The lack of delivery on Light Rail in Auckland has been a source of a lot of frustration. The most recent news was back in August when the government announced it had taken the bizarre step of pitting the NZTA and the NZ Super Fund along with their Canadian partners off against each other.
The two preferred delivery partners for Auckland light rail have been chosen and a final decision on who will build this transformational infrastructure will be made early next year, Minister of Transport Phil Twyford announced.
NZ Infra, a joint venture between the New Zealand Super Fund and Canada’s CDPQ Infra group, and the NZ Transport Agency, will further develop their proposals for Government to consider early next year.
When NZ Infra’s unsolicited proposal was first announced I was pretty excited by the idea, especially if it helped to deliver the project quickly. But it has been clear since the start that this would not be free money and that NZ Infra expect a return on their investment. One of the big questions has been just how that will work. With this post I thought I would delve into just that.
This kind of investment is obviously completely new to the NZ Super Fund but it’s not for CDPQ and so on the basis that of best predictor of future behaviour is past behaviour I’ve focused on them.
CDPQ’s current flagship project is the Réseau Express Métropolitain (REM) in Montreal, Canada. It is a CA$6.3 billion, 67km, 26 station, fully automated light metro project – although much of the length comes from a complete replacement of some existing commuter lines so much of the right of way, including tunnels under Mt Royal, already exist and just need to be upgraded. The first stages are expected to open in 2021 and the full network by 2023.
The proposal involves using 2-car light rail trains with 4-car versions at peak times running as frequently as every 4 minutes throughout most of the day. In the first year of operations about 150,000 people a day are expected to use it, about twice what the Auckland rail network does.
The project itself has many critics, some of which can be read about here, but that’s not the focus of this post. Most importantly for us, CDPQ say the model they’re using on the REM in Montreal is one they want to ultimately export to other cities, including Auckland.
As mentioned the REM is costing CA$6.3 billion to build but not all of that is private investment. CDPQ will put in just $2.95 billion, less than half of the total, with the rest coming from central and provincial governments. Yet CDPQ get a 70% stake in the resulting infrastructure. The financing structure is shown below.
My guess is we can expect something similar in Auckland with at least the central government asked to provide an upfront contribution.
Getting a return
Even though it’s less than half of the total cost, $2.95 billion is still a lot of money to get an investment back on and a few articles have highlighted exactly how they’ll do that. Instead of a set repayment, like on a loan, CDPQ will be paid 72 cents for every kilometre any passenger travels – currently about NZ 86 cents per km. I understand that this figure is also indexed to inflation.
Similarly, the same can happen if ridership is lower than the annual projections established. The annual fee the REM will receive from the government is based on a rate of 72 cents per passenger per kilometre.
If riders aren’t there, “we receive the same amount and it’s our problem,” he said. “There’s nothing (in the contracts) that account for anything other than revenue per passenger-kilometre.”
In a recent interview with Presse canadienne, a source of authority in the area of public transit noted that the portion of the annual fee paid by the Quebec government for the operation of the REM, once the network is in place, should amount to approximately $240 million, the return required by the Caisse.
In a report last week, the auditor general, Guylaine Leclerc, said the Caisse will have a “priority return,” but said “there are still some risks for the government that could lead to additional funding.” Those risks could cost another $600 million.
She said traffic projections are conservative but realistic, which could mean higher-than-projected revenue.
Tall said that under the agreements with the province, if the projections are exceeded by 15 per cent, the portion paid by government decreases, and if it exceeds 40 per cent, the REM will receive only the value of a passenger’s ticket. A clause provides that beyond this threshold of goodwill, the Caisse could ask for “an economic and financial rebalancing.”
To put things in perspective, as of just over a year ago Auckland’s rail network carried around 250 million passenger kms for the year and with fare revenue of around $48 million. An average of just over 19 cents per km so at best fares would cover only a very small proportion of these costs.
One thing that’s notable is the different tone to different audiences. The article above comes from a newspaper and talks about how there’s risk to CDPQ but the quote below to an investor newsletter is basically saying how they can’t lose as well as what kind of returns it is delivering. That same article also effectively mentions they’re looking to do this same model in NZ.
Solid return every year
“For us, (REM) will pay us 8% to 9% every year,” Mr. Sabia said. “That’s not cyclical, not affected by business cycles; that’s 8% to 9% every year. Our target return for our overall portfolio is (annualized) 6% to 6.5% long term. There’s no way this asset offsets our fiduciary responsibility. It allows us to do two things: get a significantly higher return than our long-term rate of return, and it gives us returns that are more stable, more countercyclical.”
Mr. Sabia called claims that such investments are contrary to fiduciary responsibility “kind of a paper tiger.”
Exporting investment ideas like REM has drawn attention to CDPQ as a global investor, Mr. Sabia said
On top of the all this, the deal has a 99 year term on it so CDPQ will be getting this 8-9% return for a century. An interview with the Super Fund last year suggests the same length of contract would occur here.
“That [the Montreal light rail network], is a 100 year asset. They [CDPQ] expect to own it through that. We think in that same way,” says Whineray.
To pay for their share, it appears the Quebec government are looking at how to capture the value of property price changes and development, something we need to do here too.
Running the trains
For paying per passenger km you’d think that would include operating the trains too but it appears is another cost to be covered by central/local government. Again to that newspaper article above
The REM is a $7-billion project that is privately owned by the Caisse de dépôt, but half of that amount will be paid for with public funds. Quebec and the municipalities will finance the REM’s operation, and estimates suggest it could pay out more than $11 billion over 20 years.
So about $550 million annually on top of that construction cost just to run the trains.
If all of this doesn’t work and ridership isn’t higher enough then there’s ….
The out clause
Another newspaper article talks about there being risk of lower than expected ridership, so what happens if that occurs.
The Caisse de dépôt has the right to sell off Montreal’s REM commuter light-rail network five years after it comes into full operation in 2023 if it feels one of the line’s four branches is underperforming, with Quebec having first rights to acquire the train line.
Alternatively they could just replace it a branch with a bus
It would be up to the Quebec government to decide if a sale is required, or if other options, such as compensating the caisse for an underperforming line, or simply closing it down and replacing it with bus service, was more warranted.
Everything we’ve seen so far indicates the NZ Infra proposal is pretty much identical to what’s happening in Montreal. As a summary, here’s what we can expect should the government choose them to deliver Light Rail.
- A significant upfront government contribution will be required, likely around half the cost of the project.
- Ongoing payments of perhaps around 90 cents for every kilometre travelled by every passenger on the train for the next 100 years – committing multiple generations to significant ongoing costs. The exact figure will need to be enough to provide them an annual 8-9% return on their investment and much of that will be to fund pensions in Canada.
- Auckland Transport will still to have to pay to run the trains on this new network.
- If it all fails they can sell and walk away.
Why any government would even consider signing up to a deal like this, especially they’ve just reported a huge surplus, they’re below their debt targets and bond rates are down around 1% is beyond me. If you do think it’s a good deal, I’ve got some snake oil to sell you.
What’s more, had they not been distracted by this absurd offer, we could have had light rail construction started already.