ICIS Power Foresight: Why is the French Q1 2024 power price twice as much as it should be from a fundamental perspective?

This story has originally been published for ICIS Power Foresight subscribers  and was written by Luca Urbanucci, EU Power Analyst.

Our ICIS Power Foresight customers have access to extensive modelling of different options and proposals. Our long-term price forecast now also extends to 2050 across European countries. If of interest, please get in contact with Lead Analyst Matthew Jones (matthew.jones@icis.com). 

At the end of last year, the French power market witnessed a rapid drop in the Q1 ’23 price as the delivery period approached, certainly driven by the reduction of the natural gas price, but also by the prompt disappearance of a significant risk premium, which had been in turn mainly boosted by uncertainties over nuclear generation and fear of potential cold spells.

Recent discoveries of new cracks at French reactors, together with a general uncertainty about future nuclear availability that has never disappeared, have inflated the Q1 ’24 price for next winter, similarly to what happened last year.

In this analysis, ICIS adopts a ‘theoretical price-setting plant’ approach to investigate the extent to which future French electricity prices are impacted by risk premium.

First, what happened last year to the Q1 ’23 product is reviewed, and then the current situation of the Q1 ’24 contract is analysed and the reasons behind the risk premium are discussed.

A second analysis on the topic will follow, investigating different scenarios on French power for next winter, and presenting the results from the soon-to-be launched ICIS Power Foresight Model.

Background

Along with the gas crisis that has affected the whole of Europe, France has also faced problems related to their ageing nuclear reactors.

• • In 2022, the national nuclear generation plummeted to its lowest level since the 1980s at around 278TWh, mainly due to corrosion problems.
  • The generation was below 30GW for a large part of the year (from April to November) but improved in December when returned above the 40GW level.
• The outlook for 2023 was generally more positive, with an expected generation by EDF of 300-330TWh.
  • Yet, the recent discovery of new corrosion cracks and EDF’s need to review the fleet control strategy accordingly have brought uncertainty back, with a consequent impact on markets.
  • At present, however, EDF has not revised down the expected annual generation, and news on the updated control and maintenance plan for reactors is awaited.
• As already mentioned, the general uncertainty and the news about future nuclear availability deeply impacts the French price curve.
  • Nevertheless, this analysis does not go into the details of French nuclear issues (which will be dealt with in a separate analysis), but rather focusses on assessing the risk premium priced in the French Q1 ’23 and ’24 products and discussing the reasons behind it.

Methodology

• The methodology herein adopted is based on the evaluation of the efficiency of the theoretical gas plant that would set the power price in the delivery period.
  • Given the prices of natural gas, carbon, and power for a certain future contract, the efficiency of the above-mentioned theoretical gas plant can be easily evaluated.
  • The efficiency can then be compared to historical averages and to the efficiency of actual plants to assess the “fairness” of the contract price.
• The main advantage of this methodology is that the analysis is stripped of gas and carbon prices, as they can vary (significantly) over time, thus making it possible to assess the price of electricity net of these two drivers.
• Moreover, the methodology can also be reserved, i.e., by evaluating the electricity price implied by a gas plant (set the efficiency) given the natural gas and carbon prices.
  • In this case, the implied price can then be compared to the contract price.

Q1 2023 product

In this section we apply the methodology to analyse the progress of the French first-quarter 2023 baseload power contract, for the whole period of trading (from 4 January to 30 December 2022).

• The efficiency of the theoretical price-setting gas plant was around 35-40% in the first months of the year, slightly below the average value actually realised in 2022 (i.e., 41% – evaluated considering day-ahead market prices).
• From April onwards, largely as a reaction to growing concerns about reduced nuclear generation and potential cold spells, the theoretical efficiency of the price-setting gas plant dropped significantly and hovered around 15-25% until December, when nuclear availability returned above the 40 GW threshold, restoring confidence to the market.
  • Comparing the efficiency of the theoretical price-setting plant to that of the least efficient French gas plant (around 32% according to the ICIS analytics database), it is clear how the price of French Q1 2023 power was for a significant part of the year inflated by market risk premium, well above the level suggested by fundamentals.
  • Moreover, the efficiency of the theoretical plant that has set the price so far in 2023 is around 51%, clearly showing how the market substantially overestimated the value of the contract.

• The chart that displays the Q1 2023 contract price, along with the implied prices of gas plants with different efficiencies, helps to assess how much the market overestimated the value of the contract during last year.
  • Considering as reference the 2022 average efficiency of the theoretical price-setting gas plant (41%), the price was on average around €275/MWh higher (+81%) than it should have been.
  • In particular, the price was consistently overestimated by more than €400/MWh during all of August, September, and October, largely due to the uncertainties surrounding the gas supply and fear of potential demand curtailments.

Q1 2024 product

This section focusses on the French first-quarter 2024 baseload power contract, from day one of trading (3 January 2023) to date.

• The efficiency of the theoretical price-setting gas plant hovered around 30% from the beginning of the year until the first week of March, when it fell below 25% following the discovery of an unexpected crack at the Penly 1 reactor, and the subsequent request by the French Nuclear Safety Authority (ASN) to EDF to revise its maintenance and repair programme.
  • The efficiency has always been well below the 2022 realised average (41%), but it is now even below the threshold of the least efficient plant (32%), as was the case for the Q1 ’23 product during much of last year.

• The comparison between the Q1 ’24 price and the implied-price sensitivity considering a range of plant efficiencies shows how the market is currently overestimating the value of the contract above reasonable levels, short of a massive reduction in nuclear availability leading to security of supply issues (see section below).
• As of 21 March, the French Q1 ’24 is assessed at €309/MWh, while the implied price from 40%- and 50-% efficient plants is €164 and €132/MWh, respectively.
• Should the efficiency of the theoretical price-setting plant in Q1 ’23 be in line with the average of what was realised in 2022 (41%), the delivery price would be about half of its current market value.
  • If the average efficiency realised during the first quarter of 2022 is considered (45%), a current market overestimation of around 110% is assessed.

What is the risk that is being priced in?

Thus, the question arises: what is behind the risk premium that the market deems to be priced in the French first-quarter 2024 product?

• Certainly, a general risk around gas prices (and to an extent around carbon prices) exists. Yet, the future price of gas (and carbon) already accounts for this risk, and, in any case, this applies across all Europe, whereas the French product is the only one trading so far above the fundamentals.
• This means that the premium is entirely due to the perceived risk around reduced nuclear availability. However, the market’s fear does not seem to be that a low nuclear generation may just push the power system to the end of the French merit order (which is a 32%-efficient gas plant) or even to the end of the merit order in connected countries (e.g., a 21%-efficient coal plant in Germany – as the price would likely be set by imports if nuclear generation would fall significantly).
  • Indeed, the current price would imply that these low-efficient plants would set the price in almost every hour of the quarter, which is highly unrealistic.
• To justify these price levels, we have to get into the realms of supply shortages, where scarcity pricing can come into play, accounting for very high values in certain hours.
  • This applied as well to the Q1 23 product, but scarcity prices have not materialised so far this year, as the highest hourly value has been €270/MWh.
  • Yet, a perfect storm of simultaneous and prolonged low nuclear availability and cold weather could trigger security of supply issues next winter, resulting in skyrocketing prices (at least for some hours).
• In the end, it is key to assess the likelihood of this scenario occurring.
  • In the next analyst update on the topic, we will investigate how low French nuclear availability must drop to endanger supply security, and compare such a scenario with that adopted in the ICIS Foresight Model.

Conclusions

• The ‘theoretical price-setting gas plant’ analysis suggests that the current price of French Q1 ’24 is largely inflated by risk premium, as the resulting efficiency is around 20% (well below that of the least efficient plant, which is around 32%).
  • As a result, the French Q1 ’24 is currently about twice as much as it should be, if the efficiency realised in 2022 is considered as reference.
• Such a high price level cannot be justified by the fundamentals alone and is the result of the market’s fear of scarcity pricing scenarios for next winter, resulting from security of supply issues that may occur if nuclear availability were to be very low in conjunction with cold spells.
• Looking back to what happened to the Q1 2023 contract, the analysis shows that the price has been on average overestimated by around 80% during 2022 (adjusted for gas and carbon price effects). Indeed, fears of supply shortages for this winter have not become material so far.
• In the next analyst update on the topic, we will investigate how low French nuclear availability must drop to endanger supply security, and compare such a scenario with that adopted in the ICIS Foresight Model.