Is the NBN Strategic Review's terminal value terminally wrong?
As the country waits for the results of the NBN cost-benefit analysis (CBA) it's worth looking back over the NBN Strategic Review – and reconsidering whether its assumptions about the long-term value of NBN Co aren't excessively optimistic.
Now that communications minister Malcolm Turnbull has put the national broadband network (NBN) and the entire telecoms industry into idle – while we wait for the results of the Telstra renegotiations and the findings of the Vertigan cost-benefit analysis (CBA) review – we have time to revisit and reassess the findings of the last official output we have seen from the current government.
That, of course, is the NBN Strategic Review (NBNSR) – completed in six weeks and handed down as gospel last December. This highly speculative document has been constructed and feted with an air of authority, yet was torn down in the interim report delivered by the Senate Select Committee on the NBN back in March – which was, predictably and despite its rather voluminous size, both refuted by the committee's Coalition members and completely ignored by the government.
Among the many points that are buried within the NBNSR is a financial model that seems to have been tailor-made to justify the multi-technology mix (MTM) model promoted by the review as being more cost-effective and easier to complete quickly.
That whole objective was all but rubbished by NBN Co chairman Ziggy Switkowski, who has publicly conceded that the project needs to ramp up so aggressively as to be all but impossible to achieve under current guidelines. The Coalition has yet to offer a clear strategy about how and when the company is going to ramp up to roll out what Switkowski has conceded must be 100,000 premises per month for the next eight years.
Turnbull, as has become his habit lately, has failed to produce any real answers that don't directly indict Labor for everything wrong with the NBN today – even, seemingly, the 2020 rollout deadline to which the Coalition committed itself with the NBNSR.
The NBN Strategic Review has already been discredited on numerous counts (see, for example, here and here). Yet if we are to take its promised 2020 deadline seriously – and all indications are that Turnbull still expects us to do so – we must also take seriously the NBNSR's projections around the long-term value of the NBN Co business.
That projection relates to the terminal value of the NBN – that figure used in accounting in the same sense that The King and I's Yul Brynner used the words 'et cetera, et cetera'. In other words, the terminal value of the Coalition's NBN represents everything that it expects will happen in the future but can't be bothered detailing.
The NBN Strategic Review addresses terminal-value calculations on just one page (p107), explaining that its internal rate of return (IRR) calculation of 5.3 percent is based on a terminal value of $45 billion in FY2040.
The NBN Strategic Review has already been discredited on numerous counts. Yet if we are to take its promised 2020 deadline seriously – and all indications are that Turnbull still expects us to do so – we must also take seriously the NBN Strategic Review's projections around the long-term value of the NBN Co business.
This valuation is “mechanically” calculated as being six times EBITDA for the theoretical revenues of NBN Co some 25 years from now – and 15 years after 2025, when by Ziggy Switkowski's own reckoning NBN Co will have already started to transition to a FTTP model.
By 2040, the actual NBN – the one based on the fibre that the Coalition refuses to pay for now – will have been built and will be based fully on fibre. Real revenues will therefore be based not on the services the government wants to build over its MTM model now, but on those that will have become capable over the FTTP network as it comes online.
While the NBNSR doesn't highlight the expected schedule for the transition to fibre, its revenue projections over time leave us with enough information to assume that we can figure out when it will happen. Those projections forecast a drop in cashflow during FY2029 and FY2030 – a period when the MTM model will be seven to eight years past completion and starting to show signs of technological and physical obsolescence.
This will therefore drive an investment in the fibre network that the government still acts like we won't need – pushing cashflow back up in FY2031 and helping it rise through FY3036 – after which another two-year decline will likely accompany another period of decision-making as the current infrastructure once again shows signs of age.
We cannot expect a complete upgrade at that point, however: judging by the information in the NBNSR, the government expects that its network – upgraded twice after the current belt-and-braces upgrade – will have a market worth of around $45 billion in FY2040.
The Abbott government needs this figure to be correct, because it is directly tied to the potential return on its investment. That return is, of course, the combined total of all revenues through 2040 plus the value of the remnant operation at that time, as against the expenses required to build and run the business to that date. A lower terminal value would reduce the overall return – thereby trimming that 5.3 percent IRR figure down to something far less attractive.
That would be unacceptable to the current government because it would suggest that the MTM model would not generate as much economic value for Australia as they're saying it will. By extension, it would also narrow the claimed financial improvements over the rival all-FTTP model. This would, necessarily, raise questions about whether MTM is indeed the best way forward, as the Coalition government has unilaterally decided.
The question to ask, then, is simple: Does the $45 billion terminal value assigned in the NBN Strategic Review seem reasonable?
To arrive at an answer, consider comparable large-scale transactions that offer relative values for comparison. For example, Telstra's EBITDA in FY1997, just before one-third of its shares were sold off in the $14.3 billion T1 share offer, was $6.597b. Using the six-times-EBITDA measure we would derive a total value for the company of $39.582 billion in 1997 dollars, meaning that one-third of the company would have been worth around $13 billion and the government did very well because the market actually valued Telstra at $43 billion (6.5 times EBITDA).
In FY1999, leading up to the T2 offer, Telstra's EBITDA sat at $5.849 billion, giving a six-times-EBITDA value of $35.094 billion. By this measure, the 16 percent of shares sold during the T2 offer would have been worth $5.615b, but they generated $16 billion for the government – suggesting that the market actually valued the whole of Telstra at $100b (17 times EBITDA) at the time.
Judging by the information in the NBNSR, the government expects that its network – upgraded twice after the current belt-and-braces upgrade – will have a market worth of around $45 billion in FY2040... The Abbott government needs this figure to be correct, because it is directly tied to the potential return on its investment....A lower terminal value would reduce the overall return – thereby trimming that 5.3 percent IRR figure down to something far less attractive.
The T3 offer didn't fare so well: in 2006, the government sold 34 percent of Telstra's remaining shares for $15.5 billion; this valued the company at $45.58 billion (8.2x EBITDA) overall. Telstra's FY2006 EBITDA was just $5.497 billion – providing a six-times-EBITDA value of $32.982 billion.
In other words, in 1997, 1999 and 2006 the market valued Telstra at 6.5, 17 and 8.2 times EBITDA to give it market values of $43 billion, $100 billion, and $33 billion on revenues of $15,983b, $17.571b, and $22.75b respectively.
These figures would seem to support the 6-times-EBITDA metric – but can we therefore assume that NBN Co in 2040 will have an EBITDA of $7.5 billion?
A more recent overseas infrastructure purchase – the $US45b ($A48.52b) purchase of second-place US cable provider Time Warner Cable (with 11m subscribers) by number-one Comcast (with 30m) – gives us a point of comparison because the numbers are conveniently similar.
Comcast is paying 6.7 times EBIDTA for Time Warner Cable in a move that will increase its subscriber base by one-third.
Those figures, and the numbers assigned to Telstra above, might seem to vindicate the government's assessment of a MTM-based NBN Co based on a six-times-EBITDA valuation. However, they don't offer much guidance about whether the revenue projections for 2040 are anything more than fantasy.
Chief among those is the fact that NBN Co is not a retail provider, and will not be able to benefit from increasing markup on new products into the future; its business, under current design, is providing a wholesale bitpipe and nothing else.
Without the opportunity to value-add on top of its services, NBN Co will face gradual price attrition over time – driving down revenues by up to 30 percent against the previous FTTP model by 2021, as NBN Co modelling revealed late last year. Because the MTM technology would limit the type of services that can be offered, there is limited potential to add new revenue-generating services in the future.
Can a nationwide, mixed-technology, wholesale-only network be reasonably expected to generate similar revenues to Time Warner Cable? Remember that we are talking about a diversified TV-broadband-phone-and-services market with 11m subscribers and strong ties to not one but two content-distribution empires; there is an integrated infrastructure-and-overlays story in that purchase of which NBN Co can only dream.
Can the NBN therefore be worth as much as the NBNSR believes it will? It seems like a fair stretch, given that the Coalition's IRR projections depend entirely on the terminal value that they have assigned the business at that distant point in the future.
That terminal value basically depends on NBN Co becoming a money-spinner on the scale of what Telstra was when it was sold off to the market. And that's a big ask – but a necessary one if the NBNSR is to hold water. Were the actual terminal value to be, say, 25 percent lower by that point, the Coalition's entire revenue model would be thrown into chaos – and its decision to forego a more-profitable FTTP network revealed as being highly questionable indeed.
What do you think? Are you comfortable with the Strategic Review's terminal-value calculations? Can NBN Co become as big a venture as Telstra or Time Warner Cable without the same retail margins and breadth of services? And, what does all this mean for the Coalition's RoI calculations?