Why do we price each MWh as if it was produced using gas even though most of the generation comes from other energy sources which are currently much...
Uniform pricing, part 1
Would other market designs have smoothened the increase in prices seen in the last months? Gain a deeper understanding of how prices are formed and explore potential alternatives.
The fall guy of the energy crisis
Recent increases in electricity prices (e.g. average day-ahead prices have increased by more than +200% in Germany between January 21 and January 22) have pushed people to question the European electricity market design, and in particular, the way prices are formed on the day-ahead market that serves as a reference price. More specifically, it is the so-called uniform pricing (also called pay-as-cleared) mechanism of the day-ahead auction that was sometimes pointed out as a cause of the significant price increase witnessed last year.
Would other market designs have smoothened the increase in prices seen in the last months? In the first of our two-part series and to gain a deeper understanding of how prices are formed and explore potential alternatives, we will have a look at the current design of day-ahead markets.
The uniform pricing of day-ahead markets
In a nutshell, the day-ahead auction sets the price of electricity for delivery the next day by matching supply and demand in each hour of the day. On the supply side, the set of generation units that will contribute to the supply of electricity is determined by the merit order. This means that less expensive generation units will be called first followed by increasingly more expensive generation until market equilibrium is reached, i.e. at the clearing point of supply and demand curves.
As illustrated below, renewables typically have a marginal cost close to zero and thus come first in the merit order, usually followed by nuclear, coal, gas, and oil generation units in order of increasing marginal cost. The marginal price is the marginal cost of the last accepted generation unit, and it is also the price that all accepted generation units will get for their production. Similarly, on the supply side it is the price that consumers will pay for their accepted offers, hence the designation “pay as-cleared”: the same price applies to each market participant that succeeded in the auction, no matter its original bid or offer.
With gas rising more than 300% in 1 year - even before the start of the conflict in Ukraine- and gas-fired power plants being the marginal generation technology in many European countries, we saw electricity prices surge significantly even in countries where gas represents only a small share of the generation mix. When comparing average day-ahead prices between January 2021 and January in France, for instance, they rose by 250% in 2022, even though gas-fired generation accounted for only 11% of the production mix in January 2022. It is thus legitimate to ask why a small share of the generation mix creates an upward trend for the whole market. In other words, why is the reference price for electricity set at the marginal price? After all, renewables and nuclear have not become more expensive and in countries like France, they account for a majority of the electricity production.
Pay-as-cleared versus pay-as-bid
A natural alternative to pay-as-cleared is the pay-as-bid mechanism, where participants receive the price of their bid/offer. A natural guess is that the pay-as-bid would decrease the price of electricity, as the reference price would become the average cost of production instead of the marginal one. In other words, the price of electricity would not be set by the most expensive technology, but by the full mix of technologies used in the generation mix.
This however presupposes that market participants will not change their bidding strategy if the market design changes. Unfortunately, economists have every reason to think that they will. Under uniform pricing, suppliers have every reason to bid at their marginal cost. They know that if any of their bids are rejected because there are lower bids sufficient to satisfy the demand, they will be better off because they will not have committed themselves to sales at prices that fail to cover their avoidable costs. More importantly, they know also that on their accepted bids they will receive the full benefit of whatever price above that level is necessary to equate demand and supply in the market, regardless of the level of their own bids. The difference between the clearing price and the bid price of a generation unit is called the inframarginal rent and serves to cover fixed costs and, hopefully for the investors, to make a profit over the long term.
A sufficient reason why in a pay-as-bid mechanism, suppliers would never bid at their marginal price is that they would not recover their fixed costs. Renewable energy sources, which usually have close to zero marginal cost but high investment costs would never be in the money. Similarly, for gas-fired power plants which nowadays tend to be rather far in the merit order, it is not unusual to be turned off for several days due to low prices. Those plants, therefore, need high generation prices when called upon, e.g. in low renewables situations, to recover their fixed costs.
For this reason, and to maximize profit, in a pay-as-bid mechanism suppliers would bid at what they expect will be the market-clearing price, so the marginal price of the most expensive unit needed to cover demand. In a situation where all market participants would have a perfect forecast of this price, the total cost needed to cover the demand would be exactly equal to the cost incurred in a pay-as-cleared mechanism. This will however never hold true (as forecasts are by definition subject to errors) and in the second part of our series, we will discover that a pay-as-bid mechanism would likely reduce efficiency, disadvantage small players and decrease investments in generation capacity.
Source: Alfred E.Kahn, 2001