Bain eng

Toll Road Forecasting Risk:Study Review & Update 2003: 

Toll Road Forecasting Risk: Study Review & Update 2003 Robert Bain Associate, Infrastructure Finance Ratings Standard & Poor’s, London Global Development Learning Network on Road Financing Session 2: Scope for Private Sector Finance The World Bank, Washington DC

Toll Road Forecasting Risk: 

Toll Road Forecasting Risk Highlights from the 2002 research In the interim… 2003 Results Data disaggregation Error drivers (updated) The key role of GDP assumptions Future forecasting challenges

2002 Research: Highlights: 

2002 Research: Highlights 32 toll case studies (retrospective on forecast performance) toll roads, bridges & tunnels user-paid & shadow tolls international review of forecast performance versus outturn results Key findings Systematic errors (optimism bias) Supported by other studies Error-drivers Development of Traffic Risk Index Guidelines for stress testing

Traffic Risk Index (page 1 of 3): 

Traffic Risk Index (page 1 of 3)

In The Interim…: 

In The Interim… Traffic Risk Index being used for risk assessment/communication Continue to collate toll road traffic forecasting data Industry participants offered additional data All data employed = second-sourced Continue to compile/research reasons for forecast inaccuracy Now have 67 case studies (and growing) Analysis supports earlier conclusions Allows for disaggregated data analysis

2003 Results: 

2003 Results 32  67 Case Studies Mean remains unchanged @ ~ 70% Spread increased. Range now: 18% - 146% !

Data Disaggregation: Tolling Experience: 

Data Disaggregation: Tolling Experience No tolling experience: actual traffic = 56% of forecast Tolling experience: actual traffic = 87% of forecast (with narrower ‘spread’)

Error Drivers (2002 + 2003): 

Error Drivers (2002 + 2003) High toll tariffs and a miscalculation regarding users’ WTP: especially frequent users (ie. commuters) Recession/economic downturn Future-year land use scenarios that never transpired Time savings less than expected Improvements to competitive (toll-free) routes Less usage by trucks Less off-peak and/or weekend traffic Sheer complexity of the deal (hence modelling process) Underestimate of ramp-up period (severity and duration) Underestimate of Value of Time (use of average figures) Longer-term traffic forecasts very sensitive to GDP assumptions

Compound Impact of Alternative GDP Assumptions: 

Compound Impact of Alternative GDP Assumptions +57% +149%

Future Forecasting Challenges: 

Future Forecasting Challenges Point-of-use charging in developing/transitioning economies Ability to pay/willingness to pay Toll collection technologies Reliability, take-up, back-office etc. Sophisticated pricing Discounts (frequent user programmes, resident discount schemes), peak/off-peak pricing, day-of-week, season-of-year, etc. Value/congestion pricing (by level-of-service) – dynamic? Urban congestion charging All of the above…and more!!

Contact Details: 

Contact Details Robert Bain robert_bain@standardandpoors.com +44 20 7826 3520