1.8.2 Identifying and Quantifying Fiscal Commitments to a PPP Project

A government’s fiscal commitments – both direct and contingent – will be established by the PPP contracts. The value of direct liabilities will be relatively simple to quantify. In many cases its value will be explicitly expressed in the contract. Valuing contingent liabilities is more complicated and requires a good understanding of both the size of the potential liability and the likelihood of its occurring.

Direct liabilities

During the appraisal stage, the value of the direct fiscal commitments required can be estimated from the project financial model (described further in chapter 4). The value of these direct payment commitments is driven by the project costs and any non-government revenues. The value of the direct fiscal contribution required is usually the difference between the cost of the project (including a commercial return on capital invested) and the revenue the project can expect to earn from non-government sources such as user fees.

The fiscal cost can be measured in different ways.

• Estimated payments in each year: The amount that the government expects to have to pay in each year of the contract, given the most likely project outcomes. This is the most useful measure when considering the budget impact of the project; and
• Net present value of payments: If the government is committed to a stream of payments over the lifetime of the contract such as availability payments it is often helpful to calculate the net present value of that payment stream. This measure captures the government’s total financial commitment to the project, and it is often used if incorporating the PPP in financial reporting and analysis (such as debt sustainability analysis). Calculating the net present value requires choosing an appropriate discount rate – the choice of discount rate to apply when assessing PPP projects has been a subject of much debate;[124]

It is generally helpful to estimate how the payments might vary. For example, payments may be linked to demand, inflation, or they may be denominated in a foreign currency (and therefore be subject to exchange rate changes). The effect on payment obligations of changes in these variables should be assessed.

Contingent liabilities

Assessing the cost of contingent liabilities is more difficult than for direct liabilities, since the need for, timing and value of such payments are uncertain. Broadly speaking, there are two possible approaches.[125]

• Scenario analysis: Scenario analysis involves making assumptions about the outcome of any events or variables that affect the value of the contingent liability, and calculating the cost given those assumptions. For example, this could include working out the cost to the government in a “worst case” scenario, such as default by the private party at various points in the contract. It could also include calculating the cost of a guarantee on a particular variable, for instance demand – for different levels of demand outturns; and
• Probabilistic analysis: An alternative approach is to use a formula to define how the variables that affect the value of the contingent liability will behave. A combination of mathematics and computer modelling is then used to calculate the resultant costs. This enables analysts to estimate the distribution of possible costs, and then calculate measures such as the median (most likely) cost, the mean (average) cost, and various percentiles (for example, the range of values within which the cost is 90 percent of the time). To be useful, probabilistic models need reliable data from which to estimate the probability distributions of the underlying risk variables.

Box 2.18 provides examples of approaches to assessing contingent liabilities across jurisdictions.