Wednesday, January 27, 2016

Fast Start to 2013 Broadband posts

Primer: fast start to Stevej-on-bband, pre-election 2013 posts

Taking on Three NBN Memes of the Mainstream Media

Why you want Fibre to your premises:
The full Fibre NBN is the biggest free kick we'll be offered in our lives: Grab it now or lose it forever.
A longer version of “Why Fibre?”.
Problems with the Coalition Plan.

One of my first pitches for “Fibre to the Farm”.
Today I’d include a wholesale Mobile Service [esp for ‘blackspots’] and wholesale public WiFi access network.

NBN Financials: Hidden costs of a VDSL Fibre to the Node network
A long piece addressing specific differences between Copper and Fibre that have raised with me and highlighting additional costs, problems and deficiencies with a DSL/FTTN network.
The Rivers of Gold Scenario: What if NBN Co does better than forecast?
Ain’t nobody looking at the Upside :)
Why, apparently, did NO business journalist actually try to “sanity check” the Turnbull / Ellis spreadsheet?
If I can do it at all, then expert journalists with decade(s) of experience could’ve cranked the numbers in less than a day.

Sanity Checks of Turnbull 2013 Plan: too many questions

On the 8.9M DSL services, they were to save $1910 per line, valuing the Fibre install at $2810/line, or a total of $34.3 billion vs the $28.5 billion of the NBN Co 2012 Corporate Plan.
The Coaition estimate could never have been less than $26.5 billion.

Is the real Coalition NBN plan going to need CapEx of $20.5 billion or $35.6 billion, a 75% increase on their first estimate??
The electorate deserves an update.

Unpicking Coalition Claims - failing Sanity Checks.
The claimed $20.5 billion CapEx also fails basic sanity checks.
$8.1 billion in CapEx is traded for Opex and only 28% of FTTP network is built or $5.5 billion of $17.2 billion after (common) Access costs.
Then $8.1 billion has to be added for an FTTN.
That's $3.6 billion less in construction and $8.1 billion swapped for OpEx, for total CapEx savings of $11.5 billion, yet they turn a real $5.5 billion difference into a claimed $17 billion saving.
Conveniently $6 billion of "Other CapEx" has been ignored.
The sub-projets are now identified in detail by NBN Co.
Because the breakdown of CapEx and OpEx has been deliberately withheld, nothing is included in the plan.

Prime NBN Questions:

The Ninety Billion Dollar Swindle:
  • What does the NBN ‘Cost’?
    • Wrong question! It's an investment, it makes money for the taxpayer.
  • What was the extra cost to the taxpayer of Full Fibre:
    • around $10M-$20M/year.
  • Where’s the NBN in the Federal Budget?
    • Nowhere to be seen! The NBN is an INVESTMENT, not an Budget Expense:
      For any investment, need to look at the Profit & Returns, not ‘Capital Cost’, plus the Risks.
The MTM is the Ultimate “road-block” - the Balkanised Network with no road back to Fast Broadband (fibre).
  • The MTM can never be upgraded and, just like all the incompatible State Rail Networks, only becomes more unprofitable and more expensive to fix as time goes on.
  • After the 2013 debacle, NO mainstream political party can ever afford to back Full Fibre again.
  • Abbott's outstanding success was making the issue of Fibre Broadband a toxic issue & instant election loser.
"No going Back” or “Throwing Good Money After Bad” - what happens next?
  • Turnbull's “Four Horsemen of the Apocalypse” apply, in spades, to his own MTM-NBN. It’s started to unravel already - running over time & under spec… We’re onto the fourth radical increase in ‘cost'
    • The ALP may suggest “can’t undo this mess”, but the MTM is already a massive failed project that cannot be retrieved.
    • Overbuilding with Fibre is the ONLY commercially viable outcome: Do NOT throw good money after bad, it’s a waste
The commercial risks of the MTM-NBN and failure scenarios:
  • Crystallising Losses, forcing them onto the Federal Govt Budget and for the taxpayer to immediately pick up the tab. The ALP will be blamed for Turnbull's deliberate financial destruction of the NBN.
The importance of the AusSat story:
  • How in a decade was ten-plus billion transferred from the taxpayer to private hands, as “the only rational thing to do”?
    • The current value of Optus with 30% of the market should be around 25% of Telstra's $65B.
  • AUSSAT accumulated $600M in debt because it was wasn't allowed to fully compete with Telstra and was deliberately undercapitalised, forced to load up on debt at record high rates.
    • AUSSAT was given away for its debt,
    • plus a Free unrestricted Telco license,
    • plus 5 years competition-free, the "Duopoly".
Who really needs Fibre?
  • A: The Bush.
  • Key Communications Policy questions are:
    • Funding NBN directly by subscribers and SuperFund Bonds, Fibre-to-the-Farm and a Shared Mobile Blackspot & Public WiFi network
Did the ALP ‘throw’ the 2013 election to permanently solve the Rudd-Gillard schism?
  • We may never know.
  • The NBN was an election winning issue, yet the ALP didn’t leverage it.
  • All Turnbull's arguments were easy to knock over:
    • "How much will it more cost ON BUDGET?"
  • Nor did the ALP ask the PBO to cost “Direct Action”, NBN or “Border Protection”.
  • Why? “too hard” or “too true”?
PBO claimed, in a personal communication, "it would've cost millions" to cost both plans.
  • PBO is tasked ONLY with reporting "Budget Fiscal Impact", implied is "over 4-year forward estimates".
  • the ONLY Budget Impact of BOTH plans was the interest paid on Equity and Revenue received back.
  • To 2020, economic impacts of both NBN plans would be very similar.
  • Trivial, as in 1 days' work, to create comprehensive spreadsheets of both plans for Budget Fiscal Impact.

Either the PBO staff are incompetent, inexperienced or were running an Agenda.

Background pieces:

20+ years of OTC financials. What Telstra doesn’t do with Marketing & Technology:
  • for 3 decades OTC rode Moore’s Law and
  • dropped real prices of phone calls every year,
  • making increasing profits and gross margins…

Malcolm Turnbull 'regrets' staffer's outburst in email to technology blogger
Oliver Milman, The Guardian. 07-Aug-2013
A nice contemporary report of my exchange with Stephen Ellis.

Comparing Plans.

Capital Expenditure (CapEx):
                $20.5B (28% FTTP, 65%? FTTN, 0% HFC),
versus $37.4B for 93% Fibre
Magically, Turnbull was saving $17.4B out of the $17-$18 Fibre Rollout sub-Project
[ONLY FTTP was to be changed & Telstra still got paid per service.]
yet he was still spending ~$5.5B on Fibre and $8.1B on FTTN.
That’s $13B-$14B out right there [pretty much the 1st revised figure]
Peak Funding:
                $29.5B “and not a penny more”
Notable because of the ZERO IRR, FTTN needed to be Fully Funded by Equity with NO additional borrowings.
Breaking this promise or premise is very, very bad news for NBN Co.

How will they still be “more affordable” but earn extra money to gain & payback private sector debt?]
versus $44.05B, of $30.4B Gov Equity + $14B commercial debt,
due to be raised at a time when they’d have strong cash-flows and an established P&L + Balance Sheet.
Internal Rate of Return & Positive Cash Flow year:
                “above zero in real terms for all ‘new’ investment”. i.e. NIL RETURN by design.
                A very poor public investment.
                Cash-flow positive FY 2020-2021 [with 1st rollout (of two)]

versus 7.1% on $30.4B, or $65B over 30yrs (notional $$$, not NPV)
                Cash-flow positive, IIRC, FY 2019/20. maybe 2018.

I’d already deduced the cost-per-premise of “passing” & “connecting" premises (~$1500 & $1100 per-premise install, most to Telstra to purchase ‘lead-in’).
[The NBN Co data released in April 2013 was $2450 & $1100, IIRC.]
There was also a Common Cost for all networks, incl FTTN: the “transit network” plus POI’s and the Software & Systems.
Turnbull always had to pay for the Telstra “disconnection” on top of the supposed $900/port connection cost, yet it doesn’t seem to be counted.

Estimate of cost-per-premises and common costs.

Reverse-engineered spreadsheet of Turnbull’s ‘Policy’.

Other Points on early analyses of Turnbull 2013 Plan:
  • The early year ARPU’s ($22) were less than the minimum $24 AVC charge (12/1Mbps).
    • The ULLS charge was $16.
    • How could they do this when they never noted their changed AVC fee schedule?
  • There seemed to be a swap of OpEx for CapEx (rent not buy). I couldn't properly estimate it.
    • Despite this, the OpEx for the known-to-be-expensive Copper was much lower than for FTTP, known to be much lower OpEx.
    • How could that be???
  • There was an implicit, never explicit, implication that “full Fibre (GPON)” was the final end-point of BOTH LNP & ALP policies.
    • As it was never stated, only implied by ‘the listener’, there is technically no ‘broken promise’. Just misleading & deceptive conduct.
  • The acknowledged deliberate waste of the investment [never included in the spreadsheet]:
    • 50% CapEx “reuse” for upgrade from VDSL to Fibre.
    • This was around a $4B charge, half the 8.9M premises at $900/premise CapEx figure.
  • The FTTN design silently & deliberately shifts a substantial cost to subscriber.
    • Over 25yrs, around $500 up-front + $50-$100/year, yet this is never accounted for, not even in the latest two financial plans.
    • The supply, install and maintenance of the CPE (Customer Premises Equipment) - the VDSL2 modem - is now the subscribers cost.
    • These non-standard VDSL2 modems also had significantly lower functionality:
      • single VLAN-port, not 4-port as in Fibre NTU, and
      • no obvious “Multicast” capability. [This should be the ‘sleeper’ part of the Network. Scheduled b’casts over Fibre at multiple bitrates.]
  • I failed to notice & cost an implicit major Network Upgrade. The “25Mbps you all get 25Mbps, upgraded mostly to 100Mbps by ??? (2018)”
    • I had tried to estimate the number of Nodes needed.
    • They increase quadratically with access rate.
    • Twice the rate == FOUR times the number of Nodes, at $30k/node…
    • Keeping the "Going Faster” promise would turn a dodgy project into a financial quagmire.
  • I also failed to notice & cost two large Telstra related charges:
    • the additional cost of Acquiring the Copper + Pillar (turns out it wasn’t a payment to Telstra, but a reduction in costs they had to bear)
    • the cost of Software & Systems Upgrades and Network Design. Turns out to be many billions.

No comments:

Post a Comment