The Project Finance Law Review: Government Funding
While governments look to encourage project financings in a whole range of sectors, they tend to have an intense focus on public infrastructure projects (e.g., roads, airports and public buildings). Ensuring that a country has sufficient and appropriate infrastructure, and maintaining this infrastructure, is seen by most citizens as a fundamental role of that country's government. Therefore, political pressure ensures that governments remain focused on infrastructure projects. This does not necessarily mean that the government has to finance and fund the infrastructure itself, but that if it does not, it still has to ensure that the conditions are in place such that the required infrastructure is developed by others. If this is the role of governments, then most governments continue to fail, with infrastructure development gaps reported in substantially all countries, regions and indeed globally for decades.
The Global Infrastructure Outlook, which is a G20 initiative, identifies a US$15 trillion global infrastructure gap (a comparison of current investment trends to investment need through to 2040 if countries were to match the performance of their best performing peer) on a baseline of US$79 trillion of current investment trends. There are countries identified in this outlook that have a small gap, for example France with a gap of just US$10 billion on a projected US$1.8 trillion; countries with modest gaps, for example the United Kingdom with a gap of US$148 billion on a projected US$1.7 trillion; and countries with wider gaps, such as the United States with a gap of US$3.8 trillion on a projected US$8.5 billion, Brazil with a gap of US$1.2 trillion on a projected US$1.5 trillion and Nigeria with a gap of US$221 billion on a projected US$657 billion. These gaps become even larger if all infrastructure necessary to meet the sustainable development goals set in 2015 by all United Nations Member States is included.
The numbers quoted for infrastructure gaps are huge for countries at very different stages of development, but that is not to say that infrastructure gaps in different countries and regions are similar in type or effect, as the infrastructure gap that leaves (by most estimates) 50 to 60 per cent of people in sub-Saharan Africa without access to electricity has a greater impact on quality of life and gross domestic product (GDP) than the delayed upgrades of road infrastructure in Western Europe. While sustainable economic development does not necessarily follow from all infrastructure investments, critical basic infrastructure (e.g., sufficient power and availability for industry and consumers) is a necessity for economic development and so is a priority for governments of developing countries in particular.
II Financing and funding
The terms 'financing' and 'funding' are often used interchangeably. However, when considering how governments need to enable infrastructure development, it can be helpful to make a distinction between financing of infrastructure (meaning how the construction costs and other development costs are paid) and funding of infrastructure (meaning how the built infrastructure is paid for over its life). Government support for infrastructure development may be necessary or desirable at both the financing and funding stages. To use an example, financing for a road from any source (public or private) is unlikely to be available unless it is clear how maintenance costs will be paid and, in the case of private financing, that the funding monetary stream for both maintenance and repayment of the financing must be clear and adequate. In the United Kingdom, the vast majority of funding of road infrastructure projects is from taxation (with only a relatively small portion from user charges such as tolls). On the other hand, the funding of energy and utilities projects in the United Kingdom is almost all from user charges. Other countries differ on these details, but financing for projects is most challenging in countries where the funding source is not clear (e.g., unknown demand), or not sufficiently creditworthy or reliable (e.g., a thinly capitalised offtaker).
This critical distinction between financing a project and funding a project goes some way to explaining why, despite estimates of huge amounts of private capital being ready to be deployed to appropriate infrastructure projects (including capital specifically focused on developing markets), the infrastructure gaps in a multitude of different countries have remained while capital chases the more limited number of 'bankable projects'.
The funding challenge explains why large amounts of private finance has been available to develop oil and gas projects in sub-Saharan African jurisdictions such as Nigeria and Mozambique at competitive interest rates, but the same liquidity has typically not been available for power projects in sub-Saharan Africa at any cost, despite this infrastructure being critical to the development of the countries in the region. From a funding perspective, oil and gas projects usually have substantial US dollar revenues being paid by creditworthy offtakers into offshore bank accounts, giving great comfort to the financiers, whereas the funding for power projects would tend to look to rely on local currency payments by offtakers with more limited creditworthiness.
On the basis that governments in countries with the most pressing infrastructure needs will be unable to finance the necessary infrastructure out of general taxation, such countries must focus on actions that enable non-government financiers to become comfortable with the funding side of a project (i.e., its revenue streams and other support). An example of a successful power project financing in Nigeria is the financing of the 459MW Azura-Edo IPP Project, where the Ministry of Finance in Nigeria entered into a put-call option agreement that ensured that the Ministry of Finance stood behind the thinly capitalised newly created government-owned offtaker (the put-call option agreement was an innovative structure, but similar in effect from a lender perspective to a more conventional sovereign guarantee). The Ministry of Finance indemnified the International Bank for Reconstruction and Development for issuance of partial risk guarantees and the Central Bank of Nigeria further gave some comfort as to currency exchange rates. This combination of government support for the Azura-Edo IPP Project unlocked US$900 million in financing from banks and equity providers. Nevertheless, there are challenges for governments in emerging markets in giving government guarantees (or instruments akin to a guarantee) to support funding, as the credibility and value of these guarantees will be diluted if they are given out on many projects given the size of their infrastructure deficits, as discussed further in Section V.
III Government financing
At its most basic, there are two ways to finance infrastructure projects: publicly or privately. However, it is also important to distinguish between whether the financed infrastructure will be publicly or privately owned. For example, private project financings have occurred in Turkey in the past decade for both public and private hospital facilities under the encouragement of, and frameworks set out by, the government of Turkey in 2010 when it launched its health public–private partnership programme.
Public financing for publicly owned infrastructure comes primarily from taxation or public borrowing and will typically appear on the public sector balance sheet as debt. Private financing for publicly owned infrastructure comes from a whole range of financiers, ranging from commercial banks, institutional investors such as insurance companies and pension funds, and also development banks, export credit agencies and other institutions (for these purposes, these sources constitute private finance as even to the extent these monies may originate from governments, they are not from the government of the country where the infrastructure is being developed). The encouragement of private project financings for publicly owned infrastructure in the United Kingdom was launched in earnest in 1992 under the moniker 'private finance initiative' (PFI) as part of a wider privatisation and PPP drive to efficiently improve infrastructure, while also not having such infrastructure on the United Kingdom's balance sheet. This model has been used as an example (both on a positive and 'lessons learnt' basis) by many countries in subsequent years. In 2018, the UK government stated that it would no longer use the PFI programme.
Public financing for privately owned infrastructure projects could also theoretically occur. However, in practice, this is relatively rare and financing for privately owned infrastructure is typically by private financing (either project financing or corporate financing). Nevertheless, given the governmental interest in ensuring that such infrastructure is developed, governments may provide guarantees or funding injections, or both, to ensure that critical infrastructure is developed, as further discussed in Section V. It is worth noting that there are relatively few countries where large amounts of essential infrastructure have been privatised, therefore the above considerations are most applicable to countries such as the United Kingdom that have undergone an extensive privatisation of essential infrastructure programme. This means that, for example, utilities are privately owned, and thus are privately funded. Even in countries that have a high level of privatisation, there are still publicly owned infrastructure projects that do not naturally lend themselves to private financing – such as flood related infrastructure – and so this is often entirely publicly funded.
IV Government bonds
'Government bonds' is a term used for many different types of bonds in different jurisdictions. For the purposes of our use in this chapter, we are focused on bonds for which the use of proceeds is specifically to finance infrastructure and stated as such in bond documentation. We exclude bonds where the use of proceeds is stated as being for 'budgetary and general funding purposes' or similar, albeit these proceeds may in fact be used for infrastructure development (and the bonds may, in reality, be issued for this purpose). With that distinction made, the next critical distinguisher is at what level of government a bond is issued, be it at a country or federal level; state, province or county level; or city or local level. For example, the credit of a bond issued by the federal government of Nigeria (an FGN bond) is typically seen as different to a bond issued by Lagos State in Nigeria, as each has access to different revenue streams to stand behind its obligations.
Within the category of government bonds that are to finance specifically stated infrastructure, there is a further key distinction that goes to the fundamentals of how investors will analyse the risk of investing in the bonds, namely if such bonds are 'general obligation bonds' or 'revenue bonds'. General obligation bonds are fully backed by the issuing government, with all the government's assets and revenues standing behind such bonds, and so an investor will be primarily looking at the creditworthiness of the issuer government as a whole rather than the infrastructure asset that the proceeds will develop. Revenue bonds are not fully backed by the issuing government, rather payments to investors are directly linked to the financed infrastructure project, and so an investor will need to analyse the revenue stream from the infrastructure project in the ordinary project finance way. A key advantage of a revenue bond for the government is that it typically should not feature on the government's balance sheet as a debt liability, in the same way as project financings generally are intended to be off balance sheet.
The United States has the most sophisticated and developed government (project) bonds market, so it is useful to focus on this market as an example. In the United States, these project-specific bonds are typically issued by state and local governments and referred to as municipal bonds, whereas the term 'government bonds' often is used to refer to treasury bonds only. This is a useful distinction and it is becoming international practice to refer to bonds issued at a lower level of government than the sovereign as 'municipal bonds'.
In the United States, municipal bonds issued by state and local governments have been used extensively to finance a wide range of public projects such as ports, airports, highways, sewerage infrastructure, hospitals and colleges. There is a long history of municipal bonds in the United States, with the City of New York issuing a general obligation bond in 1812 to fund a canal project. Today, almost three-quarters of all core infrastructure in the United States is financed by municipal bonds, with a market estimated at around US$3.9 trillion. A key feature of US municipal bonds is that interest or other investment earnings on them is usually excluded from gross income of the investor for federal income tax purposes. US municipal bonds can either be revenue bonds or general obligation bonds, although the majority issued to date are revenue bonds. Nearly half of all municipal bonds are held by individual investors and another quarter by mutual funds. The highly liquid market that has developed over the years ensures that bondholders can easily find other buyers and means that US local governments are not dependent on bank lending.
The concept of municipal bonds has been employed in other parts of the world, particularly in countries with well-developed private capital markets and creditworthy subnational governments, such as France. Other countries are also moving towards subnational debt as a means to fund the infrastructure gap. For example, the United Kingdom has developed a local government municipal bond market where a centralised UK Municipal Bonds Agency (UKMBA) can issue bonds with proceeds lent to local authorities. This allows a local authority to borrow at a lower cost than if it borrows from the central government (via the Public Works Loan Board (PWLB), which increased its interest rate in 2019 to discourage borrowing from the central government). The first bond in 2020 was a £350 million five-year bond priced at 0.8 per cent above SONIA (significantly lower than the PWLB five-year maturity rate) and issued on behalf of Lancashire County Council, which was to be used to facilitate the council's capital funding agenda. However, future municipal bonds could be more focused to specific local infrastructure projects. The UKMBA does, however, reflect a divergence from US practice, in that not only are the bonds general obligation bonds, but the other local authorities that participate in the UKMBA also collectively stand behind the obligations and not just the local authority where the infrastructure is to be developed. The United Kingdom is not alone in not following the revenue bond approach at this stage and, in practice, substantially all of the municipal bonds outside of the United States are general obligation bonds.
Municipal bonds have also been used in some developing countries where there are developed capital markets, such as South Africa and India, but they have not been used as extensively as in the United States and other developed markets. For example, Johannesburg in South Africa issued its first municipal bond in 2004, but has only successfully launched four municipal bonds despite having a credit rating from two agencies. These bonds were to finance infrastructure capital expenditure generally and were general obligation bonds. The last bond by Johannesburg in 2014 was a green bond and use of proceeds accordingly limited to appropriate infrastructure, but it also required partial credit guarantee coverage by the International Finance Corporation and Development Bank of South Africa, which is indicative of the creditworthiness challenge that many subnational governments have in developing countries to issuing municipal bonds attractive to international investors. Counterexamples of deep liquidity in developing markets tend to be found where there is sufficient domestic demand. For example, in December 2021, Lagos State was oversubscribed for the largest ever general obligation bond raised in Nigeria by a subnational government. Strong domestic demand can lead to success for municipal bonds issued by local government-owned companies, with the financial backing of the local government being implicit rather than necessarily explicit (or enforceable legally). For example, in India, Ahmedabad Municipal Corporation was the first to make a bond issuance in 1998 and, since then, many local entities in different cities have accessed the capital markets through municipal bonds. China also has a strong municipal bond market of infrastructure-focused bonds issued by provinces, which are often heavily oversubscribed due to local investors not perceiving a significant risk difference between province-level debt and sovereign debt, which has led to only a narrow spread between the two.
Nevertheless, there is a limit to how many general obligation bonds can be issued, with many subnational governments in developing countries already having high debt compared to their taxes and other revenues. For municipal bonds to constitute a significant portion of financing for infrastructure in developing markets, it is likely that there will need to be a move towards the revenue bonds that dominate the municipal bonds market in the United States, which in turn will put the pressure on ensuring that the infrastructure projects are 'bankable' in a project financing sense.
With this adjustment in approach, municipal bonds should be able to take advantage of the bond markets in developing markets as a whole. The largest relative increase in bond issuance generally this century has occurred in sub-Saharan Africa, which has quadrupled since 2000, and maturities have increased. Until recently, maturities of bonds in sub-Saharan Africa were rarely longer than 10 years (for Eurobond issuances at least). However, in February 2020 the government of Ghana issued a US$3 billion 40-year Eurobond, the longest for a sub-Saharan African government to date, which was almost five times oversubscribed. Nevertheless, while it is publicly understood that the proceeds of that bond will be used in relation to infrastructure development (particularly road and rail projects such as the Tema-Mpakadan Railway Project), the use of proceeds is not so specific and the main mentioned purpose is funding Ghana's fiscal deficit. If these long tenors can be combined with a move to revenue bonds, municipal bonds could start to have the potential to play the same role in developing countries as they already play in the United States.
V Other government support
i Government guarantees
Given the importance of critical infrastructure to economic development, governments in developing countries are, on occasion, willing to provide guarantees. When issued by the government of the country rather than at a subnational level, these are usually referred to as sovereign guarantees. Sovereign guarantees are often seen where critical infrastructure is being developed by state-owned companies and the state-owned companies are looking to raise financing on a project financing or corporate level. Another typical situation would be where a government-linked entity is a key offtaker for a project.
Government guarantees are not a binary and can be given in a wide range of circumstances covering a large range of varying risks, but all involve the government where a project is based assuring financiers that the government will take certain actions or will refrain from taking certain actions that impact the project. The theoretical basis of requesting a government guarantee is that the government should accept risks that it controls or it is better able to manage, or both, such as exchange rate-related risks, political risk or tax change-related risks.
ii Types of sovereign guarantees
There are generally two types of sovereign guarantees: financial (or credit) guarantees and performance guarantees.
Financial guarantees involve the government guaranteeing the debt service obligations of the borrower in the event of a default, such as where the revenue from the project does not meet the repayment obligations. As the government is putting its creditworthiness behind the project, this impacts the government's balance sheet and borrowing limits. These guarantees place all the risk on the government and are not linked to performance indicators for the private sector parties involved. As a result, the government could be liable for all political and commercial risks, from natural disasters to contractor defaults. Partial credit guarantees covering part of the debt service can also be used by governments, which ensure that the other project participants retain sufficient risk to align interests.
Performance guarantees are used to manage specific project risks and can come in a number of forms, including:
- revenue guarantees to manage demand risk;
- exchange rate guarantees, transferring the risk of volatile exchange rates to the host government; and
- change in law guarantees, where the host government agrees that, in the event of a domestic change in law which materially alters the economics of the project, the government would put the private investors back into the position as if such a change of law had not occurred.
iii Benefits of sovereign guarantees
Sovereign guarantees can increase investor and financier confidence in a project as they demonstrate government commitment and support, and increase the amount and sources of financing available. They can allow governments to defer their spending obligations to a future date as they do not require immediate funding obligations and otherwise the government may have to finance particularly critical projects itself. Sovereign guarantees can also be used to shorten the delivery time of projects as governments assume the risks that the project participants would have had to investigate thoroughly and assess as part of a due diligence process. Where negotiations have stalled due to certain political or regulatory factors, sovereign guarantees can be helpful to unblock a project, particularly if general policies are difficult to change.
Sovereign guarantees can also be used as a form of credit enhancement, improving the credit rating of the project through the credit rating of the government. This lowers the risk profile of the project, which can improve the terms of the financing, by either reducing the cost of financing, increasing the amount available or lengthening the tenor. For example, a guarantee in respect of payment obligations of a state-owned electricity offtaker could mean private sector investors and financiers would be satisfied with lower returns and lower financing costs.
Sovereign guarantees can also encourage the private sector to invest in, and finance, sectors or projects where private sector entities would normally not have invested or financed, or where the perceived risks are usually too high for the private sector (for example, supply of electricity to a rural area).
iv Risks of sovereign guarantees
Sovereign guarantees bring clear advantages to a project and its financiers. Nevertheless, they do present a number of risks to the host government that issues them. There has been increased focus in recent years on the treatment of contingent liabilities on a country's balance sheet, with by way of example the International Monetary Fund highlighting in 2016 undisclosed power sector guarantees (minimum demand or revenue) in Kenya worth US$3.4 billion (over 5 per cent of GDP). Excessive guarantees can impact a country's financial position and take away from the benefit from a government perspective of projects being undertaken on a project finance basis. Accordingly, a number of developing country governments are resisting calls to provide these types of guarantees and recognising that, if given for one preferred project, future projects may expect guarantees also.
One example quoted in World Bank materials of the fiscal risks associated with generous sovereign guarantees are the government guarantees provided to Hehe Power Ltd in respect of the Sha Jiao B power plant, covering raw materials supply, operating revenue and foreign exchange rates, causing the government to assume many project risks. The result was that the government was liable for HK$16.8 billion of exchange rate losses and 4.7 billion yuan for fuel cost escalation losses, while the private investor's internal rate of return reached 38.8 per cent. It is clearly the case that, if sovereign guarantees are very broad such as full credit guarantees, this can create a moral hazard scenario where the incentive for private sector investors and financiers to fully assess risk and due diligence is reduced as they do not have to assume as much risk.
For governments with low sovereign debt ratings or without a rating, which is often the case in jurisdictions where projects need the most support, sovereign guarantees may not be sufficient on their own unless combined with support from development banks or other international organisations, which may lead to the unfortunate situation that the country's credit is put on the line, while the projects still have to pay the fees involved in securing further external support.
Governments have looked for alternatives, with a commonly preferred alternative being a comfort letter or letter of support from the host government rather than a sovereign guarantee. This is intended to provide a clear indication of government support and future intentions towards the project or matters that may impact the project, while at the same time not being intended to be legally enforceable (and rather involving governments acknowledging a reputational or moral obligation). The scope of comfort letters can vary widely and can provide assurances that the host government will undertake or maintain certain actions or policies to support the success of the project. There is, however, no standard text for comfort letters and, while they may be intentionally drafted as binding or non-binding, they are often deliberately unclear or ambiguous in an attempt to keep both parties happy, often including a promise to 'resolve a problem' or 'prevent an economic loss'. There is precedent for arbitral tribunals ruling against host governments based on simple letters of support, with letters of support issued by some sub-Saharan African countries being construed as enforceable against their views. There is thus the risk that comfort letters be effectively concealed – or unknown or unquantifiable – contingent liabilities on a country's balance sheet.
On the more innovative side, the project financing of the Azura-Edo IPP Project in Nigeria (referred to in Section II) involved the Nigerian government's resistance to providing a guarantee in respect of termination payments if the power purchase agreement was terminated. Instead, the government entered a put-and-call option agreement in which the government agreed to purchase the power plant (the put) at a certain price in the event that the offtaker defaulted and also had the option to purchase the power plant (the call) under certain circumstances, such as a concessionaire's default. While this model is understood to have no effect on Nigeria's balance sheet, it arguably does have a contingent liability aspect similar to a guarantee.
Notwithstanding the concerns with sovereign guarantees, the critical infrastructure gaps in many developing countries in particular mean that governments continue to feel compelled to provide what is needed to make projects 'bankable'. In a recent example of this pressure, in 2021, Nigeria announced plans to increase the use of sovereign guarantees to fund infrastructure by raising the maximum value to 5 per cent of GDP (from 1.5 per cent in 2019) with the aim of reducing the need for raising debt for projects and, therefore, reducing public debts.
1 Adrian Lawrence is a partner and Grahame Fischer is a senior associate at Ashurst LLP.