Telstra’s NBN cash cow: Kohler

Telstra’s deal with the National Broadband Network Company is effectively a sale and leaseback in which it knows what it’s getting for the sale but doesn’t know what it’s paying in the “lease”.

The 30-year “sale” agreement is for a total of $47 billion – not $9 billion, as announced. That figure is a lesson in spin-doctoring: it is the net present value of a stream of cash payments using an absurdly high discount rate (10% after tax). The $35.9 billion capital cost of the NBN, meanwhile, is actual nominal cash, not NPV.

A more realistic discount rate of 7% produces a NPV for Telstra’s cash stream of $20 billion – these are, after all, payments from a government-owned entity, although the polls suggest the government is going to get tossed out, so there’s some risk, obviously.

The actual cash amount to be paid to Telstra is not disclosed, but must be about $47 billion – roughly $750 million a year for 30 years, increasing each year at the rate of CPI – to get access to Telstra’s infrastructure (total, according to my trusty spreadsheet: $37.5 billion) and roughly $750 million a year for 10 years, also adjusted for inflation, as “decommissioning” payments – that is, compensation for Telstra switching off its copper network and transferring customers to the NBN (total $9.6 billion).

By the way, this is counted as operating expenditure so you can’t add it to the $35.9 billion capital budget to get a total NBN cost of $83 billion. But you should add the operating losses, which include what’s to be paid to Telstra. They total $2.5 billion over the first 10 years.

So if Telstra gets $47 billion over 30 years for leasing/closing its own network, what does it pay to lease back the NBN? Not only is this not disclosed, Telstra doesn’t even know what it is.

NBN Co has agreed not to charge Telstra any more than $24 per month per customer for the basic service offering, which is 12 megabits per second up and 1mbps down. There are 8.5 million lines in service, so that works out at $2.4 billion a year, cash out the door.

But that’s roughly what ADSL delivers now; the NBN is all about 100mbps and up to 1 gigabit per second. The average pricing assumptions by the NBN are not disclosed but are thought to be about $40 per customer per month. That sort of cost to Telstra would build up $4 billion a year after eight years.

That means that over 20 years, Telstra will pay the NBN about $90 billion in “lease” payments. Using the 10% discount rate used for the NPV of payments to Telstra, that produces an equivalent NPV of $38 billion.

So, $90 billion cash out; $47 billion cash in. NPVs, respectively, of $38 billion and $9 billion. Is that a good deal? It’s impossible to tell. The cash in is actual revenue; the cash out is a business cost on which Telstra will make a profit margin. All of its competitors will be paying the same so there’s no reason to suspect it won’t do well out of it.

If that’s true, Telstra becomes a profitable wholesaler and retailer of NBN capacity and the recipient of two cash streams of about $750 million a year each – one for 10 years and one for 30 years.

Only Telstra can receive the decommissioning payments, but the company can and should sell the infrastructure rental stream to a superannuation fund: a 30-year income stream like that is a perfect asset for a super fund, and very hard to find. In an auction, it would fetch at least twice the $5 billion NPV attributed to it in last week’s announcement.

If it did that, Telstra would have an unparalleled war chest for capital management and acquisitions.

Even if the infrastructure is not sold, the cash stream will make Telstra the richest media and telecommunications firm in Australia by a long way. It could, and probably will, become the nation’s dominant media company.