What English gardens, monkeys and the Titanic have to do with bidding zones

On the 18^th of June 2020, the Florence School of Regulation (FSR) hosted its second online policy workshop “The Configuration of Bidding Zones, Liquidity and Competition in the Electricity Market” directed by Alberto Pototschnig (FSR Part-time Professor and former Director of ACER), as part of the FSR Regulatory Policy Workshop Series 2019-2020.

What English gardens, monkeys and the Titanic have to do with bidding zones

The configuration of bidding zones has been a heated debate in the European electricity scene for many years. Currently, most bidding zones borders align with country borders. This was fine at a time of predictable generation and load patterns, strong national networks and less developed interconnectors. Today, many bottlenecks appear within national bidding zones. This is mostly due to a significant increase in renewable generation that is installed at a much faster pace than grids can be reinforced.[1]

As a consequence, there is a significant volume of unscheduled flows (mostly loop flows). These unscheduled flows are typically prioritised and come at the expense of the opportunity to trade across (bidding zone) borders. On top, an increasing amount of costly remedial actions is required to reconcile the market outcome with the capability of the network, leading to high redispatching costs and potential gaming opportunities.[2]

One rather straightforward solution to address these issues is to redraw the bidding zone borders in such a way that structural congestion is dealt with through the market, instead of being hidden within bidding zones. A more appropriate bidding zone configuration allows for a more efficient use of the grid and thus a welfare increase due to improved trading opportunities. Also, by better aligning the market outcome with the grid capabilities, operational security is enhanced. It is rather straightforward that having “a more efficient bidding zone configuration” implies splitting up large bidding zones into smaller zones. However, some stakeholders claim that market liquidity would suffer from bidding zone splits and that smaller bidding zones would lead to lower levels of competition. In this workshop, we discussed these two claims.

Bidding zone size and market liquidity – You will only find a perfect hedge in an English garden

The first part of the workshop focused on the relationship between the bidding zones size and market liquidity. The main takeaway from this session:

• The claim that smaller bidding zones would lead to lower market liquidity is not thoroughly substantiated and mainly relies on illustrative or anecdotal evidence that is rather inconclusive. What is believed to be a more important condition to have a liquid forward market is that market participants trust price formation in the day-ahead market (which serves as a reference price for forward contracts). In that regard, the current large bidding zones do not reflect well the underlying physics and impede robust spot price formation rather than stimulate it. It is true that smaller bidding zones might make it harder to construct the perfect hedge, but this does not mean that striving for the perfect hedge at all costs is justified.

#FSRPolicyWorksop
You find perfect hedges only in English gardens!

— Gerard Doorman (@gerardd77) June 18, 2020

More detailed points:

• Market liquidity is important to consider when discussing the revision of bidding zones. A low level of liquidity might lead to higher transaction costs, higher risks and hedging costs, which may translate into higher barriers to entry into the market.
• Market liquidity is not easy to measure. Many indicators exist. There is no consensus on which indicator is most appropriate. Important indicators are trade volume, the bid-ask spread, the churn rate, the risk-premium and market depth.
• Market liquidity is mainly a concern in forward markets. In real-time, the physics need to be respected in every node. In the spot markets, better-defined bidding zones will lead to a more efficient use of the grid, and market liquidity is expected to increase (on average) beyond the impact of the ‘’de-netting effect’’.[3] There could be hours when a certain smaller bidding zone is isolated, but these hours are easier to detect when bidding zones are properly defined. Regarding forward markets, there is a fear that smaller bidding zones would suffer from low liquidity, i.e. there would be fewer counterparties to trade each other’s risk with.
• There are “natural experiments’’ to learn from. For example, the bidding zone split in Sweden in 2011, the recent Italian bidding zone review, the German-Austrian bidding zone split in 2018. Also, beyond Europe, the transition from zonal to nodal pricing in the US (PJM, ERCOT, CAISO, etc.) and even in New Zealand, Chile and Argentina. The data are there, but so far very few in-depth impact analyses have been performed.
• The Nordic experience with splitting zones is overall positive. Costs for remedial actions are marginal, there is ample retail competition in each bidding zone and the performance of the market for hedging products to cover cross-zonal risk (EPADs) has improved. However, the perfect hedge for some bidding zones is hard to achieve due to low local liquidity of EPADs. But the correlation between the prices in such zones and the system price is high. This makes it questionable whether having a perfect hedge at all costs is essential. It was acknowledged that the volume of hedging products referring to the Nordic system price has decreased over the years, but also that many factors aside from a bidding zones split could be causing this trend. Examples are an increase in bilateral PPAs for renewables and exogenous financial shocks.
• The German-Austrian split showed that the uncertainty around the split initially led to a reduction in the volumes in forward contracts. However, quickly after the split, volumes picked up again fast in Germany. On the contrary, trading activity of Austrian forward contracts has remained low. Again, this issue can be mitigated. First, the correlation between the liquid German market and the Austrian market is high, making it possible to use a German forward contract as a proxy for the Austrian market. Second, to improve the hedge further, long-term cross-zonal transmission rights between both countries are available.
• US power systems with nodal pricing in place have the most liquid forward electricity markets in the world. Liquidity is pooled in hubs. Hub-to-node risk is hedged with financial transmission rights.
• An important point made was about the link between competition in the spot markets and market liquidity in the forward markets. The reference for forward markets is the day-ahead price. To have a liquid forward market, market participants need to trust price formation in the day-ahead market. The relationship between the bidding zone size and competition was further discussed in the second part of the workshop.

Bidding zone size and competition – If you dress a monkey in silk it will still be a monkey

The second part of the workshop focused on the relationship between the size of bidding zones and competition. Main takeaway from this session:

• Smaller bidding zones do not lead to lower levels of competition. Changing the bidding zone configuration, whether it is merging, splitting or switching to nodal pricing is not a major determinant for market power issues. Hence, if you dress a monkey in silk it will still be a monkey. Competition is a function of the market structure and of network capacity. Within large bidding zones, market power is pushed to real-time as physics always wins. Market players anticipating real-time will adjust their bidding strategy in preceding markets to profit from the inconsistency between markets and physics. With smaller bidding zones, possibilities to exercise market power will be more transparent and, as such, easier to mitigate.

More detailed points:

• Current large bidding zones often lead to infeasible results after the day-ahead market clearing. Significant redispatching volumes subsequently need to be activated. Market players can anticipate the fact that they might be redispatched and reflect this opportunity in their day-ahead market bids. Such behaviour leads to an overall welfare loss. This type of gaming is often coined the inc-dec game and was an important reason to restructure US markets. This game is likely to appear in European markets or is already present.[4]
• Merging bidding zones does not increase competition, nor does splitting bidding zones or implementing nodal pricing remove market power. However, better defined (often smaller) bidding zones do improve the consistency between timeframes (day-ahead, intraday and real-time balancing). As such, incentives for engaging in the inc-dec game are reduced. It can be that for certain hours a zone or node is isolated, and a market participant has the possibility to exercise market power. With smaller zones or nodal pricing, these possibilities to exercise market power will be easier to monitor in the spot markets.
• In several US power systems, there is nodal pricing in place. The US experience shows that going from zonal pricing to nodal pricing did not have a detrimental effect on competition. To counter market power abuse, there are Local Market Power Mitigation (LMPM) measures put in place. Put simply, LMPM monitor market power and subject market players to regulated prices for a limited time in case it is believed that there is a threat that a market party could exercise market power. There is a long experience with LMPM, and many variants in terms of the implementation of LMPM measures exist.

The bidding zones revision – Are we rearranging the deck chairs on the Titanic?

Overall, it was concluded that the bidding zone revision in the EU contains a political dimension. Evidence clearly suggests that revising the bidding zones, with the splitting of some existing large bidding zones, will lead to an increase of welfare, including the consideration of effects on liquidity and competition. The main added value that academics can bring is to demonstrate how costly it is not to correctly reflect grid constraints into electricity markets to aid decision-makers to make the right choices.

The “70-% rule’’, originating from the Clean Energy Package, is expected to become an important instrument to revise the bidding zone configuration. With the “70-% rule’’ in place, congestion cannot be “pushed to the border’’ anymore. As a consequence, redispatching costs will increase in many Member States and Member States might become more inclined to split up their uniform bidding zone. Some attendees argued that the “70%-rule’’ is the only way to achieve a more efficient bidding zones configuration, others believed that this is a very costly approach as consumers will have to pay these increased redispatching costs through the national tariffs.

Derogations from the “70%-rule’’ are possible and some Member States are indeed going for this option. However, from 2025 onwards these derogations are not possible anymore. From that moment onwards, if Member States cannot comply with the “70%-rule” and do not agree on how to reconfigure the bidding zones, the file will land on the desk of the European Commission.

An important last remark is that it is not obvious how to reconfigure bidding zones due to fast-changing generation and demand patterns. A new bidding zone configuration can be outdated quickly. Hence, are we rearranging the deck chairs on the Titanic? It might be worth thinking about a more permanent solution such as nodal pricing as implemented in several US power systems. Nodal pricing would have distributional consequences –just as a reconfiguration of bidding zones has– but there are ways to mitigate these. Examples include using the revenue from long-term transmission rights to compensate negatively impacted parties or the grouping of nodes on the consumer-side.[5] Also, experience in the US has shown that the complexity in computing prices in thousands of nodes at the same time can be handled. What is expected to be the real blocking factor for such a radical change is its impact on the governance of the power system – the changes of roles and responsibilities of the different actors.

Dig in deeper

Newspaper article:

• The Geographical Structure of the Electricity Market: Some Frequent Misconceptions (Online Article on Bidding Zones by Alberto Pototschnig published on the 19th of June 2020 in Energetika.NET)

Links to relevant FSR Training Courses:

• Evolution of electricity markets in Europe
• Executive Course to master Electricity Markets

Relevant FSR Publications:

• The Importance of a Sound Bidding-Zone Review for the Efficient Functioning of the Internal Electricity Market
• The EU electricity network codes (2019 ed.)

Relevant other events:

• Policy Workshop on Electricity Pricing and Trading in the Decarbonised Energy Sector 
• FSR Online Talk: In the Shoes of an Electricity Transmission System Operator
• FSR Online Talk: Imperfect Markets and Imperfect Regulation

Notes

[1] In the future, we might expect even more unpredictable generation and load patterns with a more important role of sector coupling. On the other hand, sector coupling can also allow for more flexibility that can mitigate grid issues.

[2] Also, TSOs in smaller bidding zones are pushed to install phase shifters to manage loop flows originating from large bidding zones. Phase shifters, if not operated in a coordinated manner, could jeopardize the integration of markets.

[3] The ‘’de-netting effect’’ means that liquidity in smaller zones would ‘’artificially’’ increase compared to the situation before the split due to the fact that with a larger zone market parties could provide a netted position (cons-prod; buy-sell) while with smaller zones formerly intra-zonal positions could become inter-zonal positions (because of the rearrangement of the zones) that need to be reflected in the day-ahead auction.

[4] Evidence for the Italian market is discussed in Graf, C., Quaglia, F., & Wolak, F. A. (2020). Simplified electricity market models with significant intermittent renewable capacity: Evidence from Italy. NBER Working Paper No. 27262. Available here.

[5] In Italy, generators are subject to prices which can differ from one bidding zone to another, while consumers are subject to a weighted average price which is uniform at a country-scale. Similarly, in the US, retail energy prices for consumers are often a weighted average of a set of nodes. An academic reference about how to use income from long-term transmission rights to compensate negatively impacted market parties due to a switch from zonal to nodal pricing is the following: Kunz, F., Neuhoff, K., & Rosellón, J. (2016). FTR allocations to ease transition to nodal pricing: An application to the German power system. Energy Economics, 60, 176-185. Available here.