Energy Markets Overview

01

Understanding Electricity Prices

Before diving into electricity prices, it is important to understand what makes electricity different from most other commodities. The key difference is that electricity cannot be stored easily at a large scale. Furthermore, the power grid must be balanced at every moment, meaning that supply must continuously match demand. This requirement makes electricity specific to both a delivery period and a location. It is therefore important to clarify which electricity price we are referring to. In short-term power markets, the main reference point is generally the corresponding day-ahead auction price for that particular market and delivery period.

With that in mind, the price of electricity is set by the marginal cost of the last accepted unit of energy needed to meet demand. For example, suppose demand is 100 MWh. Wind can supply 40 MWh at €0/MWh, nuclear can supply 40 MWh at €20/MWh, and gas can supply the remaining 20 MWh at €80/MWh. Because the gas plant is the last accepted generator needed to meet demand, the market-clearing price becomes €80/MWh for all accepted generation.

This is a good model for many reasons, but the most important one is that it ensures enough electricity is produced. If we used another model with lower prices, expensive generators would not be incentivized to produce because they would make losses.

This combination of marginal pricing and the physical need to balance supply and demand at all times gives electricity prices several interesting statistical characteristics:

• Strong seasonality: Prices show regular intraday, weekly, and yearly patterns. For example, one common intraday pattern is higher prices during morning and evening demand peaks and a midday depression due to high solar production.

• Price spikes: Prices can suddenly jump to very high levels for short periods when demand is high or supply is tight, due to outages or low production.

• Negative prices: Prices can go below zero when supply is very high and demand is low, especially during periods of strong renewable generation.

• Mean reversion: After extreme movements, prices tend to move back toward their mean/normal levels which can change over time.

• Heavy tails: Extreme price movements happen more often than they would under a normal distribution.

• Volatility clustering: High-volatility periods tend to be followed by high-volatility periods, while low-volatility periods tend to be followed by low-volatility periods.

To understand why the above statistical properties arise, it is important to discuss the main drivers of short-term energy prices. I would group them as follows:

Load and residual load

Higher residual load usually leads to higher prices because more expensive generation, such as dispatchable energy, is needed to meet demand and vice versa.

Solar and wind production

Higher solar and wind production usually leads to lower prices because they have low marginal costs and reduce residual load.

• Solar: Solar mainly affects midday prices because solar production is concentrated during daylight hours.

• Wind: Wind affects prices more evenly throughout the day because wind production happens during both day and night.

Hydro and nuclear production

Higher hydro and nuclear availability usually leads to lower prices because more low-marginal-cost supply is available. Some countries, such as France, are highly reliant on nuclear energy.

Gas / coal / carbon prices

Higher gas, coal, or carbon prices usually lead to higher electricity prices because gas and coal plants often set the marginal cost. The marginal costs of gas and coal power plants can be approximated as:

Using typical assumptions:

This is only an approximation, and additional operational costs must also be taken into consideration. These formulas also show why carbon prices are important.

Cross-border flows and congestion

More flows from cheaper areas usually lower prices and also help prices converge between markets. This is, of course, contingent on there being no congestion. When there is congestion cheap electricity cannot flow freely which can increase local prices and make prices diverge. This is why congestion is also important.

Lastly, for all of the above, we note two important things:

• Weather affects all of the above, either directly or indirectly. This explains why weather forecasting is of very high importance in this industry.
• These factors affect not only the price level but also the volatility and therefore the distribution of possible prices.

02

Understanding EU Short-Term Power Markets

EU short-term power markets bring together participants with different roles in the power system and different reasons for trading. According to EPEX, the main participants are:

• Generators
• Aggregators
• Utilities
• Large consumers
• Traders
• Transmission System Operators

These participants meet mainly on organized spot power exchanges, with the two main ones being EPEX SPOT and Nord Pool.

We now move on to the different types of short-term power-trading markets in the EU. The main ones are:

• Day-ahead auction: There is one day-ahead auction. It is the primary scheduling market and has the highest liquidity. The main products traded are 96 quarter-hour physical electricity contracts. The auction closes on D-1 at 12:00 CET or CEST for delivery on day D.
• Intraday auctions: Three scheduled auctions take place after the day-ahead auction. They were introduced in 2024 to create clearer intraday prices and allocate limited cross-border capacity more efficiently.
• Continuous intraday market: The continuous intraday market is a continuous order-book market used to correct and optimize positions close to delivery. The main products traded are 168 contracts, made up of 96 quarter-hour contracts, 48 half-hour contracts, and 24 hourly contracts. Trading usually starts on D-1 at around 15:00 and continues until close to delivery.
• Balancing market: The balancing market is the market where Transmission System Operators correct system imbalances in real time.

Here, it is also worth mentioning that the continuous intraday market has recently become increasingly important and liquid, mainly due to more renewables, greater market coupling, more batteries, and increased automation. This development has made market making more important in short-term power trading. We have also seen an increasing number of trading companies registering on EPEX SPOT, which has increased liquidity even further. This brings us to the next part.

03

The Role of Traders in the Market

Trading companies are not only speculators. They also perform several functions that help short-term electricity markets operate more efficiently. There are many ways in which trading firms help the markets, but the three main ones, in my opinion, are:

• Provide liquidity: Trading companies make it easier for participants to buy and sell electricity by continuously placing bids and offers in the continuous intraday market.
• Improve price discovery: Trading companies arbitrage away mispricings, helping prices reflect available information and market conditions more accurately.
• Enable the energy transition: Trading companies make renewable investments more attractive by efficiently managing renewable assets.

These activities also create several possible sources of revenue.

• Market making: Here, trading firms earn revenue from the bid–ask spread. For more on market making, see below.
• Forecast-driven systematic / directional trading: Trading firms use forecasting, modelling, and other methods to take directional positions in different spreads, such as in:
  • Day-ahead vs. intraday spreads
  • Intraday vs. imbalance spreads
  • Cross-border spreads
  • Time spreads
  • Granularity spreads

  This is mostly proprietary trading, meaning trading without assets. However, it can also be asset-backed because having assets is greatly beneficial, since those assets can be used as hedges for closing positions and lowering overall risk.

• Asset optimization: Trading firms manage their own assets or provide asset management for their clients. This basically means using forecasting and optimization to squeeze as much value as possible out of the assets by looking for opportunities in all markets.

Note that in both forecast-driven trading and asset optimization, assets can be used as hedges for closing positions, adding flexibility and, therefore, lowering risk. This explains the rise in asset-owning power-trading companies.

Making money through these activities as a trading firm requires an edge. This edge can come from many sources:

• Data, forecasting, and models: E.g., weather forecasting is of utmost importance because it can provide an edge across all sources of revenue.
• Execution / latency edge: One way this can help is through:
  • Less adverse selection because bids and asks are updated faster
  • Better queue priority in the orderbook
  • Reacting to a signal faster
• Microstructure modelling: Microstructure modelling is very important in the context of market making.
• Flexibility and optionality: One simple example is the added flexibility of owning assets with respect to hedging or closing out a position.
• People: Change is the only constant. You need a great team.

As market making becomes an increasingly important revenue stream for trading companies, I will also briefly explain the main risks faced by market makers, for which the bid–ask spread can be seen as compensation. The main risks are:

• Inventory risk: The risk of being directional (not flat) and prices moving against you.
• Adverse-selection risk: See below.
• Stale-quote risk: The risk of not updating quotes fast enough. This is one version of adverse selection.
• Liquidity risk: The risk of not being able to exit or hedge a position at a good price.

Adverse-selection risk is a particularly important one. It represents the loss a market maker faces when informed flow picks off the market maker’s quotes. This implies that the fair value of the underlying lies outside the bid–ask spread. Such informed flow can come from asymmetric information, where traders have information the market maker does not; latency advantages, where they update their quotes faster; or better models, where they estimate fair value more accurately.