As has been widely reported in the industry, there has been a spate of gas price arbitrations in which buyers have largely succeeded in having contract prices revised in their favour, sometimes by hundreds of millions of dollars.1

This  article  looks  at  the  theory  underlying price variation clauses, and how to approach a question about the meaning of an individual clause. The approach that a tribunal may take to such clauses may be informed by going back to basics, looking at the very reasons for having such a clause, before considering some of the mechanisms that the parties might employ. Practically, though, it should be borne in mind that what ultimately matters is the specific wording of the clause (or draft clause) in front of you, rather than what was decided about a different clause in another award.

Why have a price which changes?

A typical gas sales agreement (“GSA”) provides for the sale and purchase of a minimum quantity of gas each year for 15-20 years, or the producing life of the field, with a provision for the price to change in certain circumstances. In theory, however, a producer could agree to supply, and a buyer to buy, a given quantity of gas every year for the next twenty years for a fixed price.

One obvious reason why parties might instead want to provide for the price to change is to account for inflation. The cost of labour and materials to operate, maintain and eventually decommission the field can be expected to increase in line with background inflation. The profits of a supplier that is paid the same price throughout, will tend to fall inversely to this inflation.

This is addressed by providing for price to be adjusted periodically according to a published measure of inflation. Overall, the effect should be neutral. The buyer pays more for gas, but will also be charging its customers more. Disputes will only arise if the relevant index ceases to be published, or the way it is calculated changes.

Assume, then, a contract which still provides for a fixed price at the outset, but also makes some provision for it to increase in line with inflation.   It can be seen that, if, in the next 20 years, the realisable value of the gas to the buyer rises relative to the contract price, the buyer will reap all the rewards. If the realisable value of the gas to the buyer falls relative to the price, the buyer will suffer all the loss.

What would  the  initial  price be? The producer  must  charge enough to cover:

  1. the cost of capital assets used to produce gas (wells, production hardware, pipelines);
  2. interest on loans it used to buy those assets; and
  3. the likely cost of operating, maintaining and decommissioning those assets (with a contingency for geological, operational and regulatory risk).

That is the minimum price any (sensible) producer will agree to. The result is a low risk for the producer. It will earn a guaranteed, inflation-adjusted minimum return for the next twenty years, insulated against any loss. Since the producer’s risk is low, it can expect a commensurately low return. The initial price will be set at such a level as to provide the supplier only a small profit.

A producer might have more appetite for risk, and wish to share in the benefit of any increase in the realisable value of gas. The only way to reallocate the risks and rewards is to provide for the price which the buyer must pay the producer to change as realisable value changes. That value depends on what the gas is for.

What is the gas for?

When looking at a price variation clause and trying to work out what the parties’ bargain was it is also relevant to ask what the buyer wanted:

  1. A domestic gas supplier buys gas to sell to the domestic end consumer to use for heating and cooking. Such a company usually has long term commitments in terms of infrastructure and maintenance. Its customers’ demand for gas is relatively price inelastic – they cannot easily shift to using electricity if the relative price of gas increases. Domestic gas supply may also be regulated, to prevent the supplier passing the whole of any increase in the cost on to the consumer.  If the cost of gas falls, it may be forced to pass part of its saving on to the consumer.
  2. A  domestic  power  supplier  uses  gas  to  generate  electricity which it then supplies to the domestic end consumer.  It must compete with those who generate electricity using other fuels. Its customers’ demand is more price elastic. If the price of gas increases relative to those fuels, it can only pass on so much of this increase before the customer shifts to another supplier which uses a cheaper fuel.
  3. Wholesale power suppliers and wholesale gas suppliers supply gas and electricity to domestic power and gas suppliers. They may buy the gas with a view to satisfying their obligations under long term contracts or they may buy it speculatively, in the hope of being able to sell it at a profit on a much shorter term spot market or futures market.


To provide for re-allocation of changes in the realisable value of gas, a GSA usually includes some form of indexation. A basic indexation formula would be:

Contract price = initial fixed price x (realisable value of gas in review period / realisable value of gas in base period)

If the initial fixed price was 40p and the value was 50p, then realisable value is split 80/20 between producer and the buyer. If realisable value increases 10p to 60p, the price becomes 48p = 40* (60/50). It can be seen that the extra 10p of realisable value is being split between the parties in the original 80/20 ratio: 8p to the producer and 2p to the buyer.

Sometimes the contract will only allow the price to move between a minimum “floor” and a maximum “ceiling” price, or only to move by a maximum amount each review period. Alternatively, the formula might provide for the price to increase at a lower rate above a certain point, and to decrease at a lower rate below a certain point (“S-curve” formula).

The reason for departing from the linear relationship at high and low prices is that: (i) there will be a price above which the buyer will be unable to make a profit; and (ii) there will be a price below which the producer will be unable to make a profit. Changing the formula which applies at the upper and lower bounds prevents or limits the potential loss.

Occasionally there will be disputes about the application of the formula, but these are uncommon – it is just a mathematical exercise. Historically the principal problem with indexation formulae has instead been in finding a reliable objective measure (or proxy measure) for the gas’s ‘realisable value’. The value an individual buyer can realise depends, in part, on its cost of sales which in turn depends on how it manages and structures its business. Producers will not want to make their remuneration depend upon how efficiently or otherwise the buyer manages its business. Working out the real realisable value would also involve a prohibitively frequent, costly and complex accounting exercise, which would itself give rise to disputes.

The solution  is to  use the  ‘market price’  of gas  to represent realisable value. In the past, however, there was little gas to gas competition, and so no ‘market price’. Gas was less a global commodity, but was most often extracted to be sold into the local market, for a price negotiated with the local buyer. The buyer would often be a national or regional monopoly. As a result, many older contracts used changes in the price of an alternative fuel – such as oil or coal - or changes in wholesale electricity prices, or some kind of blended figure, as a proxy for changes in the value of gas.

Gas has since become more of a global commodity, with a price which is less tied to oil.  There has been a downward pressure on gas prices due, (for example) to:

  1. increased demand for gas as environmental regulation has made it relatively less expensive compared to more polluting coal;
  2. regulation discouraged flaring, meaning more gas is recovered and brought to market;
  3. more pipeline capacity constructed,  allowing gas to be used economically further from the well head;
  4. more use of LNG technology allowing bulk shipping of gas when there is no pipeline;
  5. break-up of national monopoly suppliers;
  6. less  effective  cartel  behaviour  by  gas  producers  than  by  oil producers;
  7. increased the supply of gas due to use of hydraulic fracturing in shale, which has not increased to oil supply to the same degree.

Purchasers under older contracts have therefore been forced to pay a price which has escalated with the oil price, when the real market price of gas has fallen. This inflated cost obviously makes the purchaser less competitive and profitable, and vulnerable to competition from rivals who are able to take advantage of the low market price to undercut the purchaser. It appears to have been this disconnect between oil price and gas price which has prompted the recent spate of claims under price-reopener clauses.

Price re-opener clauses

A price re-opener clause typically:

  1. identifies ‘trigger’ criteria which cause or permit  the  review procedure to be invoked;
  2. sets out a procedure for negotiation, and provides for dispute resolution (typically arbitration or expert determination) if this is unsuccessful; and
  3. provides  criteria  for  determining  how  the  formula  is  to  be changed.

Examples of trigger criteria are:

  1. A contract might give either party the right to initiate a price review periodically - say every 3 years or 5 years.
  2. A contract might give each party the right to initiate a limited number of price reviews at any time during the contract term (so-called “wild cards”).
  3. Review might be triggered if some objective benchmark is met – such as if the reference price changes more than a certain % in a given period.
  4. Frequently the trigger will be less precisely defined, e.g. if there is:

“any substantial change in  the economic circumstances relating to this Agreement and either Party feels that such change is causing it to suffer economic hardship”2,


"if at any time either party considers that economic circumstances in Spain beyond the control of the parties, while exercising due diligence, have substantially changed as compared to what it reasonably expected when entering into this Contract … and the Contract Price … does not reflect the value of Natural Gas in the Buyer’s end user market”.3

With triggers in this last category, it may not just be changes in the reference price which trigger a review. To give one example, a party’s operating costs might increase (through taxation or regulation) such that the floor and ceiling prices, or S-curve formula no longer protects that party against loss as it did before, and a party relies on this as “substantial hardship” or a “material change in economic circumstances”.

Depending on the detail of the clause there may therefore be a threshold question about whether trigger criteria have been met, and/or whether the right to a review has been properly exercised (for example, whether any notice requirements have been complied with).4

Assuming the trigger criteria have been met, the contract will set out wording prescribing how the parties want the formula to be changed. An example might be:

“The revision of the price shall consist in adapting it in a reasonable and fair manner to the economic circumstances then prevailing on the imported Natural Gas market and on the market for the other imported energy supplies competing with this production in the East Coast and Gulf Coast areas of the United States of America within the framework of long term contracts. The parties shall take into account the individual characteristics of each of the above products including the quality, the continuity of deliveries, the production and transportation costs, etc ...”5

Sometimes the clause will expressly require the pricing formula to be changed so as to restore the original relationship between the contract price and the market price, or – if the index is no longer appropriate – for an appropriate index to be substituted. An alternative wording is to require that the gas be re-priced so as to allow it to be marketed “economically” in the end market.

Gas price arbitrations

Often the contract will provide for a form of structured negotiation. If the parties are unable to agree on what change the contract  wording  requires,  the  contract  will  usually  provide  for arbitration (less often expert determination and rarely litigation).

There is sometimes an issue as to the scope of the decision maker’s authority - in particular as to which variables the tribunal can change, and from what date. Similarly, some clauses impose upper and lower limits on what changes the tribunal can make. This can be by reference to the position which the parties took in any negotiation. For example, in a ‘pendulum’ or ‘baseball’ arbitration, the tribunal is restricted to a choice between the parties’ proposed prices. There may also be an issue as to whether a tribunal can take into account events which postdate the trigger notice, or whether it is restricted to a ‘snapshot’ as of the date the review was initiated.

Occasionally, an arbitration clause will provide for the tribunal to decide the issue ex aequo et bono or will provide for a governing law which excuses non-performance or allows bargains to be (in effect) re-written on grounds of supervening hardship.6 More commonly, though, a tribunal will simply be concerned to interpret and apply the contract – specifically that part wherein the parties agreed when, and how, their price or formula by which it is to be determined should be revised.  As noted above, various wordings are used,  varying from an  express  requirement  that  the  pricing formula be changed so as to restore the original relationship to a bare provision with provision for the formula to be revised if it is no longer ‘fair’ or ‘equitable’ or ‘reasonable’.

From the perspective of an English lawyer, many price re-opener clauses really seem to be aimed at essentially the same thing – giving effect to the substance of the bargain which the parties originally struck. The best guide to what the parties mean by “fair”, “equitable” etc. is what they were freely agreed to. The starting point, then, is a detailed analysis of the original bargain, and of the parties’ respective positions:

  1. What were the parties seeking to achieve when they agreed a price adjustment formula, rather than a contract which just fixed the price for the term?
  2. Was the input, or function, which is in contention just a means to achieving that end, or was it an integral part of the bargain?
  3. If so, would the parties’ original end now be better achieved by changing that input, or formula?

It may be much harder to answer these questions convincingly than to ask them. By way of an illustration, though, consider an older contract which provided for indexation by reference to the price of oil. It might be argued that the parties did not really intend to tie the price to oil – rather oil was a proxy for the market price of gas, at a time when there was no market price for gas.  The oil price was just the best available means to that end, and now there is a new measure which better represents what the parties wanted.

But suppose the exact same contract had been entered ten years later, when a comparator gas price was readily available. Now the linking of the price to oil begins to look like a deliberate choice by the parties – an integral part of their bargain, or an end in itself - rather than a means to some other end. The contract looks more like a deliberate 20 year speculation on the relationship between oil and gas prices.

There is some parallel here with the (historical)7 test for an implied term. If an officious bystander had said to  the  parties “what if the connection between the oil price and the gas price were severed, and there was a market price for gas readily available – would you use that price instead?”, would the parties have said “of course, we’d use that price instead”.

It seems traditional to conclude any article about price reopener clauses, and the recent spate of awards, with a rhetorical question about whether these are a cause for concern and whether arbitrators should be allowed to ‘re-write the parties’ bargain. The truth, though, is that any price re-opener clause is, itself, part of the parties’ bargain. There is no reason to think that a tribunal is any less able to determine, and give effect to, the parties’ intention as expressed in a price re-opener clause than in any other clause. Moreover, no one has imposed these clauses. They are of the parties’ own making, and the parties are free to set the limits of the tribunal’s authority at the outset.