into revenue for Eskom.

Governance issues, including state capture, corruption and poor management, have resulted in over staffing along with poor repair and maintenance at power stations, leading to erratic power supply and high operating costs. Another explanation, outside of Eskom’s purview, is electricity theft, non-payment and default by municipalities, resulting in close to R30 billion in arrears. These problems all add to the scale of Eskom’s ballooning debt.

Figure 1: Eskom’s Debt Over Time



This mounting debt, as well as the poor management of Eskom, have a direct consequence on the electricity tariff charged to consumers,inter alia. Eskom’s tariff structure is calculated using the Multi-Year Price Determination (MYPD) methodology of allowable revenue. Based on this calculation, a relatively significant proportion of the tariff is explained by expenditure and primary energy costs, with Eskom’s debt finding its way into the tariff calculation in the Weighted Average Cost of Capital (WACC) component of allowable revenue (see below).

Figure 2: Eskom’s Tariff structure

In the above figure, each percentage relates to the proportion of Eskom’s charged tariff that is driven by a particular allowable revenue item. A non-negligible proportion of the tariff (10.33%) is driven by Eskom’s returns, which are defined as the regulated asset base (RAB) multiplied by Eskom’s weighted average cost of capital (WACC). In estimating the WACC, the utilities various sources of capital including stock, bonds and long-term debt are considered when calculating the average rate it expects to pay to finance its assets. Thus, given Eskom’s debt, a portion of this will be transferred directly to the customer through tariff increases.

The largest driver of Eskom’s tariffs, however, is its operating expenditure, which is driven largely by salaries and wages. According to AfricaCheck (2018), approximately 59% of Eskom’s operating expenditure was accounted for by its wage bill in 2017, while Eskom estimates this to be between 40% and 50%. Numerous commentaries, including a World Bank study cited in Eskom’s 2018 Integrated Report, estimates that Eskom is over staffed by 66%, requiring only 14,244 employees to be run effectively (Eskom, 2019). While Eskom agree that this wage bill is unsustainable, it does acknowledge that some cutbacks (estimating overstaffing at 33%) are neccesary. This means that with or without the unbundling, Eskom will have to reduce their staffing component by at least a third. While such a move may possibly lead to a decrease in tariffs (or at least slower tariff increases), it could come with severe political and social ramifications, especially considering Eskom’s (and, the government’s) assurances that employment levels would remain the same post-unbundling. On the plus side, such a move could improve the utility’s credibility, translating to a marginal improvement in its sovereign rating outlook and improve costs of servicing debt.

Thus, even if operating expenses can be moderated, Eskom’s possibly increasing debt (given the upward trend of this debt over the last ten years) might continue to put upward pressure on tariffs, with the net effect on tariffs likely to be marginal at best (either up, or down). This is aided, of course, by the strict oversight of NERSA, which ensures that tariffs are set in an efficient manner.

1.4 Unbundling the debt

In order to successfully unbundle such an integral utility, the state should,inter alia, understand how Eskom’s debt will be allocated to the three separate entities. As such, part of this process entails allocating the debt across the new operating arms, including the expenses incurred due to the restructuring. In our view, there are three ways in which this can be done through the unbundling process:


  1. Proportionally allocate the lion’s share of the debt to the Generation arm of Eskom, making this arm unattractive to outside investment, but leaving the investment prospects of the other two arms unchanged.
  2. Equally apportion the debt to each arm, spreading the risk, and making the general prospects of investment into all three of the arms moderately or severely weaker.
  3. Create a fourth arm of Eskom in which all of Eskom’s bad assets and debt will be housed, making the other three arms (Generation, Distribution and Transmission) relatively more viable (Bischof-Niemz, 2019).


Proportionally allocating the debt would be informed by each arms contribution to revenue, asset base or costs. In other words, since 72% (±R549bn) of the regulatory asset base are Generation assets, it then follows that a similar proportion of the entity’s debt will be allocated to the new Generation arm. In this scenario, the state would likely continue to finance the debt and operational inefficiencies of the Generation arm. Alternatively, this scenario may see enough direct investment into the other two arms to cross-subsidise a portion of the debt accrued by Generation (which is not funded by the state). As such, it may be feasible to pursue this option if this cross-subsidisation (plus state funding) can eliminate sufficiently large parts of this debt while maintaining their own profitability (and,ipso facto, the integrity of the national sovereign rating outlook).

In scenario two, equally allocating debt across the three subsidiaries will weaken the balance-sheet integrity of both Transmission and Distribution arms at the cost of improving the integrity of Generation’s. As mentioned previously, given that Generation accounts for 72% of Eskom’s regulatory asset base, the equal apportionment will make the other two arms either relatively or wholly unattractive depending on their relative revenues and whether or not the state takes on large portions of this debt. This avenue will only be economically viable if the investment into Generation exceeds the potential loss of investment into Distribution and Transmission under scenario one.

Finally, creating a 4th arm of Eskom entails the pooling of all poor Generation assets, including all debt stocks across all three arms into one so-called “bad bank”. This scenario ensures that Generation, Distribution and Transmission become more viable investments. However, two possible problems make this a difficult solution:


  1. Because the new entity will not be profitable, the onus will fall onto the state to completely fund the operations of this arm. This burden, given current constraints on the fiscus, would be pushed directly onto the consumer in the form of tax hikes, whilst simultaneously putting the sovereign credit rating at risk.
  2. The creation of this new entity would not influence how the remaining three arms are managed and operated, and this could result in new borrowings by the three relatively lucrative and established entities. The state would then need to put in measures that ensure that all four arms are managed properly, and do not result in both a bad bank and an ailing set of Eskom sub-companies to take care of.


At this stage, it is difficult to say which option carries the greatest risk, but it is apparent that the state should be cognisant of its role in minimizing the risk to the economy as a whole. Additionally, government needs to face up to the possibility that the poor management of operations at Eskom could be deeply institutionalised, such that, despite the unbundling and bailout relief, management issues may continue to plague the utility after unbundling. Unless these management issues are fully addressed, debt levels could continue to creep up as operational performance deteriorates across all unbundled entities.

Moreover, since the state will be backing a large portion of Eskom’s debt, this will likely be funded through increased taxes (see Sen, et al. (2016), which examines the case in India). This could negatively impact South Africa’s sovereign debt ratings and lead to higher increased interest rates (itself, a contractionary policy which hurts consumers and producers alike). Even in the event that the state accepts a large portion of Eskom’s debt, this process by itself will not improve the utility’s management and institutional capacity.

That is not to say, however, that unbundling does not come with positive spill-over effects. Should the unbundling create an environment conducive for increased generation, the country will benefit from improved access to electricity, and reductions in tariffs in the long-term. Furthermore, if the increase in generation is met through greener solutions, this will have longer-term gains in the context of an international movement towards de-carbonising.

Beyond this, it may be that the unbundling process makes each arm relatively easier to manage and operate, bringing with it some institutional and organizational benefits. Whether this will be the case or not is, however, only going to come to the fore after the unbundling has taken place.

1.5 Closing Remarks on Unbundling

In order to unbundle Eskom successfully, the South African government has to coordinate the interests of a myriad of public and private stakeholders in what can only be described as a mammoth task. In unbundling Eskom, electricity costs might increase in the short to medium term in an attempt to deflate the debt bubble. More unfortunate are the retrenchments that will probably accompany this unbundling, especially given the SOC’s (and the state’s) assurances otherwise.

The long-term prospects of South African power security, however, may be positive. The unbundling could bring about efficiency gains; while increasing generation capacity and stimulating competition through alternative sources of energy will decrease the occurrence of unplanned and planned power cuts alike. With this comes improved competitiveness of local industries, and improved long-term production linked to long-term reductions in tariffs. Alternative energy sources will also bring about new industries and support services, creating employment in industries that previously did not exist.

The toughest question for Government to answer right now, is whether these benefits are likely to outweigh the costs. If so, the unbundling process should urgently proceed. But in doing so, this process must include a well-consideredaction planthat ensures that the ailing utility becomes sustainable as three (or, four) separate entities, whilst simultaneously keeping tariffs and taxes in check and boosting electricity production. If all of this can be done, then the answer for South Africa must be to bite the bullet, proceed with the unbundling of Eskom, and desperately hope that the shell does not fire.


AfricaCheck, 2018. Eskom and the Viral Infographic: Do the Numbers Add Up?. [Online]

Available at: https://africacheck.org/reports/eskom-and-the-viral-infographic-do-the-numbers-add-up/

[Accessed 10 April 2019].

Bischof-Niemz, T., 2019. Will Unbundling Eskom Ensure a Sustainable Electricity System?, Cape Town: EE Publishers.

Bloomberg, 2018. 4 Graphs that Show why Fixing Eskom is Ramaphosa’s Biggest Challenge. [Online]

Available at: https://businesstech.co.za/news/energy/276667/4-graphs-that-show-why-fixing-eskom-is-ramaphosas-biggest-challenge/

[Accessed 8 April 2019].

Bloomberg, 2019. Power shortages in South Africa could get a whole lot worse. [Online]

Available at: https://www.engineeringnews.co.za/article/power-shortages-in-south-africa-could-get-a-whole-lot-worse-2019-04-08/rep_id:4136

BloomberNEF, 2017. Brazil Power Wizard Takes Center Stage in Eletrobras Shock. [Online]

Available at: https://about.newenergyfinance.com/blog/brazil-power-wizard-takes-center-stage-in-eletrobras-shock/

[Accessed 8 April 2019].

Crompton, R., 2019. Explained: Why Eskom is in so much trouble. [Online]

Available at: https://www.iol.co.za/news/opinion/explained-why-eskom-is-in-so-much-trouble-19238470

Donnelly, L., 2019. Medupi and Kusile: Costly and faulty. [Online]

Available at: https://mg.co.za/article/2019-02-15-00-medupi-and-kusile-costly-and-faulty

[Accessed 15 July 2019].

Eberhard, A., 2013. An Overview of the Restructuring of the South African Electricity Distribution Industry (EDI), s.l.: s.n.

Eskom, 2017. Eskom 2018/2019 Revenue Application, Bloemfontein: NERSA Public Hearings.

Eskom, 2019. 2019/20 Tariffs and Charges. [Online]

Available at: http://www.eskom.co.za/CustomerCare/TariffsAndCharges/Pages/Tariffs_And_Charges.aspx

[Accessed 8 April 2019].

Eskom, 2019. Integrated report 31 March 2018, s.l.: Eskom.

Gosling, M., 2019. Eskom and the multi-billion rand mega projects that could have saved SA. [Online]

Available at: https://www.fin24.com/Economy/eskom-and-the-multi-billion-rand-mega-projects-that-could-have-saved-sa-20190213

[Accessed 3 July 2019].

International Finance Corporation, 2017. Creating Markets in Turkey’s Power Sector, s.l.: World Bank Group.

Mboweni, T., 2019. Minister Tito Mboweni: 2019 Budget Speech. [Online]

Available at: https://www.gov.za/speeches/budget_vote

Merten, M., 2019. BizNews. [Online]

Available at: https://www.biznews.com/budget/2019/02/18/eskom-risks-debt-gdp-ratio-60

[Accessed 8 April 2019].

Omarjee, L., 2019. Eskom gets R69bn in financial support over 3 years, but with conditions – Mboweni. [Online]

Available at: https://www.fin24.com/Budget/eskom-gets-r69bn-in-financial-support-over-3-years-but-with-conditions-mboweni-20190220

Ramaphosa, C., 2019. President Cyril Ramaphosa: 2019 State of the Nation Address. [Online]

Available at: https://www.gov.za/speeches/president-cyril-ramaphosa-2019-state-nation-address-7-feb-2019-0000

Schumpeter, J., 1942. Creative Destruction. Capitalism, Socialism and Democracy, Volume 825, pp. 82-85.

Sen, A., Nepal, R. & Jamasbm, T., 2016. Rethinking Electricity Sector Reform in Developing Asia: Balancing Economic and Environmental Objectives. ASARC Working Paper Series, Volume 2016/06, pp. 1-34.

Vagliasindi, M. & Besant-Jones, J., 2013. Power Market Strucutre: Revisiting Policy Options. Washington, D.C.: The World Bank.