Should regulators index the RAB?

Economic regulators setting cost-based prices often consider the returns on a regulatory asset base (RAB). The RAB is the cumulative capital expenditure on the regulated asset, less its accumulated depreciation (or "run-off"), and often, but not in all cases, it will also include an indexation amount.

Is this indexation necessary and, if so, in which circumstances?

Summary of methods

There are two main methods of valuing the RAB. In both cases the regulated revenue is set such that the operator will receive, from its investment:

• a return on capital (ROCE) and
• a return of capital (depreciation allowance).

Historic cost accounting (HCA) method

In this method, used, for example, by some EU telecoms regulators:

• the RAB is not indexed;
• depreciation and ROCE are calculated using historic values; and
• the weighted average cost of capital (WACC) used is the nominal rate.

Indexation method

In this method, commonly used by utilities network regulators:

• the RAB is inflated each year, typically by the consumer prices index (CPI) (although other valuation methods may be used);
• depreciation and ROCE are calculated using the inflated values; and
• the WACC used is the real rate.

A variant on the indexation method, used for example by Ofcom in the UK, is to use the nominal WACC applied to the inflated RAB, but to deduct a "holding gain" from the return.

Simple example

A worked example, using straight line depreciation for simplicity, will illustrate this.

Assume a capital asset that costs 1,000 currency units, with a five year life and nominal WACC of 12.0%. To keep the modelling simple, I assume the depreciation occurs on the final day of the year, so the average net book value (NBV) of the asset for the year is the same as its opening value, and the ROCE is thus calculated on this opening value. I also assume that there are no operating costs, and no taxes.

HCA method

An HCA model gives the following result:

Indexation method

Now we will use an indexed method, assuming an inflation rate of 5.0% and thus a real WACC of 6.7% (i.e. 112%/105%-1). This gives the following result:

The annual depreciation charge is proportional to the gross value, so increases each year by the price index. The indexation amount is the opening net book value multiplied by the price index.

Comparison of results

Superficially, it might appear that indexing the RAB value increases the total returns to the regulated operator, but in fact it leaves the net present value of the project unchanged.

In the example, discounting the revenues back to the start of the period, using the nominal WACC as the discount rate, gives the following result:

Both approaches thus give discounted revenues equal to the initial investment, i.e the operator earns exactly its WACC on the investment under either approach.

Alternative indexation method

Now consider the variation, which calculates the ROCE using the nominal WACC on the inflated value, but deducts the holding gain from the allowed revenue. The results of this are as follows:

This gives the identical result each year to the standard indexation method (provided, in this simple case, that the ROCE is calculated on the opening NBV before annual indexation).

Choice of approach

Given that the operator should be indifferent to the methodology adopted, why do regulators adopt different approaches in practice?

Reasons for using the HCA approach

• Using an HCA RAB is simple to calculate and easy to communicate.
• HCA is the standard used in financial accounting, facilitating reconciliation of regulatory financial information to published statutory statements and the ledgers that underly these.
• For policy reasons a regulator may wish to "front load" the charges, perhaps if the long term recoverability of the assets is in doubt, in which case the risk to the operator, and hence its WACC, should be reduced by using HCA.

Reasons for indexing the RAB

• An indexed RAB will give the prices that would be derived from a network of equivalent age but at current costs (assuming an appropriate price index), so may encourage other competing operators to enter the market, if this is feasible and an objective of the regulator.
• It will allow prices to rise more closely in line with inflation than an unindexed RAB, which may be important for the operator's customers, particularly if they are consumers in retail markets.
• It also reduces the impact of a price shock (step change in annual allowed revenue) when a depreciated asset is renewed by an equivalent asset at current prices. This effect will be mitigated in any case though if there are a large number of assets in the population and they are renewed evenly over time.

Tilted annuity method

The points above are however only partly mitigated by using the indexation approach in place of HCA, if straight line depreciation is used. To further mitigate these, the tilted annuity method of calculating allowed revenue may be used. The approach adds more complexity, but it ensures that the annual revenue does rise with the price index, while the discounted revenue still equals the investment value. In the current example this would be:

The three approaches above (HCA, indexed and tilted annuity) can be shown graphically, for the worked example, as follows:

Conclusion

The above is a summary of the main issues for RAB indexation using a highly simplified model. In practice the advantages and disadvantages of each model to the interested stakeholders will depend upon many points of detail, such as:

• the objectives of the price regulation;
• level of market competition;
• the life stage of the technology;
• relevant product volumes;
• certainty of forecasts;
• incentive mechanisms;
• phasing of expenditure;
• treatment of repairs and renewals costs; and
• the tax regime for investments.

Each case needs to be considered individually to ensure that the most appropriate treatment is adopted for its circumstances.