segunda-feira, abril 26, 2010

Traffic forecast error key in financing PPP

Why are project traffic forecasts proven frequently so wrong, with actual traffic failing to match the base case scenarios? 
This is the question posed to the members of the new  PPP Network, .

Forecasting is a very difficult thing to do, especially about the future …

That said, by Mark Twain, who gambled on riverboats, not on riverboat traffic, one must also look at who is doing the forecast and who is taking the risk of the traffic forecasting. With very highly leveraged public service projects, sponsors have relatively little to loose on the downside, and the Government can usually be counted on to step in with a bailout or a retroactive guarantee in case of default, in order to maintain the public service delivery.

Once-burned investors become twice-shy, refusing traffic risk especially in greenfield projects in new traffic corridors which depend heavily on induced traffic or inter-modal transfer.  Forecasting traffic in well-established brownfield corridors is seen as having a  lower forecast error, and therefore lower risk for the investor.

Some of the work presented at an EIB Traffic Risk seminar in June 2008, covering dozens of projects, found a quasi-normal distribution of the ratio of actual traffic to the base case scenarios (traffic forecast error):

About 25% of the projects reported actual traffic well below the base scenarios (-25-40%) , 25% reported actual traffic well above the base scenarios, (+25 +40% ) and and the other 50% of the projects were either somewhat below or somewhat above the base case.

In replicating this study with Portuguese SCUT shadow toll roads, it appears that they followed a similar bell-shaped distribution. SCUT roads, “sem cobrança ao utilizador”, have no user tolls, but the public sector pays as a function of actual traffic counts, subject to bands of minimum and maximum traffic risk to the sponsors.

Going further, we can look for elements that would help to predict the forecast error, that is, that might help identify the optimism bias a priori, because hindsight may be 20/20 but is not very helpful. Some of the relevant elements would be the location of the traffic corridors (greenfield or not), the year of financial close, the project sponsors, the bank financing conditions, pricing, tenor, coverage ratios, etc.

The one thing that seems to correlate most closely with the traffic forecast error is the ADSCR, the Annual Debt Service Coverage Ratio (Cash available for debt service to Debt service, year by year). The weaker projects, which reported the highest traffic shortfalls 5-7 years after financial close, had evidently been spotted earlier by the banks that required ADSCR above 1.40X . On the more robust projects, the ADSCR covenants were set closer to 1.20X or even below.

If this can be shown to be true generally, it seems to indicates that bank due diligence adds value to the PPP infrastructure financing process, by ferreting out the more optimistic forecasts. After all, banks are the project parties with the most at risk in PPP projects, normally around 80% of total project costs. And experienced bank credit committees, underwriting as "principals" for their own balance sheet, will identify and price the risk better than even the most professional advisor or third party “agent”.

The sensitivity of the ADSCR to optimism bias also suggests the need to reconsider and be cautious about the current fashion of availability payments, which keep the traffic risk with the public sector. Availability payments are  more and more frequent now , both  in new greenfield infrastructure projects, but also as a result of re-allocation of project risks upon  project renegotiation when the expected traffic ramp-up failed to materialize.  This trend may be seen as "de-risking",  due  the growing risk aversion as investors and creditors  become much more selective  in the current financial crisis.

But the re-allocation of traffic risk from the concessionaire back to the public partner or "Concedente" in private one-on-one renegotiations, with little or no disclosure of actual to forecasted traffic trends and of change in contractual terms and conditions, is a practice that faills short of  the minimum requirements for transparency in public procurement. 

Thus, it may turn out that the less the traffic risk taken by the sponsors and the banks, the more the traffic risk in the project overall, and thus the more the traffic risk to the taxpayers.

The next question is whether traffic forecast errors are accidental or intentional.  Studies indicate that the higher the probability of renegotiation, the higher the incentives for promoters to engage in "strategic behaviour", in the conviction that they will be spared most of the downside risks.
Mariana Abrantes de Sousa,  PPP Lusofonia, Portugal, 26-Abril-2010

See also: 
Traffic risk and PPP financing in the financial crisis, Risco de tráfego e financiamento de projectos PPP em tempo de crise,

PPP bankability and budget sustainability in the financial crisis, Como compatibilizar bancabilidade para do projecto com sustentabilidade orçamental em tempo de crise financeira,

Risk reallocation and PPP renegotiations, Partilha de riscos e renegociação,

Renegotiation of port terminal concession with risk reallocation, Renegociação de concessão portuária de terminal de contentores

Public sector guarantees for PPP project financing, Garantias públicas para financiamento privado de  projectos PPP

PPP, risk-less capital, and public debt, PPP, capital sem risco e divida pública

Budget sustainability, and PPP project arbitration,  Sustentabilidade orçamental de contratos PPP e a arbitragem

Irwin on how to manage public guarantees for PPP projects, Irwin sobre como gerir garantias públicas para projectos PPP,

Ho on game theory and renegotiations of PPP contracts, S. Ping Ho sobre a teoria de jogos e a renegociação de contratos PPP

Vassallo on the causes of accidental or intentional, traffic overestimation in PPP projects, Vassallo sobe as causas de sobre-estimação de tráfego , acidental ou intencional, das previsões de tráfego em contratos PPP ,

Grandes Obras Públicas

VER relacionado mini-curso em PPP