Why an ESLA-backed Energy Bond must not be 'pure' public debt - Seth Terkper

An issue under current discussions is whether Ghana should add the ESLA-backed Energy Bond it is to issue to public debt, instead of remaining on VRA’s Balance Sheet. ESLA means Energy Sector Levy Act, 2015 (Act 899). Ghana has issued an ESLA Bond Prospectus and roadshows are underway for investors at home and abroad. We understand that the IMF takes a public debt view for the impending 5th ECF Program Review. Our opposing view is that, even if this multilateral view prevails, Ghana should keep to plan and not classify such bonds or loans as public debt in the Public Accounts sent to Parliament—provided we continue to take certain concrete steps under the new debt management policy.

As noted later, we have precedent as guide but good reason to forcefully argue against this retrogressive step since the bond is backed by a special levy imposed by Parliament. In the other instances, in our medium-term plan for debt sustainability, we are using our oil revenues to set up appropriate financial market structures and instruments. An example is the use of the Sinking Fund to partially take off the 2007 Sovereign Bond. The better option, which we did with the earlier 2016 ESLA loan, is to assist Ghana to consolidate the verifiable steps it is taking to classify such Bonds as “contingent liabilities” that crystalizes upon default.

If we do not issue an ESLA Bond in excess of revenue estimates from the levy—current debt service flows that must add to escrows—the levy is an implicit guarantee by taxpayers and power consumers. Hence, if the Bond is made a pure public debt, it add an explicit (sovereign) guarantee dimension that leads to its over-securitization or extreme state undertaking. In technical terms, it is “double counting” since, even if the Bond value exceeds the ESLA and other proceeds, in computing our Debt Sustainability Analysis (DSA), the multilaterals must add only proportionate excess amount to public debt until we default.

Why we had to enact ESLA

The Government took the ESLA Bill to Parliament to (a) raise funds to pay energy-sector SOEs debt, notably for VRA; (b) minimize the non-performing loan (NPL) impact of such debt on domestic and foreign banks as well as suppliers; and (c) improve the Balance Sheet of the SOEs to enable them play their envisaged future roles in an oil-and-gas era. These roles include stable power supply and, in particular, meeting their financial obligations to banks and independent power producers (IPPs) under the World Bank Partial Risk Guarantee (PRG).

We must stress that the PRG is an alternative to a sovereign guarantee and “contingent liability” backed by our IDA resources. Hence, the guarantee will be recalled—implying use of those IDA resources to settle any default—when the SOEs fail to collect the bills for power consumed to pay for gas supplied from the Sankofa Fields. We also note that the Millennium Challenge Corporation (MCC) Compact II is complementary in enabling ECG and downstream energy SOEs meet their PRG and other obligations for power supplied by VRA, GhanaGas and private sector IPPs such as Sunon-Asogli.

We note further that the weak SOE balance sheets are due to unpaid subsidies for power consumed and operational inefficiencies that must be resolved in ongoing restructuring plans. While the subsidy conundrum is not new, the related outstanding SOE debt from unpaid direct loans and letters of credit (LCs) continue to weigh heavily on domestic banks as NPLs. Suppliers also threaten to call on explicit (e.g., promissory notes) or implied (i.e., shareholder) guarantees for unpaid bills for gas supplies. The two-and-half years disruption in gas supply from Nigeria had worsened the situation.

We must be aware of precedents

When we entered the ECF program in 2014, the IMF mission insisted that BOG must “sweep” the Sinking Fund (set up with flows from the Stabilization Fund under the PRMA) into the Consolidated Fund for general use and to reduce the budget deficit. Ghana argued that the opposition to the Sinking Fund and the Ghana Infrastructure Investment Fund (GIIF) means we muse use our oil revenues for consumption and not to manage our debt and improve infrastructure.

To date the Sinking Fund (i.e. asset) is technically a negative “budget-financing” or “deficit-reduction” item, not a verifiable asset (cash) or reserve used to offset total Public Debt (i.e., liability) in the Public Accounts prepared by the Controller. Our plan is to change this non-accounting classification by coding our debt stock and flows in the Chart of Accounts (COA) under the Ghana Integrated Financial Management System (GIFMIS) reforms.

Secondly, the US$ 1.0 billion Sovereign Bond (liability) issued in 2015 was never off-set against public debt in the DSA even though it is still “guaranteed’ fully by the World Bank. The full amount was also deposited in a verifiable BOG account (asset) in a New York Bank since we undertook to use the entire proceeds to refinance or replace existing debt. Hence, until we completed the refinancing in 2016, our public debt was “grossed-up” or inflated by the value of the Bond. This double-counting is the result of not treating the debt as a contingent liability that should crystalize only upon misapplication of the fund—considered to be remote because of the New York escrow account.

This extreme “developing country” approach to “grossing-up” debt—a form of single-entry or cash basis accounting—violates the rules which households, businesses, and (middle-income and advanced) countries use to prepare Balance Sheets—incidentally, under the IMF Government Fiscal Statistics (GFS) rules for preparing government accounts. Given Ghana’s enduring MIC status, it should be encouraged to continue with its adoption of the alternative accrual accounting approach, as best practice and structural change under the International Public Sector Accounting Standards (IPSAS) that the Controller, Auditor-General and Institute of Chartered Accountants, Ghana (ICAG) have all approved.

The way forward for Ghana

We strongly propose that, whatever treatment the multilaterals give to the ESLA-backed Bond, Ghana should continue to create the necessary “asset” accounts to move our debt methodology to an accrual or “contingent liability” basis. This “smart borrowing” approach will require that we persist in allowing the oil revenue flows into “asset” accounts such as the Sinking Fund, Debt Service Reserve Accounts (DSRA), GIIF, ESLA and other Escrows.

Secondly, some of the Escrow accounts fall under the “self-financing” rule that requires that we use flows from SOE and other commercial projects to pay service direct and guaranteed loans for such projects. We must resist the temptation of diverting these debt and infrastructure accounts into consumption, especially when we get into tight fiscal situations that may be due policy misalignments or external factors that may be beyond our control, such as fall in commodity prices.

Third, Parliament must make the ESLA law certain by placing a firm “sunset clause” on the levy and more firmly legislating its use to settle specific SOE debt. Another key issue is whether ESLA will last 10 or 15 years, consistent with the tenor of the Bond. It seems the delay in issuing the bond under an earlier shorter term syndicated loan may have increased the liability on the SOE books.

A fourth appropriate medium-term goal is to take advantage of the additional oil flows to make the “energy” bond” as enviable as the “cocoa” bond in the future.

This requires that we clarify the fiscal status of the Special Purpose Vehicle (SPV) for the ESLA-Bond, in relation to the goal of using GIIF as a sovereign wealth fund (SWF) to spearhead the commercial loan strategy for the country. Further, we must set up and strengthen the Debt Management Office (DMO) under the PFM Act, as part of the shift to accrual accounting. In line with the Chart of Accounts (COA), we must implement the codes assigned to our debt instruments and related asset or reserve accounts in the Public Accounts, under the PFM or GIFMIS reforms. This will require SOEs and CAGD to record these transactions in their books, simultaneous with deposits and disbursements at Bank of Ghana.

Conclusion

We are aware of the failure to build national consensus around the new debt management or “smart-borrowing” strategies, given the politics and chorus that made the “nominal” debt too deafening. This is the time to build this consensus in the national interest, given the “turnaround” in the economy as well as a potentially entrenched MIC status—that require we use part of the additional oil revenues to plan for the gradual but inevitable loss of access to concessional loans and grants.

Moreover, as the IMF and World Bank admonish in other contexts, we cannot afford to be put in the category of countries that are accused of dissipating their energy resources for consumption only. It is against this background that the multilateral agencies and development partners must take our debt management strategy seriously.

The right approach is to offer technical assistance to achieve this goal, as the World Bank Treasury, Commonwealth Secretariat, and IMF Fiscal Affairs Department (FAD) have been doing. This will enable Ghana to correct any flaws that may exist in administering the new debt management policies—not take the easier options of continuing with cash accounting and “grossing-up” debt.