IMANI Special Report: Ghana's “Integrated Aluminum” Fever

5th February 2010

*Key Insights*

1. A rigid devotion to developing a “closed integrated aluminum complex” in Ghana can be economically unsound since other more open models that are less driven by notions of heavy industrialization at all cost could produce more benefits for Ghana. There isn't just one, tightly-woven, approach to obtain value from our bauxite.

2. The business and political intrigue and machination going on in the aluminum sector is intertwined with developments in the oil and power sectors, but they should be no cause for worry in an open, transparent, and competitive regime.

3. The recent deal in which the Chinese purchased Ghana's key bauxite assets in Awaso should be scrutinized closely, and all terms should be disclosed since Ghana may end up the loser. It raises questions about VALCO and the economics of energy generation and distribution to power the industrialization all political parties support.

*Introduction: the allure of the white metal*

As national dreams go, none has been as compulsive in the case of Ghana as the desire of our most powerful leaders to realize the long-held ambition of erecting an integrated aluminum complex in this beloved country.

Indeed, one Ghanaian commentator, citing James Moxon's account of the geopolitics of Ghana's early aluminum ambitions, described Nkrumah's aluminum fervor as one of the primary motivations behind his promotion of the Ghana-Guinea-Mali union.

If Nkrumah was the brain behind the post-independence dream of an integrated aluminum industry in Ghana, it was Edgar Kaiser, the American industrialist, who was the financial and political muscle behind Valco and to a certain extent the Akosombo dam too (the Americans paid half the tab), and therefore the man who could most impair Nkrumah's visions. And every socialist worth his salt concludes that impair he did, with the active support of the American imperialists of course. Such a view thus readily mutate to approval of the growing Chinese prominence in the way the aluminum story is panning out here nowadays.

One supposes that more adventurous commentators feel at ease linking the recent ouster of Guinea's junta leader to certain dimensions of the geopolitical struggle over aluminum. After all, it was at the height of political confusion in that country that he decided to approve the ceding of operational control over nearly all of Guinea's bauxite reserves to a private Chinese consortium. The same Ghanaian commentator cited above argues at one point that the industrial interests behind Ghana's bauxite story from the early 60s were also the dominant forces in Guinea's case all along. And for added measure all these vested interests have been American. As far as observers of this persuasion are concerned global control over aluminum, a major input into modern industrial manufacturing, had been the endgame of Nkrumah-era African geopolitics, with only mild attenuation thereafter.

There is no doubt that serious politics continue to suffuse the aluminum story of this country. Perhaps, at least as far as Ghana itself as a country is concerned, the domestic politics has superseded the geopolitical, and we emerge now upon an era of fierce contest – part healthy business competition, part political machination of the “commanding heights” type – in which the obvious prize is dynastic and clan control of a future, multi-billion dollar, integrated aluminum industry in Ghana.

There can be no doubt that there are three industrial sectors in this country that can most rapidly breed the afore-hinted dynastic and clan dominance of economics and politics usually associated with, for instance, the Asian concepts of Chaebol and Keiretsu. We are referring to an idea of giant conglomerates with influence extending beyond the realms of business and finance into culture and politics. Even more vividly, the allusion is to the imagery conjured by such names as “Rockefeller”, “Ramaphosa”, “Oppenheimer”, “Al-Sabah”, “Rothschild”, “Pierre Du Pont” etc. etc. What we have seen in Ghana in the past, in the forms associated with such names as “Addision”, “Siaw”, “Opoku”, and “Appiah-Menka” are of a level several magnitudes lower, exactly because the essential element of “resource control” as constituting the basis of an integrated wealth-power nexus in this particular usage appear to be missing in the above-mentioned Ghanaian instances.

So what are these industrial sectors? You must have guessed already. They are Gold, Oil, and of course Aluminum.

In the case of gold, the situation has been severely distorted by the colonial and anti-democratic pressures of our history. Largely passive or complicit in the face of geopolitical torrents, the elites sacrificed their long-term interests for myopic gratification and watched confused and impotent as largely external interests flourished in the gold sector. The political economy of liberal democracy throws up one interesting advantage for multi-party politics. Despite superficial indications to the contrary it actually promotes strategic thinking on the part of elites who, through the pressures of natural selection, should become more sophisticated in their pursuit of their clan and dynastic interests.

The aluminum and oil sectors, through their “providential backwardness”, are today blessing us with the opportunity to witness resource competition amongst the elite in the sophisticated setting of what our friends on the left side of politics would call “bourgeois democracy”. In recent times, the competition in the oil sector, rather tiresomely in the view of the elite, is spilling out into the open and getting uncharacteristically loud and some would even say nasty. The competition over aluminum remains more, characteristically, sublime.

In this report, we shall be shining some light on aluminum.

We shall be providing a general, semi-structured, account of major developments in the aluminum sector in recent times. While there would be attempts at historical analysis and even at scholarly critiques of strategies at governmental and enterprise levels, the primary emphasis is on education - lay public education. By the time we are through we would have succeeded in our motive if the average citizen who has sacrificed precious time by reading this is able, only by having read this report, to contextualize news items she hears about bauxite or aluminum against the backdrop of Ghana's economy or the government's industrial policy.

*1. The Opaque Bosai - Rio Tinto Deal*

For many of the people active in the aluminum story that is the subject of this account, a quiet announcement on 4th February 2010 that received little fanfare in the Ghanaian media must have cut through their morning routine like a hot knife through butter. Some must have dashed for the mobile phones and hit speed-dial, while others probably just reclined back in their expensive swivels and closed their eyes in deep reflection.

The Chinese private—backed mid-sized Chongqing-headquartered mining firm, Bosai Mineral Group Co. ltd, had, with the support of their home government, just bought out Rio Tinto's 80% majority stake in Ghana's main bauxite concern, Ghana Bauxite Company.

The going price was $30 million. No details about the negotiations had been released since the deal was first hinted 4 months ago. Outside the clique that engaged in the determination of the terms, no one knows what the offer or asking prices were. The full technical considerations which had led to approval by the Government of Ghana (GoG), the recommendations of the Minerals Commission, and the policy briefs from relevant bodies were likewise undisclosed.

Now if you are an independent observer like we are, with a public education mandate sort-of, nothing annoys you more than this type of institutional arrogance. It is understandable that the private, foreign, parties to the deal, Rio Tinto and Bosai, may choose to only release the barest of details to the public. But you can bet everything you have that their shareholders will be privy to every tiny detail of this deal. We, the citizenry of Ghana, are the government's shareholders, and it is our right to be fully informed about decisions to which GoG is party to or ought to be party to.

The issue of transparency having now been positioned at the center of this issue, one may proceed to take the license to use whatever information is available in the public domain to examine the contours of this deal.

*2. Basic Financial Terms of Deal*

From a purely financial point of view the deal appears to make no sense. Based on the 100 million tons of bauxite reserves quoted by Bosai, it is clear that the gross (naïve) valuation of the GBC asset base is, per the average historical price of bauxite, is at least $2.5 billion. Bosai sees a total production cost for the resource as $20. If we assume a 30 year lifecycle for GBC's asset base, and using the Bank of China's prime rate as the support for a discount factor, then without comprehensive IRR (internal rate of return – an investment valuation approach) analysis we should be able to mention a figure closer to $200 million as fair value for those assets in conservative net present value (NPV) terms.

The only reason why GoG would refuse to exercise any preemptive rights to acquire these patently undervalued assets, as in the case of Jubilee, should be that the Bosai offer is significantly more than just the cash on the table. GoG must have been convinced that it would not on its own be able to marshal the resources that Bosai is promising to the table, and must therefore have found in a Bosai a compelling partner.

Recall that in 2002 the Australian company, BHP Billinton, estimated Awaso's reserves alone at 500 million tons and were inclined to invest in a 10 million ton a year capacity alumina refinery.

*3. Basic Policy Context of the Deal*

There can be no ambiguity about government's policy in this regard. Former President Kufuor felt so strongly about this matter that he dedicated a part of his final sessional address before parliament to it. Both the manifestoes of the NPP and NDC allude to its criticality. And the current Minister of Trade & Industry has reaffirmed policy continuity in the matter by re-committing GoG only last year. To wit: Ghana NEEDS an integrated aluminum value chain.

The components of such a chain have always been said to include “cheap, abundant, energy”, modern bauxite ore extraction facilities, an alumina refinery, an aluminum smelting plant, and a factory system utilising aluminum, in ingot and sheet form, for a significant range of consumer goods. Necessarily, the Bosai - Rio Tinto - Ghana deal, insofar as it has the approval of the GoG, must go some way in satisfying the creation and integration of a couple of these components.

Let's see the extent to which it does. Bosai is promising to spend $512 million over the next 3 years to set up a 1 million ton capacity alumina refinery. This may be compared to the $1 billion BHP Billinton mentioned 8 years ago when it was exploring plans to build what would have been a higher-capacity alumina refinery in Ghana. Clearly, the prospects of anintegrated-aluminum industry in Ghana have sunk a bit since then. Even so, a $512 million investment would be a major feat for the $100 million turnover- per-year company. It is easy to guess that government of Ghana promised to deliver another railway line to supplant or complement the decrepit 240 km extant colonial one that has long been the bane of many a proposed bauxite project in Ejuanema, Kibi or Nyinahin. The Kibi reserves are in fact estimated to exceed those of Awaso.

There is one curious aspect of the deal. While Bosai has acknowledged that it plans to use some of the ore obtainable from the mine to fill a shortfall in its home operations, the figure quoted for ore exports from Ghana is 500,000 tons. Based on the financials, as far as they can be extrapolated, it appears however that the company anticipates a production level of 3 million tons a year (2 million of which is to feed the alumina refinery). That implies that about 500,000 tons of ore are yet to be assigned in their strategy. The suspicion that arises is that what shall actually happen would be Bosai exporting 30% of the mined ore to China even though it had conceded in the negotiations to export only 15%. The provision to export half the alumina produced is, on the other hand, plainly obvious from the production sequence and numbers stated in the announcement.

Be that as it may, many, particularly in the ruling class, would argue that an integrated aluminum industry based on 70% of locally produced ore is better than one based on 0%. And at any rate, there is always the possibility that other entrants into the space may have to import ore in significant quantities from Guinea and elsewhere to feed new alumina refineries. One suspects though that new entrants are more likely to focus on smelting than on alumina refining, in which case the already well-established alumina import trade shall be fully available to the task. But as we shall explain further subsequently, things are not as simple as they look.

*4. The Energy Challenge*

Firstly, Bosai plans to set up a rival smelter to VALCO that shall produce 200,000 tons of aluminum per annum, which is equal to the installed capacity of VALCO and considerably more than the latter's effective capacity of 80,000 metric tons prior to its shutdown.

Quite clearly, unless major new developments of which we are unaware are in the offing, VALCO has just had a spanner thrown into its strategic cog. Granting that Bosai is for real, the Chinese entrant aims to sit where VALCO has warmed.

For instance, is it the Bosai plant or VALCO's that shall benefit from the overgenerous bulk electricity tariff regime of 3.5 cents per kilowatt-hour (KWh) squeezed out of VRA to the benefit of the aluminum industry? Surely, Bosai must have been shrewd enough to insist that this sweetener was in its deal too.

Here is the complication. The liberalisation of the energy generation sector was supposed to reduce VRA's obligation to offer concessionary tariffs to industry while at the same providing sub-market rates to ECG for onward transfer to residential users. Seeing as some of the industrial customers pay significantly above generation cost for their power, it was felt that so-called independent power producers (IPPs) shall be able to compete viably with each other for contracts to supply industrial players such as the mines with unsubsidized energy. The problem we have here is that it appears the aluminum smelting industry in Ghana has become permanently addicted to dirt-cheap tariffs.

VRA is seeking to sell power to ordinary Ghanaians at 10 cents per KWh. Yet it has to sell power to decrepit VALCO at 3.5 cents. According to VRA and its experts, the most plausible balance of hydro/thermal power generation (even utilizing gas from Jubilee or Nigeria in combined cycle configurations) would probably still cost around 5 cents, not the 4 cents ECG buys its power for. That means Bosai's smelter is unlikely to see private power producers as a viable alternative to VRA, and would definitely not contemplate buying power at 7 cents (production cost plus margin) from even relatively efficient IPPs when VALCO has a deal with VRA for 3.5 cents power, notwithstanding whether VALCO is utilizing that arrangement or not. The crux of the matter is of course that VRA cannot supply two 200,000 metric ton smelters with power at 3.5 cents even if the cabinet sent over the troops to Akosombo. So something is gotta give: VALCO or the Bosai plant?

What are the other dynamics though? Supposing we stepped out of the rough and tumble world of negotiations and just look at operations, could Ghana host two smelters?

*5. The Basic Economic Equation*

The basic equation suggests that this is fully viable. Bauxite ore that can be produced for $20 per ton in Awaso, Kibi or Nyinahin, or imported at $35 (CIF) from Guinea or further afield (we are assuming that cost of production in Ghana and the alternative source country are roughly similar, even while accounting for other external supply risks in addition to insurance and freight). This ore could easily be sold for $420 on the international market (that is to say, at an appropriate discount to the benchmark London Metal Exchange (LME) price). Note that we are using an ore to metal conversion factor of 5 which presupposes a standard-efficiency electrolytic process (simply put, if your equipment is not some discarded Russian junk then you can with ease convert 5 tons of bauxite ore into one ton of nice aluminum ingots, which you can sell to Aluworks for onward processing into sheets).

Since Ghana would be producing not more than 1% of the global output of aluminum, we can safely assume that improved conditions in Ghana could not remotely affect the supply-side of price movements. Thus, the recent-historical average price will dominate in the short-term (1 to 5 years in the wake of this hypothetical scenario), and the operational challenge would largely amount to fitting in all the conversion, labour, quality control, and marketing costs into between $385 and $400 per each ton of utilized bauxite in order to break even (generous capital allowances for mines render taxation largely inert, and we are ignoring depreciation and amortization to keep the analysis lean). Let us add some standard risk management factors and on that basis assert that provided one can convert one ton of bauxite into aluminum at a cost of less than $350 then one is in fairly sound business. Risk is a vital variable considering the volatility in the aluminum market, which has seen prices drop from more than $3000 a ton in mid-June 2008 to $1250 in early 2009.

*6. Energy's Influence on Chain Economics*

Even in China, aluminum smelting takes up 4.5% of the giant Asian country's total electricity output. At its peak, VALCO consumed 65% of Ghana's electricity output. Energy is clearly a major cost factor in these considerations. Excess capacity has been shed in Europe, the USA and China, with the latter two behemoths alone having been forced to dispose of 4 million tons of production in the past decade during the oversupply-induced price slump that the metal just exited. These are the primary cost risks.

Bear in mind that if one stops at the alumina stage in the value chain - equation explored above, the cost threshold comes down from $350 to $160 (this time we are using an ore to product conversion factor of 2). That is simply to say that, having secured bauxite ore at $20 (in a fully integrated aluminum complex) or at $35 (in a partial integration regime where the ore is imported, and where the assumptions for external risks used by proponents of “integration” are true) and succeeded in converting same into alumina, one would have to fit all one's costs into between $140 and $125 to break even; or keep one's costs below either figure, as the case might be, to turn a profit (before tax).

As you probably know, or must have guessed, the huge gap between the aluminum and alumina cost thresholds is the result of one key factor – electricity. It is self-explanatory therefore, we presume, that any mishandling of the electricity component of the value chain can easily make the whole idea of an integrated aluminum production system abject nonsense. The same consideration could in the same way heavily weigh in on the side of a partial regime. Let us elaborate.

Take another look at the figures above. Note carefully the gap between $125 and $350. Now consider the starting point, which is bauxite ore. In one partial regime the ore is $35 (imported) and in another it is $20 (mined locally). Let's use the favourable regime, in the knowledge that the silica content of Ghana's ore predisposes the resource well to alumina extraction. All one has to do to sustain a healthy enterprise is to keep costs below $80 after paying for the ore in order to strive close the safe 20% gross margin. In the case of a fully integrated regime, your challenge in hitting the same margin is to keep costs below $280. It may not look too obvious at this stage since the superficial difference seems like a whopping $200. Unless you consider the primary cost factor corresponding to that figure as already hinted: electricity.

Under the ludicrously low tariff negotiated on behalf of VALCO, power goes for $35 per megawatt-hour (MWh). True, this is closer to the international cost-range of between $30 and $40, but in the current Ghanaian setting it is simply ludicrous. And recall that VALCO, in its earlier incarnation as a majority Kaiser-owned operation, struggled even with a power tariff of $11 per MWh and had to abandon the operation when government proposed a new tariff of $30 per MWh.

Any mild inefficiency in plant installation, power transmission, and system integration could lead to each ton of aluminum requiring as much as 22MWh for smelting. In such a scenario, the cost threshold above would be sorely breached.

It would mean that even without accounting for the costs of labour, quality control, and marketing, the operation is already close to loss-making. The costs of power in the equation we have used so far (generation and distribution) comes to about $170 (which is a ridiculously high $850 per ton, compared to $520 in China, itself not a particular efficient producer). This would further mean that the contribution of power to total cost of production would far exceed the global average of 25%. Recall that one of the silly pranks the erstwhile VALCO was fond of was demanding to pay international rates for the power it bought without conforming to the global standard when it came to the contribution of power to total cost of production.

Let us look at the serious import of such a situation. It would mean that of the $200 superficial nominal difference between the cost thresholds for profitability for alumina and aluminum, a crazy $170 has been taken up by electricity. Let us allow power to represent a ridiculous 66% of production costs, it would still mean that the cost of converting 500 kilograms of alumina to 200 kilograms of aluminum (integrating our conversion factors) would cost $250 on a very, very, good day, making nonsense of the need to convert at all.

In fact, it would take the heavens to intercede to keep the cost of the other factors of production this low. And do not forget the ludicrously low electricity tariffs we used at the outset. Realistic rates of power and other costs of production could easily take the cost threshold in our Equation to $500 and therefore the average cost of production per ton of aluminum to $2500, without even significant weight to Ghana's atrocious bulk transportation system. The implication in those circumstances would be that for each ton of aluminum smelt in Ghana the producers would be losing $400. And we are talking about 400,000 tons per year for both VALCO and the Bosai smelter, right? $160 million in operational losses. And even that only because aluminum prices are holding at a 2-decade high of $2200. If the price was to slip to a more stable-looking $1800 figure, we would be discussing double the losses stated above. Clearly unviable.

*7. How “Partial Integration* Could Win*

If there was any question at all about why a partial regime of integration (of the aluminum value chain, that is) may be the most economically realistic prospect in the medium-term we are sure it has been answered by the sheer terror of the numbers. A potential $320 million per year loss? The whole scheme would come tottering down.

We have of course assumed start-up inefficiencies of 20% and argued that VRA and the IPPs shall not be able to sustain power prices at $35 per MWh. There is no absolute guarantee about these predictions, reasonable risk forecasting though they are. What the whole analysis had sought to establish is that in considering an integrated aluminum complex one must strongly bear in mind how tight the margins could be, and how tautly the scheme is required to hold together if it is to survive. Another motive has been to establish that value integration could look very differently, and rightly so, under alternative conditions and considerations.

Indeed, one could easily see a situation in which rather than truncate the chain at the level of alumina the sequence could be extended instead to glass production, to complement the export of raw alumina. The electricity problems could thus be mitigated through the elimination of the smelting stage in the chain. As you must have no doubt observed, the issue is not one of the availability of power, but of the tariff regime, and so is not wholly solved, as explained above, by the availability of gas and the elimination of crude oil from the power mix.

*8. Mitigating Factors for Full-Chain Integration*

It is also perfectly justifiable to review the economics of the integrated aluminum complex by considering modernization prospects.

The British-originated, pan-African, power producer, Infraco, has a business model dedicated solely to reducing the bulk rate of power by adopting the latest in gas turbine and other energy generation technologies. New IPPs in Ghana, to the extent that they are backed by sophisticated investors, have all the justification they need to source and integrate the latest systems. So whether or not Bui and the other thermal plants come on board according to schedule is a secondary issue at this point of the discussion. Let's trust to God that in the medium-term $30 per MWh would be feasible through the use of new systems and technologies in Ghana. Let's also hope that 10MWh per ton of aluminum output could become feasible through full-automation and process-optimisation of pot-control and the enhancement of the electrolytic process by means of, say, so-called wettable cathodes. Let us even pray that all expatriate engineering staff would be replaced by locally and superbly trained, and therefore realistically waged, talent. And that government will not chicken out of its obligations to revamp bulk transport. In those circumstances, production costs could be brought down to the global standard-performance of $1100. Then even at historically low prices of $1500 per ton of the metal, a gross margin at the psychologically important mark of 20% could be modestly exceeded.

*9. Fiscal Imperatives*

And it is important that this margin be exceeded. Our mining laws allow, as our circumstances compel, rather generous capital allowances. The companies need to be making a hefty profit so they can quickly amortise their initial capital expenditure. That is to say if after Bosai has spent their $512 million building their refinery, they decide to expend another $500 million to add a smelter (like BHP Billinton promised to do and failed), they would, per standard cross-border investment practice, be hoping to pay off their $1billion capital expenditure before any thought of taxes come into the picture. Should they sell their 200,000 tons of aluminum at $2000 per ton into the international commodity markets, they should earn $400 million. At a 20% margin, their take would be $80 million.

Now let us begin to look at these figures in fiscal terms.

We start with 3 million tons of bauxite per annum, just as at the beginning of this essay. Government is entitled to 300,000 tons of that just for being government in an arrangement christened “royalty”. At the best market rate this is about $7.5 million, which Bosai would probably just pay to government in cash in order to retain rights to the physical bauxite. Some relatively marginal payments shall also be made with respect to surface rent and other charges, but these we shall ignore in order to keep the analysis compact. Recall that Bosai expects to make $15 million gross margin per year. They get to write off a part of their initial equity investment as well as a part of their initial capital investment each year. Let us say total development-based investments total $100 million (here we are constrained by the refusal of GoG to disclose the terms). Under our laws, a good accountant should be able to get a write-off of $20 million and neatly declare a loss of $5 million for Bosai for the financial year. No profits no taxes, as we all well know.

Now we move to the value-added stage. Recall that we calculated gross margin there to equal $80 million. But we are dealing at this point with a much bigger initial equity and capital investment. Here the precise capital depreciation allowance is a bit hard to pin down, even in rough terms as we have done above, since we are now more into the GIPC (Ghana Investment Promotion Council) regime than the Minerals regime. But one is completely sure that no taxes would be forthcoming in the first 5 years at least.

The above realities reinforce previous notions already entertained. The more inefficient an investment the less likely this country benefits, even if the size of the investment appear tempting. Indeed the profitability of the investment should be considered of high importance to our own fiscal circumstances. In this particular case, Ghana not only take at least 25% of the gross margin in corporate taxes, it also take 20% in dividends (if it exercises all its options), 10% in royalties and other charges and fees – provided of course that profits are substantial enough the quicken the pace of capital depreciation.

Indeed the somewhat problematic linkage between the size of an investment and its efficiency provides another channel of fiscal risk. Ghana, as a shareholder in the enterprise, is required to contribute to its share of capital investment. In this specific case Ghana would ordinarily be responsible for $200 million if it refuses to allow its stake in GBC (Ghana Bauxite Company) to be diluted in order to preserve its dividends as assumed in the preceding analysis. Unless dividends are quickly forthcoming and substantial, such an expenditure on government part could easily become a source of concern.

But if the government simply wants to avert the risks that come with a 20% participating interest, then it risks leaving what many of its supporters would regard as a “strategic national asset” in the hands of “foreigners”.

So once again, profitability is key if the idea is to prevent capital depreciation from interfering with the need to maximize Ghanaian participation (through government carried interest) at the same time as fiscal risks are being minimized. Perhaps the trick is to support private Ghanaian citizens raise capital on the international markets to participate “on Ghana's behalf”. But given our national experience this is easier said than done.

*10. The “Value” Can be the Same*

These additional points send us back to our earlier discussion of partial regimes of integration. Recall that alumina is sold at about $360 per ton using standard discount rates. It takes 2 tons of bauxite to produce 1 ton of alumina. Therefore accounting for raw material costs (per our earlier discussion) leaves you with $320. It requires an average of 3 MWh of electricity and therefore $120 thereabouts to power this conversion (once again, taking due cognizance of our earlier considerations). This leaves you with $200. Recall our profile of cost distribution, which should bring us to $80 as the gross margin. Voila.

Remember that the projected output of Bosai's alumina refinery is 1 million tons. By what looks like a feat of magic, we have arrived at the conclusion that the gross margin of producing aluminum in a full integration model is equivalent to the gross margin of producing alumina in a partial integration model. As we are sure you have noticed the partial integration regime can be a cool 50% cheaper than the full model. And bearing in mind our earlier discussion about size of incoming investment and fiscal benefits to the Ghanaian economy we arrive at a highly enlightening viewpoint about “value”.

What has been demonstrated is that under certain conditions, a single decision about the value chain could yield double the output in benefits than would otherwise have been the case.

Indeed, for ease of analysis we had been compelled to be simplistic in how we laid out the scenario. The energy differentials between aluminum and alumina production that we explored have their own extended economic implications.

Whereas a captive 125MW energy plant could supply the needs of the alumina complex we described above, it would take the whole output of the Bui dam – some 400MW – to power the full-chain aluminum complex. That could itself mean a capital expenditure differential of almost $300 million, feeding back into our concerns about investment efficiency.

*11. Prudent & “Open-Minded” Economics is Vital*

But granted the possibility that in a few years Ghana shall be grappling with energy overcapacity rather than load-shedding. Sure, it would be great to have some large industrial customers for the energy producers that shall be on the scene there, but only if the tariff regime is not overly compromised to accommodate an unwieldy regime for what can only be described as sentimental reasons. Because what we are talking about here is 3 million MWh per annum in the case of the full-chain aluminum complex. If the only viable way to sustain such an operation is through a concessionary mechanism involving state producers of power then, even in a situation where we are dealing with 100% hydro mix, there is the thorny issue of opportunity cost to consider.

That is to say, imagine that production cost for hydro at Bui, Kpong or Akosombo is an average of $25 cents per MWh. If the Bui Power Authority (BPA) and the VRA were compelled to sell this energy to the full-chain aluminum complex' operator at $35, then after accounting for transmission and other post-generation costs, only a margin of about 5 dollars per MWh shall accrue to the Republic of Ghana. That is to say some $15 million per annum. The contribution of this sum to the BPA and VRA's costs of cross-subsidizing residential users and undertaking social responsibility activities in their “catchment areas” shall be very limited.

Very soon ordinary Ghanaians shall be paying about $60 per MWh. Even accounting for higher transmission and distribution costs, one sees immediately that a margin of $20 is feasible in a situation where the power authorities sell to residential, not even to talk of commercial, customers in Ghana. Immediately, the earnings under consideration go up a whopping 400%. Bringing conventional industrial consumers into the picture sends the producers' income proceeds rocketing upwards by 1000% thereabouts and even over.

In those circumstances, when it is possible to sell the same energy at 10 times more to other folks, why would it be prudent to sell it to the full-chain aluminum complex, when as we have demonstrated above its fiscal contributions, as a result of any combination of a multitude of risk factors, could easily become negligible?

The same general principles extend to the post-metal phase of the chain where we can look at receptive industries for which the aluminum ingots are the raw material. Presently, due to VALCO's problems the companies in Ghana that produce aluminum wares of different kinds tend to import their ingots.

It would be sentimentally correct to want to see consumer goods come out at the end of this process. But most studies have found that intermediate products, especially those used in the construction industry, tend to offer better commercial returns within the value chain than many finished items. Some empirical analysis of this situation has found a 2X difference in returns favouring intermediates.

The truth is that whenever we have mentioned “integrated aluminum complex” our thoughts have more often harboured such notions as “import substitution”, “self-sufficiency”, “heavy industry”, “planned industrialization”, “commanding heights”, “conglomerates”, “national champions”, and other such images from the socialist-nationalist repertoire.

Rarely have we considered these matters in terms of “trade”, “wealth creation”, “value maximization”, and “competitive efficiency”.

As we hope has been demonstrated in this discussion, however, it is entirely possible that the second set of notions better forecasts and describes the true state of affairs in a better managed natural resource situation than does the first set. It is clear that it may be wiser in certain instances to import bauxite than mine our own, given the marginalities involved; produce alumina and export same in raw form rather than refine further into aluminum; and to fabricate intermediate products involving medium-level skills instead of going the full length to craft exquisite, premium, hollowware by utilizing tertiary/master –level skills.

Thus we may conceive the value chain in any number of ways with as many strategic gaps as possible. A closed loop is not always the most efficient mode.

Even greater support could be found for the above conclusion when consideration is given to the social and environmental concerns.

*12. Social & Environmental Considerations*

Using the Bosai – Rio Tinto - GBC deal as our backdrop, we can look at the jobs per investment dollar indicator as a probe into the social importance of the full-chain aluminum complex idea. It is hard to imagine the project creating more than 700 jobs, using enterprise models in this industry from India and Jamaica (VALCO's 1500 in days gone by was wayward).

This implies something in the region of $1.5 million per job, not even to factor in such parameters as expatriate to local staff ratios etc. Using the partial integration regime brings this figure to $900,000, a less troubling figure. We can of course look at the tip of the downstream end of the complex but the same insights would apply.

If for instance, in a particular value chain approach, the decision is made to import alumina in the context where, for instance, there is installed hydropower capacity in excess of industrial demand to feed the smelter, the pro-trade approach will mean lower environmental costs within the complex in Ghana, and could therefore be viewed as a sharing of risks between Ghana and the bauxite origin country. Of course, the issue is complicated by the fact that environmental pollution mitigation is itself a fertile area of entrepreneurship nowadays. To elaborate, consider that each ton of aluminum creates a ton of useless mud, mostly silicates. This mud is however mixed with caustic soda and other chemicals potentially harmful to the groundwater supply. If this mud could be safely disposed off in some way the economics suggest that major producers of the ore may be better equip for reasons of scale and specialization to do a better job. The same goes for CO2 emissions, which some environmentalists peg at $50 per ton for the aluminum sector; or for habitat conservation. Scale and specialization are fundamental to implementing the interventions that can be financed through the use of the sophisticated carbon and other ecological markets, most of which are in Europe and America.

And it is trite knowledge that nothing fosters scale and specialization more than trade based on the enlightened consideration of comparative advantages at all levels.

*Conclusion: Open Models & Feedback Loops – Our Protection*

As we bring this conversation to a close, we would like to reiterate that we are not necessarily opposed to a full-chain, integrated, aluminum complex. But as we have sought to demonstrate, the principal dynamics in such a system can easily get very complex indeed, and like all complex systems complete control and precise forecasting is not always possible. There is no point getting sentimental about ideal shapes and forms since it is always very easy to mistake a mirage for an oasis and vice versa.

In the same vein we are not averse to entrepreneurs who understand the reality that political power is often intertwined with economic interests. One after all one does not go into business to crash out.

But we all recollect the state of near-panic the aluminum sector was thrown into when differences between Dr. Charles Mensa, then CEO of VALCO, and Togbe Afede XIV of Strategic Initiative spilled into the open. It was because anyone remotely interested in aluminum matters in this country knew that it had very little to do with the location of a power plant and everything to do with the quietly raging contest for control over the anticipated “integrated aluminum complex”.

Indeed, all over Accra, in every private conversation amongst industry watchers in Ghana, the hot topic is about to what extent said contest was percolating into our suspense-filled nascent oil sector through the IPP conduit.

The bolder observers are even speculating that there is some unease about the Chemu Group because Chemu Power, the Group's subsidiary, represents in the mind of some strategists the seed for an aluminum complex associated with one of two major sets of political interests, whilst Sunon Asogli is seen as a placeholder for the other, opposing, set of interests.

Indeed, the rumours about deliberate sabotage that begun making the rounds when two global aluminum conglomerates backed out of a parliament-ratified partnership agreement with the then Nick Armatefio-headed VALCO Board derived its plausibility from the widespread feeling that the aluminum sector has become a hotbed of machination.

Some analysts given to comical exaggeration like to joke that the best term for describing the aluminum industry in Ghana is “incestuous”, which outrageous comment they justify by pointing to the Boards of the Bank of Ghana, Aluworks, Pioneer Aluminum , Ghana Aluminum, Dannex, and VALCO, particularly in the erstwhile dispensation.

Political and business intrigue can be titillating, but it is usually more useful to focus on the hard empirical evidence than on conspiracy theories, however much anecdotal evidence can sometimes point one towards the hard facts. The point is not to overdo it.

We are completely convinced that Ghanaians have little to be afraid of with respect to elite machinations and intrigue, or with respect to the excesses of clan and dynastical business-political arrangements. But only so long as the time-tested productive values identified above are allowed to prevail in the aluminum sector.

For the convenience of our readers, we list these values again: “transparency”, “trade”, “wealth creation”, “value maximization”, and “competitive efficiency”.

*This report is by courtesy of IMANI-Ghana and AfricanLiberty.org; to comment, remark, or seek clarification you are invited to write to: Kofi Bentil or Bright B. Simons – [email protected]. An annotated version shall be appearing on the www.imanighana.org website.*