Monday, December 14, 2015

FTTN vs FTTP (2): Fast, cheaper, sooner vs Do it once. Do it right.

NB: Update 7, 8 Jan 16
There is a calc error** that I have amended for in the below figures. See more here. Revisions indicated with **.
==Original post
Last week,  said on 7.30 Report: Leigh Sales posed an NBN question
Leigh Sales and Malcolm Turnbull, PM"So why then do you continue to back a broadband network that relies on a decrepit copper network?" ie FTTN

The Australian PM, Malcolm Turnbull in response said "under the approach we are taking to the NBN, we will get the network completed six to eight years sooner than it would be under Labor's proposed method and $30 billion cheaper or at less expense to the Government, which makes broadband more affordable."

In a previous post, I compared FTTP and FTTN assuming the networks were available at the same time. Here I would like to compare where there is a delay in the availability of FTTP at a household.

Model
I created a model to compare the costs and benefits of FTTP and FTTN, and accounted for the delay.
The CAPEX, OPEX and Revenue come from recently leaked NBN Co. estimates to replace ailing Optus HFC cable with either FTTN or FTTP. All these figures are generated for a single household. To get to a national perspective, multiply the outcome by 2.5M households ie 10M total households by 25% FTTN rollout.

I use four comparisons:
  • sales generated over time
  • cash generated / expended over time
  • profit or loss generated by NBN Co over time, and
  • GDP impact ie Sales (customer benefit), OPEX (supplier benefit) and NBN Co profit/loss (shareholder benefit).
The results of the model, show two time frames, after Year 10, and Year 20 for a variety of delays between arrival of FTTP - Yr 0 (at the same time), Yr2 , Yr 4, Yr 6, Yr 8. The last two scenarios being the ones suggested by the PM. The Year 10 and 20 scenarios assume that revenue, and OPEX are consistent over the 20 year timeframe.

FTTP Revenue: $785pa OPEX: $350pa, CAPEX: $4,400
FTTN Revenue: $640pa OPEX: $570pa. CAPEX: $2,100
(NB:** Now revised to FTTN OPEX $440pa Revenue $490pa - amended figures below.)
These figures come from the NBN Co, Optus HFC replacement slides from Delimiter, as discussed in the earlier post.

Output of the Model
All figures in thousands of dollars. Negatives/losses in brackets. Some rounding to whole numbers.
Figures derived from model linked above. Model is in OSX Numbers, and translated in xls.
Revised** numbers only affect FTTN figures. FTTN Revenue / Opex down $150pa, $130pa. But some rounding errors when adding together.

FTTP commences at Yr 8: (Old/Revised**: not amended if diff < 10%;^)
Yr 10FTTNFTTPYr 20FTTNFTTP
GDP12.8 / 10**6GDP26 / 20**20
Sales6.4 / 5**1.5Sales13 / 10**9
Cash(1.4)^(3.5)Cash(0.7) / (1.0)**0.8
Profit/(Loss)(1.4)^(1)Profit/(Loss)(0.7) / (1.0)**1.7

FTTP commences at Yr 6: (Old/Revised**)
Yr 10FTTNFTTPYr 20FTTNFTTP
GDP12.8 / 10**9GDP26 / 20**23
Sales6.4 / 5**3Sales13 / 10**11
Cash(1.4)^ (2.7)Cash(0.7) / (1.0)**1.7
Profit/(Loss)(1.4)^(0.5)Profit/(Loss)(0.7) / (1.0)**2.5

FTTP commences at Yr 4: (Old/Revised**)
Yr 10FTTNFTTPYr 20FTTNFTTP
GDP12.8/ 10**12GDP26 / 20**26
Sales6.4 / 5**5Sales13 / 10**12.5
Cash(1.4)^ (1.8)Cash(0.7) / (1.0)**2.5
Profit/(Loss)(1.4)^1Profit/(Loss)(0.7) / (1.0)**3.4

FTTP commences at Yr 2: (Old/Revised**)
Yr 10FTTNFTTPYr 20FTTNFTTP
GDP12.8/ 10**15.2GDP26 / 20**29.1
Sales6.4 / 5**6.3Sales13 / 10**14.1
Cash(1.4)^(0.9)Cash(0.7) / (1.0)**3.4
Profit/(Loss)(1.4)^1.7Profit/(Loss)(0.7) / (1.0)**4.3

FTTP commences at Yr 0: (Old/Revised**)
Yr 10FTTNFTTPYr 20FTTNFTTP
GDP12.8/ 10**18GDP26 / 20**32.2
Sales6.4 / 5**7.8Sales13 / 10**15.7
Cash(1.4)^ 0Cash(0.7) / (1.0)**4.3
Profit/(Loss)(1.4)^2.6Profit/(Loss)(0.7) / (1.0)**5.2

Analysis
What these figures show, is that:
- FTTP is much more profitable for NBN Co. than FTTN, even when running eight years later. Profit in this case is Revenue less OPEX less Depreciation, where FTTN is depreciated over ten years, and FTTP is depreciated over 25 years. FTTN never makes a profit, nor generates positive cashflow even after 20 years. There is a cash benefit relative to FTTP though. Roughly $2100^ is saved in the case where FTTP starts eight years later than FTTN. But by Yr 20, FTTP has generated $1500 ($1800**) more cash, for the Yr 8 scenario. (Unchanged by Revision**).
- FTTP generates as much GDP as FTTN, so long as FTTP is no more than four (eight**) years later [Revised **]
- the PM is correct. when FTTP is six or eight years later than FTTN, greater customers sales and GDP is generated by FTTN. But a four year delay, makes the outcome by Yr 10 and thereafter almost identical. (Revised**: FTTP generates as much GDP as FTTN, even if eight years late at Yr 20).
- FTTP is generally superior to FTTN whenever the gap between availability is less than four years.
- none of the FTTN scenarios is profitable for NBN Co.

The Yr 10 - 20 figures must be more suspect for FTTN revenue, since the likelihood of revenue stability after the ten year useful life of the FTTN network (especially relative to the FTTP network) must be suspect. The FTTP is upgradeable for the foreseeable future, and likely to Yr 20.

Next
The next step is to compare the $30B statement. Can the PM's statement that $30B is saved by using his MTM (ie FTTN, HFC) rather than FTTP really amount to a $30B saving? That will have to be addressed in the next post.


2 comments:

Anonymous said...

Based on US deployment costs of roughly ~1300USD, the estimated FTTP cost is too high by at least $1500.

Martin Hamilton said...

Re Anonymous on January 5, 2016 - Of course it's too high. Someone is shonky somewhere with the NBN. No basis in reality has been found for the Coalition's claims of the costs of a FTTP network. But we are not to let facts of rollouts on the ground get in the way of politics.