INFRASTRUCTURE FINANCE:

THE SYDNEY CROSS CITY TUNNEL

OPIM-5668 Project Risk and Cost Management

Feb 18th, 2011

Sourabhdeep Singh Khanna

Hiren Gonsai
John Celis
Winston Spencer
 Executive Summary
In 2000, Cheung Kong Infrastructure Holdings Limited (CKI) was faced with a great opportunity
to invest in a transportation project in Australia. The group was particularly interested in investing in
Australia due to the country’s stable regulatory environment and economic growth. This project
presented a good chance for the firm to continue with its globalization strategy and achieve its global
ambitions in transportation infrastructure project investment.

In September 2000, the Australian’s Roads and Traffic Authority (RTA) invited tenders for the
construction, financing and 30-year operation of the Sydney Cross City Tunnel (CCT). The primary
objectives of the CCT project were to 1) relieve traffic congestion in central Sydney, 2) improve the
reliability of public transport and, 3) provide a safer environment and improved amenities to vehicles,
cyclists and pedestrians. The CCT project comprised two stages: The first one encompassed two
east-west tunnels that would run between the eastern side of the Darling Harbour and Kings Cross.
Stage Two would take advantage of the opportunities afforded by reduced traffic congestion such as
improvements to surface roads, including new bus and bicycle lanes and other improvements to
pedestrian facilities.

CKI together with its major business partner, Bilfmger Berger Aktiengesellschaft (AG), decided
to form the CrossCity Motorway Consortium (the CCM consortium) and submitted its final proposal to
the RTA in October 2001. Although Transportation investments generate significant risks, CKl was
confident that the project risks could be mitigated through careful planning and negotiation with the
Australian government authorities.

In the hope to reap maximum returns, CM consortium’s proposal included the following changes
to the original project plan: 1) Increase the length of the tunnel by 300 meters and, 2) the depth at the
eastern tunnel by 30 meters. The changes would add US$135.7 million to the project cost, but would
increase the tunnel's daily capacity by an extra 17,000 vehicles which in turn would allow the
consortium to earn additional revenue of AU$10.98 million per year. The total projected construction
cost would therefore become AU$680 million.

The CCM consortium project team determined that the return on investment would come from the
revenues generated from tolls charged during the term of the concession. In determining the pricing
strategy, the CCM consortium took into account the following considerations:

1. Differential pricing scheme: Toll charges would depend on the size of the vehicle and the route
that that it took. Passenger vehicles such as motorbikes, sedans, station wagons, taxis and
vehicles towing trailers would pay less money than Heavy vehicles
2. Toll escalation scheme: Tolls could be increased as follows: 4% per annum until the Q2 of
2012 and then 3% per annum until Q2 of 2018. From mid-20I8, the maximum increases would
be in line with inflation.
3. Both tunnels would be electronically tolled. Users needed to buy an electronic pass or hold a
toll account and have a valid electronic tag Vehicles without electronic tolling transponders
might pay higher toll charges
4. Buses providing public transportation were not required to pay tolls.

The CCM consortium with the advise of Hyder consulting, estimated that the CCT would be used
by over 90,000 vehicles per day by 2006 and over 100,000 vehicles per day by 2016. On 27 February
2002, CCM won the contract and started the construction of the Sydney Cross City Tunnel.
 Statement of the Problem

The CCT opened for traffic on 28 August 2005, but the results were not good at all. A report
presented by the CCM in February 2006 revealed that approximately 30,000 vehicles used the tunnel
each day - about a third of the projected 90,000. Many users did not use the tunnel because they
considered the toll charge too high for a short 2.1 kilometer journey when a slightly slower but a free
alternative was available.

A number of "traffic calming" measures were introduced, but this did not solve the problem. In fact
those changes created numerous disruptions, increased traffic congestion, and provoked a large
number of complaints from road users. The CCT has started to attract significant political attention,
negative media and a strong resistance by the community. So what went wrong? It is evident that CKI
and its partners need to act quickly and implement a plan that could allow them to get this project
back on track and achieve the project goals.

Analysis of the Problem

In order to determine the best course of action, we need to first understand how we have come
to this point and why the project is not providing the expected results. The following questions will help
us in our discovery process:

1. What are the merits of having private-sector participation in financing of the Sydney CCT?

A privately financed project was a specified from of PPP that involved not only private sector
financing but also controlling ownership. PFPs differed. From the outsourcing or construction by the
government. There has been widespread adoption by Governments across the world of Public Private
Partnerships (PPPs) as a way of providing public infrastructure. Grimsey and Lewis report that the UK
version of PPPs, Private Financing Initiatives (PFIs) fund between 10 to 15 percent of public
infrastructure. In NSW, PPPs have averaged around 11% of the overall NSW capital works budget
since 1993-1994, and this percentage is expected to remain between 10% and 15% in future (NSW
Joint Select Committee on the Cross City Tunnel, 2006b, p30). PPPs are seen to have the dual
benefits of reducing the calls on the government purse as well as bringing the skills of the private
sector to bear on the delivery of important infrastructure. Given the popularity of PPPs it is not
surprising that an extensive literature has developed.

2. Why is the Sydney CCT project attractive for CKI?

CKI had a vision to become an international infrastructure enterprise. CKI was one of the leading
companies in the infrastructure sector in mainland China and Honk Kong with a turnover of HK$3.3
billion and net profits of HK$3.22 billion in 2000. CKI had a strong financial position and a good
reputation of quality infrastructure projects. In 2000, CKI was actively seeking investment
opportunities in major infrastructures.

The CCT project heralded CKI’s first transportation project outside mainland China and Hong
Kong. Under the plan, the tunnel components of the project would be operated, maintained and
repaired by the CCM consortium until they were returned to public ownership after 18 december 2035
or for 30 years and two months from the completion of the tunnels, if their completion was delayed,
analyst expected the equity IRR on CKI’s investment in this green field project would be around 15%.
With sufficient cash on hand and reasonably low gearing, CKI itself had the necessary resources to
finance this project. It was projected that in 2006, the CCT’s annual contribution to CKI’s bottom line
would be HK$28 million.

3. What are the risks associated with financing, developing and operating the Sydney CCT?

In assessing the risks associated with the project, several were identified:

1. There was a substantial risk associated with the financing, design, construction, operation,
maintenance and repair exceeding the projected budget. The real issue that the CCT has
generated is who should bear the risks for these projects. It is often forgotten that risk is included
in all pricing decisions. NSW Government does not want to be caught funding white elephants and
the private sector is also hesitant in investing hundreds of millions in a project with no great
confidence of success. The bidding model contained significant risks. In line with RTA practice at
this time, the RTA required an ‘upfront payment’ from the successful bidder. Typically this
payment is mainly a reimbursement for the Government’s project development costs. As part of
the bid requirements, and for the first time, the RTA also required an additional ‘Business
Consideration Fee’. This fee was payable by the successful bidder to the RTA for the right to
operate the business. It represents the amount each proponent was prepared to pay the RTA for
the perceived value of the project. The financial evaluation focused largely on the size of the
proposed upfront payment and that the value for money for motorists - achieving the lowest toll -
was of less concern. Additionally, a cost-benefit analysis (CBA) prepared for the modified design
in 2003 delivered a positive cost benefit ratio of 3.0 at a discount rate of 7% including design,
construction, operating, and maintenance costs (Exhibit 8). No guarantee the CBA will be met.

2. The risk that the tunnel might fail to deliver the anticipated traffic volumes or projected revenues.
The patronage level of around 30,000 cars each day was less than half the projected level of
90,000 cars each day. Lower than projected traffic volumes continued and as a result revenue
was lower than required for the owner/operators to meet debt obligations
3. The risk of over-estimation of the valued motorist would place on the tunnel’s benefits. The risk of
over-estimating patronage appears to lie only with the bidder as they would appear to bear the risk
from lower toll receipts. If fewer vehicles than expected were using the CCT the network changes
managed by the RTA on the surface roads may be inappropriate and lead to a risk of increased
congestion. If fewer vehicles than expected use the CCT, the financial viability of the project will
be at risk. In a ‘worst case’ scenario the Government may be required to intervene to keep the
tunnel operating and avoid significant traffic disruption.
4. The risks that the capacity of the tunnel was not sufficient to allow for the projected traffic
estimates. The capacity of the tunnel turned out to be lower than projected and therefore this risk
was averted.
5. The risk of over-forecasting asset use. The demand forecasts of both the Government and the
private party overstated actual demand. Even when the toll price was zero traffic demand did not
come up to forecasts, indicating that the forecasting process undertaken in the original feasibility
study could have been more robust.
6. The risk of an increase of income tax. Even though the case made no reference to an income tax
being applied by the government, the possibility of an income tax could be applied in the future to
recover lost revenues is very real.
7. The risk that work or operational and maintenance activities might be disrupted by local and
government laws or by actions of the court or tribunal. No reference made in the case but the
possibility that the government or courts could create laws that could disrupt operational and
maintenance activities exist.

4. How could the CCM allocate and mitigate risks?

For CMM to really understand the degree of risk allocation, they must understand the structure
inherent in the project where the financial risks exist. The goal of an optimal allocation of risk is to
minimize the total cost of risk on a project, not necessarily the costs to each party separately. Not
surprisingly, the financial risk is most significant at the top of the hierarchy, but risk allocation is used
to shift those financial risks down the chain of command as contractual leverage decreases. CCM
could allocate risk through insurance protection afforded either directly to their holding entity of their
own insurance carrier outside the contract documents or indirectly through insurance requirements
within the contract documents. Therefore, insurance coverage is provided either by the insurance
carrier contracted directly by CCM or by insurance carriers contracted by construction entities down
the chain of command if certain project objects are not met.

In addition to allocation, risks can also be mitigated or eliminated. CCM could accomplished
this through provisions in the contract documents, or through due diligence performed both before and
during the project. Since the financial risk typically results from the work performed by CCM, control
over the company and its work force can mitigate or eliminate the risk. The contract provisions can
also be used with the implementation of specific contract clauses, to ensure that project objectives
risks are shared. CCM should have shared the project risk to prevent the demand forecasts risk from
being transferred to them. This was evident in the late amendment in project scope which when
funded by the private sector through higher tolls, adversely affected demand and overall viability.
Vehicle forecasts needed to be recalculated for any material changes to assumed toll levels, to
alleviate CCM from assuming all the risk. However, the transfer of financial risk to CMM didn’t free the
NSW Government of the political risk associated with the project not meeting expectations. The
reputation of NSW Government and the RTA appeared to be harmed as a result of this project.

5. Why did the Sydney CCT fail to deliver the projected returns for its investors as of 2006?

Although the project was designed and constructed in a very efficient manner, coming out 2
months ahead of schedule and with very little traffic disruption, the stockholders did not receive the
anticipated results by 2006. In fact the situation seemed so grim that it seemed that the company
might end up losing a major share of their initial investment into this project.

Motorists did not use the tunnel in numbers that CCM had originally anticipated. In fact statistics
released in Feb 2006 showed that only about 30,000 vehicles were using the tunnel, which was a
fraction of the 90,000 vehicles/day originally predicted. CCM had failed to adhere to the feedback
provided by the Standard and Poor’s report regarding overestimation of traffic forecasts and as a
result they failed to manage their risks of optimism bias.

Initially when the Benefit-Cost Ratio analysis had been conducted based on the projections, it
showed that even under the worst case of 10% discount rate and including all operating, maintaining,
designing and construction cost the ratio would be 2.4 which was a very favorable number. This could
have only increased had the discount rates come down further. Even the NPV analysis showed a very
favorable return for this project at 1.4 times its investment.

This slow start wasn’t helped by the media frenzy that erupted and the politicians starting to
criticize the project, the procurement process and seemingly overly optimistic traffic forecasts. All this
led to community developing negative views about PPPs in general and perceiving them as secret
deals. All this was further worsened by the media criticizing the government for closing roads and
seemingly forcing traffic into the tunnel.

6. Using the projected traffic volumes, financial and other data, prepare a financial model of
CCM utilizing a range of economic scenarios, and comment on the insights gleaned from such
models. Revise the model based on actual tunnel patronage experienced and comment on the
tunnel's debt capacity when the term loan becomes due.

7. Because Australia is a mature democracy with a developed economy and stable institutions,
it could be argued that CKI underestimated public sentiment in Sydney and thus its impact on
the Sydney CCT patronage. What political, regulatory and other institutional risks may CKI
further envisage from the Sydney CCT's negative public image? What pre-emptive corporate
action should it take to mitigate such problems?

If the CCT project continues to have bad publicity, politicians could get involved and CKI could
run the risk of not being able to perform its operations in Australia anymore. The outcome of this
political risk could affect 1) their ability to recover money from this investment, 2) affect the firm’s
profitability and image and, 3) their ability to execute similar projects in other countries which in turn
will affect their globalization strategy.

In order to mitigate this risk, CKI should work with RTA to develop a plan that could improve the
public opinion about the new tunnels and increase patronage. For instance, develop a marketing
campaign to change the perception that people have about the tunnels.

Recommendation & Conclusion

Recommendation

Sometimes it’s important to listen to the people who will be most affected by your business
decisions. The importance of communication with stakeholders is most evident in relation to the road
changes that occurred as a result of the CCT construction. During the planning phase there was
broad public discussion, sustained ‘expert’ input, extensive community involvement and strong
support from opinion leaders. Despite this the CCT has been a source of controversy and community
dissatisfaction. This probably occurred because the community engagement post planning approval
was not seen to be reflected in how the project was delivered, and community dissatisfaction was not
managed appropriately because the end result revealed;
 • The toll was higher than many users were prepared to pay for the trip
• There also appeared to be a lack of transparency in regards to the determination of the toll
• The CCT has no clear destination and actually was seen as denying access to certain
destinations

Community consultation over road changes did occur but it appears that the consultation
process did not identify and communicate the significant resentment held by the community towards
the project as a result of the road changes.

The CCM project team should work closely with the RTA and the State Government to increase
patronage and public acceptance. This would help the CCT project to deliver the expected results and
help secure a long term sustainable business. Below are some recommendations on how to achieve
these goals:

• Develop a plan for recovery. Everyone involved in the project needs to accept that the
current approach to delivering project results is flawed and needs to be redefined. If not, they
will likely resist the steps required for recovery.
• Calculate traffic projects for the CCT using a best, most likely and worst case scenarios and
determine the revenues associated with this project.
• Develop and execute a marketing campaign which shows all the great benefits that the new
tunnels bring to the community such as reducing travel time by up to 20 minutes during peak
hours, safer and more pleasant environments with better urban designs and wider footpaths,
reduction in traffic noise levels and improvement in air quality.
• Create economic incentives to encourage patronage. Restructure the pricing model so that it
is more attractive for road-users to utilize the tunnels. Once people have seen the benefits of
using the tunnels, you can gradually increase the toll charge.
• Execute the above suggestions and closely monitor its performance.

Conclusion

The Sydney Cross Tunnel project suggests the project failed due the unrealistic traffic forecasts
that were probably affected by a relatively high toll and public resentment with a series of road
changes.

Regardless, the project delivered a high quality road project at no cost to the taxpayer. The
lessons from the project include the need for all stakeholders to consider traffic forecasts with an
increased level of skepticism. Recent systematic reviews of the accuracy of traffic forecasts should
help this process. Governments should also be more skeptical of deals that perhaps are too good to
be true and develop an increased attention to protecting the public interest in PPPs rather than just
looking at the bottom line. The last more fundamental concern suggests that the traditional pricing
model for road PPPs may be flawed and alternative models may need to be developed.