Sunday, 21 April 2013

NBN: Problems with Coalition DSL-NBN Plan

In trying to recreate the Coalition spreadsheets, I've run into problems:
  • relevant data has been withheld.
    • Nothing after 2019 is supplied, yet we are told "important things happen later".
    • We get the scantest of information about take-up rates.
    • There are no details about the changes in Operation Expenses caused by the changeover.
  • The spreadsheet provided doesn't add up and important years (FY2020/1 and FY2028) are left off.
  • Critical figures, e.g. "60,000 nodes", can't be found in the of main policy documents nor the Background document.
    • The estimate of "$900 per premise" is based on an assumed highest-cost figure of $3,600/premises for fibre, seemingly plucked randomly out of the air.
    • A related figure, the maximum distance to the node, isn't specified either to be 400m or 800m. Halving the distance doubles the speed attainable but quadruples the nodes needed.
    • Even at a low $30,000/node plus cut-over costs, that's a major uncosted variable.
      • As well, as the radius served decreases, fewer nodes can be placed in Exchange buildings where lines cost 60% less.
    • This is main policy document only vaguely states the distance rule:
  • reducing the length of the copper connection to a few hundred metres.
The most crushing problem for the Coalition is their spreadsheet doesn't add up: simple accounting rules apply. The Cashflow must be the (Change in Working Capital) + CapEx + Gross Profit (Revenue - Expenditure), a.k.a. EBITDA (Earnings before Interest, Tax, Depreciation, Amortisation), because in a significant policy all material figures are shown and no others are given.

Withheld critical data

The only reason I know for people to deliberately withhold critical data useful in decision making, like the service books of a 2nd hand car, is to disguise unpleasant facts. To make the purchaser assume that things are in a much better state than they really are.

For the first Policy launch of the campaign, this "slippery" and devious behaviour can two downsides: it creates questions about the NBN policy and will make ever other policy, no matter how complete or well detailed, come under intense questioning.

Spreadsheet Problems

The best we can assume is that the missing value is Interest or Tax as they involve real cash transfers, not journal entries. As NBN Co won't be paying tax for some time, the only candidate is Interest

We know from the Summary Financials in the NBN Co Corporate Plan, that NBN Co makes interest at the moment, slightly reducing their Cashflow. The 2010-2014 figures are also changed materially by Depreciation and Changes in Working Capital.

Prior to 2016, Turnbull has NBN Co paying Interest, after that time, there's an unaccounted for positive Cashflow appearing out of nowhere when it is still making large losses. [corrected formula]

Turnbull keeps this interest income for 2 years into his Plan variation, suggesting that the Coalition will maintain a high level of pre-funding that he's criticised.
A major part of the Labor funding model is to move from Government Equity to direct Debt funding, starting in 2015 and increasing to 100% Debt funding in 2018. The Coalition spreadsheet is silent on this crucial detail.

This undeclared cashflow and unspecified funding destroys, for me, the credibility of calculations supplied by the Coalition:
What are they hiding or manipulating, because the data supplied don't add up?
Network Design Parameters

The assumed costs and distances are very unclear and central to the Coalition making their case for "cheaper and more affordable". Since last Friday's release by NBN Co of new detailed costings, for the standing committee on the NBN, the Coalition assumptions are looking very, very contrived.

From the Background document, list of assumptions:
  • Brownfield FTTP is restricted to areas with degraded or maintenance‐intensive copper. A combination of less gold‐plated specifications, choice of sites more effective handling of contractors lowers per-premise costs from $3600 up to 2014 to $2700 from 2015‐2019 (still 10 per cent higher than the FTTP estimate in the revised Corporate Plan).
  • Satellite and wireless costs and schedule are assumed to be as articulated in the revised Corporate Plan. FTTx costs are assumed to be 25 per cent of NBN Co’s pre‐2014 FTTP cost - the same ratio as in the US/UK and higher than the 20 per cent ratio in NZ.
There's a contradiction in there: $900 is 25% of the highest price assumed of $3,600, not the $2,100-$2,500 of the pre-2014 FTTP cost. Which is dominated by the $1200-$1500 disconnection payment to Telstra. While not cheap, it was worth NBN Co making that deal with Telstra for access to their ducts etc.

In an interview by Carol Duncan, ABC Newcastle, Mr Turnbull seems less vaguely, but is still woefully imprecise. The most important datum is that on the Coalition DSL network around one in eight premises will never get 50Mbps, the can only claim a guaranteed 25Mbps service, even after 2019.
MALCOLM TURNBULL: You've got to remember that under our scheme we are replacing almost all of the copper. The only copper that would remain in the customer access network is the last four or five hundred metres to the premise, and the reason for not replacing that is that as long as it is in good condition, as long as the length is short, you can deliver very high speed broadband - up to 100 Mbps - so you can deliver very high speed broadband, certainly more than fast enough for what people want and what people value, but you save a gigantic amount.

MALCOLM TURNBULL: Let's assume that we can spend $900 on average to get a premise up to the most part 50Mbps but no-one less than 25Mbps, and we can do that now. And let's assume it's going to take us the best part of another $3,000 to get them up to 100Mbps and up to 1Gb with FTTP, but let's assume that there's not going to be any demand for that very high speed in those residential areas for, say, 10 years ...

MALCOLM TURNBULL: No, I don't believe you'll need upgrades in the very near future.

Most people will get by 2016 on the fixed line upgraded network 50Mbps or better. We've said 25 Mbps is the minimum, that is the direction that we will give NBNCo as the minimum, so they have to do it on the basis that nobody gets less than that.

Our goal, and our direction to NBNCo will be that by 2019 to ensure that at least 90% of the people on that network have not less than 50Mbps.

No comments:

Post a Comment

Note: only a member of this blog may post a comment.