Structural Separation won’t go away, however much Telstra would prefer it to

[Piece I wrote in the Exchange newsletter]

Paul Chapman’s Ex Cathedra of 14 February 2003 pointed out some flaws in the precipitate decision to pull the plug on the inquiry into structural separation of Telstra. In my opinion, the flaws run much deeper than Paul set out.

Senator Alston’s intemperate little tirade in the AFR set out four major reasons for rejecting the separation:

  • Nobody else has done it
  • It’s all too complex (because the network is complex)
  • No-one would have incentive to upgrade the network.
  • It would destroy shareholder value and increase prices for consumers

I argue that all of these are wrong.

Nobody else has done it

Leaving aside the point that if it was a good idea, there isn’t any particular problem in being first, it simply isn’t true that incumbent telcos haven’t been dismembered by their shareholders (private or government). Many have separated out their mobile arms, there is the famous case of the break-up of the US Bell system, and the degree of accounting separation mandated by European regulation requires almost as complete separation of functions (and consequent transparency of wholesale pricing) as Senator Alston tells us “can’t be done”.

It’s all too complex

An interesting point that Senator Alston shares with the Communications Union (strange bedfellows)! The simple fact is that, while many “complex” services are provided by the network infrastructure, it is still relatively easy to define a boundary: if the network provides it, it goes in the network company, and is wholesaled to all comers including the new retail company; if not, not. The many services (such as voicemail) that can be provided either way, can be provided either way.

Incentives to upgrade the network

Let’s assume that we break up Telstra into a successor retail company (let’s call it Telstra) and a network company (let’s call it Telecom Australia). Telecom has enormous economies of scale, and a near monopoly of local and long-distance network provision, sells to all comers (as above) including Telstra, and has a regulated rate of return on capital plus a USO contribution – read an almost guaranteed income stream above its WACC for any capital expenditure it makes so long as it sticks to knitting core network products. Do we think Telecom won’t invest in the network?

Shareholder value and prices

This is the biggie. It is true that Telstra and all of the banks that made submissions to the inquiry said that shareholder value would be destroyed, but even a little thought tells you that these are not disinterested parties in this game. Telstra clearly hates the idea, while all of the banks who submitted are hoping for a slice of Telstra 3, the only big payday any of them can see on the horizon, and aren’t going to royally PO the two decision makers for who gets the fees on that deal. The fact is that in all these shareholder value models the answer you get is entirely dependent on the assumptions you put in, and any good modeller can get the answers he or she is asked for within a very wide range of valuations (believe me, I’ve done a wardrobeful of these).

Despite vast apparent sophistication, the models submitted all boil done to one assumption: the actual process of separation would cost a lot, would essentially just divide Telstra’s earnings between the two successor entities, and the competition benefits wouldn’t be substantial or long-term (otherwise they would eventually create increased shareholder value more or less whatever the costs). As already noted above, the fact that many privatised telcos around the world have indeed broken themselves up to create shareholder value indicates that it is relatively easy to alter these assumptions. MMO2 was broken out of a fully integrated British Telecom in under two years and at relatively low cost, and a group of venture capitalists subsequently submitted a bid to create more value by breaking the network out of the remainder.

If you change the balance of these assumptions only a little, there is no net present value loss of long-term cashflows, and thus no need for increased prices (or compensation to shareholders). The key elements of a value-creating break-up are keeping the one-off costs down – primarily ensured by building the achievement of this into the compensation packages of those tasked with the job – and getting some real competitive benefit in the wider market.

My prime assumptions for a model that would deliver increased shareholder value and benefits to customers are very straightforward:

  • Telecom is a regulated utility with a good if not spectacular return, and almost risk-free debt profile, a substantial fraction of Telstra’s current revenues, and is set fair to replace Commonwealth bonds as the bellwether of the market. Whether in private or Government hands, it is subject to substantial direction of its affairs, which costs us some discount on its share price, but would still be a “must-have” to most investment funds.
  • The new Telstra has ALL of Telstra’s current revenues, but a higher cost profile (as say half of its earnings are now Telecom’s). It has an almost unregulated monopoly in many parts of the market, and its (still) multi-billion dollar profits are made with a low capital and debt profile. It will be a stellar performer in the next bull market, and its managers, freed from the horrendous complexities of the Capital Planning cycle and all that engineering stuff can work hard to make it even better.
  • The other retail players can finally get real discussions with Telecom on planning horizons and wholesale pricing, knowing they are no longer opening the kimono to their biggest competitor. Telecom, as a result, finally gets better forecasting data to plan its own operations than has been available to it for more than a decade.

Anyone really prepared to believe with Senator Alston that this future is unachievable?