Zambia: Selected Issues

Abstract

Zambia: Selected Issues

Pricing of Petroleum Products in Zambia1

A. Introduction

1. Zambia imports all of its petroleum products. The bulk of fuel imports have traditionally been piped from the port of Dar-es-Salaam to the Indeni refinery in Ndola using the 1,710 km Tazama pipeline. These imports are in the form of spiked petroleum feedstock,2 which is refined at Indeni to obtain the final market mix. Due to capacity constraints of the pipeline and refinery, refined products are also imported directly by road, with such imports now accounting for about half of the total.3 All procurement is handled by government and both the Tazama pipeline and the Indeni refinery are fully government-owned.

2. Using the cost plus pricing (CPP) model employed by the Energy Regulation Board (ERB), this paper evaluates the extent to which retail fuel prices in Zambia have reflected the full cost. The analysis is based on data covering the volume and price of 35 cargos of petroleum feedstock delivered between January 2010 and March 2015.

B. Pricing Mechanism

3. The CPP model has been in effect since January 2008 and operates on the principle that the final local currency price of petroleum products should cover all costs in the supply chain plus a fair profit margin. The model involves two stages in the price buildup—wholesale and retail—each covering different costs incurred in the supply chain (Tables 1 and 2). As the margins at the different stages of the price buildup are largely fixed, changes in the overall cost are mainly determined by two variables: the international oil price and the exchange rate of the Zambian kwacha.

Table 1.

Wholesale Price Build-up

(as of June 2014)

article image
Table 2.

Price Build up for the Retail Price

(as of June 2014)

article image

4. The CPP model is applied to each shipment of petroleum feedstock going to the Indeni refinery with a view to determining the cost-reflective final sale price. Direct imports of finished products have not been reflected in the model, a potentially important source of error in establishing the overall cost-reflective price given the growing role of such imports. The ERB employs a trigger band of 2.5 percent for price reviews. This means that, in principle, the retail fuel price should be adjusted whenever the computed change is beyond the trigger band.

5. Actual retail price changes have been relatively infrequent in recent years (Figure 1). Petroleum feedstock is imported roughly every six to eight weeks and the associated cost-reflective price has in most cases changed by more than 2.5 percent—a consequence of the volatility in feedstock prices and the kwacha exchange rate. Pump prices, however, were only changed on four occasions during 2010–11 and only two times from the start of 2012 to mid-2014. With international oil prices falling sharply in the second half of 2014, retail prices were adjusted down in late November and early December, 2014, and again in mid-January 2015.

Figure 1.
Figure 1.

Zambia: Retail Petroleum Prices

(Kwacha per liter)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Sources: ERB; and Bank of Zambia.

6. The failure to systematically adjust retail prices according to cost has resulted in implicit fuel subsidies being the norm. Between early 2010 and mid-2014, retail prices in kwacha rose by close to 60 percent for both petrol and diesel. Measured in U.S. dollars, however, the price increases were lower at just under 20 percent, only about half of the almost 40 percent increase in the price of crude oil during the same period (Figures 2 and 3). In the second half of 2014, the drop in international oil prices led to the cost reflective price being below the retail price for some shipments. In January 2015, however, retail price reductions of over 20 percent were larger than warranted, resulting in the reemergence of subsidies.

Figure 2.
Figure 2.

Zambia: Retail Petroleum Prices

(U.S. dollar per liter)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Source: ERB.
Figure 3.
Figure 3.

International Petroleum Prices

(U.S. dollar per barrel)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Sources: ERB; and IMF commodity price unit.

C. Costs

7. The 35 cargoes of petroleum feedstock delivered to Dar-es-Salaam had a total CIF cost of US$2,912 million (Table 3). The cost of this feedstock broadly mirrored but was consistently higher than the international market prices for crude oil (Figure 3). Zambia’s purchase price was on average $26/barrel or 29 percent higher than the average price of Dubai crude in the month that the shipment was delivered. This price premium over Dubai crude reflected extra costs associated with transport to Dar-es-Salaam and the partially refined nature of the feedstock. The premium has varied from as little as $7 to as much as $52 per barrel, however, implying that other factors have been important too.

Table 3.

Zambia: Cost-Plus Model by Cargo

article image
Sources: ERB and authors’ estimates.

Liquid Petroleum Gas (LPG), Heavy Fuel Oil (HFO), Jet A1, and Bitumen.

LPG, HFO, Jet A1, and Bitumen to wholesale level only.

8. The CPP model shows how the different components contribute to the cost buildup. Feedstock has typically accounted for less than half of the cost of petrol and diesel but close to two-thirds of the cost of kerosene (Figure 4). Taxes have been the second largest cost component for petrol, smaller for diesel, and zero for kerosene. The data also indicate that the implicit subsidy, defined as the difference between the cost-reflective and the actual price, has for petrol consistently been smaller than the amount of taxes collected, whereas the subsidy for diesel has at times slightly exceeded the amount of taxes collected.

Figure 4.
Figure 4.

Zambia: Breakdown of Cost-Reflective Retail Prices

(U.S. dollar per liter)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Sources: ERB; and authors’ estimates.

9. Overall, using the parameters of the CPP model, we estimate the cost-reflective sales price of the final products coming out of the 35 cargos of feedstock at US$5,266 million.4,5 Based on actual prices, we estimate the realized sales value at US$4,755 million. The shortfall from full cost recovery is thus estimated at US$511 million (on average about 0.4 percent of GDP). Figure 5 shows this revenue-shortfall by cargo, revealing how the subsidies reached over $30 million for some shipments in 2012–13. The overall dollar value of subsidies for the two shipments in early 2015 were smaller at about $20 million but, given lower international oil prices, this represented almost 40 percent of the cost of the feedstock, similar to peak levels in 2012–13.

Figure 5.
Figure 5.

Estimated Retail Revenue Shortfall by Cargo

(Millions of U.S. dollars)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Sources: ERB; and authors’ estimates.

10. Given the fixed margins in the supply chain, any shortfall between the cost-reflective and actual sales value eventually has to be covered by the Zambian Government. From the start of 2010 to March 2015 the Ministry of Finance released a total of about $620 million for fuel subsidies, $109 million more than the $511 million estimated revenue shortfall. In addition, in early 2015 it was reported that as of end-February, there were subsidy payments due of close to $260 million, adding to the gap between actual and estimated subsidies. This gap could stem from our estimates not covering subsidies related to direct import of refined fuel products or unrecovered costs beyond the wholesale level for LPG, HFO, Jet A1, and bitumen. Interest charges associated with payment arrears would also have contributed.

D. Zambia and South Africa Comparative Fuel Prices

11. Despite the history of subsidies, fuel prices in Zambia are higher than elsewhere in the region. Zambian retail prices of both petrol and diesel have from the start of 2010 to March 2015 averaged about $0.33 per liter more than the inland price in South Africa, even though Zambian subsidies averaged about $0.16 per liter and prices in South Africa are not subsidized. The price premium only dropped below $0.10 per liter in May 2012 and March 2013 when Zambian subsidies peaked at close to $0.40 per liter (Figure 6).

Figure 6.
Figure 6.

Zambia Price Premium Over RSA

(U.S. dollar per liter)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Sources: ERB; and RSA Department of Energy.

12. That fuel prices are higher in Zambia than in South Africa reflects a combination of higher taxes and a more expensive supply chain. Retail fuel prices in South Africa are set in a manner similar to that implied by Zambia’s CPP model, which allows for direct comparison of fuel costs in the two countries (Figure 7). This comparison shows that the Zambian cost premium for petrol of $0.45 per liter in March 2015 can be attributed to $0.20 per liter more expensive wholesale fuel and $0.27 per liter higher taxes. The cost premium for diesel of $0.37 per liter reflects a difference in wholesale prices of $0.22, a difference in the retail margin of $0.18, and a tax difference of $0.11.6 While high fuel taxes are a policy choice, the high wholesale price in Zambia is indicative of an inefficient supply chain.

Figure 7.
Figure 7.

Breakdown of Cost-Reflective Price

(U.S. dollar per liter, March 2015)

Citation: IMF Staff Country Reports 2015, 153; 10.5089/9781513556253.002.A004

Sources: ERB; RSADE; and authors’ estimates.

E. Concluding Observations

13. Retail fuel prices below the cost-reflective level have resulted in large bills to government in recent years. The key to avoiding such costs is to regularly adjust fuel prices in line with the existing CPP model. Fuel price increases are never popular, but keeping retail prices unchanged in the face of higher crude prices or depreciation of the kwacha does not make the costs go away and only increases the size of the adjustment that eventually takes place. Regular and automatic adjustments would mean smaller and less painful price changes and would also help depoliticize the question of fuel pricing.

14. Outside the CPP model, retail fuel costs are best reduced by addressing the reasons for the high costs in the supply chain. Aging infrastructure and the associated outdated and inappropriate mode of fuel processing and transport, as well as inefficient procurement and general lack of competition, are contributing significantly to high fuel prices in Zambia. If addressing these obstacles would cut the premium of Zambian over South African wholesale prices in half, as should be possible given the similar distance from fuel export points, then this would allow for a roughly 10 percentage point reduction in retail prices.7

1

Prepared by Grivas Chiyaba and Tobias Rasmussen.

2

A blend of crude oil, condensate, naphtha, and gasoil (diesel).

3

The Tazama pipeline and Indeni refinery date back to the late 1960s and early 1970s and are near the end of their lifespans.

4

This estimate tracks petrol, diesel, and kerosene (which account for about 80 percent Indeni’s total production by volume) to the final retail stage. The remaining products (liquid petroleum gas (LPG), heavy fuel oil (HFO), Jet A1, and bitumen) are only tracked to the wholesale stage.

5

This is an estimate and not a precise figure, as there are some gaps in the available data. First, we do not have the actual volume of the final products produced at Indeni, but rely on theoretical conversion factors. A second unknown element is the applicable exchange rate, as payments for feedstock take place after the time of delivery. Here we assume that the actual exchange rate used is the one that applies at the time the next shipment is delivered. Third, we base our estimates on the transport margin applicable from Ndola to Lusaka as an approximation of the country-wide average. Forth, Zambia only adopted the system of country-wide uniform pump prices in September 2010 and for the period up to then we simply applied the Lusaka price. These data gaps could reduce or increase the estimated cost-reflective value of production but sensitivity analysis indicates that the estimated total shortfall would not change by more than at most $50 million. Finally, for LPG, HFO, Jet A1, and bitumen, we do not have information to determine if there are subsidies incurred beyond the wholesale stage.

6

The “Other” category in Figure 6 reflects the Road Accident Fund charge in RSA and the Strategic Reserves Fund fee in Zambia.

7

Alan Whitworth argued in 2011 that removing the monopoly enjoyed by the Tazama pipeline and Indeni refinery and allowing competition from finished products transported by road and rail would reduce import costs by as much as 10–15 percent and have the added benefit of increasing the reliability of fuel supplies. (http://www.zipar.org.zm/documents/How%20to%20Reduce%20Zambia%E2%80%99s%20Fuel%20Costs.pdf)

Zambia: Selected Issues
Author: International Monetary Fund. African Dept.