OT19220 - PRT: Valuation - Appendices

PRT: Text of letters sent to oil industry representative bodies regarding Brent reference pricing dated 02/11/2004

Letter of 2nd November 2004;

Brent Pricing

Following Peter Harrington’s letters of 11 November and 23 December 2003 (copies attached), this letter sets out our proposal to move entirely to new methodology (Price Reporting Agency or “PRA” methodology) for Brent from 1 January 2005. I mentioned this at the last quarterly meeting with industry representatives on 13 October 2004.

It has generally been getting more difficult to calculate the statutory value for Brent using the strict statutory method because of the increasing lack of contract information on which to found our database. This is caused by the decline in the physical and forward Brent markets, as Brent blend production declines and hedging activity moves more into the futures market.

With industry agreement, we have calculated the Brent value for the two pilot periods to 31 December 2003 and 30 June 2004 using the PRA methodology and compared the results to the Brent value that would have been produced using the old method (see Table 1 and Table 2). This has shown that there are no significant differences between the results produced by the two methods, as our initial research predicted.

Our intention to move to the PRA method is well known and we have received no adverse comment on either the methodology or the resultant valuations. We now therefore propose to move formally to using the PRA method for the Brent value with effect from 1 January 2005. We also propose that the pilot be extended to cover the Chargeable period to 31 December 2004 (2H04). A proposed timetable is shown at Appendix 1.

We would propose to continue to provide figures for comparison purposes for 1H and 2H 2005 and therefore to require PRT1A’s at least for those periods but with the aim of dropping the requirement to report Brent crude at an appropriate time.

The details of the PRA method are shown in Appendix 2.

Please let me have any comments on this proposal by 30 November 2004. This will allow us time to put into place revised guidance and make any procedural changes required by the new methodology.

SP14

We believe that there will be considerably less need for arrangements under SP14 as a result of moving to agency pricing. We will discuss with individual companies their own positions in relation to agreements currently in force.

Other Blends

We want to pilot agency pricing for other blends following the success of the Brent pilot. We will write to you again shortly with the detail.

Malcolm Phelps

Assistant Director

Energy Group; Petroleum Revenue Tax

Table 1: Brent 2H03 comparison results

2H03 Brent statutory values (estimated) with interpolations 
All prices in US$
Month:JulyAugustSeptemberOctoberNovemberDecember
Sch 3 para 2(2C) value27.9228.9328.7328.1629.1429.24
PRA value27.7928.8828.9728.1229.1329.17
?-0.13-0.050.24-0.04-0.01-0.07


Mean of differences: $-0.01

95% confidence limit: 0.13521

Since the mean is LESS than the 95% confidence limit there is no significant difference between the two methods


Table 2: Brent 1H04 comparison results.

1H04 Brent statutory values (with interpolations) comparisonAll prices in US$









MonthJanuaryFebruaryMarchAprilMayJune
Sch 3 para 2 value30.3730.7231.8633.1034.5736.95
PRA value30.3430.6631.8633.0834.5836.96
?0.030.0600.02-0.01-0.01

Mean of differences: $0.015

95% confidence limit: 0.0287

Since the mean is LESS than the 95% confidence limit there is no significant difference between the two methods

Appendix 1: Proposed timetable:

30/11/04:Time limit for Industry response to this letter.
31/12/04:Procedural changes made and guidance updated. 1/1/05: Formal transfer to PRA method for Brent. OTO calculates Brent PRA value for 2H04 and publishes it to Industry [Date?].
28/2/05:PRT1 and 1A returns filed for 2H04. OTO calculates 2H04 Brent using old method for comparison purposes only and shares this with Industry by about mid-April. OTO also calculates PRA values for non-Brent grades and shares the comparison with industry at the same time.
30/6/05:OTO calculates PRA Brent value for 1H05 and publishes this by the first week of July.
31/8/05:1H05 PRT1 and 1A returns filed. OTO calculates 1H05 non-Brent using PRA method for comparison purposes, alongside the old method values for non-Brent grades. Comparison shared with industry by about mid-October.
October to December 2005:dialogue between OTO and industry about prospects for adoption of PRA methods for non-Brent grades.

Appendix 2: The PRA method:

PRT has its own specific transfer pricing legislation – for oil these are Paragraphs 2, 2A and 3 Schedule 3 OTA 1975

Paragraph 2(2) Sch 3 OTA 1975 sets out the basic rule which describe how the value of oil for any particular month is to be arrived at; "…the price at which oil…might reasonably have been expected to be sold under a contract of sale…" under stringent conditions including that the contract at arm's length to a willing buyer and is for delivery of oil (ie a physical cargo).

Para 2(2A) then requires a monthly value to be produced and explains how to do it, using the concept of a hypothetical contract for the sale of oil. The value is arrived at by taking a volume weighted average for each business day in the valuation reference period (VRP). The legislation requires contract information for each business day in the VRP. Only contracts that comply with all of the requirements of the hypothetical contract can be used.

The legislation provides alternative rules in prescribed circumstances. Para 2(2D) allows alternative methods " if or in so far as the Board are satisfied that it is impracticable or inappropriate to determine……..the price of oil in any month …..”. Reasons can include insufficient information to calculate a value on this basis, or the nature of the market, or for any other reason. The legislation provides for two other methods;

  • para 2(2D)(a) brings in other contracts so far as it is practical and appropriate to do so.
  • So far as it is not practical or appropriate to use para 2(2D)(a) then para 2(2D)(b) requires the Board to determine values “….in such other manner as appears….appropriate in the circumstances”.

There are a number of reasons why we propose to change our Brent method now;

  1. We do not have sufficient data to support the para 2(2A) method. We have increasingly been forced to rely upon interpolations to fill the gaps in the data. This situation can only get worse, probably in the short- rather than the long-term.

  2. For ALL crudes, if we are to change methods, we feel we need to be able to demonstrate that any new method would give similar results (in terms of absolute numbers) to the old method. With falling liquidity it will become increasingly difficult to demonstrate this.

  3. For Brent there are more changes likely in the near future. We will need to be able to follow these changes rapidly and easily if we are to produce values that are an appropriate reflection of real market prices.

These demonstrate that the current method for Brent is not sustainable and is therefore inappropriate. Given the weight of these reasons, we think it is reasonable to conclude that NOW is the appropriate time to change our current methodology to one that will be able to cope with all foreseeable future changes in North Sea oil pricing practice.

So:

  • We are satisfied that it is inappropriate to use the strict method (paragraph 2(2A) Schedule 3) for Brent because of increasing reliance upon interpolations;
  • We are also satisfied, for the same reasons, that methods relying on paragraph 2(2D)(a) of Schedule 3 will become impracticable and are inappropriate; .
  • Consequently, we propose to use the PRA method determined under Paragraph 2(2D)(b) Schedule 3 and described in more detail below. Having consulted widely, we regard the PRA method as appropriate in the circumstances and one that will remain practicable for the foreseeable future.

The PRA methodology:

From 1 January 2005, the Brent pricing for the purposes of para 2 Schedule 3 OTA 1975 will be carried out using a daily average of the prices produced by the three pricing agencies – Platts, LOR and Argus. This daily average is then used to calculate the average price over the Valuation Reference Period (VRP).

The VRP consists of M-1 and half of M (number of days used depends on total number of days in M). So for a price in Feb, the VRP would be all of January and half of February (up to 14th).

Production of a daily price series throughout the VRP:

In the month before delivery (January in this example) the prices used to calculate the daily averages should be the assessed M+1 values (that is, in January, assessments for delivery in February). At the start of the delivery month (Feb in example) the values used should be M (that is assessments for the price of oil to be delivered in the same month, February in this example). M values are only quoted for 6 or 7 days at the start of the month. M values will be used until all three publications stop quoting them. For example, if LOR and Argus only quote M until the 6 but Platts quote until the 7, values of zero would be used for LOR and Argus on the 7. Once ALL M values have stopped, dated values are used for all publications. This principle will also be used where one of the agencies does not provide an assessment (usually on a Bank Holiday); the daily average would then be an average of two, or possibly even one assessment.

Calculation of the value for the month:

Once a daily price series has been produced, the value for the month is found by calculating the arithmetic mean of all the daily values.

Letter of 23rd December 2004:

Brent pricing

This is a short note to inform you that, further to my letter of 2 November 2004, we will now move to the new Price Reporting Agency (PRA) method for calculating our Brent statutory values. We will formally adopt this method with effect from 1 January 2004, under Sch 3 Para 2 (2D)(b) OTA1975.

New guidance material will be published shortly setting out the methodology in detail, as described in Appendix 2 to my 2 November letter, also providing a worked example.

A Brent value calculated using the old method will continue to be provided, for comparison purposes only, until further notice.

Yours sincerely

Malcolm Phelps

Assistant Director

Energy Group; Petroleum Revenue Tax