PPP caught in the gully
In February 2013, the NZTA evaluation team eagerly opened five expressions of interest for Transmission Gully, New Zealand’s largest PPP. On July 29 2014, the project reached financial close. By Geoff Daley, who led the financial advisory team on the project, head of advisory for the Australian Structured Finance Office at Bank of Tokyo-Mitsubishi UFJ Ltd.
To see the full digital edition of the PFI Yearbook 2015, please click here .
To purchase printed copies or a PDF of this report, please email firstname.lastname@example.org
The project itself is relatively simple. Under the NZ Government’s Roads of National Significance programme, major upgrades are planned for the highway system north of the nation’s capital Wellington. A key element of this is Transmission Gully – an alternative to the existing coastal route north from Wellington. The existing road is highly vulnerable to any future seismic events, and a bottleneck currently for commuter, freight and holiday traffic into and out of Wellington.
Transmission Gully (so called as part of the route goes up a gully where transmission lines now run) will be a 27km divided highway, with three grade separated interchanges and 28 bridges. There are significant engineering challenges involved in building a motorway standard road through such terrain, in a seismically active area crossing the Ohariu Fault.
The construction JV of Leighton Contractors and HEB Construction have a job ahead of them. However, this is a finance journal, so without diminishing the engineering task, we’ll focus on those aspects most relevant to the project financing.
The New Zealand Transport Agency (NZTA) will make availability payments for 25 years upon completion of the road, which is scheduled for April 2020. The construction phase is well over five years, elongated by the distinct seasons typical of the Wellington area. In other markets, these availability payments would support a long-term financing structure; but in New Zealand, like Australia, mini-perm structures are the norm from the bank market (and no functioning project bond market exists).
Accordingly, the majority of the debt commitments (NZ$915m – all amounts approximate, and may not add up due to rounding) were provided by a group of international banks via a construction facility and a debt service reserve facility each with a seven-year tenor.
*Mandated lead arranger s – National Australia Bank (through both BNZ in Auckland and NAB in Sydney – NZ$190m), Commonwealth Bank of Australia (NZ$190m), United Overseas Bank (NZ$170m) and Scotiabank (NZ$115m).
* Senior lenders– ANZ (though a joint Auckland/Melbourne team – NZ$155m) and ICBC (through its Sydney and Auckland branches – NZ$95m)
One of the sponsors, the Accident Compensation Corporation, had a keen interest to provide long-tenor NZ dollar debt to the project. This reflects the ACC’s long-term liability profile, providing comprehensive accident insurance across New Zealand, which it wants to match with long-term assets. BTMU and ACC structured a long-tenor (29-year) NZ$125m debt instrument, which ranks pari passu with senior debt (subject to voting conflict provisions reflecting ACC’s dual roles as senior lender and equity investor).
The inclusion of this debt piece reduces the refinancing risk for the project, and provides a template for additional tranches of long-tenor debt to be included in the project at later refinancing points. The 29-year loan is fixed-rate, which required agreement with NZTA to establish a forward curve, as there are no observable rates in NZ dollar for instruments of that tenor. Equally, the fixed-rate instrument is excluded from the base rate adjustment mechanism that applies at each refinancing point for the mini-perm loans, thus reducing ongoing exposure to base rate risk on that part of the capital structure.
Equity of NZ$130m will be contributed by the equity investors, with the majority of equity LC-backed and injected towards the back end of construction for financing efficiency. However, the ACC’s contribution was made at financial close by way of a convertible note instrument, due to its inability (under its enabling legislation) to raise a letter of credit to secure its equity. Conversely, ACC’s ongoing obligation to respond to progressive debt drawdowns alongside the other lenders is unsecured – demonstrating the banks’ willingness to treat ACC as a lender on the one hand, but requiring either cash or LC for ACC’s equity.
The NZTA is an independent statutory authority, with a hypothecated revenue stream of fuel excises and licence/registration fees of circa NZ$3bn, via the National Land Transport Fund. Given this significant revenue base, the NZTA has financial independence from government, allowing it to make decisions in the best interests of the road network across New Zealand (with the intent that it frees such decisions from political interference).
The NZTA therefore intended to enter into the PPP in its own right as the procuring authority. This was at odds with the expectation from project finance banks, which commonly see such authorities guaranteed by the Crown in Australian PPPs1. Of particular concern was the potential for NZTA to be liable for a termination payment, or a self-insurance payment (see discussion later), which would represent a significant proportion of its revenues. Further, over a 30-year term, the structure and/or revenues of the NZTA could change.
This was the subject of much negotiation. Ultimately, the Crown provided an indemnity to the NZTA that was enforceable by Wellington Gateway Partnership through privity of contract legislation2. This brought the New Zealand government’s AAA credit rating to the deal, and covered the majority of payments that the NZTA could be liable to pay over the operating period of the project (including, importantly, all availability and termination payments).
Financing structure – early completion bonus
The interplay between construction delay and its financial consequences in a project financing – liquidated damages (LDs) – are well understood. For every day a project is late, there are no revenues, but financing costs continue, and it is the D&C contractor that bears this risk through its obligation to pay LDs. In many PPP projects this risk is heightened, because for every day the project is late, the operating term shortens (ie, the forgone revenues are not delayed, but lost forever). Thus, the D&C contractor needs to compensate the project vehicle for all the revenue forgone, as it has been modelled that this revenue will be used to repay the capital providers (both debt and equity).
There were a number of factors in the Transmission Gully project that led to a slightly different approach. First, the NZTA set a 25-year operating term from April 2020 – so was expecting to pay 25 years of availability payments from that date, and calculated the net present cost (NPC) accordingly. Setting a target completion date earlier than this would bring forward the payments – thus increasing the NPC (unless some form of adjustment was made). However, setting a target completion date earlier would also increase the likelihood of the project running over time – ie, less buffer in the construction programme for delays – particularly weather delays (it’s not called Windy Wellington for nothing!).
In many PPPs, there is a lack of symmetry in the financial consequences for the D&C contractor between finishing early and finishing late. Typically, late completion will lead to LDs for the financing costs to be paid – but early completion rarely results in a gain to the D&C contractor for the financing costs saved.
BTMU and Leighton Contractors developed a strategy to balance these considerations, and reduce the NPC of the project. If the D&C contractor finished later than the April 2020 date for completion, it would have to pay LDs in the usual way (and given the size of the financing, these were considerable at around a quarter of a million dollars a day).
Conversely, if the D&C contractor finished earlier than this date, there would be a saving to the project vehicle as it would result in lower interest during construction (IDC and the facilities would not capitalise to the full commitments. Nonetheless, the availability payments from NZTA were calculated on the assumption that all the debt was drawn down. This means there is capital available to pay the D&C contractor a bonus – representing the IDC saved – if it finishes the project early.
This created symmetry – the D&C contractor pays LDs if it is late, but receives an early completion bonus if it is early. This wasn’t just a nice upside for the D&C contractor – far from it – in order to deliver the lowest cost project and reach the NZTA’s affordability threshold, the D&C contractor took a view as to how many months earlier than April 2020 it could finish, and then bid the value of the corresponding early completion bonus into the project. That is, if it didn’t finish that early, it would receive a lower bonus than it had banked on. The value of the bonus banked by the D&C contractor, and therefore reduced from its bid price, was considerable.
This structure was new for the D&C contractor – equally, it was new for the banks, which are used to the protection of LDs on the downside, but do not typically allow for drawdown of remaining commitments if completion is achieved early. The banking group ultimately understood the value offered to the bid (and thus winning the project) by agreeing to this innovative element of the deal. Complexities surrounding the mismatch of interest rate hedging profiles if the project commenced early (and therefore started amortising debt earlier than scheduled) were resolved. An exaggerated schematic of the structure is set out below in Figure 1.
One of the key drivers for the NZTA was for Transmission Gully to be an exemplar of safety. The NZ government’s “Safer Journeys” programme seeks to minimise death and serious injury on New Zealand’s roads, particularly across its motorway network.
The NZTA sought an alignment of interests between all the participants in the project in the pursuit of safety objectives. Importantly, it wanted to see the owners of the road – the equity investors – with a financial stake in the road’s safety performance. The KPI regime includes charges levied on the project vehicle if a death or a serious injury occurs on Transmission Gully. This is almost irrespective of cause; that is, the road should be designed and maintained such that it minimises harm, even if a driver is speeding, or under the influence of alcohol.
This risk was therefore one that had to be shared across the consortium – to the D&C contractor for any defect in the design or construction; to the O&M contractor for any factor in the operations and maintenance of the road; and, critically, to the equity investors for any residual risk. (To anyone who has read the fiendishly complicated risk allocation involved, please forgive this gross over-simplification!)
The intent of the NZTA was clear – to provide alignment, and incentive, for all parties to ensure safety. If the owners of the road are liable for charges for a death or serious injury, indeed higher charges if there are repeat occurrences, then there is incentive for them to re-invest in the road if that is necessary.
For Wellington Gateway Partnership’s equity investors, this was new territory. However, with the Accident Compensation Corporation as one of the investors, there was also alignment to their corporate objectives. Moreover, the equity investors took note of the statistical evidence from New Zealand’s motorway network, that the incidence of fatalities or serious injuries on well-designed and built motorways is in fact very low (no doubt due to all the appropriate safety features – divided highway, safety barriers, sight lines, road geometry, etc).
This, however, was not a risk the banks were used to seeing, and ultimately it was not a risk that they were prepared to get comfortable with, as they viewed it as ultimately uncontrollable. The end result was that a cap on charges in any year was agreed – far above what the NZTA or equity investors thought would ever be relevant – but such that the banks could ignore this risk (above the cap) for the purposes of their downside analysis.
Despite the cap on liability for charges where safety is not delivered, the project has forged new territory for risk allocation, with D&C and O&M contractors taking their share of this risk, as well as equity investors putting a portion of their return at risk. The starting point for the next NZTA PPP will be interesting to see.
One rationale for building Transmission Gully is to provide an alternative route to and from Wellington, as the coastal route is vulnerable to seismic activity: in many places it is a single lane, undivided highway with steep slopes down to the coast.
The costs of insuring the new Transmission Gully roadway during its operating term were proving significant during the bid phase, not just due to general seismic risk (with the memory of the disasters in Christchurch still vivid), but also as Transmission Gully crosses the Ohariu Fault. Thus, the NZTA offered to bidders that it would effectively insure Transmission Gully, just as it self-insures all New Zealand roads (ie, there is no insurance on those roads; the NZTA bears the costs of any repairs). The guiding principle of such insurance was that the private sector would be in no better and no worse position than if it had effected commercial insurance.
This was largely undocumented during the bid phase, and given it is relatively unusual in PPPs, occupied much time in the negotiation phase. Key to the difficulties in such a regime is that the NZTA was both the client for the project (making payments for delivery of a road that is available for use) as well as the insurer for the project (making payments in some circumstances that make the road unavailable for use).
There was thus a complex interface between paying insurance, or granting relief (and thus continuing to pay availability payments). This was in some cases exacerbated by the absence of the commercial tension that exists between insurer and insured (eg, increasing premia, or refusing insurance, if the behaviour of the insured increases risk or the effects of risk).
Ultimately, in hindsight, the support from the NZTA may have been better structured as protection for seismic risk only, with commercial insurances placed for anything else (at a modest cost with seismic risk excluded). In any event, this is unlikely to be a material issue on future projects unless the road is located in a seismically active area.
Another rationale for the construction of Transmission Gully was increasing travel time reliability. With the existing coastal route only one lane in some places, there are times when congestion is particularly bad. In the RFP, the NZTA sought a guarantee of the travel time from the private sector.
Whilst a traffic model will provide a theoretical travel time given expected traffic volumes, experience shows that driver behaviour, actual performance of interchanges and variable traffic volumes over time mean that an absolute travel time cannot be reliably predicted – particularly over a 25-year operating period. Moreover, if there was a difference between the theoretical and actual outcomes, the KPI regime would apply abatements for the entire operating period. It was just not a risk the private sector could take.
Instead, Wellington Gateway Partnership and the NZTA developed a regime centred on the “Fundamental Diagram”, which describes the relationship between flow rate and density for traffic flow on a motorway. It plots the flow of travel (vehicles per hour) against the traffic density (vehicles per km). Any driver will attest that if there are too many cars on the road then congestion results. It takes a jamologist to prove this!
The fundamental diagram for Transmission Gully will be measured over the first year (with anomalies removed from the dataset). The KPI will then apply where the flow is outside of the envelope for a given vehicle density (ie, where the speed is inhibited despite the number of cars on the road). This is designed to provide real incentive to the O&M contractor to manage the road to avoid delays. Maintenance activities will need to be undertaken in such a way to avoid delays (eg, where traffic volumes are low, outside peak hours, nights, weekends). This is clearly what drivers, and NZTA, would like to see.
A successful PPP
New Zealand’s first PPP in the transport sector, and by far its largest, is a successful application of PPP concepts from abroad, adapted to the New Zealand context. PPPs in New Zealand are neither UK, Canadian nor Australian – they are Kiwi. With three projects now complete, and a strong central PPP unit within NZ Treasury, future PPPs will be based on the precedent contract now further developed on Transmission Gully.
With over a billion dollars in capital raised, Transmission Gully has placed New Zealand as a major actor on the world stage, and will see it continue to attract international bidders and global capital for its projects.
The views expressed in this article are those of the author alone and do not represent the views of Bank of Tokyo-Mitsubishi UFJ Ltd.
1 - See K Nishinari – Jamology: physics of self-driven particles and toward solution of all jams, Distributed Autonomous Robotic Systems 8, 2009, Springer
2 - Eg, under the Public Authorities (Financial Arrangements) Act 1987 (NSW)
3 - Contracts Privity Act 1982 (NZ)