NBN: Viable or Not?

There has been a lot of bellyaching of late about whether or not the National Broadband Network (NBN) will be a viable proposition or not. Certainly, the purported cost of $43b is a lot of money, and when dealing with a sum of that magnitude, a certain level of caution is very obviously quite prudent.

Supporters of the network point to the potential flow-on effects to the Australian economy – increased productivity, environmental benefits – the list is almost as limitless as the imagination can muster. Opponents say that the huge price tag is far too much, and that a full cost-benefits analysis needs to be performed before we earmark the funding for the project. Others say the analysis is not needed.

These are strong arguments either way. For myself, the biggest thing missing from the debate is any simple, rational exploration of what the numbers might be. So I’m going to give it a bash!

The starting numbers look like this: approximately eight million premises to lie within the fibre footprint of the network over eight years – (or ninety six months) – and that is growth of about 83,333 premises within the footprint per month.

Assuming an uptake of around 70% of premises within the fibre footprint opting to use the NBN, the subscriber base will grow by around 58,333 every single month across the eight year rollout. Assuming rough parity with current wholesale access pricing for our existing ADSL/ADSL2+ network(s) of around $25.00 – (about what is expected) – the revenue base grows by $1,458,333 every month.

Presuming that the numbers are maintained across the eight year life of the rollout, the total number of subscribers connected in the 96th month will be 5,600,000, generating $140,000,000 of revenue per calendar month. Or if you like, $1,680,000,000 per year. Yes, that is 1.68 BILLION dollars.


Ever wondered why Telstra delivers such massive profits – ($3.94b in the year to June 30, 2010 for example) – year after year after year?

Adding the monthly revenues across the ninety six months, at the end of the eight year rollout, there will already have been $6.79b of revenue raised, just because the network existed and was raising revenue from customers.

Of course, I can’t factor operational costs into the numbers – as that is something there is no information about, and I doubt NBN Co themselves can do that yet – so this is only a revenue figure, not a profit figure.

It is however reasonable to suggest that whatever operational profit comes out of that $6.79b revenue, that a certain amount of that will have been turned back to offset the cost of the build throughout the eight years, reducing the $43b price tag.

Again, not allowing for costs, subtracting the $6.79b from the $43b means we have $36.21b still to be “recovered”. Since – presuming the wholesale cost to supply each premise doesn’t change (unlikely) – the network would be raising $1.68b per year, it would take a little over twenty-one-and-a-half more years to recover the $43b.

Any increase in the wholesale per-premise cost will obviously shorten the time frames.

What of the costs? Well, that’s a fair point, but are those costs offset by the benefit of having the NBN in place over those almost 30 years? Quite possibly, and in my opinion, absolutely.

It does seem that uptake numbers will be very important, so lets take another look at the numbers.

Since the copper network will be decommissioned, every premise within the NBN fibre footprint that currently has any style of active xDSL connection – or even just a POTS line – provisioned to it, will need to have that connection transferred onto the NBN. The copper network will be gone, so there is no choice.

In that respect, uptake numbers will actually be much much closer to 100% than the 70% used in my example above. Further, given that every NBN connection to a premise will be able to carry more than one service across it – (as an example, you might have one for internet connectivity, one for a POTS connection, and one for subscription television) – actual uptake might be more than 100%.

That’s a story for another day however.

For argument sake, lets change our figures to 95% uptake, and a slight increase in monthly wholesale cost per premise to $30.00.

The new numbers look like this: approximately eight million premises to lie within the footprint of the network over eight years / ninety six months – that is growth of about 83,333 premises within the footprint per month. An uptake of around 95% of premises within the footprint opting to use the NBN, the subscriber base will now grow by around 79,167 every single month across the eight year rollout, with the revenue base growing by $2,375,000 every month.

The total number of subscribers connected in the 96th month will be 7,600,000, generating $228,000,000 of revenue per calendar month, or $2.736b per year.


The network will have already generated $11.058b of revenue, bringing the remainder to be recouped down to $31.942b. At $2.736b per year, it will now take only a little over eleven-and-a-half years to get the raised revenue up to $43b – a saving of TEN WHOLE YEARS over the earlier example.

Theoretically the costs to run the network won’t change from the first example, because the network will be physically the same.

Even with relatively poor uptake levels, the NBN will be able to generate annual revenue measured in the billions.

Add to that the economic benefits of just having the network – and importantly, having this infrastructure before our trading partners – and the benefits of spending this money on the NBN start to look mindblowing.

And I didn’t need convincing before I ran these numbers!

By no means have I just done a full cost-benefits analysis, and admittedly I have made a number of assumptions in tabulating my numbers. Remember though, that anyone trying to analyse this proposal needs to make similar assumptions.

However even my cursory look at the potential seems to suggest that a lot of the FUD about massive financial blackholes that are out there right now are decidedly shaky.

  • lets use tassie as a snapshot………..NBN rolled out to over 4000 homes, uptake 70 homes……less than 3%. i think it might take a bit longer than you predict if uptake is as slow as in tassie (source:http://www.abc.net.au/unleashed/40054.html)

    • Three very small towns. There are a might more people living in the major mainland centres.

  • Pingback: Delinker()

  • Using your figures:

    $30 with 95% uptake:

    $2.7 billion revenues from $43 total cost = 6.3% return

    Doesn’t include the following;

    – Operating Expenses (NBN Staff, etc) = $1 billion /year (McKinsey report)
    – Capital Expenses (GPON, Wireless, Satellite equipment replacement) = estimate $700 million/year average

    Brings your profits down to $1 billion a year (2.3% return). Which would not be anywhere near enough to cover the interest on $43 billion worth of bonds raised by the government.

    Then you actually have spend some money to actually pay off the principal since even Fibre cabling doesn’t last forever.

  • I hate to admit it Michael but I’m starting to warm to the idea of this NBN.
    Keep up the good work.

  • That will be why I said this “doesn’t include operating costs” – but any organisation pulling revenues of $2.7b, and can’t make a profit is doing something severely wrong.

    Given Mike Quigley’s industry record – I find it VERY difficult to believe that he couldn’t build an organisation that can pull this off. Also, given that the pilot build has come in under budget, and ahead of schedule, the numbers look better again.

    • I agree there’s social benefits involved. Its just your statement that ‘it will now take only a little over eleven-and-a-half years to get the raised revenue up to $43b’ is very misleading. It implies that the NBN will completely pay for itself and there is no actual cost to the government.

      If there is a cost, then calculate it and put it on the Federal budget. Afterall, when you build freeways you put it in the budget do you not ?

      • I quite clearly state that these figures do not include costs…this was an exercise in demonstrating the REVENUE potential. NBN themselves will not be able quantify costs at this stage, let alone me.

        The NBN is in the budget – $330M was just allocated to NBN Co from the government purse – I’m quite sure you’ll find that as a budget line item come budget night.

        • The costs is somewhat quantified in the McKinsey report. Same with the revenues. There’s disagreement about the accuracy of the report of course, but its a starting point. Hope Conroy releases the NBN Co business plan.

          Revenue potential alone is quite useless in measuring the overall financials of any venture. Its trivial to make billion dollars of revenue selling pretty much any product/service (eg, cars for $1,000 each). The trick is the cost of doing so must be reasonable and thats the hard part.

          What you’re doing, using revenues alone to measure payback period of capital expenditure doesn’t really add value to the NBN discussion and is quite misleading. Yes, I know you’ve said you haven’t counted costs.. but costs do matter a great deal when its in the billions. Unlike say solar panels which have virtually no running cost.

          And the NBN by the way is an off-budget item. Funding is allocated mainly through bonds (like that $330m), but that is not measured as an expense item.

          http://m.zdnet.com.au/costings-claim-900m-coalition-nbn-hole-339305723.htm

          The expense item allocated for the NBN in the next 4 years is $1.5 billion worth of public debt interest (interest on NBN bonds) as measured by Treasury. Thus the cost of NBN itself (about $20 billion over the next 4 years) is counted as an equity investment and is ‘off-budget’.

        • I also see that you mentioned that Telstra has a net profit of $4 billion.

          Well, thats actually from revenues of a staggering $25 billion (or about 10 times of your NBN’s revenue projections). Making a total ‘cost’ of $21 billion for that year.

          Costs really matter.

  • You are also stuck in the thinking that this is a “commercial” project. It is not – this is a government infrastructure project. Do government funded freeways require the ability to create a profit? No – they are funded to create economic stimulus, just as the NBN will provide. Stimulus to the rest of the economy.

  • Pingback: links for 2010-10-21 | Leefe rates the world...()

  • Pingback: NBN: Telecommuting and Productivity()

  • While I agree with your analysis, the problem the Government has is that they’re intending to run NBNCo as a commercial enterprise earning commercial returns attractive to commercial investors, and that’s where all the wheels fall off.

    Your second scenario, 95% takeup, features $2.74b worth of revenue p.a. If the build is going to cost $43b (and we have to assume it will, because the McKinsey report said that’s almost exactly what it’ll cost, and that’s the only third-party verified data we have), then $2.74b represents about 6.3% return on capital.

    That’s without paying-down any of the capital cost (which might be okay, depending on the business model) and, more fatally, without including any allowance for operational costs, particularly repairs and maintenance.

    Considering the fact that the ONLY reason the copper network is in a state of disrepair now is because Telstra has neglected repairs and maintenance, it’s probably a good idea to not repeat that mistake. If you make a rational allowance for those, you end up with a ROI somewhere south of 5% per annum.

    If I have a billion bucks, I could get more than 5% by putting it into a term deposit (!). So why on earth would I invest it in NBNCo?

    This is the challenge the Government faces. They’ve vowed and declared that they’re going to raise 50% of the funding from private sources, and sell the whole lot to private investors 5 years after it’s built; But there’s no conceivable way that the network will earn enough money to be attractive to private investors — UNLESS they hike prices into the stratosphere to turn your revenue projections into something that looks like a real commercial return.

    That’s why I think the network should be funded entirely by the Commonwealth and never sold. http://www.abc.net.au/unleashed/36078.html

    • Agree completely that the network should not be sold…selling it would repeat the mistakes of the past…

  • See my cost benefit analysis on the NBN here – http://valman.blogspot.com/2010/07/my-cost-benefit-analysis-on-nbn.html

    I found that to achieve a positive NPV, the NBN needs to add between 0.5 and 1.0% to GDP per annum. Approx $7 billion per annum, proportional to % build complete from network 50% complete. I add NPV for NBN Co, including projected revenue, depreciation, operating costs, but no interest.

  • Pingback: NBN: Not Mobile Wireless Broadband()

  • Pingback: NBN Putting the Consumer In Charge()

  • Pingback: NBN: Congestion, Telecommuting and Productivity()

  • Pingback: NBN Legislation Passes – Now Get On With It!()

  • Pingback: The Sheep Is Dead – Long Live the Sheep()