Electricity markets: Can the EU deal with forward hedging when taxing infra-marginal profits?

In a leaked first draft the EU proposes in its own words a

Price cap for inframarginal technologies for the benefit of consumers

Sounds similar to my recent proposal. A difference to my main proposal is that the EU does not want to reduce the spot market prices, but suggests that governments rather collect the difference between a technology’s price cap and the actual spot market as a tax that can be used to finance other support measures (e.g. lump-sum transfers to consumers).

To better understand the motivation for taxing infra-marginal profits, let me show an updated version of a plot from my previous post (now incorporating the complete data for August 2022, a description of assumptions and data source can be found here):

The area of the colored bars shows the estimated average variable cost of electricity production per MWh and the area under the black line the average price per MWh. The difference between the two areas roughly describes the producers’ rents (mainly infra-marginal rents plus some rents for the most expensive marginal plant to cover its fixed costs). These rents must be big enough to cover investment costs and other fixed costs of power plants.

In the time period 2015-2020 (on the left) these rents were on average 27.4 Euro / MWh and we may assume that this was roughly large enough to cover investment and fixed costs of power plants. For August 2022, I compute average rents of 371 Euro / MWh, i.e. rents increased by a factor of 13.6 (in contrast average variable cost increased “only” by a factor of 6.57 from 14.9 Euro / MWh to 98.2 Euro / MWh).

These numbers suggest that there should be sufficient scope to reduce power plants’ rents in order to relieve consumers’ burden while keeping enough money in the system for investments and other fixed costs. Of course, not all electricity is traded on spot markets, but German baseload power future prices for 2023 trade today even higher at 515 Euro / MWh and peakload futures at 790 Euro / MWh (and those future prices probably already incorporate a positive probability that politics takes measures to reduce next year’s spot prices).

Is there a way to shift a substantial part of these rents to consumers without disastrous, unintended consequences?

The issue of hedged power plant operators with forward / future positions

The risk of unintended consequences would probably be much lower if all power plants sold all energy on the spot market and performed no hedging. We could then introduce either a maximum price that non-gas power plants receive when selling electricity on the spot market, or determine the size of the tax based on the difference between spot price and estimated variable cost of power plants (calculated for each 15 minute trading interval separately).

But consider the following situation. A coal power plant owner hedged its price risk by selling in 2020 cash settled future contracts for his expected electricity delivery in 2023 say at a price of 80 Euro / MWh. (I don’t know the actual historic future prices).

Now take a trading period in 2023 and assume spot market prices are 600 Euro / MWh. The coal power plant has to pay the buyer of the future 520 Euro / MWh, the difference between the 2023 spot market price and the delivery price specified in the future. But, without a price cap or tax, the coal power plant can just sell its electricity for 600 Euro / MWh on the spot market. So in total the coal plan operator gets 600 - 520 = 80 Euro / MWh, the same amount as specified in the future.

But now assume that the coal power plant faces a price cap in the spot market and is allowed to receive only 250 Euro / MWh (or equivalently it gets the 600 Euro / MWh but has to pay a tax of 350 Euro / MWh). Then the coal plant would lose 350-80 = 270 Euro / MWh plus its variable cost from the transaction and may quickly become insolvent.

It would clearly be very bad from an economic and legal perspective if this unintended side effect of the regulation causes power plant owners with a sound hedging strategy to become bankrupt.

I don’t know how prevalent this case is, but unless it is a very rare exception (which I doubt), it must be addressed.

What could one do?

I guess the core idea would be that for hedged electricity producers one must determine the “actual price” at which they sell electricity, taking into account future and forward positions.

That may be a very tricky task as there may be different instruments for hedging (cash settled futures, forward contracts with physical delivery, options (?), or more complex contractual structures). Also hedging probably does not take place at the level of a single power plant but at the level of a firm which may have different types of power plants.

[UPDATE 2022-09-05] Note that many electricity producers probably also have fixed price contracts with retail consumers. Indeed, some firms (e.g. “Stadtwerke”) may mainly have retail contracts but squired power plants or shares in power plants as a hedge to reduce their exposure to wholesale prices. Such retail commitments probably should be treated similar to hedging in future or forward markets.

So one probably needs some rule of how company wide hedging positions are boiled down to “actual prices” that the different producing plants receive.

OK, assume we have such a rule that maps hedging positions into prices for power plants. Now firms might have strong incentives to exploit that rule in order to reduce their tax burden. And history shows that firms can be quite creative in inventing schemes that reduce tax payments.

To limit such exploitation risk, one may need to look at firms’ hedging positions at a date lying sufficiently far away in the past when firms had not yet incentives to manipulate their positions in order to reduce new taxes on infra-marginal rents.

Can it be done without too much excess burden? I don’t know, as I have very little detailed institutional knowledge.

Is the hedging problem really so severe?

Again I lack institutional details to assess how severe the hedging problem actually is for the goal of shifting infra-marginal profits to consumers. Possibly some factors might make the hedging problem less severe:

• If I understand correctly commercial wind and solar farms are often financed by Power Purchase Agreements (PPA) that fix a long term price for each MWh electricity produced from that actual wind or solar farm. Under PPA it might be simply to know the price a solar and wind farm actually achieves: it’s just the price specified in the PPA. (But maybe PPA are more complicated than I think.)

• Coal power plant owners hopefully have not sold forward all their potential production. If the price cap is sufficiently loose, they may still make substantial profits on their excess production sold in the spot market which balance potential losses from their hedged positions. I mean, in the end someone probably gets high infra-marginal profits (unless all spot market buyers hedged themselves at low prices in which case there would be no problem). It seems unlikely that some electricity producers were fully hedged and others not at all.

• Also a varied plant portfolio of power plant companies might help to balance profits and losses from hedged positions.

So if we are lucky, a simple rule might suffice to ensure that producers with perhaps 95% of market capacity will not get financial troubles when taking away infra-marginal profits and one has to apply more complex procedures only for a small share of companies. Are we lucky? I don’t know.

[Update 2022-09-03] Assessment from a market expert

I already got valuable feedback. On twitter Christoph Maurer provided a great link that provides some data into the hedging practice of RWE.

In addition, he send me an email with a personal assessment of the problem that I consider very insightful and plausible. He allowed me to share his assessment. Below is first the original version in German, followed by a Deepl translation.

Einschätzung von Christoph Maurer

Die Lösung/Entschärfung des Hedging-Problems ist meiner Meinung nach eine zwingende Voraussetzung, wenn man eine Lösung „im Strommarkt” finden will (und nicht auf eine „simple” Steuer auf erklärte Gewinne geht).

Deshalb nachfolgend ein paar Punkte, die mir relevant erscheinen.

1. Bei dem Punkt muss man sehr deutlich zwischen EE und konventionellen Kraftwerken unterscheiden.

   1. EE sind im Moment vermutlich die mit den höchsten Renten, weil sie überwiegend spotvermarktet sind.

      1. EE im Fördersystem der gleitenden Marktprämie (ab 100 kW aufwärts sind das die allermeisten Anlagen) sind in der Vergangenheit fast ausschließlich am Spotmarkt vermarktet worden, weil das die risikolose Vermarktungsform war (Prämienzahlung hat Differenz zum durchschnittlichen Spotmarktwert erstattet).

      2. Allerdings könnte sich das in den letzten Monaten geändert haben. Wenn die erwartete Prämienzahlung 0 ist, spielt das keine Rolle mehr. Dennoch haben EE keinen vollen Hedge gegen die Standardterminmarktprodukte Base und Peak, was Spotvermarktung attraktiver macht.

      3. PPAs sind m. E. nicht das große Problem. Das sind in der Regel sehr langfristige Verträge, die eher auf Basis von Durchschnittskosten + Marge abgeschlossen werden.

   2. Bei konventionellen Kraftwerken ist das Problem hingegen riesig, weil insbesondere die von variablen Kosten günstigen Kraftwerke fast vollständig auf Termin vermarktet werden. Das sind aber auch die, wo es evtl. Renten gibt.

      1. Pressemeldungen deuten aber darauf hin, dass dort aktuell noch keine besonders hohen Renten anfallen, weil Vermarktung noch vor Krise erfolgte.

      2. Allerdings könnte sich das im Laufe der Krise ändern. Aktuell verkaufen die durch die Anlagen abgesicherte, sehr teure Terminkontrakte. Wenn die zur Erfüllung gelangen, werden sehr hohe Renten anfallen.

      3. Problem kann dabei sein, dass zu diesem Zeitpunkt die Krise auf dem Spot längst vorbei ist. Die Anlagen speisen im Extremfall sogar gar nicht ein. Dennoch verdienen die Betreiber aus den Futuresgeschäften.

      4. Wenn man hier Fehler macht, kann man Unternehmen potenziell in die Insolvenz treiben.

2. Struktur des Terminmarktes macht es vermutlich sehr schwer, Lösungen zu finden.

   1. Kontrakte sind nicht anlagenscharf.

   2. Typischerweise werden Terminpositionen mehrfach geöffnet und geschlossen (das Handelsvolumen am deutschen Terminmarkt für ein Erfüllungsjahr beträgt ein Mehrfaches des Stromverbrauchs).

Deepl Translation of Christoph Maurer’s Assessment

Solving the hedging problem is, in my opinion, a mandatory requirement if one wants to find a solution “in the electricity market” (and not go for a “simple” tax on declared profits).

Therefore, below are a few points that seem relevant to me.

1. on this point, one has to distinguish very clearly between RE and conventional power plants.

   1. RES are probably the ones with the highest rents at the moment, because they are mostly spot-marketed.

      1. renewable energies in the support system of the sliding market premium (from 100 kW upwards, these are the vast majority of plants) have been marketed almost exclusively on the spot market in the past, because this was the risk-free form of marketing (premium payment reimbursed the difference to the average spot market value).

      2. However, this may have changed in recent months. If the expected premium payment is 0, it no longer matters. Nevertheless, RE does not have a full hedge against the standard forward market products Base and Peak, which makes spot marketing more attractive.

      3. PPAs are not the big problem in my opinion. These are usually very long term contracts that tend to be based on average cost + margin.

   2. For conventional power plants, on the other hand, the problem is huge, because especially the power plants that are favorable from variable costs are almost completely marketed forward. But these are also the ones where there might be annuities.

      1. press reports indicate, however, that there are currently no particularly high annuities, because marketing took place before the crisis.

      2. however, this could change in the course of the crisis. Currently, they are selling very expensive forward contracts hedged by the assets. When they are fulfilled, very high pensions will accrue.

      3. problem may be that by that time the crisis on the spot will be long over. In extreme cases, the plants may not even feed in at all. Nevertheless, the operators earn from the futures transactions.

      4. making mistakes here can potentially drive companies into insolvency.

2. structure of the futures market probably makes it very difficult to find solutions.

   1. contracts are not asset specific.

   2. forward positions are typically opened and closed several times (the trading volume on the German forward market for a fulfillment year is several times the electricity consumption).

Feedback

Given my limited knowledge of institutional details, I would appreciate feedback very much. In particular, if you the know the reality of electricity markets much better than me and have suggestions please don’t hesitate to send me an email. I try to update the post regularly if I get relevant new insights over time.

Author: Sebastian Kranz, Ulm University