Ofwat home > Publications > Letters to Regulatory Directors

RD 14/07


To Regulatory Directors of all
water and sewerage companies
and water only companies

• price caps
• revenue caps, and
• revenue corrected price caps which continue with the current price control, but which correct for past revenue over or under recovery at each periodic review.

• remove the disincentive for companies to promote water efficiency by ensuring that companies essentially receive a fixed amount of revenue regardless of customer behaviour;
• remove the scope for out or under performance by correcting revenue over or under recovery in net present value terms after the event. Companies and investors can be confident of expected revenue over the longer term;
• remove the incentive for companies to understate expected revenues, because companies will end up with the same amount of revenue regardless of their forecasts; and
• still incentivise companies to bill all properties. We would include an adjustment to correct for companies that billed more or fewer properties than we had expected. We would multiply the property variance by an allowance based on the costs we judged an efficient company incurred in finding and billing properties, plus an element to act as an incentive. Price caps provide an incentive far above the normal costs incurred in maintaining accurate billing records and therefore over compensate water companies.

• will keep prices more stable for the following reasons:
  - it averages annual revenue over and under recovery from one five year period over the next five year period. Years that have over recovery will offset years that have under recovery. And the equal spreading of the correction over five years will reduce price volatility. A revenue cap would lead to less stable prices for customers. For example if, in a drought, metered customers respond to requests to save water, the reduction in their bills would directly feed through to increased bills in a following year.
  - greater than expected demand reductions from metered customers will lead to reduced revenue in the short term, but should lead to lower investment in water resources in the long term. Under a revenue cap customers would experience higher bills in the short term and have to wait until after the next price review for lower bills from reduced investment needs. Under a revenue corrected price cap these two impacts would happen at the same time. Ideally the reduction in customers' bills from reduced investment would outweigh the increase in customers' bills required to compensate companies for lost revenue.
• is simple to understand and implement because it is a variation of the current process. As the correction would be done at a periodic review there would be no change to our annual processes and little or no increased regulatory burden on companies. Under a revenue cap there would be an increase in both our work and companies' work as we would have to forecast revenue annually. A revenue corrected price cap also does not require a licence modification which would add uncertainty to the 2009 periodic review process.

Stochastic method

We have used a program, @Risk, that allows a model to be automatically run many times changing defined inputs in a random way. This unveils the extent of possible options and provides a useful indicator of the range of possible outcomes. We adjusted the following factors in using this method:

• Inaccuracy in forecasting volumes
• Inaccuracy in forecasting new properties
• Inaccuracy in forecasting optional metering
• Metering characteristics
• Extent of water efficiency
• Number of existing properties identified and added to the billing system

This analysis helps to highlight the impact of a particular input. We have run the model with a central case varying each of the following factors in turn for each form of price control:

• Inaccuracy in forecasting
• Extent of water efficiency
• Number of existing properties identified and added to the billing system

• Incentive for water efficiency
• Reduction in companies undue revenue outperformance
• Impact on customers' bills
• Incentive for companies to bill all properties where it is economic to do so

Form of price control	Variant	Aims and key features
REVENUE CAPS
Fixed revenue control	No correction & tariff setting based on forecast data	For companies to get a specified revenue in each year that we determine at a periodic review or subsequent interim determination
Symmetrical fixed revenue control	Symmetrical correction & tariff setting based on forecast data	As above, but the determined revenue in future years is adjusted so that the cumulative revenue is the same as determined. (In this case forecasts are in NPV terms).
Asymmetrical fixed revenue control	Asymmetrical correction & tariff setting based on forecast data	As above, but the determined revenue in future years is adjusted so that the cumulative revenue is no more than determined (ie it can be less than determined).
Dynamic revenue control	With property driver and symmetrical correction & tariff setting based on forecast data	As the symmetrical fixed revenue control, but the yield is also changed by a fixed amount per the number of properties.
Dynamic revenue control	With property and volume drivers and symmetrical correction & tariff setting based on forecast data	As above, but the yield is also changed by a fixed amount per unit of water delivered.
PRICE CAPS
Current cap	As now with tariff setting based on historic data	For companies to set charges so that they get a fixed amount of revenue per unmetered property and a limit on metered charges.
Average revenue yield per household customer and average price cap for non-household customers.	No correction & tariff setting based on historic data	For companies to set charges so that they get a fixed amount of revenue per household and a limit on average charges for non-household customers.
Average price cap	Tariff setting based on historic data	For companies to set charges so that they are beneath a specified limit on the average charge.
REVENUE CORRECTED PRICE CAPS
Revenue corrected price cap (annual)	As now but with annual correction to the price limit.	As the current cap, but the price limit would be changed annually for past over and under performance in NPV terms.
Revenue corrected price cap (periodic review)	As now but with correction at periodic reviews	As the current cap, but past over and under performance taken into account at the next periodic review.
Revenue corrected price cap with property driver (periodic review)	As now but with correction at periodic reviews	As the current cap, but past over and under performance taken into account at the next periodic review, plus an adjustment for the number of properties billed

• We will prescribe in detail what we will do at PR14 before we determine PR09 price limits.

• We will continue to monitor companies as we do at present.
• In addition to this, we will report on the revenue correction that we will make at PR14 so that consumers, companies and investors are kept informed.

• We will reduce or increase the revenue requirement for the next review period by the over or under performance by companies between 2010 and 2015. We will annualise the amount over five years. In calculating this we will use the tariff basket model used at the previous review, except that we will use companies' actual price increases each year for each service, rather than RPI and the indicative price limits we assumed at the Periodic Review. This will correct for any interim determinations, intentionally deferred K or changes to RPI. We will make an adjustment if companies want to return revenue outperformance to customers in advance of PR14.
• We only plan to correct for tariff basket revenue streams. We believe that the reasons to correct for other revenue streams are weaker and there is a greater scope for perverse incentives or creating barriers to competition. We will keep the method we use to correct for revenue over or under performance under review to avoid the method of correction introducing barriers to competition. A change to the competition regime may lead us to change the method of correction between periodic reviews.
• Companies' revenue forecasts for PR09 will be based on the customer base in the base year and assume that customers will remain either in or out of the tariff basket for the whole period. Each year companies will need to report volumes and revenue of customers that move in or out of the tariff basket and we will take this into account in the adjustment. We will not change any other assumptions in the tariff basket model.
• We will adjust the revenue correction to take account of the companies' performance on billing. If the companies either bill more or fewer properties than we expect we will multiply the variance by a factor, the Efficient Billing Factor, that we believe covers the costs of finding and billing extra properties by an efficient company, plus an amount to act as an incentive. We will determine what this is at PR09. The variance in properties will only be based on properties with a bill above the Efficient Billing Factor to avoid a perverse incentive for companies to bill uneconomic properties. In addition to this we will increase the adjustment for incremental outperformance or underperformance between two years by the factor (4-t), where t is the remaining years of the price review period. This will help to maintain an incentive to bill all properties throughout the period.
• We will correct for the fifth year of the period based on forecast information. We will also correct for any outperformance or underperformance against this forecast at the following review period.
• We will calculate the correction and billing adjustments in net present value terms so that the timing of revenue has no effect on the company. We will use the same discount rate as we use for other logging up and down. This will prevent perverse incentives for the company to try to affect the timing of revenue out or under performance.