Running the Numbers: ADB and OpenOil launch report on government modeling

Geneva, 19 October 2017 – The African Development Bank (AfDB) and OpenOil, a Berlin-based financial analysis firm, have jointly produced a new report on how African governments use financial models to manage oil & gas and mining projects. The report was launched at the 13th Annual General Meeting of the Intergovernmental Forum on Mining, Minerals, Metals and Sustainable Development (IGF) in Geneva.

Over 150 experts and representatives from international development institutions, governments, civil society and extractives companies attended the launch, including World Bank, United Nations Environment Programme (UNEP), United Nations Development Programme (UNDP), United Nations Conference on Trade and Development (UNCTAD), Organisation for Economic Co-operation and Development (OECD) as well as mining companies and miners’ associations such as Newmont Mining Corporation, AngloGold Ashanti, Anglo American and International Council on Mining and Metals (ICMM). The joint report was presented by Pietro Toigo of the African Natural Resources Center of AfDB and Olumide Abimbola from OpenOil. “This report is the first of its kind in Africa and we hope that it will stir debate within the continent’s mining sector and contribute to countries getting more out of their mining projects,” they told the participants from governments, bilateral and multilateral organisations, private mining companies, and civil society groups.

The report, Running the Numbers: How African Governments Model Extractive Projects,  analyses the capacity of 19 African resource-rich countries to use financial models, which simulate a simplified version of a real-world project in order to determine their financial benefits to the countries. AfDB and Open Oil conducted a survey of nearly 50 government officials to illustrate not only how widespread use of financial models is, but also how their results are utilised to inform policy, but also how they are used.

“Financial models are essential throughout the life-cycle of extractive projects,” said Johnny West, Director of OpenOil. “They are not just important during the development of the fiscal regime, but also for the negotiation of fiscal terms with companies, for revenue forecasting, and for auditing and tax-gap analysis.”

“This report not only stresses the need for African Governments to make efforts to close the information gap with extractive companies, but also shows where there are capacity gaps and how those gaps could be addressed, calling for development partners to invest further in capacity building.” recognized Modibo Traore, Officer in Charge of the African Natural Resource Center of the AfDB,

African countries also face a substantial gap in access to data that are key inputs for financial models, with the largest gaps seen in assessing information on capital costs and operating costs of projects.

In addition to the need for governments to build in-house financial modelling capacity, the report suggests that governments need to improve internal business processes and address the large gap that the report shows exist between information available to different agencies, departments and ministries.

“This study forms a crucial part of the Center’s support to African countries in realising the full potential of their natural resources”, said Modibo Traore, “How are countries supposed to enter, for example, negotiations with extraction companies -who all use financial models- if the government has not used the latest and best models to calculate what a potential project is worth?” added Pietro Toigo, macroeconomist at the African Natural Resources Center.

The report also encourages development partners to make capacity building in financial modeling a more significant part of their support to the management of extractive resources. Those partners who already do are encouraged to not just supply financial models as part of isolated technical assistance, but to also invest in equipping government officials with skills to create and use models.

Category: Blogs, OpenOil blogs · Tags:

Yaoure tax holiday costs Cote d’Ivoire $120 million

The publication of the financial model of Yaoure gold mine shows what can be done: both in the tricky area of what tax incentives offer, and in terms of building government analytical capacity.

OpenOil, with support from the German Corporation for International Cooperation (GIZ), worked with colleagues from Cote d’Ivoire’s Ministry of Mines to build the financial model and accompanying report. We believe this is the first time that officials themselves have substantially led the development of financial analysis.

The results are revealing at a couple of levels. Cote d’Ivoire has had a policy of a holiday on Corporate Income Tax in place for at least 20 years. During that time, many people have had much to say about the policy: successive governments have defended it as necessary, those opposed have denounced it for giving too much away – and sometimes effectively accused those who maintain the policy of having ulterior motives.

First of all, nobody had ever costed the tax holiday, and the model solves that.

Against a base scenario (explained in detail in the report), over $120 million of tax revenues will be foregone in the Yaoure gold mine in its first five years of production. That’s equivalent to more than a quarter of government revenues from the entire sector in 2015.

But the model goes a stage further. Because even knowing how much the incentive cost does not mean you can prove it was unnecessary. The case that if the holiday was not on offer the investor would walk has to be addressed.

This is where financial analysis comes into its own. The whole point of a government, or a CSO, building a financial model is to assess the business position of the investor – so as to know what constitutes a “fair” deal, what revenues governments can expect, and so on. This is done through estimating the rate of return, both before and after tax.

So, in the base scenario, the company (originally Amara, but Amara was then sold to Perseus Mining in Australia), earned nearly 30% rate of return after tax. Investors adjust for “country risk” and Cote d’Ivoire would certainly carry some of that. Nevertheless, it is a handsome return by any standard.

But that is with the tax holiday. What the Yaoure model then answers is: what would the investor’s rate of return be if the tax holiday were cancelled? And the answer is: a still fairly handsome post-tax Internal Rate of Return of 25%. This builds a strong case that the tax holiday is, in the case of the Yaoure mine, not necessary to provide an attractive enough incentive.

In other words it is money left on the table.

It does not follow, of course, that that will be the case in every mine. Elsewhere the economics could be different. Or not. What this project has done is deliver a tool to the government which enables it to run this kind of analysis, project by project, and lead an informed discussion about tax incentives in the country.

It is possible the results will need revision. We have based this model on data published by the company itself, but mainly in 2014 (A publication in 2016 was not comprehensive enough in an internally consistent manner to be able to build a model off). Costs and prices could both require adjustment which could change the overall numbers on the tax holiday question.

It is unlikely, though, that the numbers would change so much as to alter the basic conclusion – that the tax holiday at the Yaoure gold mine seems unnecessary. And if it did, it is the publication of the model which will have triggered the release of better data proving that case. Just as that better data could be easily slotted into this existing model.

The government in Abidjan is now considering how to deploy financial analysis more broadly. And maybe to require companies bidding for licenses to “show the model”, just as they are required to do with other documents.

This kind of financial modeling is available, and within the technical grasp of any government which would like to develop it. As the results from this one mine alone show, billions of dollars could be in play.

Category: Blogs, OpenOil blogs, Uncategorized · Tags:

Where are your country’s oil projects on the “Climate Change Supply Curve”?

The latest report from Carbon Tracker Initiative, a non-profit analysis house in London, suggest that up to a third of currently planned oil and gas projects may now never go ahead because of climate change policy.

In “2D of Separation”, CTI presents the potential risks to shareholders and investors in the world’s largest fossil fuel companies of “stranded assets”: the idea that if world leaders do what they said they would in the Paris Climate Change agreement, fossil fuel production will tail off. And that in turn means that plans for investment in new projects are riskier. And that in turn means companies who invest in new projects that would produce more carbon than be accommodated in the world’s carbon budget risk wasting their shareholders money.

We know that companies have been factoring in this risk of stranded asset for many years now. Global majors such as BP, Shell, Total and Chevron have been applying carbon prices internally to determine which potential projects are most at risk. The question is: who is going to do the same for the many countries around the world whose economies and public finance depend on fossil fuel production?

There are lots of factors in play, of course, price being the main one. New investment has taken a hit since the oil price crash of 2014, rebounding a little as the US shale industry has figured out how to cut costs and present more “agile” supply to the market. And there is no certainty that either policy or the ever dropping costs of green energy will guarantee that the commitments made by governments in the Paris Agreement will be honoured.

Nevertheless, if the Kashagan oil field in Kazakhstan gets stranded, stranding any investment by Shell, the government of Kazakhstan will see its benefits from the project stranded. The same with ExxonMobil’s investment in the Bonga field in Nigeria, or Oil Search’s potential investment in Papua New Guinea. All around the world governments are making plans for economic growth and revenues to Treasury which are at risk, and there is little sign that they are receiving professional advice to help them plan for that. The case is even more pronounced with LNG, presenting risk to governments like Mozambique and Tanzania, which are making plans on the basis of vast new offshore projects.

And it is not just that some projects may not go ahead. There will be other projects where carbon budget considerations make the project more marginal but potentially viable – if the companies can convince countries to restructure terms. Companies could propose deals which maximised revenues to them in the short term, offering governments a larger share of revenues as the project progresses, some time down the road. But if governments are not aware of the overall business environment, what happens if by the time the government’s higher share kicks in, the whole game has moved on? That high return in the future will never, in fact, arrive.

Category: Blogs, OpenOil blogs · Tags:

Reputable Mining Investors Wanted in Sierra Leone: How to Improve Integrity Due Diligence in Mining License Management

On the first anniversary of the Panama Papers, the National Minerals Agency (NMA) of Sierra Leone wants to know who they are doing business with. The NMA is the government agency responsible for issuing mining licenses. Beny Steinmetz of Simandou fame, is one such mining investor who has prompted the NMA to strengthen integrity due diligence. The Panama Papers implicated Steinmetz in a complex chain of ownership of Koidu Holdings, Sierra Leone’s largest diamond mine. By checking the reputation and track record of applicants, the NMA hopes to weed out shady investors, and instead attract reputable companies to transform Sierra Leone’s mineral wealth. This ambition is echoed by the EITI’s call for disclosure of the ultimate owners of extractive companies, as well as the launch of a global register of beneficial owners by OpenOwnership.

But, is it realistic to expect a resource constrained country such as Sierra Leone to be able to reliably test the ‘integrity’ of mining investors?

The short answer is yes. From October 2016 to March 2017, OpenOil was contracted by the Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH to work with the NMA to strengthen integrity due diligence in mining license management. We found that prioritisation of resources, smart use of commercial products, and the ability to judge risk, enable the NMA to build a reasonably comprehensive picture of the integrity (or not) of potential investors. The expectation should not be that the NMA will always uncover the ultimate owner, or chase down the last dollar. The goal is to make informed decisions about who gets access to Sierra Leone’s natural resources.

Here are four lessons we learned that may be useful to other countries.

1. Prioritise mining license holders for ongoing integrity checks.

Developing country governments may not have enough staff to run integrity checks on all mining license holders. The NMA has four compliance staff to oversee 175 industrial mining licenses, and 2,078 trading licenses. Thankfully, not all license holders pose the same level of risk to the sector, or the wider economy. With the NMA, we developed a way of ‘triaging’ high-risk license holders for regular integrity checks. Regular integrity checks are important because the business profile of a license holder may change over time (e.g. sale of asset).

Here’s what we did:

a) First, we identified a set of risk indicators to measure all license holders against, based on data the NMA already collects via its Mining Cadastre System (e.g. contribution to total non-tax mining revenue);

b) We gave each indicator a weight based on level of risk, and reliability. For example, variation in sale price has been given less weight than license type, or presence in a low-tax jurisdiction, this is because there may be numerous legitimate reasons why a company’s sale price would be lower than the market price (e.g. differences in quality), making this indicator less highly correlated to risk;

c) Finally, we proposed an algorithm and scoring system for each indicator. The NMA is currently embedding these rules into its data analytics system. The result will be an automated list of 10-12 ‘watch list’ companies that requiring ongoing monitoring by NMA compliance staff.

Our aim was to establish an automated system that can generate a reasonable assessment of whether an existing mining license holder should be subject to ongoing integrity checks. Where the results appear incorrect, NMA staff can add or subtract ‘watch list’ companies.


Integrity Due Diligence Risk Matrix


2. Commercial databases must be evaluated based on their specific application to the mining industry in the source country.

There may be limits to what open source data can deduce about an investor (e.g. where an investor is wholly owned by a company in a financial secrecy jurisdiction). The response to this challenge is usually that countries should invest in a commercial database from Reuters, or Bureau Van Djik. Yet, for many developing countries, including Sierra Leone, the cost of a database (anywhere from £5000 to £25,000 per year) is hard to justify. To decide whether to invest in a subscription, government agencies should test the specific application of the database, to the mining industry in their country. In doing so, they should consider the following issues:

  • The volume of mining companies that need monitoring. It may be that government would use only a fraction of the information on the database, despite paying for the whole resource;
  • How much information the database contains on privately held companies, that is not publicly available from the local registry. The main reason to subscribe to a database is to aid due diligence of privately held companies. Yet, if the database is so scant that government would need to purchase extra data elsewhere (e.g. from a consultancy), it may not be a good investment;
  • The extent to which the database uses public information to identify Politically Exposed Persons. It may be that this information comes from media and news websites, company websites, and sanctions lists, all of which are publicly available;
  • If the mining regulatory agency can establish the corporate ownership structure of mining investors i.e. identify all legal entities, directors, officers, shareholders etc. Some databases rely on users first knowing who the legal entities, and individuals involved are, to be able to search them. Developing countries should avoid databases that need a high level of prior knowledge of corporate structures.

A commercial database might be feasible if the EITI Secretariat and donor partners were to buy a collective subscription to license out to EITI implementing countries. This would be more cost effective, as well as help countries to implement the EITI standard on beneficial ownership.

3. Countries with a small number of mining investors may find purchasing one-off due diligence reports more cost effective than subscribing to a database.

If a government agency finds closed data necessary to conduct due diligence, a more cost effective and targeted solution is to buy one-off due diligence reports for specific mining license applicants. However, there are costs involved with this approach, therefore it should only be done for high-risk privately held companies (e.g. investors linked to State Owned Enterprises), once the compliance team has exhausted all possible means of researching the company. An advantage to this approach is that it ‘kills two birds with one stone’ by getting a report, as well as translation of local registry documents and media.

Depending on the level of reporting, standard reports range between £100-£800. Very sophisticated reports can go up to $30,000. The cost comes down to how much source commentary, or ‘on-the-ground’ investigation is involved. In most cases, a standard due diligence report (company registration, adverse media results, sanctions and enforcements, basic information on directors and shareholders) should be enough for integrity checks, unless there are very serious concerns, or challenges to accessing company records.

For a country like Sierra Leone, where there is a handful of significant mining operations, buying one-off reports is undoubtedly more cost effective than subscribing to a database. This may not be the case for other countries with a more developed mining industry. However, even these countries may find it strategic to purchase one-off reports given databases can be variable on privately held companies.

4. Countries seeking disclosure of ultimate owners require a legal framework for beneficial ownership. In the meantime, government agencies can use other data points to inform integrity due diligence.

Whilst the NMA has ambitions to know who the ultimate owners of companies are, as it stands, the law only requires disclosure of shareholders who own 5% or more of the issued share capital. There is a legal review currently taking place, which will likely result in comprehensive beneficial ownership legislation. However, integrity due diligence does not have to wait until legal reform takes place. A lot can be deduced from information already collected by government agencies. For example, the NMA can already detect some PEPs issues based on the shareholder and related party data it collects from license applicants. Bearing in mind their current legal framework, government agencies should update existing disclosure requirements to maximise the due diligence relevant information they are already entitled to.

* * *

Effective due diligence of investors is critical to countering corruption, tax abuse, and criminal activity in the mining sector. Due diligence solutions should be simple and cost effective, considering the resources and capabilities of the government agency responsible for administering them. The mark of success is not a government agency knowing every legal and natural person involved in an investment. But, whether that agency can put forward an informed and reasoned view of the level of risk associated with an investor, such that a decision to grant a license (or not) can be made, and areas of concern flagged for ongoing monitoring.

Alexandra Readhead was the OpenOil Team Leader for the NMA due diligence project. She is an independent consultant in international taxation and extractive industries.

Category: Blogs, OpenOil blogs · Tags:

Is Tanzania’s export ban a high stakes game of poker with Acacia?

OpenOil’s Excel model on the Bulyanhulu mine  is available to view the workings of the conclusions in this post. 

Tanzania’s decision to impose a ban on concentrate exports from its mining industry has gained wide coverage among industry watchers. The question is: what does the government hope to achieve with it, and what will companies’ reaction be to it?

In this piece, we explore the possibility that the ban particularly targets Acacia Mining’s gold production in Tanzania and that the objective is to pursue re-negotiation to increase revenues to government from Acacia’s three large gold mines, which generated turnover of over a billion dollars in 2016 but have yet to pay any corporate income tax.

We further consider the impact of the ban on Acacia, and on government revenues. Financial analysis suggests the following preliminary conclusions:

  • The government will lose $1 million or more per month in revenues, as falling sales translate into lower royalties.
  • The economics of the Bulyanhulu mine, which were marginal anyway, could be pushed over into operating loss (depending on gold prices), leaving the company with a tough decision about whether to shutter the mine, and that decision point could come in the next few months. Of Acacia’s three operating mines, Buzwagi could also face difficulties while North Mara lies at the other end of the spectrum and could continue to run at operating profit.
  • Acacia’s overall financial position could be materially affected as lower revenues and profits affect its leverage, and erode investor confidence. Shares dropped by 20 percent in the week following the government’s announcement of the closure and have only partially recovered.

Each of these points is dealt with in detail below.

The background to the ban on concentrates

Exports of concentrates and ores of all metallic minerals were banned on Tanzania on March 3rd.

Thomas Scurfield from the Natural Resource Governance Institute has rigorously laid out how the ban on mineral concentrates fits schematically into a broader push by many producing countries to increase beneficiation. In theory, more processing of raw commodities in-country helps add value and convert the mining industry into a broader engine of wealth creation. Against that has to be balanced the amount of capital and time needed to create such facilities in Tanzania from scratch and estimates of how much extra value would be created in this instance – from these mines, and with these commodities.

Scurfield’s article shows that in the Tanzanian case the specifics look challenging: it would take two to three years to establish a facility, a previous feasibility study suggested the economies of scale were not there to make a copper smelting plant viable, and the level of value added in the case of copper is also debatable.

At least according to media reports, the government’s decision seemed related to another issue, that Tanzania was not receiving enough income from its minerals sector, and that better physical auditing was needed as part of measures to redress this.

Acacia’s mines are the only scaled operations which are immediately affected by the ban. The other main producer of gold, Anglo-Gold Ashanti, does not export concentrates. There are plans for several large nickel projects in the future, which could also be affected, but these are several years away from execution and actual export.

Separately to the ban, the government of Tanzania has had a long-standing tax dispute with Acacia over payment of corporate income tax. The Tanzanian Revenue Authority raised a case against Acacia for $41 million withholding tax on dividends paid out to shareholders of the UK-based holding company. OpenOil issued an analysis of the tax dispute last June. Acacia’s last public comment on the case was in October 2016, when they said they would refile an appeal in the Tanzanian court system against the ruling. Because the Bulyanhulu contract has not been published, it is not immediately clear if the Tanzanian courts are the ultimate arbitrator in the dispute, or if some other arbitration or adjudication procedure is also specified.

Tanzanian loss of income: over a million dollar a month

The government of Tanzania is losing $1-1.2 million a month in royalty payments as a result of the export ban.

Acacia has stated that the export ban on concentrates affects 30% of its gold production. In 2016, the company reported paying royalties of $47 million. The simplest approximate calculation of the effect of the ban, then, is to assume a straightforward correlation between the percentage of production shut in by the ban and value in the market: $14.1 million a year, or just over a million dollars a month at current market prices.

If the ban continues for some time this revenue is unlikely to be recoverable as and when it is lifted, since there is a limit to how much concentrate Acacia can store at the mines, so it is likely to curtail production.

Will Bulyanhulu tip into operating losses?

Of Acacia’s three mines, the ban has the most drastic effect on Bulyanhulu where, according to Acacia, concentrates account for 45% of the mine’s production. The ban could push the mine over the economic limit, where Acacia would be running operating losses in order to keep producing at all.

To some extent the marginal economics of the mine, relative to Acacia’s other holdings, has become clearer this month with the publication of the company’s 2016 annual report. These have led us to revise near term future forecasts for revenues and positive cash flows down from our first publication of financial analysis of Bulyanhulu, which were based on an investor presentation published by the company in March 2015.

Acacia has for some time been advertising capital investment and management overhaul strategies to make the mine more profitable. In the 2015 investor presentation, still the company’s last public statement of forward looking estimates, it anticipated the pay off from these efforts to be a rise in production from 195,000 ounces in 2013 to 380,000 ounces by 2017, and a drop in cash costs from $890 per ounce produced in 2013 to just $480 per ounce in 2017. It was these two effects combined which would push the mine, at long last, into free cash flow.

But the company’s annual reporting since has given a different picture (see table below).

Bulyanhulu report vs estimated

Production has risen, and costs have dropped, but by considerably less than was predicted two years ago. In particular 2016 cash costs at the mine were $722 per ounce, substantially higher than the $540 per ounce anticipated in the 2015 investor review (we assume here that the reporting basis of these metrics remains constant, since the company has provided no information to suggest the contrary, as would be expected under normal reporting practices).

Accordingly, OpenOil has revised down revenue and earnings forecasts from Bulyanhulu. We still model the general trend of the company’s assertions about future performance at the mine – that production will go up and costs will come down – but by less and over a longer timeline than originally assumed (see table below).

Bulyahnulu financial model assumptions

If Acacia’s 2015 forward estimates had held the mine could have expected to generate around $150 million in 2017 in cash flow for the company at current market prices. But under revised assumptions this drops to just $25 million. This could rise to $75 million by 2019 and go over $130 million in 2020. In terms of the tax issue, these lower assumptions would push back the point at which Acacia would pay income tax from 2020, projected in last year’s iteration of the model, to 2022.

And that is before the concentrates export ban.

If we assume that Acacia’s statement that 45% of gold at Bulyanhulu is in concentrate form means the mine will simply produce and sell 45% less (something which remains to be confirmed from a more precise understanding of the geology of the mine), Bulanhulu’s cash flows turn negative. At current market prices, the company could run an operating loss of $40 million in 2017 from the mine. It seems unlikely the export ban is going to remain in place for years without movement one way or another – either a partial or total lifting of it, or, by contrast, the move to build facilities and therefore unlock the concentrate-based production. However, theoretically, if it did, Bulyanhulu might not get out from under operating losses until the next decade.

Interestingly, under the same assumptions about the impact of the ban on production, but keeping to Acacia’s original 2015 forward looking estimates on production and costs, Bulyanhulu could stay cash positive over the next few years with a ban – but only just. Positive cash flows could be in the region of $30 million a year for the next three years. The company would therefore not face a question of whether it needed to shutter the mine as it approached the Economic Limit.

Bulyanhulu Cash flow scenarios

North Mara and Buzwagi

We have not conducted full financial analysis on Acacia’s other two mines. One unresolved question here is how much production from concentrates come from these two mines, since Acacia has not specified this for them, unlike Bulyanhulu. Piecing the company’s statements together, it would seem as though an additional 115,000 ounces of gold production at the two mines could be locked in by the ban (since according to the company 30% of overall production is affected – about 245,000 ounces in 2016, of which Bulyanhulu, with gold production from concentrates at 45%, accounts for 130,000 ounces).

Nevertheless, North Mara’s clear status as the group’s cash cow, with the highest production and a materially lower cost base, makes it unlikely Acacia would face the same issue of operating losses there.

Buzwagi at first sight also could be precarious. There is concentrate production there but the proportion was not immediately known. The mine is approaching end of life, but has the highest all-in costs of the three mines – although its vulnerability to price volatility may be lower since the company locked in pricing agreements which apparently covered all of first quarter 2017’s production at a floor price of $1,150 per ounce.

Nevertheless further analysis needs to be done to determine the vulnerability of both these mines to a continued concentrates export ban.

Bulyanhulu rev scenarios

Acacia’s overall position

Clearly the concentrates export ban, if it shuts in 240,000 ounces or more of gold production in 2017, could materially affect Acacia’s financials as a whole. At current market prices it would lead to a loss in gross revenues of $300 million.

One of the reasons Bulyanhulu could be pushed into net operating losses is because the mine operation is still leveraged. Acacia’s 2014 annual report includes a reference to a $142 million loan. Terms of that loan are not available, but it is possible Bulyanhulu Gold Mine Limited, the Tanzanian entity operating the mine, could be repaying between $15 million and $30 million in interest and principal. Further research would need to be done to determine the company’s overall leverage. Acacia reported a net cash position of $218 million at the end of 2016, considerably better than 2015’s $105 million, so it is better placed to face loss of cash flows than it was a year ago.

Despite Acacia’s stronger 2016 performance, the ban still comes at a delicate time for the company. Market rumours persisted throughout 2016 of talks with South African firms of a possible sale of the company, whose majority position (63%) is held by Barrick Gold. When news of the ban first broke, shares in Acacia dropped 20% within a week. In the following three weeks they recovered half of that. But the impact of the ban itself is likely to make investors skittish, both in and of itself and in relation to what it might portend in terms of ongoing relations between the company and the Tanzanian government.

Category: Blogs, OpenOil blogs · Tags: