Financing the Mozal Project in Mozambique

by Maximilian Wegener (Author) Ivan Ivanov (Author) Frederik Schreus (Author)

Elaboration 2013 16 Pages

Business economics - Investment and Finance



We go back to the year 1997 - the year in which two companies, a base metal company and a electricity company, several development banks and government development organizations, commercial banks and governments itself are in talks about the building of a new aluminum smelter in Mozambique. The project US dollar (USD) volume amounts to a total of $1.4 billion USD.

Mozambique is one of the poorest countries in sub-Saharan Africa, which just recently ended their civil war. Additionally, Mozambique has the lowest per capita GDP of $90 USD a year in their peer group, a debt burden resulting from the previous civil war years of 355% of GDP. It ranks 166 out of 174 countries in the United Nations Human Development index.

Eskom, a south-African energy company, which supplies 95% of the energy in south-Africa and 50% of the continents power, wanted to establish hydroelectric generators at the Zamibi river and rebuild the worn out electricity network, to supply power at competitive prices. The cooperation of Eskom and Alusaf secured Eskom a buyer for its excess power and Alusaf secured a critical factor of production at competitive prices.

Alusaf, an aluminum subsidiary of Glencore, a south-African company operating in the precious metals and base metals industry, and IDC, a governmental owned development bank, are the main sponsors with an equal $125 million USD investment each. (A $125 million USD investment, amounts to 8.92% of the project value and 25% of total equity). The project is supposed to be financed with a 50% debt and 50% equity. Alusaf already built and operated a larger project of the same type, an aluminum smelter with double the volume - the Hillside smelter in Richards Bay with a yearly capacity of 500,000 tons. It accomplished it 21% below budget and 4 months before planned completion.

Senior debt holders would provide $680 million USD or 50% of total financing. The loans were supposed to be covered against commercial risk and political risk, such as expropriation, war, breach of contract etc. However, they did not guard against “hidden” expropriation such as creeping expropriation i.e. raised taxes by the host government.

To reduce and overcome the political risk, the International Finance Corporation (IFC) was asked to analyze the project and to be a quasi sponsor of the project. They are supposed to give a subordinated loan of $65 million USD with the interest rate rising as sales accelerate and profits rise, but with a option to the project to suspend interest and principal payments in bad times. The other $85 million USD of the $150 million USD subordinated loan coming from another source to be determined. In addition, IFC invests $55 million USD in senior loans.

The new smelter was planned to have a capital cost of $4,750 USD per ton compared to $4,850 USD on average. Capital cost refer to the money it costs to construct the capacity to produce one ton of aluminum. The plant is located in a free trade area with a 1% tax rate. Moreover, the cost of producing a ton of aluminum including depreciation and financing cost amounts to $1,493 USD in the first year, declining to $1,070 USD in the 11th year. This compares to a peer company average of $1,750 USD before depreciation and financing. The plant would have relatively stable breakeven quantities, since the alumina price, a input compromising approximately 33% of input per ton of aluminum is indexed to the aluminum price. Furthermore, the other major input, electricity, which compromises 25% of total cost was negotiated with Eskom at competitive prices in the first years and also indexed to aluminum prices in the later years. This implies that 58% of the costs move with the market price of the output.

Question 1 (a): Should Alusaf and Gencor invest in the Mozalproject?

Alusaf should evaluate the project on the basis of financial indicators, such as the NPV, compare the IRR to the required return on equity and consider qualitatively, by judgment, the different options, risk mitigations in contracts and due to the separation of power. Before a comparison of the IRR and the required return (RR) can be done, the rate of return is estimated with a modified version of the original CAPM model, which can is shown by the equation RR = Rf +ß* (Market risk premium + country risk premium) (CFAI, 2013). Thereby, the market risk premium is the excess return of the market, which is the US equity market risk premium, over the risk-free rate, which is estimated as the 10-year TIPS rate. The inflation-protected rate, the real rate, is thought as the correct one, since historically commodity products are highly correlated to inflation. The prices of these products tend to increase when inflation is high and tend to decrease due to deflationary pressures. The country risk premium (CRP) is determined by the credit spread of 10-year US treasuries and Nigerian Brady bonds adjusted to liquidity. With this approach, the CRP is assumed to vary according to the project’s risk. Nigerian Brady bonds are a good proxy, since their country risk rating was determined as 14.8 (as of March 1997) by the institutional investors magazine, while Mozambique’s rating was determined as 14.9 (as of March 1997).

Brady bonds are dollar denominated bonds of emerging countries or least developed countries. Therefore, they are comparable since they are both denominated in US dollars and internationally traded. However, we adjust the country risk premium downwards for liquidity reasons. The assumption here, that Nigerian bonds, even though denominated in US dollar, might carry a liquidity premium because they definitely have lower volume and trading activity than US treasuries, which are one of the most frequently traded securities in the world. In the previous year the yield on Nigerian Brady Bonds varied in a band of 13.3 % to 15%. We decided to use the lower end boundary since recently the country’s institutional investor risk rating doubled, improving from March 1990 until March 1997 by 7.4 to 14.9. This resulted from the end of the war, since GDP and FDI were increasing, while inflation was falling. However, as already mentioned, 200 basis points get deducted for liquidity reasons from the overall spread.

In the case of the Mozal project, the asset beta, which is obtained as the average of comparable US companies, is not relevered. The asset beta explains the sensitivity of all assets of the firm to the market. Normally, a beta gets relevered to reflect differences in the capital structure of a company since the cost of equity increases with a higher debt burden because of higher distress cost and asymmetric information. However, the distressed cost in this case is highly reduced, because with an interest coverage ratio steadily greater than 1.5 for senior debt, rising in later years towards 12, which is not likely to deteriorate due to changes is input and output prices because they are indexed. Furthermore, the distress costs are reduced because of the use of project finance since it simplifies restructuring issues because of only a few participants, especially a low number of debt holders.

Moreover, the asset beta consists of a weighted average of debt and equity betas: ßasset = ßpmiitv t + ßrfpht- *7—: . However, in our case we assume that the asset beta is requuy (i_t) *d+E raeoi (1-T) D+E equal to the equity beta, since there is much less risk involved in the debt and the debt is not very sensitive to market movements. On the one hand, the debt is highly collateralized and on the other hand insured. Therefore, beta debt is assumed to be zero.

The CAPM calculation (including the CRP) gives a required return (RR) of 11.21% = 3.57% + 0.78[5% + (13.35% - 6.56% - 2%].

A more sophisticated way, but out of scope of this project would be, to adjust the credit spread of the Nigerian Brady Bonds by the ratio of the annualized standard deviation of the equity index in the developing country divided by the annualized standard deviation of the sovereign bond market in terms of developed market currency. This ratio adjusts the country risk premium for volatility differences in the different asset classes and countries.

Next, we examine the projected cash flows and flows to the equity holders. As one can see from the table below, two scenarios are calculated. Both take into account the cash flows to equity holders and “quasi” equity holder. The quasi equity holder, as mentioned earlier, hold subordinated debt, structured to have some features of equity such as participation in profits in good times or suspension of interest and principal payments in bad times. Therefore, the cash flow to equity holders is compromised of dividends, interest on subordinated debt, and any principal payments to subordinated debt holders or equity buybacks (there are no equity buybacks in our case, so the company/project doesn’t return profits in this form). We include the initial cash inflows into the project as negative cash flows for equity holders. This allows us to calculate the IRR. The IRR is the rate of return that sets the NPV of the project equal to zero and it assumes that all cash flows can be reinvested at the IRR. Normally, a discount rate lower than the IRR makes the project a profitable one, every discount rate that exceeds the IRR makes the NPV negative. Since our discount rate is the RR, there is a large gap between the IRR of 7.23% and the RR of 11.21%. Therefore, the project should not be undertaken from a purely financial point of view.

The second scenario includes the terminal value as of 2012 of the project calculated as Ц#[3]. As the case does not state if the plant is scrapped after the 2012 period we assumed going concern in the future. We also assumed that the CF2013 is equal to CF2012 = $170m because the cash flows are not expected to grow or be lower since all investments and principal payments are completed. In that scenario an IRR of 12.67% would result. This rate is higher than the RR of 11.21 %.

illustration not visible in this excerpt

However when dropping the going concern assumption after 2012 and assuming again scenario one, which is more realistic given the long time horizon and the uncertainty inherent in the project, the discount rate determined with the CRP from scenario one does not include the fact that IFC’s involvement reduces much of the country risk. Research confirms, that organizations that are associated with the World Bank, involved in projects in developing countries, reduce country specific risk. This results from governments’ fear that the World Bank, as the lender of last resort for developing countries, might take any counteractions that harm their economic ability.

If one estimates another CAPM equation without the CRP one arrives at a discount rate of 7.47% = 3.57% + 0.78 [5%\. This is very close to the IRR of 7.23%.

However, Alusaf should take several other factors into account when deciding whether or not to undertake the project. They use the same suppliers for other input materials than labor, electricity and alumina. This implies that Alusaf will become a larger customer to these suppliers, which increases Alusafs bargaining power. Furthermore, Alusaf will manage the day-to-day operations of the plant. Therefore, additional management fees for Alusaf will be generated, but more importantly control over operations is obtained. The growth option to expand the plant to a capacity of 500,000 tons a year, making it as big as the Hillside smelter, is another impact Alusaf should consider. This real option is valuable, especially in light of the already established capacity for a 500,000-ton plant at the berth, where most of the inputs and outputs arrive or depart via sea freight. In addition to the proximity of the estimate, in the past Alusaf was able to complete similar projects 21% under budget and 4 months earlier. Moreover, analysts forecast new industry demand of five million tons in the next ten years, in addition to a 2-3% yearly demand growth. Growing demand implies higher prices in the short run, since it takes time to build up industry demand in a capital-intensive industry.

From the discussion and valuation above one can conclude, that Alusaf should invest in the project as long as the IFC is involved since the project’s IRR is very close to the required return, has some valuable growth options and capitalizes on Alusafs core competencies and experience.

Question 1(b): What are the greatest risks? Have they been adequately addressed?

Undertaking a project like the Mozal Project in a different country bears many risks and exposes investors to different country level threats that have to be dealt with in advance in order to secure enough financing funds. In general, one can differentiate between pre-completion risk, post­completion risk and the, in this case most important, sovereign risk.

Firstly, the pre-completion risk deals with the general circumstances before the project actually starts and influences the set-up of it. In the case at hand, the project should be fulfilled in Mozambique, which is one of the poorest countries worldwide, and is therefore quite risky. The infrastructure within Mozambique, especially after 17-years of civil war, is not developed and can delay the time of completion due to inefficiencies in transportation or electricity. Moreover, there are highly complicated bureaucracies combined with a high degree of corruption (Corruption Perception Index, 2012) hampering a smooth process of obtaining the required permits. Another critical risk concerns the financing of the project. The sponsors of the project plan to use the generated cash from the start-up phase for a further investment of $35 million, which can be very risky. However, they counteracted this risk by planning a $75 million contingency budget for the construction period as a safety net.



ISBN (eBook)
ISBN (Book)
File size
1.2 MB
Catalog Number
Institution / College
Maastricht University
mozal case project finance



Title: Financing the Mozal Project in Mozambique