Operating toll road assets are an attractive low volatility sector for infrastructure investors. They benefit from high operating margins and long concession lives which leads to high quality forecastable cash flows. History has demonstrated that revenue volatility of existing toll roads is very low. Qualities that lead to their low risk profile are:
Monopoly type characteristics with long term concessions from Government,
Low demand elasticity and resilience to economic fluctuations,
Predictable mechanisms for tariff increases to sustain revenues over the long term, and
Typically well-established competitive dynamics with other methods of transportation.
Over the last decade the market has lost appetite for patronage risk on new build toll roads. The Governments response was to move to availability payment structures such as on Peninsula Link in Victoria and Transmission Gully in New Zealand. For Westconnex the New South Wales Government has built a portion of Stage 1 (M4 West) before selling the rest of the concession. The procurement of Westconnex seems to be a partial movement back to the private sector adopting a level of patronage risk.
Transurban have majority ownership of all other existing Sydney toll roads - Lane Cove, Cross City tunnel, M2, M7 (50%), M5 (50%) and the M1 Eastern Distributor (75%). Given their ownership of the existing toll road network they would have the best visibility on potential patronage over all other bidders. They are also the only bidder with potential synergies with an existing network (both in terms of operating costs as well as from a cost of capital perspective). The traffic history of the M4 West and M5 West and East available to Transurban, which significantly reduces patronage forecast risk.
In terms of the deal, the Transurban consortium will acquired 51% of WestConnex for $9.3 billion with the remainder staying in ownership of the NSW Government. Total enterprise value is $18.2 billion. The Project is predominately a greenfield project and widening of the existing M4 and M5 motorway as well as linking the two motorways together around the Sydney CBD, Sydney Airport and Port Botany. The Project has a concession to the end of 2060 and is for 388 km of lane highway (or 33 km of motorway). The first full year of operations is expected to be in 2028. Toll escalation under the concession is a greater of CPI or 4% until 2040 and the greater of CPI or 0% after.
The major components of the concession are broken into 3 stages:
1. Stage 1: M4 concession
a. M4 west (widening) – completed July 2017
b. M4 east (new) - expected completion late FY 2019
2. Stage 2: M5 concession
a. M5 east (new) – expected mid FY 2021
b. M5 west (existing) - concession to be included in Westconnex from 2026. Currently owned by Interlink Roads (which includes Transurban and other Superfunds)
3. Stage 3: M4-M5 link concession and Rozelle Interchange
a. M4-M5 link (new) - expected late FY 2023
b. Rozelle interchange (new) – expected FY 2025
Transurban has given guidance that they expect the Westconnex concession to almost double the amount of revenue from their existing Sydney toll roads. The level of forecast EBITDA has not been disclosed but guidance has been provided that it is somewhere in line with their past acquisitions. The Cross City Tunnel was acquired in March 2014 at 18x EBITDA and Airportlink M7 was acquired in November 2015 at 28x. Presumably the acquisition multiple is somewhere around 28x – although exactly how to calculate an EBITDA multiple on a project that doesn’t reach full operations for almost a decade is slightly tricky.
Transurban claim that 6% of forecast revenue is entirely greenfield with the remainder an overlap of existing brownfield sites or well developed travel patterns. The greenfield site is contained within stage 3 where there seems to be significant demand given the connections to Port Botany and Sydney Airport. In theory Transurban should have all the data available to it given it runs the existing brownfield network.
It will be fascinating to watch how this plays out over the years. Infradebt’s view is that 28x is a very high multiple to pay particularly into the headwind of rising base rates (a headwind for all infrastructure in general), but against this there are several compelling features in relation to the transaction structure, and certainly whilst high this comparable to other core infrastructure transactions completed of late.