ACCC letting NBN descend into retail “market failure”, says Macquarie


news Macquarie Telecom has accused the ACCC of letting the NBN broadband market descend into “market failure” due to the lack of retail competition, in the wake of comments made by the competition regulator that it was not planning re-examine its controversial decision to set the number of points of interconnect with the NBN at 121.

In 2011 the ACCC set the number of points at which retail broadband providers could interconnect with the National Broadband Network at 121.

The move received extremely heavy criticism from the Government, from industry and from the NBN company itself, as it forced small to medium-sized ISPs to set up infrastructure connecting to each of those points if they wanted to compete for NBN customers on a national basis. This model strongly advantaged major companies such as Telstra and Optus, which already dominated Australia’s broadband market.

Internode founder Simon Hackett described the model as “insane” and said it would lead to a rapid consolidation of the the broadband market.

Over the past several years, that prediction has come true, with many medium-sized ISPs such as iiNet, Internode and others having been acquired, and only three major players — Telstra, Optus and TPG — still remaining in the market. Their closest competitor is the smaller M2.

It is believed that the ACCC commented in the Communications Day industry newsletter this week that it would not re-examine the situation.

In response, Macquarie Telecom national executive of industry and policy, Matt Healy, said the ACCC had little idea what was really happening in the competitive broadband market.

Healy said the ACCC’s view on the emerging market for aggregation services on the NBN was “premature at best, wrongheaded at worst, and betrayed a fundamental failure to appreciate what was actually happening in communication markets

Healy said the 12 Points of Interconnect decision had led to “market failure”, with customers being migrated to the NBN actually having fewer choices of provider than they would have had on existing infrastructure such as ADSL2+ broadband.

“The Commission needs to consider what is actually going on in wholesale aggregation markets to ensure this does not become a crisis in 2016 as more customers are forced to move to the NBN,” said Healy.

“The Commission is relying on statements of intent from companies who say they will offer a full suite of wholesale aggregation services at some point in the future, not responding to market reality. Even those assurances are too far into the future to protect competition as more and more customers move to NBN in the coming months,” he said.

Healy said Macquarie Telecom had from the outset of the NBN reforms championed a strong independent wholesale aggregation model as the best long term industry structure. This would help alleviate the 121 POI issue.

But for a variety of reasons, the market had not yet developed as envisaged, Healy said.

“There are two circumstances the Commission should be concerned to properly understand – the transition while the NBN is being built, and the on going industry structure after it is completed.”

“The transition right now is clearly in serious trouble, and the situation many warned about as early as 2010 – that the build out period was a time when incumbent players would be motivated to squeeze out retail competition – seems to be playing out. The longer term industry structure emerging looks like one dominated by a handful of vertically integrated companies, which is not a formula for an independent wholesale market.”

“At the very least the Commission needs to take a proper look at the current market and move to put in place some interim backhaul arrangement, or encourage NBN to do so pending actual release of products so far only talked about by wholesalers.”

“Instead, the Commission’s comments suggest it has lost touch with what is actually happening in communications market at the worst possible time,” Healy said.


  1. Extending NBNCo reach even further is going to help with infrastructure competition?

    Compare the AVC with competiive infrastructure pre-NBN or CVC vs backhaul rates today (including submarine cable connections). What a policy.

    Turnbull to blame?

    • Richard, Infrastructure competition for last mile service delivery is downright wasteful, that is why we dont have multiple PSTN’s, multiple gas mains, multiple power lines or multiple water/sewage mains!

      When are you neo-con’s going to accept reality?

        • I do, I used it as an insult as the US neo-con’s were behind the two Gulf wars.

          To be fair you are more of a Neo-liberal.

          • @do used it as an insult because I was somehow behind the two gulf wars (or in some way similar to those that were)? You people are insane.

          • Oh c’mon Richard, surely you are aware of guilt by association?

            Anyway, it’s pointless debating you, you frequently take contradictory positions and clearly fail at understanding the economics of nation building infrastructure – clearly you are happy living in your little “Corporate bubble” world dreaming about imaginary constructs like “free markets” and “infrastructure competition”.

          • @do I’m guilty because I associate with those that started the two gulf wars? Who?

            Keep digging.

          • So Richard you honestly believe your conservative views have nothing in common with those of the Republican conservatives in the USA?

            Wow, some serious disconnect from reality there mate considering your favorite Australian party has essentially become the Australian branch of the GoP!

          • Once again more contradictions and innocent victim sobs from Richard… does anything ever change FFS?

            Apparently not….

            So let’s get this straight Richard, it’s not ok for Derek to throw you into the one basket of conservatives, along with the Dubbya’s and Abbott’s of this world, but it was quite ok for you to throw all of us into your totally obscure resurrection of the Brisbane Line basket?

            Seriously, although I do not agree with your views whatsoever and respect your right to voice an opinion, the one thing I never actually believed you were was a hypocrite.

      • I like he had a go at the most obviously unrelated part of your statement and then ignored the parts where he has no leg to stand on.

        • @ho perhaps he should have stopped before exposing his comical misunderstanding.

          Also incorrect to assert competition isn’t a policy objective in natural monopolies. Failures and cost is acknowledged.

          Here the proposal is to extend further a wasteful, inefficient GBE even further beyond the last mile. Again lets look at bandwidth cost pre/post NBN in the fixed internet space.

          • “lets look at bandwidth cost pre/post NBN in the fixed internet space.”
            Finally coming down about 2 years prior to Labor being elected with thanks in part (at the very least least) to talk about NBN, cheap as shit during the FTTP rollout and then skyrocketed after Malcolm Turnbulls Mess and certain ACCC rulings were introduced.

        • It’s always the way Hotcakes…

          What posters are better to do is hone in on, and simply comment upon, the crux and don’t elaborate (which I am also guilty of).

          Because once you give someone who has no basis in facts and who is here not for meaningful and friendly communication but to push his own ideological and/or personal financial agenda/crusade… a reason to divert from the facts, you are presenting to easily shoot them down, they will of course ignore the real issue/facts and do as Richard just did here…

          • Yup definitely noticed that over the years in most ‘discussions’. Sadly, I’ve also noticed it tends to work – most seem to be ‘won’ over by such asinine comebacketry. Yes, that is now a word.

          • Richards biggest flaw is that his arguments are based on ideology, not facts.

            It’s also why his points come across as being so disjointed…he doesn’t build an over arching case, he just snipes at individual points with “facts” drawn from reports/CBA’s/reviews where Malcolm has paid for a specific outcome.

    • @R I haven’t the foggiest Idea what part of the article you’re trying to respond to? much less your point (or why MT needs mentioning) :(.

    • Richard what happen to Turnbull policy to allow Optus and Telstra to complete with there HFC?

      • @jk the mtm today doesn’t encourage infrastructure competition, actually overbuilding competitors. You might like to ask Turnbull why.

        • But it’s the policy you could have writen. So I am asking you. Turnbull was all for infrastructure competition before the election. But then we have seen how he has tried to stop TPG.

          • @jk nice misquoting, even then I still don’t see how you’d then extend that to execution. Perhaps you’ve neither written nor executed a business strategy?

          • Yes what is that business strategy. Step one pretend to be for infrastructure competition (yet it’s a policy you could have writen) pretend to have a fully costed plan $29B and claim it all could be done in 3 year.

            Then find out your own figures on cost and time are wrong but where according to you ambitious even the cost for FTTP $90B wrong. Over estimated the cost of FTTP and then under estimate the cost of MTM in the SR with figures redacted lacking the transparency of a policy you could have writen.

            When a redacted CP16 with even more cost blowouts with a counter factual which isn’t a counter factual which also claim the best estimates are in the SR when you compare the S1 $15B cheaper S2 $8B cheaper S4 cost more lol but then you dismiss it claiming it’s not apples vs apples even though the CP16.

            You start claiming the upto $94B and y26-28 is labors new FTTP costing which it isn’t. You didn’t think it was strange that both had the same cost blowout. But as morrow has said to the seanate that’s if they had to stop and go back to FTTP which then explains this not counter factual with the same cost blowouts and increase till due to the delay of switching back.

            You claim that the NBN won’t make revenue due to cross sub but then Turnbull looking at charging other infrastructure competition $6. But that it due to not being able to supply faster speeds.

            You are now in other threads taking about the SOE. But in the SOE NBN needs to deliver “at least 25Mbps” which according you from the past it can’t even do that.

          • Yes thanks alain…

            CP 2016 from the cost blow out and timeframe extension kings (not to mention the HFC kings…LOL) is not a great doc to refer to.

            But as it’s it’s all you have, go your hardest son…

          • @ Richard please.

            Jason may have used different words but the gist is there, so enough of the desperately deflective pedantics.

            And as such I agree with him… why don’t you answer considering…

            Richard: “The Liberal policy document includes everything I’ve been posting about for years; use of existing infrastructure, priority for areas where market failed, ACCESS TO INFRASTRUCTURE FOR COMPETITION, review of NBNCo past activities, CBA, etc It is almost as if they commissioned me to write it…”

            Use of existing infrastructure (copper to be replaced by copper & Optus HFC rooted possible overbuild/switch off) = FAIL.

            Priority for areas where market failed… WHAT.. Markets fail? Noooo…ROFL.

            Sorry after that guffaw –

            Priority of areas where markets failed. What like poor little backwater market failures such as Warringah & North Sydney (which when chosen, were/are Abbott’s and Hockey’s electorates) = FAIL.


            Access to infrastructure competition = FAIL (you just admitted so)…

            Review of past NBNCo activities (by the likes of Ergas & Co – Lib donators/sympathisers and FttP critics) = FAIL

            CBA (oh the independent CBA to be carried out by the PC ) = FAIL.

            100% FAIL Richard, well done.

          • Breaking news, the UnOz is reporting that copper remediation costs have exploded from the SR predicted $2k per node to $26,115 per node to fix Telstra’s copper lines to ensure 25mbps minimum speeds.

            At an average of 178 premises per node, the bumps up the CPP by another $146 putting the real FTTN cpp at $2,246!

            Compare that to $2,700 for FTTdp and the question needs to be asked, wtf are we not just spending an extra $450 per premises for a far superior not obsolete solution?

          • So basically we have the following scenario:

            MtM costs going up


            FTTP costs that where going down

            Just goes to show the Libs so called economic credentials are mere propaganda!!!


            The company building the National Broadband Network is facing a ten-fold blowout on the cost of fixing the copper network it bought from Telstra.

            According to The Australian newspaper, confidential documents show a $640 million bill to repair or replace parts of the lines to ensure the network will deliver the promised high speeds.

            The NBN expects to spend $26,115 per node to fix the copper connections – up from the $2685 assumption in a 2013 strategic review prepared after the Coalition won power.

            Labor’s communications spokesman Jason Clare said Malcolm Turnbull had made a mess of the NBN as both communications minister and now prime minister.

            “He’s doubled the cost, doubled the time it will take to build it and now we find out the cost of upgrading the copper has blown out by almost 900 per cent,” he told reporters in Canberra on Thursday.

            “Most people don’t get promoted for blowing their budget by almost 900 per cent, most people get the sack.”

          • Tim
            Well reality/alian according to his previous post cost reductions in FTTP wireless and sat is the cause of the $15B cost blowout. So this cost 900% increase must be saving money to him.

          • @do great post as is the list of those that fell for it. The leaked document is background for the CP16 figures (experience gleamed from trials).

            Unsurprising failure to comprehend, The Australian source is. Funny the fanboys didn’t dismiss the source (continuing insight).

            But continue…

  2. What we are experiencing as we transition to the NBN is the almost complete destruction of the healthy ISP competition environment. It’s as if the past 15 years of progress, fruitful consumer journeys through research and blossoming of excellent smaller ISPs never happened. Presided over by the so-called Australian Competition and Consumer Commission. This agency has clearly failed in its mission.

  3. Maybe if the ACCC had the foresight to let NBN build their 14 POI’s to begin with, and eventually roll out the other 107 as the number of people connected to the NBN increased, this wouldn’t be such a complete and utter farce…


    • @Josh I don’t think you’re quite comprehending the why’s and wherefores behind POI’s ;)

      POI = point of interconnect. It’s a critical part of the back haul network so you can’t just pick 14 then increase the number later to get to 121. How many you have fundamentally changes your entire fibre back haul infrastructure and design.

      That is where RSP’s connect to the NBN network to get access to their customers. (where Wholesale meets Retail)

      • I’m aware of their purpose, but I also know that interim POI’s such as 2ULT exist which once a permanent POI that covers that area is complete, users are migrated onto.

        • I think @Josh is getting at the fact that the transition period between when POIs come into existence with vastly less than their 100K or so connections and when POIs are fully populated is a bigger burden on smaller RSPs.

          Smaller RSPs need a closer-to-fully populated POI to achieve critical mass which means they have to wait years between the time a POI is first commissioned and when they can make a profit. That’s a tough ask on capital-starved small fry.

          A small RSP’s world would be a vastly different place if the rollout had been designed to completely populate a single POI, then move onto the next one.

          For political (and possibly technical and logistical) reasons that is never going to occur, so how can that be simulated?

          @Josh suggests one option. A variant might be to offer at-cost inter-POI transit until a POI achieves a certain connection density then re-price that transit to at-market (or withdraw it completely).

          • I’m with you now! That is why I think ppl when they did the maths said the small fry actually need to be quite large with 250k customers to weather the storm and be able to compete.

          • Yep @MarkD. A small RSP’s world would be a vastly different place if it was done like this, and CVC charges wouldn’t be quite as big of a ballooning problem as it is now.

            Unfortunately, the ACCC messed this one up bigtime, to put it politely!

    • A better question is, What was the logic behind the high number of interconnects?

      RSPs should just have to link their networks with nbn’s at major capital cities in each state. Instead, they have to connect at 121 all over the place. This is because the ACCC thinks this will increase competition, or something. How was that supposed to work, anyway?

      • Gavin, it had nothing to do with competition and everything to do with protection of existing business interests and their existing investments.

        Imo this is entirely against the ACCC’s mandate and should have been overruled by the alp gov.

        • Telstra and Optus were lobbying hard so their fibre to T$ exchange locations wouldn’t become redundant. The more (they wanted 300) locations means the bigger advantage they have over every other player as well.

          Those two were threatening high court action over 14 POI!

      • Is there a rule that says RSP’s have to use all the POI?

        This is something I’ve never really understood about the low vs high number of POI arguments.

        Surely it’s better to have a large number of POI and a small RSP servicing a particular area by connecting to just a few POI in their area?

        What am I missing here?

        • If the RSP only operates within one city or region then yes, it only has to pay for the links to that city or region’s POIs.

          But 121 of them nationwide leaves small companies with national reach on ADSL, like Exetel and SpinTel to name two, out in the cold on the NBN and effectively prevents small ISPs from growing market share. Imagine if Exetel could not grow beyond Sydney. The fledgling company just would not get anywhere to threaten the big players into keeping pricing and value keen. It also would not get the cheaper wholesale agreements that would allow the company to offer more competitive plan pricing.

          It’s a glass ceiling and is an inexplicable decision by an agency that claims to facilitate competition.

        • 121 POIs: “What am I missing here?”
          When NBN first became available here mid 2014 it meant my sole option was to either sign up with Optus or remain on ADSL.
          Even now 18 months later I still can’t sign up with my preferred NBN provider unless I relocate into the nearby POI area 20K North of here.

          • It’s just “a” real world consequence Martin.

            It also means a small start up telco doesn’t need to come up with the funding to support an Australia wide organisation straight up, so it’s all swings and roundabouts.

    • Low (14) number of POI means it requires far less investment for a startup RSP to try and compete in the market on their own (ie they aren’t reselling someone else’s wholesale product). RSP could eve be state based and potentially viable.

      121 POI means you’ll need 250k customers to be able to afford the backhaul and CVC charges (estimated number from an industry expert who owned their own ISP at the time).

      Good for competition bad for Optus and T$ who own Fibre to just about all of what would be redundant exchange locations.

      • Ah ok. I think I had that wrong in my head. Is there any reading available about the POI and how it costs etc?

  4. “…that prediction has come true, with many medium-sized ISPs such as iiNet, Internode and others having been acquired…”

    Ummmm. In comparison with, say, TPG? how much wholesale data transmission did iiNet have? But wait!

    First, who acquired (and when did it happen) Internode and others? shhhhh! whisper: maybe it was iiNet? maybe it happened BEFORE NBN?

    OK. Now let’s look at iiNet’s wholsale transmission networks. “… this would be unlikely to materially alter the level of competition in the market for the supply of wholesale data transmission (wholesale transmission) services…” (Section 5, ‘Statement of Issues’ 11 June 2015 TPG Telecom Limited – proposed acquisition of iiNet Limited)

    There is a lot of good coming from the 121 PoI mandate, at least in WA: Sandgropers now have access to much more competition than previously, even if it does live on the Far East Coast. At least the competition is Australian.

    Macquarie’s whingeing seems to be just a complaint that it now has to face up to some competition from smaller ISP wannabes.

    • Gordo not sure what you’re on about!

      “maybe it happened BEFORE NBN?”

      Sorry but it didn’t. SH sold Internode due in part to figuring out even Internode wasn’t big enough to compete with 121 POI and the costs involved. That decision basically relegated Node back to garage status (that’s how big a deal/screw up that number of POI is) and I’d say SH wasn’t keen on doing it all over again after all the fighting of the last 15-20 years.

      “There is a lot of good coming from the 121 PoI mandate”

      feel free to let us know of any specifically (or just one)!

      “at least in WA: Sandgropers now have access to much more competition than previously”

      Say what?? how many ISP’s could you get services from now that you couldn’t before??

      “smaller ISP wannabes.”

      Like who exactly? and how are they benefitting from 121 POI (you know what a POI is right)?

    • what?
      The competition you are seeing at the retail level is a result of iinet, optus and telstra being connected to all the POI’s already, 90% of the other RSP’s are just reselling AAPT/M2.
      If you had 14 POI’s then all those small RSP’s would be buying services direct from NBNco and offering differentiated services.
      ACCC wanted competition in the *backhaul* market, which from the perspective of end users is just pointless, the incremental cost for NBNco to run 121/6 fibres back to a central point in each state is practically nonexistant. The cost for a small RSP to run 121 fibres is insane 10’s to hundreds of millions of $, and they want each RSP to duplicate that and everybody to have lots of fibre running at 1% capacity?
      ACCC could have just mandated NBNco buy carriage from the 121 POI’s to the 6 major POI’s and achieved the same outcome.
      Morrow is actually pushing for this now, to offer an aggregated service, be interesting to see what ACCC says about that.
      Particularly with M2 now offering a L2 aggregated bitstream service hauling all the POI’s back to the customers place of preference.

      • “…competition in the *backhaul* market, which from the perspective of end users is just pointless,”

        Do a search on Whingepool for “backhaul”, all forums.

        But I guess for some people, it really is not important for end users.

        However, using my recently acquired 20/20 hindsight, I do appreciate that some things could have been done better. Maybe a more planned approach to the NBN from Julia Rudd’s administration would have helped. Maybe John Howard’s administration could have been more helpful than it was. But then we don’t live in UnicornLand :(

    • Says there’s more competition in one paragraph (but doesn’t explain how) and then goes on to admit all small ISPs don’t stand a chance in the next paragraph.

      Doesn’t pause for reflection before hitting Post.

  5. @SImon M: “Say what??…”
    “There are 103 service providers that may be selling services over the nbn™ network in your area.” ( looking at my area.)

    “(you know what a POI is right)?” Last time I looked, PoI = Point of Interconnect.

    “Like who exactly?”
    … there’s no way I’m posting 99 addresses here.

    “Healy said the 12 (sic) Points of Interconnect decision had led to “market failure”, with customers being migrated to the NBN actually having fewer choices of provider than they would have had on existing infrastructure such as ADSL2+ broadband.” Para 9 above.
    What Healy misses is that there’s no way known we could have 103 (I)SPs seeking our custom on ADSL.

    You do know what a PoI is for, right?

    • You realise that multiple ISPs can connect through each POI. the lower number of POIs would drastically reduce the amount of backhaul and cvc charges an isp would face to provide a whole australia service.

      You could have thousands of suppliers at each POI.

      I would guess those lovely small ISPs are just reselling other providers.

        • I initially didn’t think it was a big deal on a commercial level (on a technical level it’s just dumb and wasteful), but it took me a while to see the full commercial implications and Simon was right about them.

        • Except that guy is full of shit. having a large number of pois and trying to serve all of those regions requires an investment in cvc and backhaul to each and every one of those POIs.

          He would be correct if you could change which POI a customer connected to and consolidate into a single one but it doesn’t work like that. Internode attempting to serve an australian wide infrastructure would be required to have backhaul to each and every POI or at least every single one that they have a single customer on.

    • Take the T$ ADSL/CAN network now each exchange building is acting like a POI. You want to access the customers behind that POI you pay for back haul to it or you resell a wholesaled product that comes with access (and has a single/redundant POI somewhere easier for you to get too).

      I’m sorry but how am I supposed to believe that connecting to just 14 locations across the country is harder and more expensive than than connecting to ~17 per state? Its not like you can pick or choose you either connect to them or you don’t get access to the customers behind the POI (redundancies aside).

      Want to be a state based ISP then connect to 2 locations vs 17 now (you couldn’t decide on 2 of those 17 now because a POI isn’t optional like an exchange isn’t with ADSL if you want the customers).

  6. Rod Simms has directly led to competition reducing mistakes.

    The gfc bank mergers is another, 3 & Vodafone. I dare say, if one were to look there would be a lot more. Was he in charge when bilo and franklins were eaten?

  7. Rather ironic that Simon is on the board of nbnco, the company that effectively caused the demise of his ISP business.

    • Yet he is the only board member that sold his shares before joining most of the other still have Telstra shares.

    • Trying to fix it on the inside rather than sit on the outside and complain about it. How effective that has been I guess only time will tell. It has meant he’s not been public commenting if nothing else.

  8. So speaking as someone who knows little about this whole POI business other than what’s been described in this and similar threads, how does an RSP like Skymesh fit in? They only serve specific regions, so does this mean that they are only hooked up to select POIs? Or is there some other reason for that?

    Otherwise, what’s to stop other smaller RSPs from cherry picking parts of Australia in a similar fashion? Admittedly still eating into their profits by way of not having the whole of Australia opened up to your service when only choosing to hook up to (say) 50 POIs, but it does still mean smaller RSPs can get a start, right? The 250k figure is based on serving the whole of Australia, right?

  9. 121 PoIs really is insanity. But it’s worth noting the intervention only happened at all because of the bizarre initial proposal – 14 PoIs in just seven capital cities, and NBN grossly overreaching on how much of the end-to-end network they control. Very low latency is just as likely to be a future requirement as very high speeds, so for a supposedly next-generation network in a country the size of Australia the 14 PoI plan was completely inappropriate. I realise from the plush NBN offices in Melbourne (where the entire state is within spitting distance) it doesn’t seem a huge issue.

    Apparently Telstra et al. wanted even more than 121 PoIs, so maybe it was considered a “compromise” but the solution that actually would’ve been appropriate IMO is somewhere between 25 and 50 PoIs (depending if they’re still in pairs), geographically dispersed and with backhaul competition added if necessary. I can’t help but feel if this were done, all involved would be content and ACCC meddling would have been avoided in the first place.

Comments are closed.