General measures are mainly related to the design of the PPP contract and to the inclusion of such provisions as this may reduce the probability of a conflict situation, on the one hand, and the setting of rules which may provide for the efficient management of distress situations once they have actually taken place.
In this section, we will consider the information available in the PPI Database for the purpose of analyzing to what extent different mechanisms affect the incidence of distress on PPI contracts. We base our analysis on four variables: government support, support by multinational agency, listing at the stock exchange, and initial payment to the government. In principle, it is reasonable to expect that these instruments might be useful to reduce the risks faced by the project, and therefore, they should contribute to reduce the incidence of the cases of distress.
The government's direct support to a particular project may be provided through different channels including direct subsidies, exchange rate guarantee, payment guarantee, debt guarantee, etc. BOX 36 contains a description of all the types of direct financial support provided for in the PPI Database.
BOX 36: Government Support
| Government support: Hosting governments provide financial support to or reduce the financial risk of a project in many ways. The forms of government support tracked by the PPI database tracks are the following: Cash subsidy - This is when a government agrees to provide cash subsidy to a project. It can be a total lump sum or a fixed amount per new connection, and payments can be either in installments or all at once. Cash subsidies are included in the "investments in physical assets" total for projects in which the private party takes some investment risk/decisions: concessions, divestitures and greenfields. Payment Guarantee - This is when a government agrees to fulfill the obligations of a purchaser (typically a state-owned-enterprise) with respect to the private entity in the case of non-performance by the purchaser. The most common example of this is when a government guarantees the fixed payment of an off-take agreement (e.g. Power Purchase Agreement (PPA), Water Purchase Agreement (WPA)) between a private entity and the state-owned enterprise. Debt Guarantee - This is when a government secures the borrowings of a private entity. That is, a government guarantees repayment to creditors in the case of a default by a private entity. Revenue Guarantee - This is when a government sets a minimum variable income for the private operator; typically this income is from user fee payments by end-use customers. This form of guarantee is most common in roads with minimum traffic or revenue set by a government. Exchange Rate Guarantee - This is when a government protects a private entity from fluctuations in the value of the local currency. For example, the government will agree to reimburse the private entity for losses on debt services if the value of the local currency dips by, say, 20 percent or greater. Construction Cost Guarantee - This is when a government protects a private entity from potential cost overruns in the construction phase of a project. |
Source: PPI Database Glossary
In principle, all the government's direct support mechanisms to any PPP Project entail a reduction in the risks undertaken by the private sector. Therefore, this type of support would probable reduce the incidence of distress situations.
The second variable considered is the project support by multilateral institutions. This support may range from a direct investment (equity) to guarantees against political risks. BOX 37 provides a breakdown of all the types of financial support provided by multilateral institutions provided for in the PPI Database.
BOX 37: Development Bank Support
| Development Bank Support: This is financial support that a multilateral institution has given to a project. The types of financial support tracked are: Equity - multilateral institutions are allowed to invest in equity except for IADB, IBRD and IDA Guarantees - two types of guarantees are covered: • political risk coverage against currency inconvertibility, expropriation, war/civil disturbance and breach of contract • partial credit guarantees, which turn medium-term finance into a longer-term arrangement by guaranteeing longer maturity or offering liquidity guarantees in the form of put options and take-out financing Loan - direct loan using the multilateral institution funds (also referred to as A-loan) Quasi-equity - these products have both debt and equity characteristics and some of them are convertible debt, subordinated loan investments, and preferred stock and income note investments (also referred to as C-loan) Risk management - the risk management products, or derivatives, allow project companies to hedge currency, interest rate, or commodity price exposure. Some of them are currency and interest rate swap, options and forward contracts and derivatives. Syndication - a multilateral institution arranges the financing with the resources of other investors, but the institution is always the lender-of-record (also referred to as B-loan) |
In principle, the support provided to the project by multilateral institutions has two consequences. In the first place, it contributes to decrease the risks undertaken by the private party upon having explicit guarantees to face some of the risks of the project. In the second place, whenever the support consists in loans or direct investment, such support would probably act in part as a restriction and mechanism to commit the public sector through the need to maintain its reputation before international institutions.
The third variable included in the analysis is the existence of initial payments as part of the process of award of the PPP contract. This would state a specific and concrete commitment of the private sector toward the project, and therefore, it should reduce the risk that the project may be cancelled or submitted to arbitration. In fact, Guasch (2004)150 finds that higher-payment bids entail a renegotiation probability which is materially lower than those in which the award is made on the basis of the lower tariff charged to end users.
The PPI database defines the initial payments (Payments to the government formerly known as Investment in Government Assets) as the "Resources the project company spends on acquiring government assets such as state-owned enterprises, rights to provide services in a specific area or the use of specific radio spectrums. License fees, canon payments or divestiture revenues are the common revenue collection mechanisms."
The fourth variable under analysis consists in assessing whether the company holder of the PPI contract is listed at the stock exchange. As it was analyzed in section V.b, in principle this should act on the probability that the project may be under distress on the one hand, via the existence of higher information requirements and the existence of clearer supervision mechanisms. On the other hand, the distribution of a less concentrated ownership should act in favor of the project's stability since they induce less opportunism from the government.
Table 28 shows an estimation of the incidence on the distress probability (cancellation and arbitration) of these four variables, as assessed in terms of number of projects for each of the four sectors under analysis. Even though in respect of the exception of the variable that the project may be listed at the stock exchange, the remaining variables are quantitative variables (that is to say, they have the support amount), in this analysis we have focused on treating them as a dichotomy (whether or not such particular support exists).
The government support has positive effects on the energy, telecommunications, and transportation sectors, while in the water and sewerage sector, the existence of government support increases the probability of distress (from 8.5% to 21.2%). The support provided by multilateral institutions has strong positive effects on transportation and it has moderate positive effects on the telecommunications sector. Its effect on water and sewerage and on the energy sector is clearly negative.
Table 28: Incidence of the projects under distress per type of support (in number of projects)
|
| Energy | Telecommunications | Transportation | Water and sewerage | ||||
| YES | NO | YES | NO | YES | NO | YES | NO | |
| Government Support | 5.0% | 6.0% | 0.0% | 6.6% | 1.3% | 6.1% | 21.2% | 8.5% |
| IMA Support | 9.1% | 5.3% | 6.0% | 6.7% | 1.8% | 6.2% | 19.0% | 9.0% |
| Payments to the government | 8.8% | 4.9% | 4.3% | 8.3% | 4.4% | 6.1% | 10.7% | 9.9% |
| Quoted in the stock market | 3.7% | 6.1% | 1.4% | 7.3% | 2.3% | 5.9% | 18.2% | 9.9% |
The same pattern of behavior is found when the variable considered is the initial payment. It has positive effects in telecommunications and transportation, and negative effects on the energy and water and sewerage sectors, though in the latter, the variation is scarcely significant. The fact that it may be quoted in the stock market, presents negative effects only in the water and sewerage sector, where the risk of distress is almost twice as much for the companies listed at the stock exchange as compared to non-listed companies. With respect to the three remaining sectors, the effects are positive and highly significant.
Table 29 shows the same analysis per sector (energy, telecommunications, transportation, and water and sewerage), as measured in terms of committed investment.
Table 29: Incidence of the projects under distress per type of support (in committed investment)
|
| Energy | Telecommunications | Transportation | Water and Sewerage | ||||
|
| YES | NO | YES | NO | YES | NO | YES | NO |
| Government Support | NA | 9.1% | 0.0% | 4.5% | 1.6% | 10.2% | 30.7% | 31.2% |
| IMA Support | 11.9% | 7.8% | 7.5% | 2.8% | 3.3% | 10.7% | 51.8% | 18.8% |
| Payments to the government | 12.4% | 6.9% | 3.3% | 9.0% | 3.5% | 12.7% | 31.4% | 30.9% |
| Quoted in the stock market | 14.6% | 8.4% | 0.5% | 7.6% | 1.3% | 10.2% | 2.6% | 39.2% |
The results, at this point, are also varied. The support provided by the government is positive for the four sectors. Government support is marginally significant only in the case of the water sector. In the water sector, the probability of projects distress with no government support amounts to 31.2% of the risk, and it drops to 30.7% whenever the project has some kind of government support, which does not constitute a significant difference.
The support provided by international organizations becomes positive in the transportation sector (where the support provided by multilateral institutions reduces the risk of distress from 10.7% to 3.3%), but it remains negative for the other sectors. Again, the effect is even clearer in the water and health sector, where the support increases the risk of distress from 18.8% to 58.8%.
The initial payment has a positive effect with the exception of the energy sector, where the impact is negative (with the risk of distress increased from 6.9% to 12.4%). The effect is also negative, though the difference is not significant, in the water and sewerage sector. On the other edge, the most visible positive effect occurs in the transportation sector with a drop in the risk of distress from 12.7% to 3.5%, seconded by the telecommunications sector (from 9.0% to 3.3%).
Quotation at the stock exchange has positive and highly visible effects in all sectors, except for the energy sector in which the effect is negative. This would reflect the positive effect that the market supervision generates on projects thus contributing to reduce the incidence of distress.
It is worth highlighting that the information contained in the chart discloses the wide variety of effects among sectors. Only in the transportation sector itself, the effect is the one expected for the four variables, if we considered investment amounts and number of projects. With respect to the remaining sectors, the effects are even more varied, thus exerting a positive influence, in some cases, and showing contrary behaviors in other cases.
From the analysis of the PPI Database, as presented in Table 28 and Table 29, it may also clearly be observed that there is a strong difference in the water sector's behavior. This is the sector which the highest distress proportion, and at the same time, it is the sector where the risk mitigation mechanisms analyzed seem not to significantly contribute to reduce its incidence151.
The results obtained in relation to the effect of the government support and of the support provided by multilateral institutions on the risk of distress are grounded on some antecedents in literature. Guasch (2004) finds that the support provided by the government to the projects implemented in Latin America increases the risk of renegotiation152.
In turn, the analysis of the exit of international operators in the water and sewerage sector (IDB 2007) is highly critical of the role placed by international institutions in the sector when it allows extremely high indebtedness levels: "These companies' access to finance by multilateral institutions, such as the IDB, the World Bank and the International Financial Corporation, without the pertaining increase in equity contribution, could have led to this situation. Among other consequences, this situation influenced on a lower corporate capital risk involved and it encouraged opportunist conducts, as well as a higher exposition to devaluation, since part of the loans were guaranteed by the governments. Once the outcomes have been observed, it is clear that the contracts should have included restrictions in this sense. It would seem that all the parties underestimated the political and financial risks of the investments in the sector."
Irrespective of the antecedents to the results presented, three lines of analysis may be followed to explain why there is no positive effect in all the cases pertaining to the variables under analysis. The first argument consists in the existence of sample selection bias. The support provided by the government and by international institutions is not randomly distributed among projects; instead, it is focused on certain types of contracts. On the basis of this assumption, these contracts would have little economic feasibility at the beginning, or no economic feasibility at all, and this explains the higher incidence of distress in this subset.
The second argument is related to the existence of moral hazard problems by the institutions. Once the project has government support or once it is supported by multilateral institutions, managers have fewer incentives to behave efficiently, which leads to higher risks and results in a higher incidence of distress projects153.
The third explanation is related to the mere fact of considering certain dichotomic variables in our analysis instead of the relative significance of support in relation to the amount of the investments.
____________________________________________________________________________________
150 Guasch (2004 - Table 6.12) reports a probability of renegotiation which amounts to the 60% when a lower tariff is awarded, and which amounts just to the 11% when it is awarded following the highest payment criterion.
151 The more "political" nature of this sector would explain the results obtained at least in part.
152 Guasch 2004 page 89.
153 Laffont, Guasch and Straub (2003) develop a contract renegotiation model and they allege that contract provisions are internal; therefore, the self-selection and moral risk problems arise out in the estimations