Ausnet, and the case for a rethink on who pays for new grid connections

In Brief

Network owner Ausnet has been an OK investment. Its electricity distribution business will see prices grow at 1.7 per cent per year over the 2016 -2020 period.

There is no prospect of substantially reducing the distribution open, which is 40% of the revenue, and 43% of the distribution capex is replacement and only 11% is new connections.

As such tariff reform can’t drive much of the spending. One change that we think would be worth looking at is making all new connection capex in the form of customer contribution. That is, it’s paid up front by the property developer and as such never enters the regulatory asset base.

Ausnet would then only earn a maintenance (opex) cost on the assets. This would push more focus onto property developers to optimize their connections capex and force a proper consideration of distributed generation at the time the network is built.

Ausnet’s security price has grown from $1.46 in 2007 to $1.73 in 2017, up about 21% in total or 2% per year. AST was forced by the banks to degear during the 2009-2010 and as a result dividends for the past five years have been lower than in 2007-2008 period, flat, and just resumed growth (we note there is a new chairperson). So the gross yield’s been about 5.5% for most of the past 5 years.

Figure 1: AST DPS. Source: factset

Actually though, even ignoring tax, the yield really isn’t that good in aggregate because there have been two equity raises and some of the vast capital expenditure has been funded by dividend reinvestment. So the net yield has been <3% over the past five years.

Figure 2 AST, Cumulative and net cash flows to equity holders. Source: Company, Factset

RAB has doubled in ten years and debt’s up as well

AST’s main issue has been funding growth in its regulated asset base, particularly in electricity distribution (last mile wires and poles).

AST owns Victoria’s electricity transmission, one of its five electricity distribution  businesses and one of the gas distribution business. Over the past 10 years the total RAB and Ebitda have progressed nicely, with the electricity distribution business the biggest contributor to growth.

However financing that $4 bn of extra assets hasn’t been that easy. Essentially if we look at the cash flows the capital expenditure has taken up all of the operating cash.

RAB 10 yr annual growth rate
Transmission	4%
Electricity distribution	11%
Gas distribution	5%

Figure 4 RAB compound growth. Source: Company

However financing that $4 bn of extra assets hasn’t been that easy. Essentially if we look at the cash flows the capital expenditure has taken up all of the operating cash.

Figure 5: AST cash flows. Source: company

How AST funds its distributions

AST determines how much cash it can pay to its security holders essentially by by subtracting maintenance capex from its operating cash flow. Maintenance capex broadly equals net regulatory depreciation.  Of the $830 m of 2017 capex for instance about $330 m was treated as maintenance capex.

The growth capex is then financed mainly from debt and reinvested dividends.

Its perhaps easiest to see this from the company graph on the topic.

Figure 6 Dividend and capex funding. Source Company

The key point to understand is that for this model to be stable the debt:rab ratio has to be constant or falling. This in turn requires the return on rab (ebitda – net interest) to be constant. Under those conditions its reasonable to borrow to fund part of the annual growth capex.

Capex and opex in electricity distribution have doubled

AST’s electricity transmission and gas distribution performance would not raise too many eyebrows but electricity distribution is another story. For both gas and electricity volume per connection has declined and gas has seen an absolute decline in consumption. Still it’s the growth in electricity RAB that is the main driver of many things.

The following figure shows how capex and opex have more than doubled even though volumes are flat over the past decade.

Figure 8 AST Electricity distribution segment. Index of volume, capex and opex. Source company

The two main factors that have driven this doubling of annual cash outflow are:

1. The rollout of communicating meters in Victoria was very expensive relative to what its costing in other States.
2. Significant capex and opex have been spent on things such as vegetation management following the severe bushfires of 2009

Still ,what we see for Ausnet’s electricity distribution business in Victoria is more or less what every electricity distribution business is going through. Typically asset bases, capital expenditure and opex have doubled over the past decade but volumes have been flat.

Over the next five years electricity distribution opex will rise, capex constant

This is really the key point. Opex is rising despite flat volumes and capex is constant. AST is allowed to raise prices by about 1.7% per year (perhaps more if they win a court case) and when coupled with higher generation costs the “centralized” sector can’t do much to compete against the falling costs of distributed generation.

What could a different tariff model change

The return on capital and return of capital add to about 55% of total revenue for Victorian electricity distribution with annual opex a surprisingly high 40%.  The chart below is a summary of the AER’s allowance for the next five years.

Figure 9 Ausnet electricity distribution regulatory revenue. Source: AER

No tariff reform is going to get the opex changed and only an asset abandonment would change the regulatory depreciation.

If we look at the capex that drives the RAB growth over the next five years we see:

Figure 10: AST electricity distribution capex, Source: AER

We subtracted customer contributions from new connections. The point is that 43% of the capex is replacement. 19% of capex comes from augmentation and that could conceivably be impacted by a change in tariff method although we have our doubts.

11% of capex comes from new connections. In our view one reform might be that all connection capex be in the form of customer contributions, paid up front, and therefore not entering into the RAB. This would encourage a focus on efficient design of the network at the time it is constructed.

Wires and poles 101

Have wires and poles been a good investment? Will they continue in the future as in the past?  Ever wondered about the relationship between DORC and RAB? The AER uses the “Vanilla WACC”, is that the same as “textbook” WACC? In this section we delve into the wonderous world of wires and poles and their financing using Ausnet [AST] as a case study , so grab a beverage and a digestive biscuit and let’s dive in.

AST is a stapled security. As far as we know Macquarie Bank was the first to employ this concept when it spun Envestra  out from building materials company Boral back in the 1997. A stapled security consists of a trust and a company indivisibly stapled together.

The trust does the financing and the company the operations. The trust raises money and lends it to the company. A trust is used because  (i) trust income is taxable in the hands of the recipient not at the trust level (ii) at the time it was easier to make cash distributions in excess of accounting profits out of a trust.

Wires and poles earnings are monopoly regulated

Wires and poles are regarded as a monopoly. Most of their revenue is determined at five yearly intervals by the regulator, the Australian Energy Regulator [AER] via a proposal and counter proposal method with appeals to the Australian Competition Tribunal [ACT] and even the Federal Court if either party is dissatisfied with the outcome.

These days the AER determines a revenue cap. That is an allowed annual amount of regulatory revenue. Consumer prices for the next year are “trued up” to allow for any over or underachievement.

The allowed revenue is the sum of:

Return on Capital = Regulated asset base * Vanilla WACC [1]

Return of Capital = Regulated Depreciation = Depreciation – Asset inflation [2]

Opex

Tax allowance

The  Regulated Asset Base [RAB] grows with inflation, grows with capital expenditure and declines with depreciation. It was originally estimated by engineers as the Depreciated, optimized capital base.  Inflation is generally uncontroversial but the appropriate allowance for future capital expenditure is debated.

The Vanilla WACC is socalled because there is no allowance for the tax shield on debt. Instead the estimated tax paid by the business is a separate part of the total allowed revenue. The WACC equations are:

Where

WACCv is the vanilla wacc

Rf = Risk free rate = Govt Bond rate,

β = Equity Beta = correlation of security  with market return,

Rm = Assumed equity market risk premium,

E & D are the assumed equity and debt weight.

There is endless argument over the appropriate values for the parameters. Remember these are not necessarily the actual parameters the stock faces, but the ones the regulator assumes when calculating the allowable revenue. Specifically in recent years there has been a bunch of argument about the debt margin, the equity beta, the expected return on the market and another parameter called “theta” which is the estimated reduction in tax allowance because of dividend franking.

In practice what this means is that for a low risk asset the conventional WACC right now is less than 5%.

Figure 11 WACC measures. Source: AER, ITK

Many readers may not care too much about the cost of capital for a wires and poles business but it’s illustrative of the WACC for a renewables business that has a locked in revenue stream such as that provided by a Government PPA.

Its much lower than the discount rate typically used to calculate LCOE in academic studies. A low risk investment has a WACC of 4.7%

David Leitch is principal of ITK. He was formerly a Utility Analyst for leading investment banks over the past 30 years. The views expressed are his own. Please note our new section, Energy Markets, which will include analysis from Leitch on the energy markets and broader energy issues. And also note our live generation widget, and the APVI solar contribution.