British Gas Trading Ltd, R (on the application of) v The Gas and Electricity Markets Authority & Ors

The domestic energy market features two distinct tiers, with customers either on a standard variable rate tariff (SVT) or a default rate fixed term tariff (collectively “default tariffs”), or on an active choice fixed term tariff (“fixed tariffs”). Despite fixed tariffs tending to offer lower prices, in April 2017 60% of all customers, approximately 11 million in total, were on default tariffs. The six largest domestic suppliers, namely the Claimant (“British Gas”) and the first five Interested Parties, known as the “Big Six”, collectively serve around 90% of customers on default tariffs.

In a report published in June 2016, the Competition and Markets Authority concluded that there was ineffective competition in the energy supply market, particularly so far as the Big Six were concerned. Default tariffs were considerably higher than they would have been if competition was effective. In an attempt to remedy the situation, the Government decided to introduce a price cap. To that end, it enacted the Domestic Gas and Electricity Tariff Cap Act 2018 (“the Act”) which came into force on 19 July 2018.

Section 1 of the Act imposed an obligation on the Defendant authority, (“GEMA”), to set and implement as soon as practicable a price cap on certain tariffs for the retail supply of domestic energy, with a view to protecting existing and future domestic customers on default tariffs. “Ofgem” is the name given to the office of civil servants who carry out the functions of GEMA on a day to day basis. For the purposes of this judgment, there is no material distinction to be drawn between GEMA and Ofgem, and I shall refer to GEMA unless the context otherwise requires.

The price cap was to apply to all energy suppliers, and in setting it, GEMA was obliged to have regard to the four “needs” set out in s.1(6)(a)-(d) of the Act. The need with which this case is principally concerned is (d), “the need to ensure that holders of supply licences who operate efficiently are able to finance activities authorised by the licence.”

Before the legislation was implemented, GEMA worked on designing the price cap and addressing the numerous policy and technical issues that arose. It first consulted stakeholders on the design of the cap in a series of working papers published in March

On 6 November 2018, GEMA made the decision under challenge (“the Decision”) setting the price cap. The cap is calculated by reference to a complex formula, but in simple terms, the maximum prices that suppliers can charge domestic energy consumers on default tariffs are a function of allowances made in respect of the many categories of supplier costs, of which wholesale energy costs provide the largest element (around 40%). GEMA referred to this as the “bottom up” approach.

The claimant’s parent company Centrica Plc wrote to GEMA on 21 November 2018 seeking to persuade it to reconsider the Decision, on the basis that one aspect of its decision-making, pertaining to the wholesale energy cost allowance for the first period of the price cap (Q1 2019), was fundamentally flawed. GEMA refused to reconsider. That refusal is also the subject of the claim for judicial review, but as both parties accepted, the challenge to that decision stands or falls with the challenge to the substantive Decision.

British Gas challenges the assessment of the sufficiency of the allowance on the basis that GEMA made that assessment on an assumption about the behaviour of a

“typical” supplier, which it did not share with the suppliers or give them a fair opportunity to comment upon (and thus correct). The assumption only came to light when the Decision was published. Mr Fordham QC, on behalf of British Gas, submitted that the assumption was unfairly adopted, insufficiently investigated, and demonstrably flawed. Mr Maclean QC, on behalf of GEMA, submitted that there was no such assumption, or if there was, it was immaterial. Alternatively, if there was such an assumption, and it was material, it was sufficiently communicated to the suppliers. It was implicit in the questions asked in the September consultation, and the suppliers had an adequate opportunity to respond and provide the information to challenge it.

It is common ground that under the Act, GEMA has wide powers to adjust the cap, including to deal with an inadequate allowance in a previous period. In the light of this, British Gas does not seek to quash the price cap. Instead, it asks the court for declaratory relief in appropriate terms.

There is no doubt that in fixing a single price cap for all suppliers, with a view to protecting current and future customers on default tariffs, GEMA had a complex and difficult task to perform. It was also put under significant time constraints. For reasons which I shall explain, fixing the wholesale allowance for Q1 2019 presented special difficulties, of which GEMA was aware. It went about the task conscientiously. It was mindful of its statutory obligations, and it tried to fix the wholesale allowance at a level which would suffice to cover an efficient supplier’s costs, but which would not result in overcharging the relevant body of consumers. Nevertheless, I am persuaded that the criticisms made by British Gas are wellfounded and that it is entitled to the declaratory relief that it seeks.

Section 1 of the Act contains the following relevant provisions:

Section 2 (1)(b) of the Act provides that tariff conditions must set out how the cap is to be calculated, and may make provision about assumptions required to be made in making the calculation. Subsection 2(b) expressly precludes those conditions from making different provision for different holders of supply licences. That means that the same cap applies to everyone.

Section 2 (3) provides that:

“before making the first modifications under section 1 [GEMA] must… consult such persons as it considers appropriate on the methodology to be used for the purposes of the cap.” [Emphasis added].

Section 2(4) provides that consultation undertaken before the Act was passed is as effective for the purposes of subsection (3) as consultation undertaken afterwards. Both parties accepted that an obligation to consult on the methodology is not confined to consulting on the broad structure to be adopted, i.e. “bottom up” versus alternative approaches. The consultees should be consulted about the key ingredients of, and material assumptions adopted in, the chosen approach.

The statute makes it clear that GEMA’s primary focus when introducing the price cap and deciding how it should be calculated must be on seeking to protect consumers on default rates. It was for GEMA to balance each of the matters in s.1(6) and decide what weight to give them, bearing in mind that aim. It was obliged to have due regard to the 4 “needs”, but that is not the same thing as achieving them. So long as those matters were properly and conscientiously taken into account, and weighed in the balance, GEMA would have complied with its obligations under that subsection.

The wholesale energy market is volatile. Prices on the spot market may rapidly change by significant amounts. There is an inevitable divergence between the prices a supplier pays in the wholesale market and the retail prices it charges to its customers, as consumers do not want frequent and volatile changes to the retail prices which they pay. A key part of an energy supplier’s business is to try and set stable and competitive retail prices which reflect a reasonable expectation of its future wholesale

This is typically addressed by purchasing forward contracts for the delivery of energy (at a specified point in the future) gradually over a period of time. This hedging approach protects the suppliers from sudden or sharp changes in wholesale prices, as their costs reflect the average price over the relevant period. Hedging reduces the risk of a mismatch between the accumulated costs of wholesale market purchases relevant to delivery in a specific period, and the wholesale cost element that contributes to the competitive retail market price. Each supplier has its own specific hedging strategy that is kept confidential from its competitors. Suppliers will revise their hedging strategies from time to time in response to developments in the market.

In the light of GEMA’s obligation to set the price cap at the same level, irrespective of the position of individual suppliers, it was inevitable that some of those suppliers would be better or worse off than others when the cap was introduced. Ensuring that an efficient supplier with a very high cost customer base covered its costs would entail setting the cap at a very high level for all suppliers, and GEMA rationally considered that such an approach would be inconsistent with the objective of consumer protection.

In the executive summary of the September consultation document, GEMA made it clear that it did not consider it would be consistent with the Act’s objectives if it were to set a higher cap level for the purpose of enabling every supplier – even those who may be relatively inefficient – to compete. It therefore proposed to set the cap level based on its assessment of efficient costs and a ‘normal’ profit level (1.9%), which it considered would enable an efficient supplier to finance its activities.

GEMA referred to some of the challenges that it faced when estimating an efficient cost level, including the fact that what was an efficient level of costs would differ between suppliers, depending upon their customer bases and operating environment; the fact that not all suppliers were currently operating on a cost-efficient basis; the difficulty of predicting future costs; and the risks of overstating allowances if reliance was placed solely on data provided by the suppliers themselves. Setting an efficient benchmark needed to be based on an examination of the whole market, and not picking one specific operator as being perfectly representative.

No complaint is (or could be) made about that approach: British Gas accepted that the allowance set by GEMA would necessarily be an estimate of the costs that suppliers would incur. However, if and to the extent that the estimate was based on generalised assumptions about suppliers’ behaviour, there had to be a valid basis for making those assumptions.

GEMA’s consultation and decision-making processes involved measuring wholesale costs on the basis of an index of future supply contracts observed within a specified period (“the observation window”) for delivery during the period of the price cap. This approach would allow suppliers to adapt their hedging strategy to align their purchasing to the index in order to ensure that their actual costs would not exceed the wholesale costs allowance. A failure to align would be tantamount to gambling upon future market prices remaining in the supplier’s favour. Therefore, a prudent supplier

GEMA needed to set the observation window for the first price cap period as a transition into the new arrangements. The first cap period, Q1 2019, was going to be only three months, rather than the normal six months. Setting the cap for this period presented special challenges. One of these was that if GEMA chose an historic observation window, suppliers would be denied the opportunity of alignment, since those market prices would no longer be available.

From April 2018 onwards, the price of forward contracts for delivery of gas and electricity rose significantly. Energy bought earlier in time was cheaper than energy bought later, so an earlier observation window would give a lower allowance than a later observation window.

GEMA was aware from the results of a detailed request for information process (“RFI”) carried out in March 2017 (and responded to in April/May 2017) that historically, large energy suppliers had purchased energy at regular intervals over a period starting at least 18 months in advance of delivery (“a long-term purchasing strategy”). If a supplier adopted an 18-month rateable strategy, one-eighteenth of the energy it anticipated it required for a future period would be purchased each month for 18 months. This would result in the supplier acquiring most of the necessary energy well in advance, in this example, one third within the first 6 months of the 18month period, and two thirds after a year. It could then make any adjustments to meet its actual requirements closer to the period of delivery.

One of GEMA’s specific aims in setting the observation window for the transitional period, expressed in an internal paper prepared for a meeting of its Board dated 31 October 2018, (“the October Board paper”) was to protect customers of suppliers that had bought energy when prices were low from paying higher prices than their suppliers’ actual costs would justify.

In the May consultation, GEMA set out its proposals for setting the cap in the transition period in paragraphs 5.42 to 5.50 of the main document. At that time, it proposed to use a “6-3-12” three-month model. A 6-3-12 observation window looks at forward contracts for energy supply during a period of 12 months (“12”) purchased over a period of 6 months (“6”) ending 3 months before the energy supply begins (“3”). In May 2018, GEMA proposed to take the 6-month observation window of April-September 2018. It provided an illustrative example of how its model would work. As part of that period was still in the future, suppliers had the opportunity to adapt their hedging strategy to align with the prices in that window to some extent.

GEMA understood that the actual costs that suppliers would incur would be driven by their hedging strategy. It recognised that suppliers might attempt to follow the hedge (i.e. align with the observation window) once GEMA’s proposed approach was confirmed, but it thought the volumes involved were likely to be smaller than if it picked an observation window entirely in the future, or a shorter observation period. It anticipated that, in accordance with the long-term purchasing strategies revealed in 2017, the big suppliers would have already bought large amounts of energy for the initial cap period, and it believed that because of this, they would have less opportunity to follow the hedge. Crucially, it was not expecting suppliers to unwind existing contracts.

After it had considered the responses to the May consultation, GEMA reached the view that its proposed observation window of April-September 2018 would result in a wholesale costs allowance that over-compensated the energy suppliers and insufficiently protected consumers. In simple terms, the reasons for that view were as follows. The market had risen by around 30% from April 2018 onwards. Therefore, a supplier who had bought most of its energy for Q1 2019 prior to April 2018 would have done so at much lower prices than those charged in the observation window on which the allowance would be based (April-September). In such circumstances it seemed likely that using the April-September window would result in a huge disparity between the costs such a supplier was likely to have incurred in acquiring most of its energy, and the hypothetical costs upon which the calculations would be based.

In consequence of this, in the September consultation GEMA announced that it now proposed to use what it described as “our standard approach for a winter cap period”, observing prices offered between February and July 2018 for contracts that would be delivered between October 2018 and September 2019. GEMA said, in paragraph 2.22 of the body of the September consultation document, that it proposed to use this approach:

“to ensure the wholesale allowance better reflects the underlying costs that suppliers are likely to have incurred when purchasing energy for delivery during the first cap period. Large suppliers have most likely already bought much of the energy SVT customers will use in early 2019. Now that the wholesale market has increased since April 2018, our initial approach risked customers paying significantly more than suppliers’ underlying costs. Had market prices reduced, our proposal would have risked setting the wholesale allowance below suppliers’ actual costs, and would have needed correcting”.

By necessary implication from the final sentence, GEMA acknowledged that if it did adopt an approach which resulted in the wholesale allowance being set too low to enable the hypothetical “typical” efficient supplier to cover its costs, it would need to make an adjustment to rectify that situation.

This thinking is reflected in the section of Appendix 3 of the consultation document specifically dealing with its regard to the s.1(6)(d) need to ensure suppliers who operate efficiently are able to finance their activities. GEMA stated:

“we are minded to consider this from the perspective of setting the level of the cap at a level such that a supplier with efficient costs, noting the inherent uncertainty in assessing the efficient cost benchmark, could finance its activities. Our principal measure is the ability of an efficient supplier to make a long run normal rate of return under a default tariff environment.”

This was echoed in paragraph 4.15 of Appendix 4 to the September consultation, where GEMA stated that its primary function was to set an allowance that gives “a realistic allowance to cover the costs that suppliers might incur delivering energy.” Whilst it had indicated in paragraph 4.13 that it was proposing to adjust its approach from the proposal in May “to improve outcomes for customers and ensure the wholesale allowance reflects underlying costs more closely,” the adjustment was not intended to create a situation in which the allowance would be insufficient to cover the likely costs of an efficient supplier. Of course, that did not mean that everyone’s actual costs would be covered.

In paragraph 4.16 of Appendix 4, GEMA referred to evidence (from the RFI exercise) that large proportions of energy for delivery in early 2019 would have been bought in early 2018 or even earlier, predating any detailed discussion or consultation on the default tariff cap. In the following paragraph it acknowledged that suppliers’ existing purchasing strategies created a risk that their wholesale costs may be different from those assessed in the default tariff cap. It then said it did not consider it likely that suppliers would seek to undo previous actions to align to the model for setting the initial periods of the default tariff cap:

“suppliers are unlikely to sell all of their contracts, and then buy them back again at newly available prices. Even if they did so, this would incur costs (reducing potential profits) or realise potential losses. If market prices had increased since that supplier purchased the contract initially, they would make a gain (adjusting for transaction costs). As such, the default tariff cap would provide a higher allowance than their actual cost required. If the market had fallen since the supplier had first bought contracts, they would incur a loss. Our default tariff cap would set a lower allowance than the supplier’s actual cost required. In the latter case, we may have needed to consider a temporary adjustment to allow suppliers to recover their actual cost.”

Thus, GEMA indicated to consultees that there were two reasons for its assessment of the sufficiency of a wholesale allowance based on the proposed adjustment to the observation window, namely:

GEMA must have carried out some assessment that an allowance based on the observation window of February-July 2018 would cover (or sufficiently approximate to) the estimated costs of a typical efficient supplier who had aligned with the indicative April-September window in May. Otherwise it could not have given reason (ii). Despite this, the basis upon which GEMA had reached that conclusion does not appear on the face of the consultation document.

In paragraph 4.18 GEMA referred to the significant rise in market costs since the beginning of April 2018. It said that the effect would be that suppliers that purchased energy for customers on an SVT tariff in advance would incur significantly lower

In the concluding paragraph of that section, 4.21, GEMA said that on balance, it judged that using the February to July observation window gave “a sufficient approximation of the cost suppliers actually incur” and was the most reasonable position to adopt. For larger suppliers it considered that the default tariff cap was likely to be an overestimate, although by a lesser amount than the proposed approach in the May consultation. Again, it did not explain why.

The responses to the September consultation comprehensively disabused GEMA of the validity of its primary assumption that suppliers would not have adjusted their hedging strategy to align, so far as possible, with the observation window that had been indicated in the May consultation paper. Behaviour which GEMA had characterised as “unlikely” was flagged up as being very likely indeed, though some suppliers had reacted to the indicative window more swiftly than others. All the Big Six either told GEMA or implied that they had aligned to the April-September observation window in reaction to the May consultation.

This meant it was no longer safe for GEMA to rely on the premise that a “typical” supplier would not have taken steps to align with the indicative window in May 2018. It followed (as GEMA now accepts) that reason (i) could no longer be regarded as a sound basis for assuming that an allowance based on the February-July window would suffice to cover/approximate to the costs. The evidence of alignment therefore threw into sharp focus the importance of the validity of reason (ii).

By the time GEMA’s proposed change of approach was announced in the September consultation, suppliers that had aligned their purchasing to the May-September period within the previously indicated window would still have some energy to purchase in what remained of September 2018. They would have no option but to purchase that energy at higher prices, as the market continued to rise. Moreover, any purchases they had made in August and September 2018 would no longer be reflected in the proposed observation window (so the actual costs to them of those purchases would be higher than an allowance based on prices for the same quantities of energy sold in the February-July window). They had no opportunity to align with the new observation window, which had already passed. Any February and March 2018 contracts which an aligning supplier had sold on the market at May 2018 prices, would now be reflected in the new observation window.

In its response to the September consultation, British Gas objected to the proposed change in the observation window and, like the other major suppliers, tried to persuade GEMA to adhere to the proposal it had put forward in May. It provided GEMA with confidential information about its own hedging strategy in order to demonstrate to it that the profits it had already made from its historic forward purchases before market prices began to increase sharply in April 2018 were insufficient to offset the costs of purchasing contracts at the new higher prices. It told GEMA that it would suffer a loss of many millions of pounds if the proposed change in the observation window went ahead, and it explained why.

British Gas was the only supplier to provide that level of detail, but 11 other suppliers raised concerns that the February to July observation window could or would lead to an under-recovery of costs. As one of British Gas’s competitors put it: “Ofgem’s changing approach to wholesale market hedging assumptions for the first period of the cap (combined with rising wholesale prices) has resulted in costs that will not be recoverable under Ofgem’s proposal”.

Suppliers also complained that GEMA had not put forward any reasoned or evidenced justification for the assessment of sufficiency in the case of suppliers who had taken steps to align with the indicative window. That was fair comment. It is apparent from the internal documentation disclosed as part of GEMA’s duty of candour that, at the time of the consultation, GEMA was alive to the possibility that suppliers would challenge the evidence base for its assessment. It recognised the importance of robust analysis. On 20 September 2018 Ofgem’s Head of Tariff Cap Design recorded that: “we are likely to be particularly challenged on what is our evidence base in relation to suppliers’ hedged position for Winter 18/19”.

In the October Board paper, Ofgem set out the thinking behind the two reasons given in the September consultation for the assessment of sufficiency. They also summarised and analysed the responses to that consultation. They stated that they believed the allowance (based on the revised observation window) was on average sufficient to cover suppliers’ actual costs, and indeed would marginally overcompensate large suppliers. They said that even if suppliers had reacted to the May proposal (something which the authors of the paper belatedly acknowledged would have been the prudent thing to do) “we have calculated that they should more or less ‘break even’ overall in relation to this initial period.” Logically, that involved treating the articulated position of British Gas as being atypical.

In paragraph 17 of the paper, the authors stated that they were confident that they had recent and reliable evidence on what the suppliers serving the majority of SVT customers were doing prior to May 2018, and that evidence suggested that deviating in a major way from the prevailing approach (a long-term purchasing strategy) would be a bad decision, and therefore unlikely. This was a reference to the evidence gathered in response to the RFI in 2017, not to anything said in response to the consultation.

In paragraph 24 the authors explained why they considered that the net impact of changing proposals regarding the observation window would “marginally overcompensate” suppliers. This was done by reference to a table, which showed the stages that a large supplier would have taken if they “purchased gas contracts starting with our typical approach, switched to our May proposal in May and then reacted to our September proposal in September.” The table indicated an overcompensation of such a supplier against the allowance of 1%.

The authors of the paper acknowledged that their analysis was a simplified case, but they considered that the deviations with suppliers’ actual approaches would be “relatively minor” compared to the level of overcharging that customers would have faced had they maintained the May proposal. The comparison was simplified because they used a generalised representation of suppliers’ strategies, which they described as an “average”.

The published Decision reflected those views. In paragraph 3.143 GEMA said it considered that it had sufficient evidence and understanding of how suppliers typically purchase energy for SVT customers to assess the impact of its proposals on SVT customers and suppliers, and that it did not agree with the suggestion that it ought to have sought further evidence or information before proposing to set the allowance using an observation period between February and July 2018. In paragraph

GEMA then stated in paragraph 3.145:

“given their stated rationale for their approach to purchasing energy as provided to us in 2017, it is unlikely that the 6 largest suppliers would have deviated materially from that approach in the succeeding months (before our May consultation)”. [Emphasis added].

In other words, it assumed that the Big Six would have largely continued with a longterm purchasing strategy (of 18 months or more) until May 2018. It went on to say that, given market conditions, suppliers may have adjusted their strategy in small ways, but not in a way that changed the overall rationale, and that no development in the market since 2017 suggested a substantial change in approach before May 2018. That ignored the proposed introduction of a default tariff price cap, which GEMA itself acknowledged was not a consideration informing any supplier’s hedging strategy at the time that the data that it relied on was gathered. It is now known that this development did have a significant impact on the general nature of suppliers’ hedging strategies after October 2017.

GEMA then went on to explain how it had carried out an analysis of the impact of setting the allowances using three different approaches and combinations of them. It looked at:

As the subsequent passages in the Decision make clear, the “medley” approach assumed that the supplier continued to buy all or most of its energy using the socalled “typical” 18-month hedging strategy until the publication of the May consultation paper, when it sold its contracts at the prevailing market price and purchased just under 30% of the volume it needed. From then until the announcement of the change of proposed observation window at the beginning of September 2018 it was assumed that the supplier would have purchased enough contracts to obtain over 85% of the contracts it needed, the remainder being purchased at the prevailing market price at the beginning of September. On that basis, the calculation was made that the proposed allowance based on the earlier window would overcompensate such a supplier by just over 1%.

GEMA concluded in paragraph 3.165 that on average the analysis suggested that the wholesale allowance in the first cap period was sufficient for a supplier using a typical purchasing strategy; or for one that used a typical purchasing strategy and then adjusted it in response to the two consultations. It recognised that some suppliers might have more favourable or less favourable positions than average, but it pointed out that was an inherent outcome of the cap, as the Act requires one allowance for all suppliers.

Given that it is now accepted that no supplier would have failed to adapt its hedging strategy to the observation window indicated by GEMA in May 2018, the first part of the analysis is based on a counterfactual assumption. However, that would not matter if the second part of the analysis was sound.

Mr Fordham submitted that the “medley” approach was based on the equally invalid assumption that a “typical” supplier would not have materially deviated from its longterm forward purchasing strategy as observed in April/May 2017, until it aligned with the indicative window in May 2018. That is self-evidently not the same as the initial premise that all the suppliers would have had a long-term forward purchasing strategy in 2017, which would have led to their buying some of their energy requirements for Q1 2019 at least 18 months in advance. It presupposes that they would have largely maintained that strategy until May 2018. I accept those submissions; the October Board paper and the Decision cannot sensibly be interpreted in any other way.

That assumption (“the continuity assumption”) was not articulated until the Decision was published, even though it underpinned GEMA’s assessment in the September consultation that the net impact of changing the proposed observation window would marginally overcompensate suppliers who had aligned to the April-September window in May 2018.

Consequently, Mr Fordham submitted, the suppliers had no opportunity to provide any input into the consultation which would have shown the continuity assumption to be just as flawed as the assumption that they would not have aligned with the indicative observation window in May 2018. The continuity assumption was not ventilated or tested in the consultation process, and it was unfair of GEMA to reach a conclusion about the sufficiency of the allowance without giving the suppliers any prior warning about this key aspect of its reasoning. I shall consider those submissions later in this judgment.

The fact that the continuity assumption was unsound has been demonstrated by the direct evidence in this case from British Gas and from other big suppliers such as Npower, SSE, and Scottish Power. There is also evidence that a fifth member of the

“Big Six” provided information to GEMA in the course of the September consultation process which demonstrated that the continuity assumption was wrong so far as they were concerned. Most suppliers had moved away from the long-term forward purchasing strategies that they had used at the time of the responses to the RFI in 2017 towards shorter hedging strategies, in anticipation that the default tariff cap would be set by reference to a shorter observation period. The expert evidence explains why this made economic sense.

It is unnecessary for the purposes of this judgment to refer to the details of the suppliers’ evidence, much of which contains confidential and commercially sensitive information. Suffice it to say that I am satisfied that five out of the “Big Six” did make material changes to their hedging strategies before May 2018, and that they did so because of developments in the market after the evidence-gathering exercise which took place in 2017, particularly in reaction to the publication in October 2017 of the draft legislation relating to the introduction of the price cap. There is no evidence as to the position of the remaining member of the “Big Six” and therefore one can make no assumptions about its hedging strategy one way or the other. This does not matter, because what is now known about the behaviour of the majority is enough to establish that, as a generalised representation of the behaviour of a typical large supplier, the continuity assumption was fundamentally flawed.

The fragility of the continuity assumption was apparent to at least one person who was working for GEMA. In an internal email sent on August 10, 2018, that person made the following prescient observation: “Big issue is of course that we don’t know extent to which any given supplier will have actually hedged Q1 201[9] in 2017: given the election in May and price cap announcement in Oct this would have been quite risky.” [ Note – the email actually refers to Q1 2018, but in context this is obviously a typing error].

Miss Rossington of Ofgem also refers in her evidence to an internal email sent on 21 August 2018 that noted two reasons why suppliers might have altered their hedging strategies since 2016, one of which was the announcement of the price cap in October

However reasonable a factual assumption underlying an analysis or calculation might appear subjectively or objectively, unless it is clearly articulated to those likely to be affected by it, and they are given the opportunity to comment upon it, there is no means of testing whether it is sound. GEMA did articulate its assumption that it was unlikely that a supplier would align with what remained of the indicative AprilSeptember window, and the responses to the September consultation demonstrated that the assumption was incorrect. Had it done the same with the continuity assumption, it would have discovered that assumption to be equally incorrect.

I have already rejected Mr Maclean’s submission (in the face of overwhelming evidence to the contrary) that GEMA did not make the continuity assumption. Mr Maclean further contended that it was immaterial to the Decision whether a supplier maintained a long-term hedging strategy until May 2018. He submitted that GEMA had given a clear indication in the May consultation that (because of the long-term purchasing strategies revealed in 2017) it expected that “some suppliers have already bought large amounts of energy for the initial cap period.” No-one had contradicted that statement. If that was correct, then one would expect that basing the allowance on prices in an observation window in a period after market prices had gone up would overcompensate those suppliers, because irrespective of their individual hedging strategies, they would have paid lower prices for most of the energy they had already acquired for delivery in Q1 2019.

I agree that this would be a reasonable view to take, but whether it was right or not in the case of a supplier who aligned in May 2018 would depend, inter alia, on how much profit it made from unwinding its existing contracts and how much of that profit would be absorbed by the costs to it of alignment (including the cost of purchasing energy at the higher prices in the periods within the May-September observation window which could no longer be “matched” with the prices in February and March).

Mr Maclean pointed out that the relevant proposal that GEMA was consulting on in September 2018 was the proposal to shift the observation window. He emphasised that the purpose of this aspect of the consultation was to find out whether the impact of the proposed change to the observation window to a period entirely in the past, would still give rise to a sufficient approximation of the costs of the notional typical supplier. He accepted that GEMA assumed (on the basis that they would have purchased much of their energy for Q1 2019 before the market rose) that the big suppliers would still have sufficient long-term contracts available to unwind to cover those costs. However, only the individual suppliers would be able to tell GEMA whether that was true of them or not, and that was what the consultation was designed to discover.

Mr Maclean submitted that the suppliers were a sophisticated body of consultees who had ample opportunity to explain in their responses whether the effect of the change would be to leave them so exposed on costs that it would be inappropriate to adopt the earlier observation window. Indeed, British Gas did so. It was unnecessary for these purposes for GEMA to ask or to be told about individual hedging strategies. Even if GEMA had been supplied with that information, it would have been regarded as irrelevant for the purposes of ascertaining whether an allowance based on the earlier observation window would be enough to cover (or approximate to) a typical efficient supplier’s likely costs.

I cannot accept that line of argument, which overlooks the fact that for the purpose of making its assessment of sufficiency/approximation of costs, an assessment adopted in the Decision, GEMA chose to base its calculations of the notional costs of the typical efficient supplier on generalised assumptions about how such a supplier would have behaved, including when it would have bought its energy for Q1 2019, how much of its requirements it would have purchased (and kept) before the market prices escalated, what the average cost of purchasing that energy would have been, and whether it would have aligned in May. Even though they were generalised, those assumptions still had to be realistic in order to produce a fair basis for making the assessment.

“Typical” in this context means representative or archetypal. In the course of the hearing, I asked Mr Maclean how behaviour could be described as “typical” if it transpired that no supplier would have behaved in that way. He responded by pointing out that the exercise was not designed to precisely replicate the pattern of behaviour of any individual supplier, and that “typical” in this context really meant average. That is no answer to the objection. If there is a sufficient foundation for them, one can properly make assumptions about likely generic behaviour, e.g. that all or most suppliers would align to the observation window in response to the May consultation. One could legitimately then go on to make a further exemplar assumption for modelling purposes about when a prudent supplier might have been expected to align, without necessarily picking a date in May that any individual supplier happened to choose. But one cannot fairly operate on an assumption about the “typical” behaviour of a notional supplier that bears no resemblance to reality.

A consultee would be unable to comment on whether the proposed allowance would sufficiently approximate to the costs of a typical notional supplier if they did not have enough relevant information about how GEMA proposed to estimate those costs. It was not enough to know that GEMA assumed, on the basis of what it had been told in May 2017, that suppliers would have bought large amounts of energy for Q1 2019 a long time in advance. There had to be a link between that initial premise and the conclusion that an allowance based on an historical observation period to which it was no longer possible to align would be sufficient to cover/represent a typical supplier’s costs even if it had reacted to the May consultation by aligning, as far as possible, to the April-September observation window. That link was the continuity assumption.

The model of the typical supplier which was adopted by GEMA for the purposes of the October Board Paper took 18 months as the typical hedge period, based on the information gleaned in May 2017. That, being the shortest period indicated for a longterm hedging strategy, would result in the smallest amount of the energy required for Q1 2019 being accumulated by the notional supplier before prices started to rise. If the model showed that the costs of such a supplier were likely to be covered if the February-July observation window was used, then logically the costs of any supplier who had adopted a longer-term purchasing strategy would be covered as well. However, that would not necessarily be true of a supplier who had switched to a 12 month or 9 month hedging strategy, or a mixture of long-term and short-term.

At one point, Mr Maclean submitted that it was calculated that the point at which the notional supplier utilising an 18 month hedging strategy would “tip” from having enough long-term contracts to unwind to cover the costs it would incur by reference to the newly proposed observation window just so happened to be May 2018. That was pure coincidence. It had nothing to do with the fact that there was a consultation in May, or that suppliers would have started aligning their contracts in May to meet the indicative window.

I am unable to accept that submission, for which there was no supporting evidence. Miss Rossington’s evidence and the October Board paper appear to me to contradict it. That paper and the Decision itself demonstrate that the calculation that an allowance based on the earlier observation window would be sufficient to cover the costs of a typical large supplier who aligned in May did depend on that supplier having maintained a long-term hedging strategy until the May consultation and only then “adjusted” its strategy by aligning in reaction to the May consultation.

For the purposes of its modelling GEMA used an assumption, favourable to the suppliers, that the notional supplier would have aligned with the indicative window at the earliest opportunity after the May consultation was published. That was fair, but it only provided one factor in the calculation of the costs that a typical supplier who aligned in May would have incurred by reason of that alignment. The other critical aspects of the calculation involved an assessment of how much energy the typical supplier would have purchased for Q1 2019 (and kept) before aligning in May 2018, how much that would have cost, and how much profit the supplier would have obtained by selling it in May.

The October Board paper spells out how the continuity assumption fed into the calculations and how fundamental it was to the outcome. Paragraph 24 demonstrates that it was assumed that the supplier would have purchased 76% of the energy it needed between April 2017 and May 2018 - because it would have used an 18-month rateable strategy until 24 May. That assumption fed into the average price calculations and the calculation of the profit that the supplier would make by selling their original position in May. It also fed into the further assumptions about how much energy would have been bought during the May-September window and how much energy the supplier would still need to buy when the change of proposal was announced in early September.

Therefore, whilst I accept that what GEMA was considering, in the light of the September consultation responses, was the effect that moving the observation window back (and thereby reducing the allowance) would have upon the ability of a typical supplier to cover its likely costs, that ability was itself predicated upon an assumption that the “typical” supplier would maintain a long-term hedging strategy at least until it aligned with the originally indicated observation window (the now-abandoned reason (i) had depended on the “typical strategy” being maintained beyond May 2018). That assumption was in turn based on the misconception that nothing had happened in the market to cause suppliers to change their hedging strategies in any material respect before May 2018.

I am satisfied that GEMA chose to base that aspect of the assessment on a “typical” hedging strategy before alignment (including the continuity assumption) that was not only atypical but unrealistic. The continuity assumption was undoubtedly material to the assessment of sufficiency which underpinned the setting of the allowance for Q1 2019 based on the February-July observation window.

The short answer is no. It is not obvious on the face of either consultation paper. Paragraph 4.16 of Appendix 4 to the September consultation simply refers to the initial premise drawn from the responses to the RFI - before there was any detailed discussion of the price cap, as GEMA expressly acknowledged. The link between that premise and the conclusion that prices in the Feb-July window would be a sufficient approximation of costs incurred by a typical aligning supplier is not apparent. There is nothing in that paragraph to suggest that GEMA considered that those strategies would have been followed without material change until May 2018 (let alone the reason why it thought that).

In his oral submissions, Mr Maclean relied on paragraph 4.21 of Appendix 4 to the September consultation, but that simply conveys the conclusion reached by GEMA, not the reasons for it or the factual premises underlying it. I do not accept that the inference could fairly be drawn from any of the relevant passages in the September consultation that GEMA was assuming that its “typical” supplier would be maintaining a long-term hedging strategy of at least 18 months from April 2017 until at least the publication of the May consultation.

The suppliers’ evidence is that they had no knowledge that the assumption underpinned GEMA’s assessment of sufficiency. I accept that evidence. Indeed, they would hardly have complained of the absence of a reasoned or evidenced justification for the assessment of sufficiency in the case of suppliers who had taken steps to align with the indicative window, if this critical aspect of GEMA’s reasoning had been made apparent to them or was as obvious as Mr Maclean submitted it to be.

The continuity assumption and the underlying reason for it were only communicated to the suppliers in the Decision itself, when it was too late for them to do anything about it except ask GEMA to reconsider, which it refused to do.

This is a case which both parties accepted turns on the facts. The relevant legal principles were uncontroversial. Consultation, in accordance with basic public law standards, is required to operate so that the decision-maker’s thinking is made transparent, in order that formative stage thinking engages informed responses from the body of consultees, leading to conscientious consideration, resulting in a lawful decision.

A consultation process must be fair, in the sense that it affords a fair opportunity for those to whom the consultation is directed adequately to address the issue in question before a final decision is made: see the well-known principles adumbrated in R (Moseley) v Haringey LBC [2014] UKSC 56 [2014] 1 WLR 3947 at [25], and R (Keep the Horton General) v Oxfordshire Clinical Commissioning Group [2019] EWCA Civ 646 at [18]. The aspect of the obligation of fairness that is particularly relevant here is the requirement to provide consultees with sufficient information. As Hickinbotttom LJ put it in R (Help Refugees Ltd) v Secretary of State for the Home Department [2018] EWCA Civ 2098, at [90]:

“Consultees must be told enough – and in sufficiently clear terms – to enable them to make an intelligent response.”.

In the present case that did not happen. GEMA did not communicate the continuity assumption (or the underlying reason for making it) when consulting in September 2018 or at any other stage before it made the Decision. In my judgment, in order to meet the requirement of fairness it should have done. GEMA was consulting on a proposal to move back the observation window for the transitional period from that which it had proposed to adopt at the time of the previous consultation in May, to an historic period that suppliers would no longer be able to match. The question whether an allowance based on the newly proposed observation window of February-July would be sufficient to cover (or fairly represent) a typical large supplier’s costs was of fundamental importance to the decision whether to adopt that window.

GEMA recognised this. It also knew that if suppliers had aligned with the observation window consulted on in May, that would have some bearing on the answer. It flagged up the fact that it thought this behaviour was unlikely – which sufficed to enable the suppliers to provide information about whether that was right or wrong. However, and for no apparent reason, it chose not to share with consultees the critical factual assumption about a typical supplier’s behaviour which underpinned its assessment that even if such a supplier did align, it would be marginally overcompensated (or at least break even). It thereby left them completely in the dark as to why it had reached that view. They responded to the proposal as best they could, but they could not address this key aspect of GEMA’s reasoning for the simple reason that they were not told about it.

GEMA failed to explain that the continuity assumption formed any part of its thinking until the Decision itself. It was not mentioned in any meetings with suppliers between the consultation and the promulgation of the Decision letter. The suppliers had no chance to explain to GEMA that it was labouring under a misapprehension. The fact that the assumption was a generalisation is no answer; it was a critical factor in the assessment made by GEMA of the sufficiency of the allowance, and therefore it had to be communicated to the suppliers.

That conclusion, by itself, is sufficient to warrant the grant of relief to British Gas.

However, for the sake of completeness I shall briefly deal with Mr Fordham’s additional grounds. He contended that there was also a breach of the duty of enquiry which falls upon a decision-maker to take reasonable steps to acquaint itself with the relevant information to enable it to take a properly informed decision. It seems to me, however, that this was more a case of the decision-maker failing to communicate material information to the consultees, than of a decision-maker failing to ask them to provide it with material information or to seek further relevant evidence.

GEMA could have asked the suppliers a specific question about whether they had deviated in any material way from the strategies they had revealed in response to the RFI, but it did not need to go that far. If GEMA had said in the consultation paper what it eventually said in the Decision itself, the suppliers would have been made sufficiently aware of the continuity assumption, and of its importance to the assessment, to be able to explain to GEMA why it was wrong. The amount of information they decided to provide to GEMA to prove that its assessment was based on a fundamental misapprehension was a matter for them. It was not incumbent on GEMA to seek the details of each supplier’s hedging strategy, but it did need to gather enough information to enable it to be satisfied that its assumption was correct. The easy way to do that was to make that assumption known to the consultees.

GEMA stated that it believed it had enough reliable and recent evidence to form a view. In hindsight, that was perhaps an unwise stance to adopt given that the evidence

Mr Fordham also relied on E v Secretary of State for the Home Department [2004] EWCA Civ 49 [2004] QB 1044 as authority for the proposition that if there is a mistake of fact (for which the objecting party was not responsible) that is uncontentious and objectively verifiable, and the mistake played a material (but not necessarily decisive) part in the decision maker’s reasoning, the decision may be susceptible to judicial review. He submitted that the principle was also engaged if the decision maker formed a mistaken impression which played a material part in its reasoning, or acted upon a wrong assumption, or if the decision was taken on facts unsupported by evidence. Again, there was no issue between counsel about these principles, but only as to their application to the facts of this case.

I am satisfied on the evidence that these requirements are made out and that, irrespective of the fairness or otherwise of the consultation, British Gas would succeed on this ground also. The continuity assumption was factually incorrect, and so too was the reason given by GEMA for making it. The hedging strategy characterised as “typical” had ceased to be so; suppliers did not largely maintain longterm hedging strategies. It was not a fair basis for the assessment that GEMA made.

Mr Maclean realistically did not pursue the argument originally raised by GEMA that this was a case in which s.31(2A) of the Senior Courts Act 1981 would operate so as to lead to the refusal of relief.

I am satisfied that for the reasons stated, British Gas succeeds in its claim and is entitled to the declaratory relief that it has sought. As the decision is not to be quashed, GEMA will have to reconsider the allowance for Q1 2019 in the light of the information that it now has, and make such adjustments as it considers appropriate in the light of that reconsideration. For the avoidance of doubt, if on reflection GEMA decides that it should seek further information from the suppliers before it reconsiders the allowance, there is nothing to prevent it from doing so. It does not follow from the fact that I have preferred not to treat this as a case of a decision-maker failing to carry out necessary inquiries or gather sufficient evidence before it made the Decision, that it should not carry out further inquiries now, if it so decides. Finally, I would like to pay tribute to both legal teams for the excellence of the presentation of the parties’ legal submissions both in writing and in oral argument.