[apn day1-5] wbi clive harris understanding and managing
TRANSCRIPT
Understanding and Managing the Fiscal Risks of PPPs
Clive HarrisManager, PPPsWorld Bank Institute
Objectives of this session
1. Review the risks inherent in PPPs and understand the reasons why these risks materialize all too often
2. Identify what can be done to better understand and manage these risks– During project selection and procurement– During project implementation
0
50
100
150
200
250
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Public Option $mnPPP $ mn
Governments like PPPs because they don’t have to commit money now to get a project built – so they may ignore future costs
Year
PPPs by definition impose an obligation on governments• They may be contractually obliged to
make payments under PPPs related to asset availability or service delivery– Obligations to make payments under
PPPs are like debt (though payments usually related to performance of the asset)
• They bear risks under PPP contracts, many of them outside of their control: e.g. demand risk, foreign exchange risk on the costs of inputs, termination liabilities– Risk-bearing by governments under PPPs
produce obligations like guarantees
The existence of these obligations does not mean a PPP is a bad option
• Both types of obligations often are needed for the PPP to proceed and can represent the best way of allocating risks and getting the project done
• But it is important for governments to understand and track the extent of these obligations
• And too often risks materialize and lead to substantial fiscal consequences e.g. toll roads – Mexico (1990s): Spain (1960s-70s); power projects – Dabhol (India), Pakistan, Philippines
Why do these risks arise more often than they should?• Incentives facing government….
– To ignore long-term consequences of projects
– To absorb risks that the private sector could/should bear
• Capacities not adequate….– To analyze possible risks– To manage projects during
implementation= contract management skills are crucial
The government’s incentives• Lack of long-term focus:
– Most governments don’t have long-term budget frameworks, so PPPs involving spending in future years are often not recognized at all.
– Costs of problems likely to be incurred by next Minister/government given lead time for project to be operational
• Lack of process: – Publicly-funded projects at least have a defined
process for approval involving competition against other funding priorities
• Willingness to absorb risk:– Governments very wary of service disruption– Government per se has high ability and willingness to
absorb risk
Evidence that governments absorb risk?• High rates of renegotiation: (100%
contracts in Chile saw post-award adjustment)
• Cancellation rates very low: (Harris & Pratap(2008))
Sector Percentage of Projects With Private Sector Exit1990-2006
Energy 3.6%
Telecoms 4.4%
Transport 4.7%
Water and sewerage 8.8%
Total 4.7%
Capacities for assessing risks• Government agencies often
inexperienced in:– Assessing magnitude of risks– Negotiation…….especially when compared with private
sector and its teams of advisors• PPPs are not part of standard budgetary
process so – there may be no clear process on who
approves projects and associated obligations– or even who is informed of these
Approaches to allocating risk• Mantra is “allocate risks to those best
placed to manage them”• How to apply (following Irwin, 2007):
– Who influences risk factor?• E.g. Construction risk allocated to construction
partner– Who can anticipate or respond to risk
factor?• E.g. Demand – who forecasts (project design)
– Who can best absorb the risk?• Exchange rate movements (who can hedge?)
Managing PPP-related liabilities
• Managing means deciding whether to incur liabilities and monitoring outstanding liabilities with a view to limiting spending or at least getting an early warning of it
• Three key factors:– Who approves PPPs?– What analysis is done before
approval?– Which data is publicly reported?
Chile• The government has guaranteed the revenue
of many of the toll roads and airports• PPPs must be approved by the Minister of
Finance, who gets advice on contingent liabilities from a small Contingent Liabilities and Concessions Unit
• Among other things, the unit estimates the cost and risk of revenue guarantee using a stochastic model
• The government publishes information on contingent and direct liabilities under PPPs in annual reports on public finance and contingent liabilities
13
Chile discloses the cost of its PPP obligations(Present values of commitments in US$ million. Note: GDP is about US$110 billion)
Subsidies and annuity-type payments
Commitments made in renegotiations
Maximum guarantee spending
Expected guarantee spending
Route 5 706 62 3,482 224
Other intercity roads 204 63 962 78
… … … … …
Total 2,328 780 5624 335
Chile – annual estimation of concession guarantees
South Africa• Relatively large program of PPPs, including toll roads,
the Gautrain, and several accommodation-related projects
• There are some revenue guarantees and also obligations to make termination payments tied to debt in case of project failure
• The Treasury must approve projects at each of four stages
• The PPP unit is the key adviser in the National Treasury, but other parts of the Treasury are now involved
• Treasury requires analysis of contingent liabilities as part of project preparation
• Departments include a disclosure note on PPP in their accounts
16
Britain• Very extensive program of PPPs – many
using PFI approach of unitary payments under long-term contracts
• Britain discloses the fiscal implications of PFI projects
• Formal accounting requirements have caused many PPPs to be recognized on the government’s balance sheet– has helped shift attention from reducing
government expenditure to capturing the real potential benefits of PPPs, such as innovation and efficiency improvements
– Further changes possible depending on Eurostat rule changes
17
Britain discloses committed payments to PPPshttp://www.hm-treasury.gov.uk/ppp_pfi_stats.htm
0
1
2
3
4
5
6
7
1995 2000 2005 2010 2015 2020 2025 2030 2035Fiscal year ending
Bill
ion
poun
ds
Implications for other countries
• The biggest contributors to success in these countries is probably the quality of their public sectors
• But all highlight importance of paying attention the following:– Approval– Analysis– Reporting
• Precise solutions depend on the institutional context
Approval• Multistage review of proposed PPPs by ministry of
finance looks like a good idea for any country• But how much of the responsibility for advising on
PPPs should lie with ministry of finance? • In Victoria and South Africa, a PPP group in the
Treasury votes early and often; in Chile, review is later and focuses on contingent liabilities
• Earlier, more intensive involvement by the ministry of finance has advantages—which suggests PPP group in the ministry of finance
• These reviews can help to prevent “gold-plating” or over-specification of projects and ensure good project selection and design
Analysis• Quantifying contingent liabilities is useful when
liabilities are large, not under control of government, and estimates may affect decisions (revenue guarantees in Chile and South Africa)
• But just as important is thinking about whether the allocation of risk creates good incentives
• Often the problem can be simplified by considering the ordinary liabilities of the public sector, instead of the contingent liabilities of the government—as in Australia
Reporting• Financial reporting is surprisingly influential e.g.
adoption of IFRS is helpful in Australia • But adopting modern accrual accounting isn’t a
sensible priority for all developing countries• To try to reduce temptation to disguise debt,
countries with less advanced accounting can disclose information on PPPs and contingent liabilities:
• Simple reporting may be the best approach– list with description of project and liabilities and
quantification of total investment, committed payments, and maximum contingent payments
– Publication of PPP contracts, as in Australia, is easy and helpful
– SOEs or agencies may enter into PPP contracts so important to ensure they reflect these obligations in their accounts
Examples of reporting
• For long-term payment obligations
• For guarantee-like obligationsProject name
Project cost ()
Contingent liability that creates the largest exposure
Estimated payment required if contract had been terminated for that reason on March 31, 2009 ()
… … [e.g., termination payment for contractor default equal to 90% of outstanding debt due to senior lenders]
…
… … … …
Project name
Project cost
Actual unitary and construction-subsidy payments
Forecast unitary and construction-subsidy payments
NPV of forecast payments, 1
January 2009 (at x%)
2006 2007 2008 2009 2010 2011 2012 … ….
… … … … … … … … … … … …
… … … … … … … … … … … …
Managing unanticipated costs during implementationEven in countries with strong institutional
capacity, PPPs can give rise to “unanticipated” costs during implementation
• UK: NAO (2008): “changes to operational PFI deals are often poor value for money” GBP 180mn spent on these changes in 2006 (many changes are actually aspects taken out of contract specification pre-bid to save money!)
• Chile: additional expenditures approx. 39% of original investment, most determined via negotiations and is additional investment requested by the government
Solution: • tighten scrutiny of contract adjustment and
involve Ministry of Finance• Proactive contract management with adequate
resources
Broader sector policies matter• Quality of sector governance and performance
critical in recovering costs are recovered from users not taxpayers
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Telecom Gas Pow er Water
Ratio of Re ve nue to Cos ts
Thank You and Questions