UFE Response to ACER consultation on the proposed amendments to the HAR methodology

Topic 1: Collateral requirements :

1.1 Background :

With the introduction of long-term flow-based allocation (LTFBA), auctions will be performed simultaneously for all bidding zone borders in the CCRs where the flow-based capacity calculation approach is applied. The currently approved HAR requires that participants in an LTTR auction provide sufficient collaterals to cover the value of their bids. Having simultaneous auctions for all bidding zone borders implies that the collateral requirements may significantly increase during the auction phase if the current approach remains unchanged. Therefore, under Article 34(6) of the Proposal, TSOs propose to introduce a price cap for the calculation of the maximum payment obligations in case of flow-based allocation. The only impact of such price cap would be to limit the collateral requirements from a bid for the calculation of the maximum payment obligations. More specifically, it is proposed that if the original bid price is lower than the price cap, the bid price shall be used for the calculation, and if the original bid price is higher than or equal to the price cap, the price cap shall be used for the calculation. The TSOs propose to calculate the price cap as follows:

For yearly auctions, the average value of market spreads of the six (6) last calendar months before the publication of the final auction specification shall be used for calculation, by adding all MTUs with positive values of the market spread for a bidding zone border direction. The resulting total value shall be divided by the number of MTUs with such positive market spread.

For all auctions having a shorter product duration than yearly auctions, the average value of market spreads of the last calendar month before the publication of the final auction specification shall be used for calculation, by adding all MTUs with positive values of the market spread for a bidding zone border direction. The resulting total value shall be divided by the number of MTUs with such positive market spread.

(Note: the market spread means the difference between the hourly day-ahead prices of the two concerned bidding zones for the respective market time unit in a specific direction, as defined in the FCA Regulation art. 2(9))

ACER agrees with the TSOs that there is a need to amend the provisions on the collateral requirements for flow-based allocation of LTTRs. However, as described under point 1.2, ACER is concerned about the expected accuracy and efficiency of the cap calculation as proposed by the TSOs.

In ACER’s view, there are several ways to limit the collateral requirement in case of flow-based allocation. ACER would like to collect stakeholder views on the possible options outlined below.

1.2 Option 1: Cap option using the average value of the market spread :

According to Article 34(6) of the Proposal, the TSOs intend to use an average of the latest market spreads before an LTTR auction to define the cap. ACER in general considers that using the market spread for defining such a cap would be a simple and transparent method. While other, more complex methods may provide a higher forecast accuracy, ACER considers that the approach based on the market spreads would be easy to implement and can be expected to be in place before the go-live of the LTFBAs in the Core CCR in November 2024.

However, ACER sees some room for improving the TSOs’ proposed calculation. More specifically, only dividing the total summed-up value by the amount of MTUs with a positive value, the calculated cap might not result in an equal consideration of all bidding zone border directions and could lead to unjustified high caps for some bidding zone border directions. While bidding zone border directions with constant positive values over all MTUs would have an accurate representation of the past directional market spread, bidding zone border directions with a very small share of MTUs with a positive value would be subject to a significantly overestimated cap. Dividing the total summed-up value by the number of all MTUs within the relevant time period might be, in ACER’s view, a more accurate approach than the calculation method proposed by TSOs.

1. Do you consider Option 1, using the average value of the market spread, an acceptable solution?

• Strongly agree
• Agree
• Neutral
• Disagree
• Strongly disagree

2. In your opinion, what is the preferred method on how to address the described issue of collateral requirements, which could still be implemented by the deadline of November 2024?

While the idea of a cap on collateral seems positive, it all boils down to whether the cap actually decreases collateral burden for market participants. The cap doesn’t guarantee a lower collateral requirement: sometimes the bid price could be lower than the cap (in which case the cap is not useful), it only becomes a limit of collaterals in case of skyrocketing market prices.

In any case, looking at the current proposal, it seems clear that looking at DA spreads as a basis for the cap does not make much sense. We note however that ACER slightly improves the ENTSO-E proposal by including all prices (and not only positive price spreads) and we appreciate this minor improvement. It would be more relevant:

• To use the average observed forward spread instead during a certain period since the FTR auctions concern forward maturities (and not day-ahead), using the average observed forward spread instead during a certain period. It is important to make sure that the reference price used to compute the cap is in line with the maturities of the FTRs. We suggest Entso-E to engage into discussions with data providers in order to obtain the necessary data.
• To compute collaterals based on the final auction price. In other words, calculate the collateral amount within the allocation, that way there is no filter of bids prior the allocation. N-side has proposed a solution to calculate the collaterals within the allocation and based on the final auction price.

That being said, we want to stress that this issue is clearly not the priority compared to all the drawbacks brought by current design choice. We also want to reiterate our concern that performing an auction with only a limited set of buying orders seriously challenges the potential merit of such an auction. In order to properly assess the situation, we suggest once again performing an analysis to test to which extent the set of buying orders would be limited by collateral. To our knowledge, this assessment has not been performed and could seriously affect the relevance of the outcome of the auction.

3. Do you have any comments on the TSOs’ proposal for the cap calculation?

Guarantees are intended to cover the risk of market participants failing to meet their financial obligations arising from the forward capacity for which they are bidding. Day-ahead prices have nothing to do with the execution of this contract, and are therefore inappropriate.

1.3 Option 2: Cap option using forward prices :

Another approach to calculate the price cap could be to use available prices from the forward electricity market. It is expected that using forward prices would result in more accurate forecasts of LTTR auction results than when using the average of past day-ahead prices. Option 2 would therefore result in a more efficient cap application. One problem with the use of forward electricity prices is the availability of reliable and consistent prices that can be used for a cap calculation for all relevant bidding zone borders. Another complexity is the transformation of bi-directional market spread resulting from the available forward electricity obligation prices to a market spread per bidding-zone border direction (i.e. required for defining a cap per auctioned LTTR option). ACER is therefore concerned about the complexity of implementing this method, especially considering the required implementation by November 2024.

4. Do you consider Option 2 of using forward prices an acceptable solution?

• Strongly agree
• Agree
• Neutral
• Disagree
• Strongly disagree

5. If you agree, please provide a detailed description on how you consider the calculation of the price cap using forward prices can be done in the best way possible (i.e. how should the described problems be addressed most efficiently)

Before entering the details, we want to remind that we are opposed to the notion of bid filtering and as said in question 2, we suggest performing an analysis to estimate the impact of such filtering on the auction results. We consider that any bid-filtering would reduce the efficiency of the auction.  To avoid making an ex-ante bid filtering, we want to voice our support to the “option 3” proposed by ACER and refer to our answer to question 7 and 8, which is for us the only acceptable solution in case bid filtering has to be put in place.

As mentioned in question 2, we consider that forward prices should be used as a benchmark for the value of FTR, not historical spot prices, especially if the historical period does not match with the FTR delivery period (eg: using Q3 spot prices as a benchmark for CAL prices in the forward). One could use the historical forward quotations for the relevant delivery period during an historical period of e.g. 1 or 2 months prior to the auction. This could be used for in the money directions.

For out of the money directions, the question of price cap does not seem that relevant since the value is anyway not very high.

6. If you disagree, please clarify the reasons why you consider such solution not acceptable or not feasible

Again, we want to insist that ex ante filtering is not a proper solution and refer to our answer to questions 7 and 8.

Other options described in question 2 would be more accurate.

1.4 Option 3: A solution where bid filtering is based on the market results :

An approach, where the bid filtering is based on the market results and is not performed before the auction, might constitute an appropriate long-term solution. This approach should eliminate the need to exclude bids before the algorithm is run. Such a solution would therefore effectively address the drawback of a cap solution, where inaccurate forecasts for the calculation of the cap would lead to inaccurate assessments of the required collaterals before the auction is run. Although this approach can’t be implemented in time for the go-live of the LTFBA in the Core CCR, it could be explored at a later point in time as a potential long-term solution and ACER would therefore still like to receive input on this option. The go-live of the first LTFBA auctions in November 2024 would require another, transitory solution.

7. Do you consider that Option 3 should be further explored as a long-term solution (i.e. after the go-live of the first LTFBA auctions)

• Strongly agree
• Agree
• Neutral
• Disagree
• Strongly disagree

8. Do you have any other comments concerning Option 3?

Should a bid filtering have to be performed, then we consider option 3 as the only acceptable solution. In this approach, each Market Participant defines the maximum acceptable amount of collateral he can provide. Then an optimal bid selection is determined, while respecting the maximum amount defined by each MP. To take the language of the consultation: This approach should eliminate the need to exclude bids before the algorithm is run.

We however disagree with ACER assertion that this should be regarded as a viable long-term solution post-implementation.

We would like to emphasize that the go-live date is set by an ACER decision and not  prescribed in a primary regulation. Thus, we believe that it is more crucial to prioritize finding the most  appropriate approach for establishing Flow Based Allocation of LTTR, rather than focusing solely on meeting this deadline. It is imperative that all necessary efforts, including the allocation of adequate resources and time – should be dedicated to the project in order to ensure that we do not proceed with a project that could detrimentally impact the efficient functioning of the forward market.

We therefore urge ACER and the TSOs to investigate the implementation possibilities for such an option, possibly turning to IT/Algorithm providers if necessary. In any case, we request the publication of a more detailed impact assessment for this solution, before ruling out the possibility of a swift implementation.

1.5 Timing for publishing the calculated cap on collaterals :

According to the Proposal, the calculated price cap for collaterals in case of flow-based allocation is published with the final auction specifications, at the latest one hour before the start of the bidding period. ACER considers that it could be beneficial to publish the calculated cap on collaterals earlier, so that market participants have more time to alter their credit limit.

9. Do you have any comments on the proposed timing for publishing the cap on collaterals?

We consider, as ACER, that the cap should be published earlier, at least 3 working days before the auction.

Topic 2: Sanctioning in case of non-payment :

In the Proposal, in case a registered participant is suspended from the participation agreement due to a payment incident, they may not use their allocated LTTRs until their payment of the LTTRs is fully settled or fully secured by collaterals. This provision implies that a market participant who refused to pay its debts may regain access to its LTTRs once the market turns in its favour. To prevent such a situation and the resulting costs for the TSOs and consequently the tariff payers, ACER intends to introduce a stricter sanctioning regime in case of non-payment by market participants. ACER would propose that after a non-payment within a certain deadline for settling open positions, market participants will lose all rights on awarded capacity.

In the Proposal, a market participant, who is suspended from the participation agreement, is not able to participate in an auction until the payment of the LTTRs is fully settled or secured by collaterals. To reduce the risks of non-payment of LTTRs, ACER considers to implement a provision where a suspended market participant is excluded from all further auctions for a certain cooling-off period, e.g. minimum of three months, after the LTTRs have been fully settled after the payment incident.

10. Do you have any comments on strengthening the sanctioning regime as proposed by ACER?

In case of non payments by default market participants, other MPs should not be impacted.

Topic 3: Auction specifications :

3.1 Offered capacity with flow-based in the auction specifications :

The final offered capacity is provided in the final auction specifications. According to the Proposal, this final offered capacity in case of flow-based allocation shall consist of:

1. Max Exchanges (MaxBex) per bidding zone border directions;
2. Min Net Positions; and
3. Max Net Positions

ACER considers that it would be beneficial for the market participants to receive the full set of flow-based parameters, in order to have the opportunity to simulate the LTFBA and asses their positions. ACER considers that the final offered capacity in case of flow-based should consist of:

1. Power transfer distribution factors (PTDF) per critical network elements (CNEC) and, if applied, grouped network elements (GNEC);
2. Remaining available margin (RAM) per CNEC and GNEC;
3. External constraints (EC) per border directions, where applied;
4. ATC values per border directions, applied for evolved flow-based (EFB) approach;
5. Max Exchanges (MaxBex) per bidding zone border directions;
6. Min Net Positions; and
7. Max Net Positions

11. Do you support the proposal of providing the flow-based parameters in the final auction specifications?

• Yes
• No

12. Do you have any other comments concerning the proposal on the offered capacity with flow-based?

First we want to point out that definitions of “Max Exchange” or “Max Net Positions”, and any reference to the relevant capacity calculation methodologies, are missing. Second, we agree with ACER that publishing only “Max exchange and max net position” is not sufficient for “final offered capacity” in case of Flow Based Allocation. Indeed, the full list of required data (e.g. as published in the CORE LT CCM, annex 1, article 20.1, referring to article 3f of FCA regulation) shall be published prior to the auction. Hence, in addition with ACER list, we consider that the detailed characteristics of the CNECs (ie, not just an ID but also the various technical components) should be published such as:

1. Branch Name
2. Contingency Name
3. EIC code
4. Direction
5. Hub from / hub to
6. Substation from / substation to
7. Element type
8. TSO
9. Fmax
10. Imax
11. U
12. FRM
13. F0core
14. F_uaf
15. minRAM factor
16. AMR
17. Presolved (true/false)

Furthermore, market participants will need the Publication Handbook as currently available for CORE Day Ahead, including the description of the data and publication times.

Also, API will be needed for market participants to request data (as currently available in day ahead).

Finally, regarding the results, we need to have access to volume allocated, prices, and bidding constraints after market coupling (shadow prices).

Other comments :

13. Do you have any comments on other amendments proposed by the TSOs?

• Long-Term Flow based allocation implies a significant impact for collateral requirements:

UFE wants to highlight that the move to Flow Based Allocation implies a significant impact for collateral requirements. We understand that the collateral requirement has not been adapted/modified to the allocation of more than 20 borders at the same time: therefore, Market Participants will have to provide at once the full amount of collateral corresponding to the “sum” of all the induvial borders they are bidding. This leads to several side effects:

• It will increase the amount of collaterals requested from market participants, because they will need to provide collaterals on all borders on which they would like to bid. Subsequently, some market participants will have to select fewer borders on which they can bid and probably submit less bids, with a lower price.
  • We therefore want to reiterate our concern that performing an auction with only a limited set of buying orders challenges the potential merit of such an auction. In order to properly assess the situation, we suggest performing an analysis to test to which extent the set of buying orders would be limited by collateral.
  • In addition, we think that the constraint related to collateral (ie the limitation in terms of bidding a market participant will have to respect) should be integrated into the optimization algorithm, in order to make sure that the best combination of bids is selected (instead of an ex ante arbitrary selection).

• The important amount of collateral requested will discriminate small market participants who will not be able to gather the necessary funds to secure their participation to the auction (market entry barrier). As a result, small market participants will not have access to hedging opportunities, which is in contradiction with the FCA regulation. Only big companies will have access to LT market CORE.
• Furthermore, under NTC allocation each border auction was performed on different period which gives the time for market participants to adapt their bidding strategy based on what was allocated at first. With the LTFB allocation, as all borders are run at the same time, market participants will not have the opportunity to have a “ second chance” to get LT rights on other borders, they will have only one opportunity.
• Last but not least, we consider that  the required collateral (for both Flow Based and ATC) seems disproportionate compared to the risks TSOs are bearing. Indeed:
  • Article 22 mentions that the validity of the collateral requirements should be 30 days after the end of the Product Period;
  • Article 66 says that the payment for long term rights shall be settled before the start of the Product Period; We suggest that the validity of the collateral should end right after the payment of the acquired rights.

• UFE is opposed to the establishment of Long-Term Flow Based allocation whose added value has not been sufficiently demonstrated:

From a more general perspective on the LTFBA project, UFE is opposed to Flow Based allocation whose added value has not been sufficiently demonstrated by ACER and is hence not compliant to FCA guideline article 10.

• Most importantly, it has not been proved that FB allocation will lead to more cross-zonal capacities being allocated, which should be the ultimate goal given the need for long-term hedging under current circumstances.
• More worrying, the recent simulations performed by TSOs show that some bidding zone will have very low/zero volumes allocated at their borders. This is in contradiction with the foundations of the article 30 of FCA regulation, which states that TSOs should provide enough hedging opportunities to the market. These extreme situations, where no capacity is allocated, could also lead to operational security risks as the fallback of the Day Ahead Capacity Calculation is the capacity allocated in the Long Term timeframe.
• Should LTFBA were to be implemented anyway, it is key that TSOs investigate possible mitigation measures, such as imposing a minimum volume at each borders, ensuring that no bidding zone becomes isolated in the forward market.

• The new article 49 proposed by ENTSO-E is not compliant with the FCA Regulation and should therefore be deleted:

Regarding the addition of the new article 49, UFE would like to highlight that a cap on LTTRs remuneration is neither permitted by the FCA Regulation, nor economically justified.

First, UFE would like to remind that the main objective of financial firmness of LT capacity allocation (LTTR remuneration at day ahead market spread) is to allow the market participants to hedge position across borders.

Second, even if UFE understands TSOs concerns regarding the remuneration of LTTRs in case of decoupling (day ahead market spread VS  day ahead shadow auction prices), the recent decoupling events do not corroborate this concern. Indeed, on the recent decoupling events (2019, 2020, 2021), the total amount compensated by TSOs to market participants (as LTTRs) represents a very small part of the total revenue incomes for the TSOs across the whole year (coming from the allocation of cross zonal capacity on Long Term auctions). Furthermore, caps on the remuneration of long-term transmission rights are reserved to cases of curtailment.

Third, UFE shares ACER’s Decision 15-2021 where ACER ruled out such a proposal underlining that there is no legal basis to implement a remuneration cap in case of decoupling and that a modification of EU HAR would imply a change to the FCA Regulation.

Article 35 of the FCA Regulation lays down rules for the remuneration of LTTRs. It requires the remuneration to be equal to the market spread for implicit auctions or their fallback in day-ahead. We therefore do not understand why and disagree with the fact that this point is being brought again in the debate.

Finally, UFE would like to remind that market participants are not responsible when a decoupling occurs, it is not in market participant’s hands, and they are suffering from it. While there were several decoupling cases since 2019, the focus of TSOs and NEMOs should be on the robustness of the algorithm and the whole day ahead market coupling process. Instead of changing the EU HAR, the focus has to be on the reinforcement of the testing/improvements of the SDAC process to avoid any decoupling event in the future.

In case it happens, shadow auctions should be maintained, and training sessions like the ones organized in the last years should be maintained (these sessions however duly require the presence of all TSOs). Communication towards market participants in case of (a risk of) decoupling should also be improved.

The new article 49 is not compliant with the FCA Regulation and should therefore be deleted.