Current banking crisis:  …Why ESLA could have been used to minimize the impact

Seth Terkper

This article discusses the potential of using ESLA (Energy Sector Levies Act, 2015 [Act 899]) to substantially resolve the ongoing banking or financial sector crisis. The Act was a “smart-borrowing” initiative under the Home-Grown Policies that Cabinet and Parliament approved in the 2013 and 2014 Budgets. We note that ESLA was not among the measures under 2014 IMF Enhance Credit Facility (ECF) Agreement.

The relevance to the financial crisis is clear from the Ministry of Finance (MOF) 2016 report sent to Parliament in 2017—Annual Report on the Management of the Energy Sector Levies and Accounts (Years 2016)—which notes the objectives of ESLA as follows:

“The Act (i) consolidates existing Energy Sector Levies and defines a framework to correct imbalances in the collection, distribution, and utilization of the levies; (iii) ensures the financial viability of energy sector State-Owned Enterprises (SOEs); (iv) facilitate investments in the sector; and (v) mitigate against market, credit and liquidity risks of energy sector SOES and their counterpart creditor banks” (par. 2, p 6; emphasis added).

The 2017 ESLA report repeats the goal of using ESLA to leverage the markets to resolve the adverse direct and indirect impact of debt to creditors and counterpart banks.

“The debt overhang also increased the exposure of those institutions to credit and liquidity risk and, consequently, impacted significantly on the balance sheets of their counterpart creditor banks. To address these challenges, Government passed the ESLA in December 2015, with full implementation beginning in 2016” (see also Par. 70 to 72).

ESLA is a crisis resolution tool for SOE non-performing loans (NPLs) and arrears to suppliers (e.g., N-Gas, Sunon-Asogli); improvement in rural electrification and public lighting; stabilization of petrol pump prices; an enhanced levy for the Road Fund to clear arrears to contractors and counterpart banks. MOF’s setting up of ESLA plc in 2017 as a quasi-fiscal agency seems to compromise the non-classification of SOE debt as public debt. In contrast, the innovative 2015 ESLA mechanism “ring-fences” the inflows to pay debt on SOE balance sheets—through a mandatory “escrow” or debt service account (DSA).


In early 2016, MOF used the actual and estimated ESLA flows, in a “cascade” model, to pay US$250 million cash and restructure of GHc2.2 billion of debt on VRA’s Balance Sheet. The successful model uses the “ring-fenced” DSA flows to pay the banks directly, to settle NPLs due from VRA, TOR and BDCs.

  • Strong ESLA inflows: As the Tables in Section 5 show, about Ghc 2 billion of the total annual ESLA flows are dedicated to settling energy and road sector debt. It was possible to use these funds to ease the pressures facing the banking sector without further loans.
  • “Cascade model” helpful to banking sector: The model uses ESLA flows meant to (a) pay debt to creditors and banks; under (b) a tripartite agreement with government to (c) use the “ring-fenced” flows in “escrow” accounts to settle NPLs due counterpart banks.
  • Successful resolution of SOE and other debt in FY2016. The ESLA “cascade” debt resolution model was used in 2016 as follows:
  • VRA debt: With payment of Ghc250 million upfront, MOF successfully restructured 2 billion of VRA debt owed to 12 domestic banks—paying both principal and interest in five (5) years. By end-2016, MOF had paid two quarterly instalments of Ghc228 million.
  • Road arrears: The enhanced [ESLA] road levy was used to restructure over Ghc 1 billion of Road Fund debt to contractors and creditors, including a loan due to SSNIT since 2010.
  • TOR and BDC debt: the ESLA TOR levies for debt recovery and price stabilization were used to restructure over Ghc 1 billion of Tema Oil Refinery (TOR) and Bulk Distribution Company (BDC) debt.
  • Strategic stock initiative: this initiative supported the liberalization of fuel-pump prices, with MOF releasing nearly GhC190 million to the Bulk Oil Storage and Transit (BOST) company to (a) buy fuel for storage at low crude oil price points; and (b) releasing it at a reasonable margin for re-export or domestic use when prices increase.
  • Draft AFRIEXEM Term Sheet: At end-2016, MOF had the final draft of a Term Sheet negotiated with a consortium of banks led by AFRIEXIM Bank. The US$600 million loan with 5-year tenor and 8 percent interest rate was for restructuring additional VRA and other energy-SOE debt due to suppliers such as N-Gas and Sunon-Asogli.

New administrations take time on existing policies but the Parliamentary debate of ESLA led to an Opposition boycott of the “nuisance tax”, with a sluggish take off in 2017.

  • Delays in FY 2017 worsened the situation: The 2016 draft AFRIEXIM facility was set aside, with MOF choosing to make piecemeal payments. this delayed the restructuring of US$600 million of debt owing to creditors and counterpart banks.
  • ESLA extension for 10 years means more inflows. The 2017 ESLA Bond implies a levy extension for 7-to-10 years—compared with 5 years for the 2016 VRA facility. At an annual flow of GHc 3.2 billion, ESLA could generate over Ghc30 billion to support the estimated US$ 8 billion facility needed to resolve the banking crisis.
  • FY 2017 (Interest-only) Bond reintroduces risks: The “principal-plus-interest” 2016 VRA debt resolution is subsumed in a riskier “interest-only” 2017 ESLA Bond.
  • Liquidity risk: the longer tenor and interest-only payment means less inflows to the banks and inevitable BOG “fiscalization” of huge financial costs.
  • Bullet and roll-over risks: the payment of principal as a “bullet” in 7 or 10 years implies a major “roll-over” risk, devoid of “smart-borrowing” schemes like a Sinking Fund and Buy-Back.
  • Interest and forex risks: the 2017 ESLA Bonds converted the 8 percent 2016 VRA dollar loan—under a “swap” of existing VRA $-debt service flows—into a 19 percent cedi-loan. This increases interest and forex risks under a depreciating cedi scenario.

Ghana had used “soft amortization” structures to minimize roll-over risks for its external bonds, with the 2023 Sovereign Bond being the only “bullet” of 4 Bonds at end-2016.

  • ESLA subject to FY2017 Budget “cap”: The “capping” of various statutory funds such as DACF, GETFund and NHIL is extended to ESLA. The 2017 ESLA Report (par. 41) states:

“… a shortfall in lodgment of GCc306 million (21.1 percent) was recorded, as a result of …. Government policy on “capping” statutory funds, which reduced transfers into the Road and Energy Fund Accounts by GHc 157.23 million”.

Clearly, national policy makes ESLA secondary to schemes such as F-SHS, whereas an alignment to pay creditors in the banking, energy, and road sectors would have improved the fortunes of the financial and real sectors.

  • ESLA as mainstream revenue: Besides the “ring-fenced” debt-service amount, the 2016 Budget/Mid-Year treated some of the ESLA flows as mainstream revenue since they accrued to agencies such as Energy Board that were on central government budget. The “cap” raised these transfers from GHc … million in 2016 to GHc …. million in 2017.
  • Strategic Stocks and BOST refund: The 2016 ESLA report notes that MOF released GhC189.39 million to BOST to start the strategic oil stocks scheme (par. 34, p 14), as alternative to subsidy payments. However, the 2017 ESLA Report notes:
See Also:  March inflation drops marginally to 10.4%

“An amount of Ghc189.39 million was transferred to BOST … for the procurement of strategic oil stock in 2016 … As at end-December 2017, BOST had refunded Ghc120 million … leaving an outstanding balance of GHc69.39 million” (par. 45, p 25).

The quote is curious since the amount was to create a revolving fund for managing stocks, prices and subsidy—a primary reason for the SOE debt ESLA levies. It is not a traditional loan to be refunded.

  • Payments outside ESLA legislation: The 2017 ESLA and other reports show ESLA proceeds being used to pay (a) pension arrears [NPRA TPFA] of GHc600 million and (b) Partial Risk Guarantee (PRG) of GHc61.5 million. These were not part of the “legacy” debt in the 2015 ESLA Memorandum. The PRG also suggests ECG inability to pay for the SANKOFA “gas-to-energy” Letter of Credit (LC) waterfall payment arrangements.
  • ESLA plc and exposure to public debt: SOEs are profit-making entities whose debt is not co-mingled with the public debt. ESLA simply viewed the statutory levy as merely improving the probability of payment. Though quasi-fiscal agency, ESLA plc must not resemble a guarantee institution instead of being managers.

In 2016, MOF successfully negotiated a facility to refinance Ghc 2.2 billion of loans owed by VRA to twelve (12) domestic banks, led by the President of the Association of Bankers. Table 1 shows extracts from the Sheet as follows:

Table 1: Restructuring Agreement (2016) with 12 Banks



Amount (million) Rate (%) Period (years)  


Ghc (million) component
Loan amount Ghc 765.0 25 5 Flexible rate pegged to T-Bill rate
US$ (million) component
Loan amount US$ 357.0 8 5 Fixed rate
Common elements
Estimated VRA debt Ghc 4,200.0     Foreign component x-rate of
1st restructuring amt 2,200.0      
Upfront payment Ghc 250.0     Paid from the 1st ESLA flows
Repayment period     5  
Instalments Ghc 228.8     2 quarterly instalments (end-2016)


A second term sheet was handed over during the 2016 Election Transition process for refinancing debt owed to banks and suppliers. The US$600 million loan from an AFRIEXIM-led consortium of six (6) banks was repayable at 8 percent in five (5) years. It was not utilized and replaced with the ESLA plc bonds in August 2017 (Table 2), including a swap of the 2016 multi-currency facility (Table 1).

Table 2: Structure of ESLA bonds issued in FY 2018     



Amount (Ghc mill) Interest Period Comments
Target Realized (%) (years)  
1st Tranche  


2,408.6 19.0 7 These bonds were issued together under the same market program and include refinancing of the 2016 facility for the debt on VRA & TOR balance sheets (Table …)
2nd Tranche 2,375.4 19.5 10
Total   4,784.0    
SOE plus liability   9,393.5    
Other terms/features?  


Table 3 summarizes the status of energy-sector debt, payments, and balances from the 2017 ESLA report. With GHc 20-to-30 billion ESLA funds to back GHc6-to-8 billion of bank debt, the state should revert to paying principal plus interest to protect depositors more. Due to the delays in 2017, the SOE and BDC debt rose from US$ … million and Ghc … in FY2016 to US$1,708 million and Ghc9353—in addition to higher foreign exchange and interest rate risks.

Table 3: Post-2018 Bond Liability, Payment and Balance Status



Liability Payment Balance  



US$ m

Exch, rate  

Ghc m

Total (Ghc m)  

Ghc m


Ghc m

VRA 887.0 4.4 3,902.6       These “cedi-Bonds” (see Table ..) refinanced both foreign & domestic components of the debt under the FY2016 loan restructuring facility
      582.8 4,485.4 2,050.1 2,435.3
TOR 257.8 4.4 1,134.4      
      973.4 2,107.8 775.0 1,332.9
ECG 424.0 4.4 1,865.5      
      318.7 1,664.7 519.5 1,664.7
BDCs 140.0 4.4 616.0 616.0 616.0 0.0  
Total 1,708.8     9,353.5 3,960.6 5,432.9  



Table 4 shows the ESLA rates underlying the inflows from 2016. As noted, they could remain in force for ten (10) years to support the 2017 Bonds—unless another viable source of revenue is identified to lift the burden on taxpayers.

Table 4: Energy Sector Levies and Pricing Formula

      Products Total
Levy Collectx Agency Benefactor/


Petrol Diesel Marine Gas Fuel Oil Kero-sine LPG Electricity  
      price per litre (ppl) ppkg ppkwh  
Energy Debt Recovery GRA   0.41 0.41 0.03 0.04 0.37 1.26
o/w forex loss GRA EDS a/c 0.05 0.05           0.1
    TOR debt GRA EDS a/c 0.08 0.08 0.03 0.04 0.04     0.27
   Power Gen/Infra GRA PGISsA 0.28 0.28     0.28     0.84
Road Fund Levy GRA Road Fund 0.4 0.4           0.8
Energy Fund Levy GRA Energy Cmx 0.01 0.01   0.01   0.01   0.04
Price Stabilization NPA PSR a/c 0.12 0.1     0.1     0.32
Public lighting levy ECG/NEDCO/VRA MOP/EDCs/PGISsA             0.5% 0.005
National electrifx             0.5% 0.005

Source: ESLA Report (2016), Section 7, Appendix A

Table 5 shows estimates and actual inflows from the levies and are deemed to be enough to support a major restructuring of debt owed to suppliers, contractors, and counterpart banks.

Table 5: Programmed, collection, and actual 2016 ESLA performance

January-December 2016 (Ghc million)


Collection     Deviation 
Program Actual Lodgment Prog-Actual Act-Lodgmt % (Prog-Actual) % (Act – Lodgmt)
Energy Debt Rec 1,174.60 1,281.20 1,264.10 106.60 (17.10) 9.08 (1.33)
Price Stabilization 396.40 338.50 326.30 (57.90) (12.20) (14.61) (3.60)
Public Lighting 295.70 168.40 22.90 (127.30) (145.50) (43.05) (86.40)
Nat. Electrificx 296.90 276.90 32.70 (20.00) (244.20) (6.74) (88.19)
Road Fund 1,061.80 1,204.20 1,002.00 142.40 (202.20) 13.41 (16.79)
Energy Fund 30.90 29.80 24.70 (1.10) (5.10) (3.56) (17.11)
Total 3,256.30 3,299.00 2,672.70 42.70 (626.30) 1.31 (18.98)

Source: ESLA Report (2016), Section 7, Appendix B

The ESLA law requires that MoF sets up specific accounts to manage the inflows and outflows—amended since FY2017 to include all the accounts under the ESLA law. Tables 6 and 7 shows that there are significant deviations between the ESLA lodgments and actual utilization.

Table 6: 2016 ESLA lodgments versus utilization


Levy Accounts

Lodgment Utilization Deviation
GhC million Nominal Percent
1.0 Energy Debt Service a/c (EDSA) 478.0 311.0 (167.0) (34.9)
1.1 Tema Oil Refinery (TOR) debt nr 187.0 nr nr
1.2 Forex loss nr 124.0 nr nr
2.0 Power Gen & Infra sub-a/c (PGISsA) 841.7 787.0 (54.7) (6.5)
2.1 Debt payment nr 787.0 nr nr
2.2 Power supply nr 0.0 nr nr
2.3 Under-recovery nr 0.0 nr nr
2.4 PRG nr 0.0 nr nr
3.0 Price Stabex & Rec’vry a/c (PSRLA) 326.3 313.3 (13.0) (4.0)
3.1 Petrol under-recovery nr 0.0 nr nr
3.2 Petrol price stabilization nr 269.5 nr nr
3.3 Premix subsidy nr 43.8 nr nr
  Total utilization 1,646.0 1,411.3 (234.7) (14.3)
Source: 2017 ESLA Report, MOF; nr (not reported)


As Table 7 and 8 show, the 2017 ESLA flows were also strong, given global economic recovery and firms passing the increase in crude oil prices to consumers.

Table 7: Programmed, collection, and actual 2017 ESLA performance

Levy Nominal Amount Deviation (Nominal) Deviation (%)
Program Actual Lodgment Program – Actual Actual – Lodgment Program – Actual Actual – Lodgmentt
EDRL 1,358.1 1,293.0 1,293.0 (65.1) 0.0 (4.8) 0.0
PSRL 415.3 345.3 345.3 (70.0) 0.0 (16.9) 0.0
PLL 229.0 180.0 179.7 (49.0) (0.3) (21.4) (0.2)
NESL 189.7 151.1 151.1 (38.6) 0.0 (20.3) 0.0
RFL 1,331.4 1,152.0 1,058.1 (179.4) (93.9) (13.5) (8.9)
EFL 31.9 30.7 28.1 (1.2) (2.6) (3.7) (9.3)
Total 3,555.4 3,152.1 3,055.3 (403.3) (96.8) (80.5) (18.3)

Source: 2017 ESLA Report, MOF

The Tables continue to exhibit the same significant deviation between collection and lodgment.

Table 8: 2017 ESLA Lodgments versus utilization



Lodgmt Utilizx Deviation  


Ghc million Ghc m %  
1.0 EDSA 438.9 228.8 (210.1) (47.9)    
1.1 TOR debt 156.8 181.8 nr nr    
1.2 Forex loss 282.1 47.0 nr nr    
2.0 PG&IS 687.5 919.0 231.5 33.7 See new item 3.0  
2.1 Debt pmt 640.0 855.1 nr nr    
2.2 Power supply 19.0 0.0 nr nr    
2.3 Under-recovery 28,6 0.0 nr nr    
2.5 PRG Na 61.5 na na Not in ESLA Act


2.6 Bank charges Na 2.4 na na  
3.0 Trs: ESLA plc Na 279.8 na na 2017 ESLA Bond Prospectus  
4.0 Price Stabex 345.0 21.7 21.7      
4.1 Petrol underrec Nr 0.0 nr nr    
4.2 Petrol px stabex Nr 0.0 nr nr    
4.3 Premix subsidy Nr 21.7 nr nr    
  Total utilization 1472.3 1,449.3 43.1      
Source: 2017 ESLA Report, MOF; nr=not reported; na=not applicable


Table 9 shows additional data and inflows that are attributable to extending the coverage of reporting to all ESLA levies and the channeling of the energy debt service through ESLA plc.


On the 10th anniversary of the fall of Lehman Brothers in the US, which many experts believe to be the immediate of the global financial crisis, it is useful to remind ourselves that ESLA is designed to resolve several fiscal, real sector, and banking challenges. While remote causes such as the “sub-prime (rate) lending” debacle remained in the background, the fall of Lehman Brothers raised issues relating to idealism versus pragmatism in making public policy.

The ideal was to make Lehman fall but the contagion effect was disastrous and led to pragmatic interventions by treasuries and central banks, led by the quantitative easing measures by the US Federal Reserve and Treasury. The immediate outcome from drying global financial flows, lack of credit to business, loss of jobs, and slump in global demand for developing countries was the sharp fall in commodities prices. It has taken a decade for the world to see simultaneous recoveries in the economies of developing, emerging or BRIC and advanced economies.

ESLA and the refinancing of VRA’s legacy debt was a pragmatic way of preventing the impact of the energy and road sector debt from contagiously affecting depositors’ funds, jobs and meltdown of the Ghanaian economy—which, in 2015, was just emerging from the energy crisis caused by the damage of the West Africa gas pipeline and disruption in gas supply from Nigeria, christened or nicknamed “dumsor”. It was a tripartite agreement that improved the NPL situation in twelve (12) domestic banks.

We are seeing the contrast in allowing some banks to fall and others forced into a consolidation program. Big is not always beautiful in this context because the global financial crisis also coined the term “too big to fail”, when it came to the question of allowing bigger banks than Lehman Brothers to also fall. Already, the fiscal cost in the form of levies (ESLA) and loans has been heavy; the impact on the real sector in job losses is also real. Given the strong ESLA inflows, we still have an opportunity to redirect the levy to the core business of resolving the real and financial crisis.

We can do this by removing impediments such as the ESLA “cap” and mainstreaming of its flows, to focus better on the NPLs emanating from the energy, road, and other sectors. The banking sector also needs the ESLA inflows and, therefore, we must revert to paying principal plus interest for all ESLA-backed loans or bonds. This to avoid the “bullet” risk associated with the 2007 Sovereign Bond and use of the Sinking Fund to pay over US$500 million of petroleum revenue between 2014 and 2017 to redeem the maturing US$750 million debt.

The writer is the immediate Past Minister for Finance

Leave a Reply

Please Login to comment
Notify of