European power and carbon markets affected by COVID-19 – an early impact assessment

This analysis has originally been published in an extended version for ICIS EU carbon subscribers on 24 March 2020 at 14:38 CET.

In order to assess the fundamental impact of COVID-19 on European power and carbon markets we modelled a scenario in which we accounted for the first data on dropping electricity demand as well as assumed industry production cuts and continued travel restrictions.

We have based our power model run on some early signs of demand decrease shown in Italy and based assumptions for industry and airline demand on news reporting of planned measures and the potential impact on production and aviation emissions. Please note that this analysis is based on an early assessment of the potential impacts and requires refinement once more solid data becomes available. Therefore, this analysis needs to be treated as initial impact assessment.

The overall result is a substantial drop in EU ETS emissions of 388.8m tons for 2020 compared to a pre-COVID environment (-24.4%). EUAs would drop by €3.00/t as quarterly average this year and €4.90/t quarterly average in 2021 as a result of this scenario. European power prices will on average decline by 9% with significant variation across member states.

This is less of a price drop than what we have seen on the market already, so we would expect a slight market correction based on our fundamental findings.

However, we are aware that the situation changes with each incoming datapoint – we will monitor these developments closely and amend our forecast accordingly.

Current situation and impact on demand

• Several large European economies are in national “lockdown” to slow the spread of COVID-19:
  • Italy – since 9 March
  • Spain – since 15 March
  • France – since 17 March
  • Belgium – since 18 March
  • Great Britain – since 24 March
• Other member states have strict measures in place but avoid an official lockdown so far
• Several large industries are currently under partial or complete closure, including automotive and steel production
• Italy, as currently the most affected member state, ordered the closure of all non-essential factories
• Italian power demand dropped 3% during the first week of quarantine (9-15 March) then by 10% (16-18 March) compared to the 2015-2019 March average with latest analysis indicating that the temperature-adjusted drop might be more in the range of 16% assessing latest available data
• France has seen a similar drop compared to Italy, UK is currently in the range of minus 10% while German demand fell less than 5% compared to the relevant five-year period
• Airlines are hit severely by the travel bans and quarantine restrictions, with many flights cancelled. Lufthansa Group reports to only operate 5% of originally scheduled flights until 19 April

Market reaction

• Even though negative sentiment and fear about the economic consequences of the COVID-19 spreading has caused European carbon prices to shed some 33 percent between 11 and 20 March, the data situation on de facto decreased demand remains shallow
• European power markets have reacted with volatility since Italy entered lockdown in the attempt to price in a potential demand hit in April
  • Italian day-ahead baseload contract dropped 20% and the April ’20 baseload contract shed 25% between 9 and 20 March
  • Italian prices fell below the UK for April baseload delivery during the lockdown period. They had been trading at around an €8/MWh premium for much of the year until the last week of February
  • German and French April baseload electricity prices have fallen from around €30/MWh on 9 March to closer to €20/MWh on 20 March. Strong losses have been felt in particular on the near curve (Q2 and Q3 contracts) across all European power markets due to demand uncertainty
  • German spot prices have delivered at an unusual premium to other markets in the flow-based market coupling region (France, Netherlands, Belgium, Austria)

Scenario set-up

• We have based our first COVID-19 scenario run of our pan-European power model on some early signs of demand decreases observed in member states affected by the lockdown. Further, we based assumptions for industry and airline demand on news reporting of planned measures and the potential impact on production and aviation emissions
• This scenario setting requires refinement once more solid data becomes available – therefore, this analysis needs to be treated as initial impact assessment
• In order to provide a first indication of the potential fundamental effects of COVID-19 on EU industry production, power sector demand and the aviation sector we provide an indicative scenario setting that combines some early data evidence as well as experience from the previous financial crisis
• While we have already some early data for power demand and aviation sector impact we take assumptions for industry sectors based on announcements to conduct this analysis
  • Power
    • We use the current demand drop in Italy as an indication for assessing the impact on a European scale
    • In order to assess the fundamental impact of COVID-19 on European power markets we ran our long-term pan-European power model with a 10% decreased power demand for the period March to June
    • From July to December, we assume a steady recovery of demand, resulting in 6.2% lower demand in 2020 compared to our base case, equal to a decrease of 198TWh
  • Industry
    • At this point we have not done a sector-specific differentiation due to lack of data, acknowledging that some sectors will be hit harder than others
    • For industry we assume January and February production as normal
    • As of March we account for the increasingly stricter measures getting in place and assume a 30% production drop compared to our base assumptions
    • We see April to June as most affected, given that factory closures, short time worker schemes and liquidity as well as supply chain issues will likely hit production numbers significantly. Consequently, we assume a drop of 50% compared to our base case
    • From July 2020 to June 2021 we assume a linear recovery towards 100% base case production level
  • Aviation
    • We assume emissions to be 7.5% lower for the first quarter, the second and third quarter being affected by the very restricted travel schedule (-80% emissions compared to baseline), with an expected full recovery afterwards

Impact on emissions

• Power sector dynamics
  • The majority of the 6.2% annual demand decrease will be carried by gas power plants, lowering their generation by 17.4% compared to our pre-COVID assumption, followed by hard coal generation shedding 12.8%, and lignite generation 6.5%
  • All non-thermal generating technologies decrease their production by an average of 1.2% only compared to pre-COVID setting
  • We see a strong differentiation when it comes to power price declines between the different markets as a result of the modelled COVID-19 related demand cuts. Depending on the set-up of the member states merit order, we expect power prices to dip on average by 9 percent in 2020 compared to a pre-COVID setting
  • With the modelled scenario, European power sector emissions would decrease by 12.7%, equal to 87.6m tonnes

Figure 1: Power sector emissions by country (in million tons)

• Industry sector emissions
  • The decline in industry production of 30% in March compared to our base case, 50% between April and June and a gradual increase to 100% until June 2021 results in 265m lower emissions in 2020 and 40m lower industry emissions in 2021
  • With this approach, metal sector emissions would decrease by nearly 70m tonnes in 2020, oil & gas would shed 65m tonnes and cement & lime 51m tonnes this year

Figure 2: EU ETS industry emissions by sector (in million tons)

• Aviation emissions
  • The aviation sector would face a decline in emissions of 35.7m tonnes with the assumed restricted travel schedule
• Overall ETS emissions
  • With the described scenario, we see EU ETS emissions dropping by 388.8m tonnes compared to our pre-virus expectations (-24.4%) to 1,177.4m tonnes (from 1,566.2 pre-COVID)

Fundamental carbon and power price impact

• EUA prices
  • As a result of the reduced emissions, we see carbon prices fundamentally decrease by €3.00/tonne as an average over the four quarters during 2020 compared to a pre-COVID setting (-13.7%)
  • In 2021, we see prices to trade €4.90/tonne lower on a quarterly average compared to the pre-COVID setting (-21%), with 2022 trading €1.00/tonne lower (-3%)

• Power prices
  • We see a strong differentiation when it comes to power price declines as a result of the modelled COVID-19-related demand cut
  • We expect German power prices to decline by an average of €4/MWh in 2020 as a result of the decreased demand (-11% compared to pre-COVID modelling) and French prices to tank by 15% (or €5.1/MWh) during the same time period
  • Poland only sees an average decline of €0.7/MWh as around 11GW of hard coal and lignite plants built in the 1970s with similar variable costs set the price. A demand-related closure of a few plants would therefore not significantly affect the price
  • This is similar to Spain, which has a relatively flat merit order associated with the modelled demand flexibility, leading to a moderate price decrease of 3% (-€1/MWh)

Structural implications for the EU ETS

• Auction volume to be reduced as of 2021
  • Lower emissions in one year result in a higher total number of allowances in circulation (TNAC), forming the basis for the MSR reduction the following year
  • With significantly lower emissions in 2020 and 2021 as a result of this COVID-19 scenario, the TNAC publications in May 2021 and 2022 will show a higher value due to the higher number of “unused allowances”
  • This in turn would mean that the MSR reduces auction volume by a larger number between September 2021 and August 2022 as well as the following year compared to a pre-COVID setting

• Free allocation numbers impacted as of 2026
  • Generally, the allocation an installation can receive depends on the activity level (historic production of an installation) multiplied with the benchmark of the product, adjusted by the Carbon Leakage Exposure Factor (CLEF) as well as a cross-sectoral correction factor CSCF (if applicable)
  • The activity levels for the first allocation period (2021-2025) will be the average production values of the years 2014-2018, and will not be affected by the virus spreading
  • For the second allocation period (2026-2030), the respective average production values of the years 2019-2023 will be used as basis and here, the COVID-19 pandemic will have an effect due to lower production numbers

Marcus Ferdinand is the Head of European Carbon & Power Analytics at ICIS. He can be reached at Marcus.Ferdinand@icis.com

Our ICIS EU carbon customers have access to extensive modelling of different options and proposals. If you have not yet subscribed to our products, please get in contact with Justin Banrey (Justin.Banrey@icis.com) or Audrius Sveikys (Audrius.Sveikys@icis.com).