A risk-based approach to setting the baseline for base capital maintenance

Transcription

1 0 March 2011

2 Northumbrian Water Limited Regulatory compliance: reducing the regulatory burden A risk-based approach to setting the baseline and improving incentives for capital maintenance - a discussion document 1. Introduction We welcome the lead Ofwat is taking to reduce the regulatory burden through its Regulatory Compliance project. We also welcome the Sustainable Assets project, in which Ofwat indicates it wants to look at how wider thinking on issues such as minimising the regulatory burden can be applied to capital maintenance 1. Linked to both of these projects is Ofwat s promotion of outcome-based regulation as outlined in its recent discussion paper 2, which we also support. We are keen to work with Ofwat to achieve sustainable reductions in regulatory information and other requirements leading to more efficient use of resources by both Ofwat and companies. The area of focus of this paper is the process to set the capex baseline for base capital maintenance. At PR09, this was the single most data intensive regulatory information requirement. Northumbrian Water Limited alone submitted business cases and supporting evidence in the final business plan amounting to some 570 pages together with a detailed capex database. This was in addition to detailed reporter scrutiny of the company s systems, processes and business plan proposals. This information intensive regulatory process may have been required so that Ofwat could understand how well companies were applying the Common Framework. However, the Common Framework is now bedded-in, serviceability is largely stable and the direction of travel is towards less intrusive and outcome based regulation. We note also that there are concerns about the robustness of the Asset Management Assessment (AMA) scoring approach and associated processes to set the base capital maintenance baseline. We, therefore, believe the time is right to look at options to significantly simplify the process for determining the capital maintenance baseline. We believe this could be achieved using a risk-based approach resulting in benefits for both companies and Ofwat. We have identified a possible outline approach in this paper which we believe reduces the regulatory burden, is outcome based and at the same time strengthens incentives. 1 Extract from M Worsfold letter to C Jones AMA Review Company visits 1 December Ofwat Inputs, Outputs and Outcomes what should price limits deliver? (March 2011) 1

3 2. Background 2.1 Asset Management Assessment (AMA) and associated processes Ofwat describes the AMA as: a full assessment of the technical and managerial processes applied in developing [companies ] capital maintenance plan submissions. It takes into account both the quality of the technical data and the processes applied, and the quality of the decisions made. The AMA scoring process is basically a mechanism that evaluates inputs (processes, methodologies, tools and data) to deduce if the outputs of those business arrangements (future capital maintenance levels) are likely to be robust. It requires Ofwat to become familiar with the processes, procedures, methodologies and data for all four sub-services of every company at a very detailed level. This in turn relies on Ofwat being able to review thoroughly the large volume of information required to do this, which is resource intensive and we believe impractical in a realistic timeframe. Since it is impractical for Ofwat to properly review and assimilate all of the evidence, many companies do not have confidence in the scoring outcomes. Once Ofwat has the AMA score, it must then find some way to convert this into an adjustment to the company s proposed future capital maintenance requirements to create the baseline. We understand the difficulties this presents but in our view the approach used at PR09 was excessively arbitrary and was varied significantly over the course of the price setting process. The process also constrains companies ability to innovate. Ofwat sets out in detail what it expects from companies in terms of processes to achieve a good AMA score, which effectively drives companies planning arrangements. This is likely to restrict the identification of different and potentially better approaches (which may not attract good AMA scores). It works against Ofwat s wish that companies should take ownership of their business plans. The vast amount of information required from companies for the AMA, and the detailed level of Ofwat s scrutiny, does not fit well with Ofwat s regulatory simplification agenda or an outcome-based approach. Nor does this depth of scrutiny guarantee a robust outcome. 2.2 Efficiency and the historic bias Companies currently have differing levels of capital maintenance on a normalised basis (eg. per population supplied, per volume supplied or per length assets), some differences being significant. For example, diagram 1 overleaf shows water infrastructure maintenance investment for the five year period to normalised by length of water main. It can be 2

4 seen that Violia Water Central or Thames Water s investment for that period was around four times that of Cambridge or Bournemouth. Diagram Water IM - Investment / 1000km 2005/ / k/1000km W&S Companies Water Only Companies NWL Average Investment/1000km We believe these differences are unlikely to be wholly explained by operating environment and that some of the differences will be due to efficiency (both efficiency in planning and targeting investment and in the unit cost of delivery). We believe that the process for setting base capital maintenance levels since privatisation may have contributed to the variation in normalised investment levels. Companies had different starting positions at AMP1, when all parties had poorer assessment and review capabilities. Because the main focus of challenge at subsequent periodic reviews was the amount of uplift to historic investment levels, this meant that those companies with higher investment in AMP1 have tended to retain this through subsequent reviews and vice versa (we have called this the historical bias ). 2.3 Future approach We describe below a possible risk-based approach to setting the base capital maintenance baseline. It is designed to reduce significantly the amount of information to be audited and submitted to Ofwat, it starts to address the historic bias and positively incentivises efficient capital investment both at a company and industry level. 3

5 3. Possible basis for an approach for PR14 for discussion 3.1 Overview We propose a three step approach. First, Ofwat sets a benchmark level for future base capital maintenance for each sub-service for each company. This is done early in the process using a modelling approach. This is primarily based on historic levels of investment, adjusted to take account of relative and frontier efficiencies, including the historic bias. In the second step, companies assess their future base capital maintenance requirements for each sub-service and compare this to the relevant benchmarks. If the company is content with the benchmark investment level for a sub-service there is no scrutiny at all by Ofwat for that sub-service the benchmark allowance is included in price setting, subject to agreed serviceability reference levels and control limits. If the company is not content with the benchmark for a sub-service, that sub-service is opened up to scrutiny by Ofwat in step three. In step three, Ofwat scrutinises companies plans for base capital maintenance only for sub-services where the company has not accepted the benchmark. The approach to this scrutiny is not covered in detail this paper. We note consideration is being given to revisions to AMA in an ongoing UKWIR project 3, which may help to inform the way ahead. Our view is the scrutiny should be higher level than previously and targeted predominately at the justification for the increase above the benchmark. We propose that: proposals for enhancements to levels of service would be reviewed separately to base requirements; companies would be required to demonstrate they had identified and adjusted for any overlaps between other programmes and base capital maintenance; and any capital maintenance relating to the transfer of private drains, sewers and pumping stations would be dealt with outside of this proposed approach, though we understand that investment in this respect to bring these assets up to an appropriate standard will be categorised as an enhancement. 3 UKWIR Capital Maintenance Planning: A Common Framework Stage C Review 4

6 3.2 Proposed three step process Step 1 Ofwat sets a benchmark level for each sub service of each company Rationale The purpose of this step is for Ofwat to set a benchmark level for base capital maintenance for each sub-service for each company against which companies can compare their assessed requirements and decide on the level of scrutiny. It is not a draft baseline. It is proposed that benchmark levels are set early in the process, before companies determine their capital maintenance forecasts. The starting hypothesis for this step is that the serviceability framework is now maturing, serviceability levels are largely stable and capital maintenance levels should, in general, be stabilising. We note that this position was advocated by Mark Worsfold at a recent meeting with NWL. Ignoring efficiency issues, it could be postulated that, if serviceability is stable, then investment over the last five years is a good initial approximation of the investment required for the coming five years. However, there are efficiency related issues to take into account regarding the scope for unit cost and planning efficiencies. This includes the issue of the historical bias mentioned above. This can be illuminated by considering diagram 2 below. Diagram 2 shows capital maintenance levels for sewerage non-infrastructure maintenance for to , normalised using volume treated. It is highly unlikely that all of the difference in investment between, say, Thames and Southern will be the result of differences in operating environment. An element is likely to be due to differences in efficiency. This should be taken into account in setting the benchmark level. Diagram 2: Historic levels of sewerage non-infrastructure maintenance 160 Sewerage NIM - Investment / Volume 2005/ / k/ml/d Thames Wessex Yorkshire Anglian Northumbrian Dwr Cyrmu United Utilities Severn Trent South West Southern W&S Companies Average Investment/Volume NWL 5

7 Creating the benchmark for a company s sub-service The step 1 process to set a company s benchmark is shown in diagrams 3 and 4 below. The letters allotted to labels on the diagrams are referred to in the descriptive narrative. Diagram 3: Capital maintenance sub-service benchmark setting where company historic investment is above industry average (E) Company forward looking benchmark (C) Relative efficiency adjustment (25% towards industry average in this example) (D) Frontier efficiency adjustment (B) Industry 5 year historic average (A) Company s 5 year historic average Diagram 4: Capital maintenance sub-service benchmark setting where company historic investment is below industry average (D) Frontier efficiency adjustment (B) Industry 5 year historic average (E) Company forward looking benchmark (C) Relative efficiency adjustment (25% towards industry average in this example) (A) Company s 5 year historic average Firstly, industry and company capital maintenance for the previous five-year period is normalised to ensure comparability, using appropriate metrics such as cost per property or per metre of main/sewer (see A and B in the diagrams above). Adjustments would be made for any exceptional items identified by Ofwat in companies FD Supplementary Reports from the previous periodic review (not shown on diagrams). Where the company s five-year historic average investment is above the industry five-year average investment for that sub-service, its five-year historic actual investment would be adjusted downwards towards the industry average by a percentage in recognition of the scope to improve unit cost and planning efficiencies, including the historic bias (see C in diagram 3 above). 6

8 Companies below industry average would, in a similar way, be adjusted upwards by the same percentage (see C in diagram 4 above). This adjustment towards the industry average is similar to cost base at PR09. The outcome would be indexed to the periodic review base year prices. Because of the uncertainty regarding the magnitude of efficiency related effects versus differences in operating environment, care would need to be taken in determining the size of the adjustments applied. A suggested starting point is to close 25% of the gap between the company s five-year historic average and the industry five-year historic average (assuming efficiency and operating environment each contribute 50% of the difference and moving companies half of the 50% efficiency element towards the average). This makes a significant step in addressing the relative efficiency differences (and historic bias). Twenty five percent is a coarse assumption but is appropriate given this step is simply to establish a benchmark and companies will have the scope to put their case to Ofwat in full in step 3, if they are not content with the benchmark. Our view is that this adjustment obviates the need to use Cost Base for base capital maintenance, again reducing the regulatory burden. (Note: if the Cost Base exercise were to be retained for capital maintenance, this would contribute only part of the above adjustment: the unit cost efficiency element. An assessed adjustment for planning/targeting related efficiency would then need to applied, taking care not to double count. We doubt using Cost Base would add significantly to confidence over and above a wholly assessed adjustment because there would remain significant assessment to make and, in any case, Cost Base by its nature has limited power to explain relative efficiency). To create the benchmark, a final adjustment is then made by Ofwat to reflect the scope for future efficiency (frontier efficiency), based on available evidence (see D in the diagrams above). The final figure so derived would be the benchmark level for sub-service base capital maintenance for that company (see E on the diagrams above).. Step 2 Company accepts the benchmark level or challenges it In this step, each company would determine its future base capital maintenance investment required to manage serviceability using robust assessment processes, methodologies, tools and data and compare this to the benchmark level. The approach we describe does not absolve companies from making a proper assessment so that it knows what investment it requires to maintain stable serviceability. 7

9 Company accepts the sub-service benchmark level In the event the company undertakes its base capital maintenance assessment and believes it can operate within the benchmark investment level identified by Ofwat for that sub-service, there will be no scrutiny of that sub-service by Ofwat. The benchmark level will become the baseline and the company will factor this into its business plan. Serviceability indicator reference levels and control levels would, however, need to be agreed. There is a clear incentive, through significantly reduced regulatory assessment burden, to accept the benchmark level. Company does not accept the sub-service benchmark level A key facet of this approach is that a company may reject the benchmark level and put forward for consideration its view of base capital maintenance requirements for the sub-service. In the event the company believes it requires more investment than the benchmark level for a sub-service for any reason, it must open its assessment for that sub-service to Ofwat scrutiny, presenting its justification for the investment it proposes see step 3. Step 3 - Ofwat scrutinises sub-service proposals only where the benchmark has not been accepted In our view, the aim would be for Ofwat, in this step, to determine if the company has presented a good case for the portion of investment proposed above the benchmark level (eg. to address an adverse serviceability trend, lumpy investment or assets coming up for replacement for the first time). This could be through a modified, less detailed and prescriptive AMA type of approach that allows companies scope to justify investment in their own way. The reduced overall burden on Ofwat (because of companies accepting benchmark levels for some sub-services) would free resources for proper scrutiny of companies cases where the benchmark level had not been accepted for particular sub-services. Following scrutiny, Ofwat would include its draft decision on capital investment levels for the sub-service in the draft determination. As usual, the company would be able to respond to this in its representation. We believe this mechanism would work well in conjunction with the Capital Incentive Mechanism (CIS). In return for lighter touch regulation, companies may accept a benchmark level somewhat lower than it believes is required in the knowledge that overspends of the FD will be partially remunerated through the CIS 8

10 4. Why introduce the proposed approach? 4.1 Substantially reduced regulatory burden for Ofwat and companies On the assumption that capital maintenance investment should be levelling out, and assuming realistic efficiency adjustments are incorporated in the benchmark, it is very likely that some companies will accept the benchmark for at least some sub-services under this proposed approach. For each subservice where the company accepts the benchmark, no scrutiny at all by Ofwat will be required, save to agree reference levels and control limits. For each sub-service where the company accepts the benchmark, the company will not be required to assemble evidence and write specific business cases in a format suitable for Ofwat consideration. Reporter (or similar) audit will also be significantly reduced. (Note: the modelling process for setting the benchmarks as described above is not resource intensive). For those sub-service areas where companies challenge the benchmark, we support Ofwat scrutiny targeted predominately at the increase proposed above the benchmark with a much simplified, less prescriptive AMA type of approach. 4.2 Improved incentives to minimise capital maintenance investment If adopted, it is proposed that this approach would be introduced at PR14 and signalled as being in place for the next periodic review. This would maximise the incentive properties of the approach in terms of reducing capital maintenance expenditure. The approach is designed to benefit companies that have lower than industry average capital maintenance expenditure at each periodic review. This incentivises companies to minimise actual capital maintenance investment (whilst maintaining serviceability of course) so that they then gain at the next review. This will tend to reduce overall industry capital maintenance investment, introducing an efficiency ratchet at successive reviews. In this sense it is a mechanism that, over the longer-term, will lead to more efficient use of capital maintenance resources and lower customer bills. The incentive thus becomes focussed around beating the industry average whilst at the same time meeting stable serviceability (outcome based), rather than achieving asset planning process goals set by Ofwat (input based). This appears to be in line with Ofwat s preference for outcome based regulation. A company that does spend less than the industry average will gain under the CIS, whereas those above will receive a penalty (see worked example in Appendix A). 9

11 4.3 Safety net for companies It is acknowledged that the modelling approach to establish the benchmark is simplistic and will not always result in a capital maintenance allowance that is consistent with maintaining stable serviceability. In this case, companies have recourse to a process where it can put its full justification forward for consideration by Ofwat. 10

12 Appendix A Illustrated example Setting the AMP6 base capital maintenance baseline for water infrastructure maintenance Step 1 Five year industry average investment per m of main Company 1 Company 2 7/m 7/m Five year company average investment Five year company cost per m of main Adjusting company average (25% movement towards industry average) gives: 100m 6/m 6.25/m 200m 8/m 7.75/m Benchmark total Company own forecast Step 2 104m 103m 193.7m 200m Accept or challenge benchmark? Accept Challenge (?) Step 3 Ofwat scrutiny No Yes Notes 1. Frontier adjustment omitted for clarity 2. Accepting or challenging is the company s decision and would not be an automatic choice for either company in the above example. In the table above, company 1 has a normalised capital maintenance average historic investment lower than the industry average, attracting a 25% adjustment upwards towards the industry average. Its own forecast is higher than its fiveyear historic average but slightly lower than the benchmark. If it pegs actual investment to its own forecast or lower, it would make a CIS gain. Company 2 has a normalised capital maintenance average historic investment higher than the industry average, attracting a 25% adjustment downwards towards the industry average. Although its own forecast is the same as its fiveyear historic average, it is higher than the benchmark. It may challenge this or accept the benchmark. If company 2 accepts the benchmark it would have a reduced regulatory burden but would attract a CIS penalty if it needed to invest above the benchmark to maintain serviceability. If it challenges the benchmark, the figure Ofwat includes in the final determination after scrutiny of the company s proposals and company representations would become the baseline and CIS rewards/penalties calculated on that basis. We note that company 2 may not have received a significant challenge from Ofwat under PR09 arrangements (as it has proposed historic 11

13 levels of investment) but will do under this new approach, at least for the portion of investment it proposes above the benchmark. 12