Unravelling Ghana State-Owned Oil Refinery (TOR) Debt: A Provisional Analysis

Tuesday, January 27, 2009

By Bright Simons and Franklin Cudjoe
oil refineryNo comprehensive report has thus far been issued on the recent and present condition of the Tema Oil Refinery (TOR) that we are aware of, not by the Ministry of Energy and not by TOR itself. One hopes that this dearth of information will soon be addressed.

More to the point, one hopes that whatever account is produced will depart from the usual partial assessments restricted to certain politically defined epochs, but will instead encompass all the relevant phases of the evolving TOR situation since the establishment of the facility in 1963.

In the interim, however, in view of the public interest and comment, we have decided to try and piece together, from the fragments of news reports, government statements, and corporate reports, a brief evaluation of the refinery’s debt position and financial prospects. It goes without saying, given the limitations of this approach, that some of our calculations could be affected by third-party reporting inaccuracies. This is a constraint that cannot however be assumed to limit the substance of the analysis itself except where the logic is driven by the numbers. Which is another way of saying that the numbers are primarily illustrative. We will proceed on the back of that foregoing disclaimer.
Between the time of its establishment in 1963 and 1997, TOR underwent no significant expansion of its 25,000 barrel capacity. Now this is not as outrageous as it may seem; gasoline-thirsty USA hasn’t built a refinery in thee decades. What is certainly unsettling is that the plant also did not undergo, unlike is the case with installations in many other places, any significant retrofit to enhance the efficiency of its core operations and keep it abreast with global best practice.

In 1996, after many false starts, a definite program of reform was unveiled to revamp the moribund enterprise, and TOR was sized up for significant upgrading. The perennial inability of the refinery to recover the costs of its operations because of government interference in the marketplace was identified in addition to its inefficiency problems as deserving of policymakers’ attention. The legal framework defining the context of TOR’s operations (the Petroleum Law of 1984, Acts 544, 593 & 577, amongst others) also received some review.


In 1997, a South Korean consortium led by the energy subsidiary of the Samsung Conglomerate and the infrastructure services company SunKyong were contracted to expand the output of the refinery from 25,000 barrels to 45,000 barrels in order to take full advantage of improved economies of scale (for perspective though, note that Ghana’s total ANNUAL consumption of crude is less than twice Russia’s DAILY output). Plans for the development of an allied catalytic cracker (RFCC) were also drawn around this time.

By 2002, the construction of the catalytic converter was complete, and TOR entered a phase of partial modernisation. It will aid an understanding of the context to elaborate a bit more on the above developments.

Concerns about refinery efficiency are anchored to what is usually referred to as downstream capacity utilisation rate. The simple explanation is that the better a refinery is at converting crude oil (‘heavy’ petroleum) into finished consumables (‘lighter’ products), the more efficient it is. That is to say: given a barrel of crude oil (roughly 42 gallons, depending on the specific gravity of the grade of oil), how many gallons of end product can a refinery turn out? Secondly: of this end product how much of it is made up of light products such as gasoline (‘petrol’)?

Great advances in petroleum chemistry and in the use of computer-controlled processes have enabled some refineries in the United States, for example, to convert nearly everything in a barrel of crude oil into gasoline, the most desirable product in several markets within the union. The use of certain types of aluminium silica gels and reprocessing flow mechanisms can, amongst other specific technical interventions, rather than general managerial competence, account for the overwhelming bulk of these improvements.

In the absence of a catalytic cracker, as was the situation at TOR for several years, the oil refinery process must rely largely on the physical processes of distillation, which merely take advantage of the fact that certain desirable components of the crude oil, like gasoline, have different evaporation rates and can therefore be ‘boiled off’ the less valuable residue. Of course this is wasteful, especially in a country like Ghana with relative high emphases on gasoline and diesel. The best practice nowadays is for refineries to adopt the most sophisticated chemical processes, involving the application of additives and the liberal use of computer-controlled production programming in order to incrementally reconvert ‘residue’ to useful product.

The foregoing provide a background to the efficiency issues TOR was grappling with and, to a reduced but still significant extent, still grapples with.


Prior to this era of reform, TOR was a sprawling operation involving the financing and importation of crude, its processing, and the distribution of its derivatives to retail outlets.
The process of rationalisation involved an attempt to define a core competence for TOR and to progressively hive off peripheral operations to the private sector rather than to parastatals like BOST. By 1998, consensus at the policymaking level had virtually consolidated in favour of full cost-recovery and enhanced deregulation (the progressive privatisation of peripheral – non-refining – operations).

Then, before significant progress could be made, events took a dramatic turn for the worse. Between 1999 and 2000, fuel prices skyrocketed ($11 to $30 in one particular movement). At this point, unfortunately for all concerned, full cost recovery through a principle of ‘import parity’ (more on this later) had yet to move from lip service to definite policy.


Government, on the strength of social considerations, instructed TOR not to sell into the distribution chain at the prevailing market rate but to assume a level of government subvention. The result of this decision was that by early 2001, unfunded subsidies had piled up GHC220 million of debts on TOR books mainly in the form of overdraft facilities offered by their principal bankers, Ghana Commercial Bank.

Because TOR was also at this point the sole importer of refined petroleum products, which was necessary in order to meet the perennial shortfall in Ghana’s annual demand of 60,000 barrels (TOR’s capacity, you may recall, is 45,000 barrels), it had opened letters of credit with local and international banks for GHC110 million that were yet to mature.

Furthermore, the GHC210 million spent on the expansion and modernisation of TOR had, despite the protests of some stakeholders, been kept on the books of TOR rather than treated as an infusion of cash by the shareholders of the concern, the tax payers of Ghana.
It would thus seem that at this point, the total debt position of TOR was of the order of about GHC540 million. But as the debts piled so also did the accumulated interest. With overdraft and other loan rates close to 30% for some of the facilities, the debt position could double in less than 3 years unless both the principal and interests were serviced aggressively. Not even the downturn of prices after 2001 could yield sufficient returns to tackle the mounting debt crisis.

Government of Ghana’s approach to tackling the situation between 2001 and 2002 consisted of short-term respites and a strategic component. The viability of the latter is course the subject of the present controversy.



 Myjoyonline Ghana News Photos |
Between 2003 and 2004, with the Ghana Commercial overdraft liabilities now at GHC469 million, the Government obliged to ‘absorb’ GHC320 million of the TOR debt in a two-part process. GHC80million of this amount was in the form of direct cash defrayments, while GHC240 million was amortised into dollarized bonds at a 4.5% real rate of interest (that is adjusted for inflation). GHC150 million remained on the books, to the chagrin of some stakeholders.

At this stage our ability to follow the debt trail weakens on account of lower-quality data gleaned from the fragmentary accounts we mentioned at the beginning. It is unclear whether the GHC210 million incurred by TOR for its capital expansion between 1998 and 2002 was incorporated into the restructuring process or not. We will return to this point again.


At the core of all these matters of course is the ability of TOR to sell its products at a price that allows it to recoup its costs and make a decent margin for continuous improvement and expansion. Secondary to that is the type of financing (short-term and inflationary) it chose to use to plug the gaps once the dam of financial prudence had been breached.

The principle of ‘import parity’ broached in 1998 thus took centre-stage in 2001. The logic is simple: assume that Ghana imports all its refined products, what will be the price of said finished products on the market? In that sense then if we are going to manage a national refinery, it should be because it produces at a cost comparable to that rate or we are better off importing.

The exact approach was to identify the corresponding prices in Northwest Europe of the 10 refined products on TOR’s production list, add the requisite freight and other charges, and use this as an import parity benchmark. A slight accommodation was also made for the fact that refineries in this part of Europe operate at more favourable economies of scale than Ghana’s 45,000 barrel installation (most refineries in the developed world operate at above 250,000 barrel capacity). In order also to accommodate erratic movements in prices of said products in the more deregulated markets of Europe, it was further decided that an average of ex-refinery prices over a three month period will be used, and that local changes will be made in prices to reflect the proposed European benchmark only when the two sets of prices diverged by 2.5%.

From 2002 onwards, this import parity model was incorporated into a general pricing formula for the determination of ex-refinery and ex-pump (distribution and retail) prices. In this formula, TOR’s import-parity defined share of the final retail price was pegged at 64% of the ex-pump price. Taxes and levies amounted to 24% and in-built margins for distributors, transporters and retailers were combined into a 12% supermargin. A cross-subsidisation method was to ensure that the prices of kerosene and LPG are lower than they should be while that of gasoline is slightly higher.

In 2004, market liberalisation moved into full swing and private operators were allowed to import refined products into Ghana by a process of bidding and tendering. Further liberalisation anticipated private importation of crude oil for TOR to refine on a commercial fee basis, as well as private storage and distribution outside the traditional BOST system.

In 2005, the National Petroleum Authority was established by statutory instrument (ACT 691) and empowered by said statute to exercise full oversight over the entire downstream petroleum sector (the upstream – prospecting and production- activities remained within the ambit of the GNPC). The NPA, by this reckoning, is to set firm but general guidelines about pricing, consumer protection and sector viability, within which eventually private operators were to be free to engage in vigorous competition with each other.

In the succeeding years, political and social pressures frequently interfered with the levels of petroleum levies such as the Debt Recovery and Road Development tolls. If an average of 10% of ex-pump price is used as the historical average for the controversial Debt Recovery Levy, between $600 and $700 million dollars may have accrued between 2001 and 2008 (this is not exactly synonymous with the Petroleum Debt Service Surcharge, which for a while had been embedded in the import-parity benchmark itself to take advantage of favourable turns on the world market).


Over the past three years, it appears that not very much progress has been made despite the pursuit of reforms for more than a decade. In 2004, at the height of the reformist zeal, Government subsidies amounted to $200 million, and thus clearly in support of the thesis that the behaviour of the debt had become, to borrow a description from a CEPA Analyst, ‘quasi-fiscal’. The current debt level is said to hover around GHC1.1 billion.

About GHC180 million in debts arose in 2008 when cost recovery was frozen at $116 even though world market prices had shot up to $144. Now recall the GHC210 million CAPEX debt that we lost track of earlier, and the GHC190 million that was left on TOR’s books after the restructuring. Then there is the GHC110 million bundle of LOCs dating from the early days of the crisis. That brings us to GHC700 million. The 5-year petroleum bonds must also by now have matured. Add an additional GHC240 million (and it is a debt regardless). Now consider this interesting aspect.

The amortisation of the debt into dollarized bonds definitely did provide a respite by lengthening the maturity but it also deprived the borrowers of the ameliorating effects of exchange rate pressures (while the debt was denominated in cedis interest rates were in a tussle with exchange rate losses; less so in the case of dollars) and the use of real interest rates served to mitigate inflationary pressures to the benefit of the creditors. The sum effect of these matters is that the accumulated interest on this portion of the debt may in the end have behaved quite similarly to what obtained under the old regime. Indeed if we assume that the effect of interest was to double the principal, the TOR debt comes to about GHC1880 million, which, after you have accounted for the mitigating effects of the petroleum levies (see above), is extremely close to what has been mentioned in recent days, indicating that whatever the flaws in our calculations as a result of low-quality data, the analytical logic probably corresponds very closely to what has happened over the past 5 years when the debt management process moved into full gear.

Charging the $1 billion TOR debt on the consolidated fund is out of the question. It will balloon the fiscal deficit and blight the macroeconomic environment for years to come. Another around of amortisation would more likely than not lead down the slippery road of convergence with the general lending rate that we have described in the case of the 5-year bond. Instead, we think the debt should be securitized.

An Expert affiliated with Government of Ghana estimates that TOR’s debts are double the value of its assets. The implication is that its credit standing is somewhat precarious, as most of those debts are trading losses and therefore largely not reflected on the asset column of the balance sheet. Meanwhile the refinery’s medium-term strategic goal is to triple capacity and thus enjoy improved economies of scale. One presumes further improvement in downstream capacity utilisation will also be considered. The price tag would be in the order of $300 million. This would mean additional debt, and not on the best of terms considering the refinery’s credit standing and the global economic crisis.

With assets of $600 million and debts of 1.4 billion, will securitisation be a tall order? Only to the extent that raising finance for any project is getting tougher by the minute. After all raising loan financing wouldn’t be child’s play either. But when valuing TOR’s assets we are wont to ignore the weight of its statutory monopoly status in the refinery portion of the downstream sector. Government could tactically decide to preserve that monopoly in order to construe it as a highly valuable soft asset. Indeed the monopoly could be interpreted to represent an exclusive license to refine crude in Ghana, and could on that basis be appraised at another $600 million based on discounted future projections.

If TOR’s assets are worth $1.2 billion, then its debt position of $1.4 billion ceases to become, strictly speaking, a millstone. The conversion of TOR’s debt to equity could then be justified to creditors on account of the massive enhancement of its financial performance owing to improved economies of scale and higher turnover. The broader context is that petroleum refining, especially against a backdrop of cartelist behaviour or limited competition, is inherently viable commercially.

TOR could safely operate at a 20% net margin even at current sub-optimal output and capacity utilisation rates.

Of course the bottom-line is that any such process would probably involve an alteration in Government’s sole ownership status and a resolution by the political elite to commit, beyond lip-service, to full cost recovery. Ameliorating the social effects of price deregulation should then become a process external to the operational considerations of the crude refinery.

A voucher system based, perhaps, on the E-Zwich platform could be introduced to ensure that subsidies are guaranteed for public transport operators in the form of reimbursements during sharp upsurges on the international petroleum markets, while cross-subsidization is maintained to stabilise the prices of social-sensitive kerosene and LPG.

Otherwise TOR may be doomed.

Bright B. Simons and Franklin Cudjoe are affiliated with IMANI: The Center for Policy & Education and www.AfricanLiberty.org