The first appeal came from Northern Powergrid (NPg), an electricity DNO. The second came from British Gas Trading Limited (BGT), a buyer of DNO services, and is the first example of an appeal by a network user against an Ofgem price control. In its determination on the BGT appeal, the CMA dismissed four of its five grounds, but partly upheld one ground that challenged the adjustment that Ofgem made to the upfront rewards and penalties for DNOs in its Information Quality Incentive (IQI) scheme. The CMA’s determination reduces the amount of revenue that the DNOs can recover by around £105m in total over the RIIO-ED1 price control period (a reduction of less than 0.4%). In its determination on NPg’s appeal, the CMA dismissed two of the three grounds of appeal, but upheld one in relation to Ofgem’s adjustments to reflect potential savings available from the introduction of smart grids. The CMA’s determination increases NPg’s allowable revenue by around £11m in total (an increase of less than 0.3%). The various areas of appeal are shown in the box below.
Notably, the regime in electricity distribution is one of appeal against specific areas of a price control, which differs somewhat from a regulatory-reference approach where all areas of a price control may be re-examined—as happens, for example, in the water sector. The standard applied is also of interest—the CMA may uphold an appeal only where the particular decision appealed was ‘wrong’ (although the legal wording around ‘wrong’, as specified in the Electricity Act, covers a number of issues). However, the CMA interprets this as meaning that it is not just limited to conventional ‘judicial review’ grounds; rather, it must consider the ‘merits’ of the decision appealed (albeit by reference to the specific grounds of appeal laid down in law).
This article focuses on cost assessment and IQI issues. However, BGT’s appeal also covered two finance issues. First, BGT challenged Ofgem’s decision to introduce a transition regime for depreciation on new assets, which BGT considered provided more depreciation upfront (than otherwise) to deal with financeability issues. Second, while BGT did not challenge the principle of using an index for the allowed cost of debt, it did question Ofgem’s decision on how to calculate this index. Both of these challenges were dismissed by the CMA.
The RIIO-ED1 appeals
The issues covered by the CMA in the RIIO-ED1 cases (and the results of the appeals) were as follows.
NPg appeal (DNO)—exclusively cost assessment issues
Smart grid benefits (SGBs)—upheld
Real price effects (RPEs)—rejected
Regional labour cost adjustments (RLCAs)—rejected
BGT appeal (DNO purchaser)—exclusively incentive and finance issues
Double recovery allegation—rejected
Information Quality Incentive—partially upheld
Asset life policy change—rejected
Change in cost of debt indexation—rejected
Source: BGT determination of appeal and NPg determination of appeal.
As shown in the box, NPg’s appeal centred on three issues around cost assessment. Of these, SGBs was the main area of contention—and the only area of appeal granted to NPg.
SGBs are the forecast cost savings that result from the application of smart grid technologies. Ofgem changed its approach to the assessment of SGBs between its March 2013 Strategy Decision and its July 2014 Draft Determinations (DDs). This was because, in its view, companies had materially underestimated SGBs in their business plans—to the extent that separate adjustments were required, over and above Ofgem’s more general cost benchmarking exercise. In the DDs, Ofgem applied a downward adjustment to DNOs’ total expenditure (TOTEX), based on the difference between SGBs identified by DNOs in their business plans (‘embedded SGBs’) and the level of SGBs that Ofgem thought DNOs should achieve (‘potential SGBs’). The latter had been based on a combination of external evidence and industry modelling.
Ofgem also altered its approach to assessing embedded SGBs and potential SGBs between its Draft and Final Determinations (FDs). The scale of embedded SGBs increased markedly, as Ofgem broadened the definition of what constituted a ‘smart’ solution. However, Ofgem also altered the approach it used to derive potential SGBs, following industry criticism regarding the use of external evidence. In its FDs, the assessment was based solely on comparative benchmarking of the business plan data submitted by the DNOs. In one area, Ofgem set the benchmark at the upper quartile (UQ), or 75% of the highest identified level of DNO SGBs. In another area, the benchmark was set at the frontier (the highest level of SGBs identified by a DNO).
NPg argued that there were a number of methodological, conceptual and implementation issues with Ofgem’s approach.
The following considerations by the CMA are of relevance to other sectors:
that it was not unreasonable, or indeed an error, for Ofgem to depart from its initial (Strategy Decision) approach;
but that such a change would require a strong basis.
However, the CMA considered that Ofgem erred on the second point. Given the significant change of definition of smart solutions between DDs and FDs, and the resultant increase in embedded SGBs, the CMA considered that Ofgem should have reassessed its original judgement of an underestimation of SGBs in the DNOs’ business plans. The absence of such a reassessment did not give the CMA confidence in the new level of savings identified at FDs. Taking all the evidence into consideration, NPg’s appeal on SGBs was upheld.
In reaching this judgement, the CMA also provided some specific views on the SGB benchmarking undertaken at FDs. Some of these views also have implications for other sectors. In particular:
the use of benchmarks based on single observations. The CMA questioned the appropriateness of Ofgem’s approach to determining the level of SGB adjustment for fault-level reinforcement costs, noting that the benchmark was based on a single data point, which it considered to be clearly an outlier. The CMA further noted that this method was not ‘adequately tested’, and might not truly reflect a shortfall in cost savings;
introducing new approaches during the process. The CMA also highlighted that the benchmarking exercise was a ‘novel’ introduction to the price review process undertaken after companies submitted their business plans; that Ofgem had altered the definition of what comprised a ‘smart’ cost saving at FDs; and that there was an absence of established reporting rules. The CMA concluded that Ofgem should have exercised more care in concluding that this method reflected a true shortfall in savings.
In contrast, the appeals on RPEs and RCLAs were disallowed.
RPEs refer to the increase in input prices over a price control period, over and above general inflation. Several regulators take account of real input price inflation in their determinations. Broadly, NPg argued that Ofgem should have used data on DNOs’ actual pay settlements to calculate RPEs, rather than external indices. However, the CMA shared Ofgem’s view that the purpose of estimating RPEs is not to accurately reflect the costs faced by DNOs, but to reflect costs faced by companies in the wider economy. As such, the CMA considered that there was no need to use actual DNO pay settlements data. This also aligns with regulators’ approaches in other sectors. Indeed, the CMA ‘found no evidence, and none was put to [it], of any sector regulators using the actual wage settlements of the regulated companies subject to the price control to construct RPE forecasts for those companies’.
Identifying whether the benchmarks being used are efficient is key to any aspect of regulatory determinations on efficiency. While DNOs have incentives to ensure that pay settlements are reasonable, this is not sufficient to conclude that they are efficient. This is equally true of the general regulatory framework—despite the strong incentive properties of RPI – X regulation, regulators rarely rely on such incentives alone to ensure efficiency, and place considerable emphasis on benchmarking.
The final area of appeal was that of uncontrollable factors. A key element of an efficient cost benchmarking exercise in any sector is to adjust for costs that are beyond a company’s control. Differences in labour costs due to regional factors are one such factor. On this issue, NPg argued that Ofgem used an approach that ‘materially overstated’ the cost premium in London and the South East relative to the rest of Great Britain. This had the effect of lowering NPg’s relative efficiency (given that it operates outside London and the South East). The CMA dismissed this element of NPg’s appeal because NPg had not demonstrated that Ofgem’s approach was wrong; it only demonstrated that alternative approaches were available. As discussed above, this conclusion is specific to the legal context in which energy appeals are made—namely, that it is necessary to show that Ofgem is wrong; and that the CMA must be satisfied that there is a better approach, rather than showing that an alternative approach is possible. It is therefore unclear whether the CMA would have reached the same conclusion on this issue in another sector. This outcome does seem to indicate, however, that the legal framework for energy appeals makes it harder to improve outcomes, as it can be difficult to overcome the additional first hurdle of demonstrating that the regulator is wrong.
Information Quality Incentives
While NPg’s appeal was based exclusively on Ofgem’s approach to cost assessment, BGT’s appeal focused on finance and incentive issues. A central focus of this appeal was calibration of Ofgem’s IQI mechanism. The IQI is a form of menu regulation used by Ofgem in RIIO-ED1, and is designed to address information asymmetry through incentivising companies to submit accurate expenditure projections. In practice, the IQI is a matrix that needs to be ‘calibrated’. Decisions need to be made on three issues:
the ‘incentive rate’, which determines how much outperformance-sharing there is with consumers once prices have been set;
‘interpolation’, which determines how much weight is given to each DNO’s expenditure assessment versus that of Ofgem in setting the allowed expenditure;
the upfront reward or penalty applied to each DNO (‘additional income’), dependent on its forecasting.
BGT appealed the upfront reward/penalty component applied to the slow-track DNOs.
Ofgem first set out its intentions regarding upfront rewards/penalties in its September 2012 Strategy Consultation. First, companies that had presented the best business plans had the prospect of being fast-tracked, with an upfront reward of 2.5% of TOTEX provided through the IQI. Second, for the remaining slow-track companies, there was still the prospect of financial rewards in the draft determinations for those submitting the best forecasts in their resubmitted business plans. Specifically, a company submitting a forecast that exactly matched Ofgem’s own assessment of efficient expenditure would ‘be able to achieve a return equal to its cost of capital, if it were then to spend, over the price control period, the amount it had forecast’. In its March 2013 Strategy Decision, Ofgem elaborated on this, stating that it would ‘set the break-even point in the IQI so that a DNO that forecasts [TOTEX] in line with [its] view of the upper quartile and achieves that forecast would earn [its] cost of capital but not receive any additional reward under the IQI’.
‘Efficiency’ was therefore about UQ performance—i.e. it was assessed on a relative basis. Companies that achieved efficiency in their forecasting within the UQ would expect to break even or better. However, as noted above, following submission of final business plans by slow-track companies, Ofgem was concerned that companies had not put forward robust forecasts of RPEs or SGBs, and changed its cost assessment architecture to deal with these two issues. The consequences for the slow-track companies were then outlined in its July 2014 DDs:
We have reviewed the design of the IQI in the light of the cost adjustments we are making after setting the UQ efficiency benchmark (RPEs and smart grid savings). These adjustments mean that no DNOs are achieving our view of efficient costs and that no DNO would receive a reward according to our original design.
We think the IQI is key to encouraging better information at slow-track. We think that it is right to reward companies that have provided good information that has helped our comparative benchmarking. In the light of this we have adjusted the break-even point in the IQI matrix so that the best-performing DNO groups receive a reward. The break-even point is now an IQI score of 102.9 rather than 100. This means that a DNO group that forecasts 2.9 per cent above our efficient cost benchmark and achieve [sic] its forecast will earn its cost of capital but no additional reward or penalty.
BGT argued that this adjustment was an error, was harmful to consumers, and offered no countervailing benefit. By this stage, the companies had already submitted their business plans, and therefore any incentive effect due to the IQI would have already been achieved. Moreover, several companies benefited from the adjustment, and not just the UQ companies.
The CMA noted that BGT’s appeal was directed mainly at the question of principle (i.e. whether there should have been an adjustment), but that BGT also questioned whether the specific adjustment made was consistent with Ofgem’s stated rationale.
On the first point, the CMA accepted ‘that any incentive effect from the IQI in RIIO-ED1 had been achieved’. However, Ofgem had made it clear during the course of the appeal that it did not rely on any such incentive effect. Additionally, the CMA considered that Ofgem was right to take into account the potential implications for incentives in future price controls of sticking to its original policy intention. The CMA rejected BGT’s argument that the assessment of RPEs and SGBs means that no slow-tracked DNO should receive a reward through the IQI, since this would involve rewarding DNOs against a standard of ‘absolute’ rather than ‘relative’ efficiency. The circumstances of the RPE and SGB adjustments justified the recalibration. Ofgem’s intention was to recalibrate to maintain consistency with the UQ approach, and the CMA did not consider that making such an adjustment was wrong, taking account of the overall outcome of its cost assessment process.
However, on the second point, the CMA argued that the actual recalibration at FDs was inconsistent with the stated UQ approach. The recalibration set the cut-off point between the sixth and seventh most efficient DNOs, based on the efficiency scores at the end of the process. Under a UQ approach, a cut-off point between the fourth and fifth most efficient DNOs would have been expected. Ofgem’s approach resulted in the DNOs being significantly better off than under an approach that was consistent with its UQ reasoning. This was ‘wrong in law’. The CMA therefore upheld BGT’s appeal, albeit to this limited extent, and substituted its own decision on the appropriate level of the cut-off point. The slow-track DNOs had agreed to this remedy.
While incorporating elements of a merits-based approach in its decisions, the CMA was restricted to the specific areas being appealed, and could uphold an appeal only if it was satisfied that Ofgem’s approach was ‘wrong’—which was a challenging hurdle. Alternative approaches may therefore have been available on a particular issue that might have led to a slightly different answer, but this was not sufficient in itself to uphold an appeal. The cases do not represent a complete assessment of the merits of Ofgem’s overall RIIO-ED1 approach, and neither was this their intention.
In the case of both SGBs and IQI recalibration, there are some common themes. The first is that it is not unreasonable for a regulator to adjust its regulatory methodology when needed, in order to adhere to more fundamental regulatory principles that had been set out in advance. This is more important than the precise mechanics. However, in the case of SGBs, there was considerable uncertainty in the separate benchmarking exercise subsequently undertaken by Ofgem. The CMA effectively concluded that Ofgem was prodding in the dark for a solution to a problem that may not have existed. In the case of the IQI, the need to retain the principle—to offer rewards to UQ companies—was endorsed by the CMA. However, the practice went beyond that needed to achieve the desired intent. The lesson from this might be that process is important, and that the details matter: it ain’t what you do, it’s the way that you do it…