.jpg)
Public-private partnerships (PPPs) often leave governments holding contingent liabilities: guarantees and implied rescue promises that cost nothing until something goes wrong, then cost a great deal, as Mexico's US$7.7 billion toll road bailout showed. Governments can manage this risk by approving PPPs centrally, valuing each guarantee, reporting their exposure openly and setting aside a dedicated fund.
Not always: a public-private partnership (PPP) can keep new infrastructure off the government's balance sheet at the start, but the contract often leaves the government holding contingent risks, meaning obligations that only become payments if a specified event occurs. Because these obligations need no cash up front, cash-based budgets and accounts tend to overlook their cost, and governments preparing a PPP often ignore these risks or are unaware of them.
When such an event does occur, the cost can be large and it can arrive suddenly. The Covid-19 pandemic showed how payment risks sitting quietly in different contract clauses can be triggered at once and turn into a substantial fiscal burden: the International Monetary Fund expected many PPP companies in health, energy and transport to be hit, with force majeure clauses and minimum revenue guarantees coming into play. It also warns that the fiscal costs of guarantees tend to surface in a crisis, when they do most damage.
Because many projects would not be viable for private investors without them. Governments use PPPs to bring private sector efficiency and possible cost savings into infrastructure. Yet exchange rate movements, market risk and force majeure can make a project too risky for the private sector to carry alone, which is why the European PPP Expertise Centre lists demand and macroeconomic risks among the main reasons governments give guarantees.
Governments therefore step in with support such as revenue guarantees and exchange rate guarantees. Each form of support creates a future contingent liability, so the risks need careful analysis before a guarantee is given. Sound structuring and mitigation of risk are fundamental to a successful PPP.
The usual principle is that each risk goes to the party best placed to manage it. Where the government takes on a share of the risk, it creates substantial explicit and implicit fiscal liabilities, mostly through the PPP agreement itself. Understanding and quantifying those liabilities is essential, and the International Monetary Fund sees valuing them as central to transparency about fiscal risk.
The European PPP Expertise Centre sets out the range of state guarantees used in PPPs in its 2011 guide. They include finance guarantees and contract provisions such as revenue guarantees and termination payments. Table 1 sets out the three families of guarantee the guide describes.
Table 1: Types of state guarantee used in PPPs
The guide groups state guarantees into three families. Only the first is a guarantee given directly to lenders; the rest are written into the PPP contract or given on behalf of another public body.
Any obligation that only turns into a payment if an uncertain event occurs. The most common is a guarantee, which legally binds a government to take on an obligation if a clearly specified uncertain event occurs. Such obligations can be explicit or implicit, and can be mapped against obligations the government owes in any event.
An explicit contingent liability is written into a contract, law or policy commitment. The government guarantees payments to the PPP partner if exogenous events named in the contract occur, and a minimum revenue guarantee is the most familiar example.
Explicit liabilities come in two kinds. Some have amounts that cannot be predicted with certainty, such as the cost of acquiring land for the project (Irwin and Mokdad, 2010, page 3). Others can be specified in advance, so the PPP agreement can include a specific provision for them, as with revenue sharing agreements. Table 2 summarises the contingent liabilities that PPP projects create.
Table 2. The fiscal risk matrix for PPP projects: direct and contingent obligations
Government obligations in a PPP, split by whether they are written into a contract or only expected of the government, and whether they fall due in any event or only if something happens.
The International Monetary Fund ranks government obligations by how certain they are (Table 3). The further an obligation sits towards the uncertain end, the less likely it is to appear as a liability in the government's accounts, and PPP guarantees sit near that end.
Table 3. Government obligations ranked by how certain they are
Government obligations arranged from the most certain, such as debt, to the least certain, such as implicit support. PPP payments sit in the middle; guarantees and implicit support sit at the uncertain end.
Because no contract records them, so they are hard to spot, value and budget for. Implicit contingent liabilities arise when people expect the government to take on an obligation even though no contract or policy commits it to do so. The expectation usually rests on past or common practice, such as providing relief after uninsured natural disasters or bailing out public utilities and strategically important private infrastructure firms that get into financial difficulty.
These liabilities can pose an even greater fiscal risk than explicit ones: the International Monetary Fund finds that implicit liabilities are potentially the most costly. Their value is hard to estimate and the probability that they will be called is uncertain. Because they are not part of a PPP programme's formal commitments, governments find it difficult to make budget provision for them, and budgets seldom set aside anything to meet calls.
An implicit guarantee has a cost even when nothing is paid. A worked example by Ehrhardt and Irwin (2004) shows this: a firm funded almost entirely by debt has about a one-in-three chance of being unable to repay within a year (Figure 1).
Figure 1: How likely a highly leveraged firm is to go bankrupt within a year
The spread of values the firm could be worth after one year. The shaded area is every outcome in which the firm is worth less than the $94.5 million it owes, which happens about 31 per cent of the time.
If a state bank lends to that firm at the government's own borrowing rate, part of the loan's value passes to shareholders (Figure 2). If lenders then come to expect a rescue, the same amount passes again, this time from the government to the lenders (Figure 3).
Figure 2: A cheap public loan moves value from the lender to shareholders
The state bank lent $90 million at the risk-free rate, but a loan with a 31 per cent chance of default is worth only $84.7 million. The $5.3 million difference goes to shareholders, whose $10 million stake is worth $15.3 million. A commercial lender would have charged about 11%.
Figure 3: How an implicit guarantee moves value from the government to lenders
In the authors' example, the firm has about a 31 per cent chance of being unable to repay its debt within a year. Once lenders expect the government to step in, their claim becomes risk-free and its value rises from $84.7 million to $90.0 million. The $5.3 million gain is matched by a $5.3 million loss to the government.
Timing makes these liabilities harder still. The need for support tends to be counter-cyclical: it tends to arise in a crisis, when the government is least able to provide it. Such liabilities can have sizeable financial implications, especially when the government backstops public enterprises, public financial institutions, subnational governments and private firms.
The government ends up paying, sometimes for years and sometimes in billions of dollars. High debt is a common thread: infrastructure companies tend to borrow more than most businesses (Figure 4), so a modest shortfall in revenue can push them towards default and the government towards a rescue. Three cases show how, and Table 4 summarises them.
Figure 4. Leverage by industry among companies listed in the United States, July 2002
Every industry the authors selected carries more debt than the median listed industry (0.22). Power companies and western electric utilities sit at or near the top of the range, at 0.73 and 0.67. Project-financed infrastructure is higher still: debt averaged 65 to 76 per cent of total capital by sector for projects financed from 1997 to 2001 (same source, Table 1, page 11).
The chance of failure does not rise evenly with debt. It stays close to zero at moderate borrowing and then climbs steeply (Figure 5).
Figure 5. The chance of bankruptcy rises steeply as debt increases
With debt repayments of $50 million or less, the firm almost never fails. Beyond that, each extra dollar of debt raises the chance of failure faster: about 13 per cent at $80 million, 31 per cent at $94.5 million and 40 per cent at $100 million.
In Mexico, the government awarded more than 50 toll road concessions covering about 5,500 kilometres between 1989 and 1994. The concessions were highly leveraged, and local banks provided the debt on a floating-rate basis. Traffic came in below forecast and interest rates rose, and by 1997 the government had to restructure the whole programme and bail out the concessions. In total it took over 25 concessions and assumed about US$7.7 billion of debt, according to Ehrhardt and Irwin (2004).
In the Republic of Korea, the government guaranteed 90 per cent of forecast revenue for 20 years on a privately financed road linking Seoul to the new airport at Incheon in the 1990s. When the road opened, traffic revenue was less than half the forecast, and the government has had to pay tens of millions of dollars every year, as Irwin (2007, pages 1 to 2) records.
In the United Kingdom, National Air Traffic Services was to be paid fees based on airline traffic volumes under its PPP arrangement, and the company took on considerable debt for investment and operations. After the 11 September 2001 attacks, air traffic fell below forecast and the company was in danger of missing its debt payments. To reduce the perceived risk of disruption to the service, the government stepped in with £100 million of equity, as reported by Ehrhardt and Irwin (2004).
Table 4. Three cases where governments paid for PPP risks
Three cases in which support the government had given, in writing or by expectation, turned into large payments.
Through central approval backed by analysis and public reporting, though each country does it differently. Chile has guaranteed the revenue of many infrastructure projects, including power generation, toll roads and airports. These PPPs are approved by the minister of finance (Irwin and Mokdad, 2010, page 20) on the basis of contingent liability analysis, which includes estimating the cost and risk of revenue guarantees with a stochastic model. The government also publishes information on contingent and direct PPP liabilities in its annual reports on public finances and on contingent liabilities.
Chile's guarantees have been called, but the payments have stayed small next to the projects they supported (Irwin and Mokdad, 2010, page 18), as Table 5 shows.
Table 5. Chile's spending on revenue guarantees, 1997 to 2008
Chile's payments under its revenue guarantees rose from US$0.45 million in 2002 to US$17.37 million in 2007, then fell to US$7.44 million in 2008.
South Africa follows a similar approach. The National Treasury must approve PPP proposals at four stages before a contract is signed, and proposals must set out their contingent liabilities as part of project preparation. Line ministries then include a disclosure note on their PPPs in their accounts (Irwin and Mokdad, 2010, pages 28 to 33).
The United Kingdom, with its long experience of PPP projects, gives most of the responsibility for project development and contingent liability assessment to the relevant line ministries: the departments that procured or sponsored each project supply its data. Like Chile, it discloses the guarantees it gives under PPP arrangements, with public reporting of the fiscal implications of Private Finance Initiative projects every six months.
Irwin and Mokdad compare Chile and South Africa with the Australian state of Victoria (pages 36 to 38), another programme regarded as good practice (Table 6).
Table 6. How Chile, South Africa and Victoria manage PPP contingent liabilities
Three well-regarded PPP programmes, compared on the liabilities they carry and how they approve and analyse them.
Victoria's process shows how those checks are built in from the start: Cabinet must approve a PPP at four points (Irwin and Mokdad, 2010, pages 10 to 12) between the business case and the start of contract management (Figure 6).
Figure 6. Developing and approving a PPP in Victoria, Australia
The nine stages Victoria follows to develop and award a PPP, with Cabinet approval required at four of them: before funding is committed, twice before going to market, and before contract management begins.
In different ways: practice varies in where these liabilities are reported and in which ones are included. New Zealand recognises PPP contingent liabilities on the government's balance sheet. Other countries take a more conservative approach and disclose them in the notes to the balance sheet, as in the United States and Canada, or often in a separate statement, as in Australia and Japan.
Countries also differ in which contingent liabilities they report. New Zealand and the United States report all of them, while Hungary reports only explicit liabilities. International accounting standards require disclosure of explicit contingent liabilities but do not cover implicit ones.
With three kinds of rule: guidance on which risks to take, dedicated guarantee funds, and caps on total PPP commitments. Alongside quantitative valuation and reporting, several governments have set qualitative guidelines for managing the fiscal risk of PPP projects. Table 7 compares four of them.
Table 7. Fiscal risk rules for PPP guarantees in Chile, Colombia, Brazil and Indonesia
Brazil and Indonesia have both chosen to set up an independent guarantee fund, which is separate from government accounts, privately managed and capitalised upfront by transfers from the government. Brazil's fund was created under its 2004 Federal PPP Law (International Monetary Fund, 2006, page 18), and Indonesia's is the Indonesia Infrastructure Guarantee Fund.
Several countries have also set overall ceilings on their guarantees. In Hungary, the public finance law limits the total nominal value of multi-year PPP commitments to three per cent of government revenue. Brazil's Federal PPP Law limits the total financial commitments made under PPP contracts to a maximum of one per cent of annual net revenue.
Proposals for contingent obligations may need to be considered alongside competing instruments, and ceilings on the total issue of guarantees may need treasury approval during the budget process.
Through one accountable ministry and a clear approval process. In principle, central management should cover four things: the overall policy for approving projects; the identification, classification and recording of risk exposure; provision of funds to meet potential liabilities; and systems for monitoring the government's risk exposure.
Giving this responsibility to the Ministry of Finance, in close coordination with other key stakeholders such as the central bank, helps ensure that the most viable PPPs are selected; the International Monetary Fund likewise wants the finance ministry to take an active part in developing, reviewing and monitoring guarantees. For exchange rate and interest rate guarantees in particular, the central bank can give early insight into the potential future liability.
Controlling implicit contingent liabilities matters as much as controlling explicit ones, as the National Air Traffic Services case shows, and the International Monetary Fund treats it as a priority, if a hard one. For fiscal risks to shape decisions, the budget process needs suitable procedures for guarantees, and the right design will vary with the country and the maturity of its PPP programme. A multistage review of proposed PPPs by people with expertise in fiscal management, together with quantification of certain contingent liabilities, leads to a better value for money assessment (Irwin and Mokdad, 2010, summary, page vii).
Depending on the country, central control may mean requiring the prior approval of the minister of finance, the cabinet or the legislature, under a well-articulated policy framework that covers the justification, design, analysis and approval of guarantees.
Part of the problem is overconfidence. Irwin shows how an intuitive forecast of a risk factor such as traffic can be far narrower than the true range of outcomes (Irwin, 2007, pages 38 to 39), so a guarantee looks less likely to be called than it really is (Figure 7).
Figure 7. Why intuitive forecasts understate the risk a guarantee creates
The shaded band is where the risk factor, such as traffic or an exchange rate, will stay 98 per cent of the time. The dashed band is what an overconfident forecaster believes is the same range. It is much narrower, so real outcomes often fall outside it, as the simulated path does.
Mostly with one of two techniques, Monte Carlo simulation or the Black-Scholes option pricing formula, and a growing number of governments now use them. Several governments already value the guarantees they give to infrastructure projects, including Canada, Colombia, Chile, the Netherlands, Sweden, Turkey and the United States (Cebotari, 2008, pages 16 to 22). Their methods differ: New Zealand values the maximum possible loss, while Colombia analyses the probability of default for infrastructure projects.
Chile shows why valuation matters. Its published figures separate the most it could have to pay from what it expects to pay (Irwin and Mokdad, 2010, page 24), and the gap is large (Table 8).
Table 8. Chile's liabilities in its concessions, September 2008 (US$ million)
The most Chile could have had to pay under its revenue guarantees was US$5.8 billion. After allowing for the revenue it shares when traffic is strong, their expected net cost was US$232 million, about four per cent of that maximum.
That uncertainty covers whether the government will have to pay and, if so, when and how much. Two techniques are widely used to value guarantees, including those in PPP projects: Monte Carlo simulation and the Black-Scholes option pricing formula. Both can model guarantees such as toll revenue under a minimum revenue guarantee. The right choice depends on the structure of the guarantee and the information available about what drives guarantee payments.
In Monte Carlo simulation, the value of the underlying risky variable at any time depends on its initial value and on the mean and variance of its growth rate. Taking a large sample of outcomes for the random variable, and calculating the guarantee payment in each case, gives the probability distribution of payments and the expected payment. The value of the guarantee is the discounted present value of the expected risk-adjusted payments over its life.
The Black-Scholes formula values a guarantee as a financial option, since a guarantee gives its holder the option to claim against the government if a specified event occurs during the contract. It produces a precise valuation but can only be used for fairly simple guarantees. Monte Carlo simulation can handle more complex guarantees, but the result is only an approximation.
Governments that want a ready-made tool can use the PPP Fiscal Risk Assessment Model (PFRAM), which the International Monetary Fund and the World Bank developed to assess the fiscal costs and risks of PPP projects.
Option pricing also shows what drives the cost of an implicit guarantee. Using the same example as Figures 1 to 3, Ehrhardt and Irwin find that the cost is negligible at moderate debt and rises steeply when high debt meets high volatility (Figure 8). Table 9 compares the two valuation techniques.
Figure 8. The cost of an implicit guarantee grows with debt and risk
The height of the surface is the value of the government's unwritten promise to rescue the firm, for every mix of debt and business risk. The three coloured lines pick out low, medium and high risk. At $50 million of debt or less the promise costs almost nothing, whatever the risk; at $100 million it is worth about $0.4 million, $3.8 million and $7.5 million respectively.
Table 9. Two techniques for valuing a PPP guarantee
How the two main valuation techniques work, how precise they are and which guarantees suit each.
Not every estimate needs a stochastic model. South Africa's PPP manual shows a simpler, probability-weighted approach (Irwin and Mokdad, 2010, pages 31 to 32) that any finance ministry can apply (Table 10).
Table 10. Putting a number on construction cost risk: South Africa's worked example
Weighting five cost scenarios by their probability adds R19.25 million, about 19 per cent, to a R100 million base estimate of construction cost.
Yes: developing and maintaining a PPP contingent liability fund, preferably held by the treasury or the central bank, is a good way forward. Colombia already pays the expected cost of its guarantees into such a fund, and the European PPP Expertise Centre suggests a fund as one way to limit exposure to state guarantees.
The PPP agreement could require the private partner to contribute to the fund. On a road PPP, for example, if toll traffic exceeds the expected level by more than an agreed percentage, the private partner could pay an agreed share of the extra toll revenue into the fund, much as Chile's concessionaires share revenue above a set threshold. The government could hold that money for future contingent liabilities on the same project or on other PPPs. The central bank, a private party or a financial institution could manage the fund so that the money is not spent on other purposes.
Indonesia has taken this route. The Indonesia Infrastructure Guarantee Fund was established on 30 December 2009 as a state-owned enterprise under the Ministry of Finance.
To set up a fund of this kind, a government can assess the potential contingent risk exposure from all its PPP contracts for a given fiscal year and size the fund with suitable tools, such as Monte Carlo simulation or the Black-Scholes formula; Colombia sets its appropriations to cover 95 per cent of possible outcomes. An annual review of how the fund is used can then guide further allocation and management. Independent PPP bodies such as the World Association of PPP Units & Professionals could act as gatekeepers, balancing the risk the government takes on against the allocation from the fund to each PPP project.
I was delighted that MCC's work was completed on time, and within budget, helping us deliver important changes and improvements, to the benefit of our stakeholders. MCC's report is published on the CCC website.
I am delighted to recommend MCC Economics. Specifically, I worked closely with PJ, who helped us with our Nuclear and CCUS projects. PJ helped us develop new policies and answer questions from our stakeholders. His support helped us deliver important changes and improvements, to the benefit of our stakeholders.
MCC Economics has helped us better understand the most important issues for our stakeholders, including: charges, shareholder returns, debt payments and inflation impacts.
I am delighted to recommend PJ and his team at MCC Economics. We've been working together on National Policy Statements to help meet net zero targets for 2030 and 2050. We initially appointed MCC Economics to support us on offshore wind consultation analysis and have recently reappointed MCC Economics to undertake a larger consultation analysis role across all sectors, including hydrogen, CCUS and networks. I can confirm that PJ and his team have shown excellent spreadsheet skills, alongside very good project management, planning and analysis skills, helping us deliver important changes, and continuous improvements, to the benefit of our stakeholders.
I am delighted to recommend PJ and his team from MCC Economics. They helped us with our price controls for Heathrow airport and for NATS (En Route) plc (the air traffic services provider). Specifically, the MCC team helped us deliver important changes and improvements to our financial models and supporting policy documents, to the benefit of our stakeholders.
I am delighted to confirm that I worked with PJ on a retail project in 2015. The project helped stakeholders understand electricity costs and charges. Specifically, the project helped us explain to stakeholders, internally and externally, why electricity charges differed across the regions (GB, NI & Ireland). PJ was a key member on the project team, which helped deliver changes and improvements in the understanding of energy retail.
.png)
Evidence-based analysis for transparent, defensible, and effective decision-making.
© All rights reserved – MCC Economics 2026.
.jpg)