Industrial energy exposure within a common market

Eurostat's non-household electricity dataset records prices of 0.2264 euros per kilowatt-hour for Germany and 0.1534 for France in the second half of 2025 for annual consumption of 500 to 1,999 megawatt-hours. These observations exclude VAT and other recoverable taxes and levies. Both are drawn from the same dataset, consumption band, currency, period and tax definition, as available in the October 19, 2026 data vintage. Eurostat, non-household electricity prices, band IC.

The difference is 73 euros per megawatt-hour, or approximately 47.6 percent of the French observation. Converting between the units multiplies the kilowatt-hour price by one thousand; it does not create another independent observation. A hypothetical purchase of one thousand megawatt-hours at those unit prices would differ by 73,000 euros before other circumstances were considered.

This comparison is deliberately narrower than a claim about all German and French manufacturers. The band contains consumers with different load profiles, contracts and activities. Larger energy-intensive establishments can fall in other bands, and recoverability differs from a statement that all taxes have been removed. The harmonized observation establishes a price difference for a defined statistical category, rather than the invoice every firm would obtain.

Germany's network regulator reports that renewable sources supplied 59.0 percent of net generation entering the public grid in 2024. Its stated coverage excludes industrial and railway networks outside that boundary. It also reports a German annual average day-ahead wholesale price of 78.51 euros per megawatt-hour for that year. These are separate generation and wholesale-market observations, not the delivered price in the preceding table. Bundesnetzagentur, 2024 electricity market data.

France's energy regulator records nuclear generation of 361.7 terawatt-hours in 2024 in its wholesale-market surveillance report. That figure describes an important part of French supply, using the report's stated electricity balance. It does not reveal the full price paid by each industrial consumer. CRE, 2024 wholesale electricity and gas market surveillance report.

The generation evidence refers to 2024, whereas the harmonized retail comparison refers to the second half of 2025. No difference between the two retail prices is attributed mechanically to those earlier generation figures. The national generation series also have different coverage, so their aggregate levels are not presented as a like-for-like production ranking.

Generation can affect available supply, price-setting conditions and exposure to fuels or weather. Connecting those mechanisms to a customer's bill nevertheless requires the relevant delivery period, market arrangements and contract. A low operating cost at one generator does not imply that every purchaser receives electricity at that cost. Similarly, an annual average wholesale price does not identify the timing or quantity of a plant's contracted purchases.

A useful invoice decomposition separates energy procurement, network charges and nonrecoverable taxes or levies. The components should be measured using the same physical volume and delivery period. A change in one component is then applied only to its exposed share, rather than to the entire invoice.

Consider a hypothetical baseline electricity bill indexed to one hundred. Energy procurement accounts for sixty, network charges for twenty-five and nonrecoverable taxes for fifteen. Assume the market price relevant to exposed energy procurement increases by 50 percent while network and tax charges remain unchanged. These are illustrative assumptions, not the actual component weights in either country.

A buyer with 80 percent of its energy component fixed for the delivery period has an exposed energy component of twelve. The market increase adds six to its total bill, producing an index of 106. A buyer with only 20 percent fixed has forty-eight exposed; its bill rises by twenty-four to 124. The calculation is 60 × exposed share × 0.50, added to the original bill.

The same market-price movement produces different current bills without requiring different physical generation at the customer's location. The fixed arrangement delays exposure; it does not remove the question of the price available when the contract expires. Comparing one buyer's protected period with another buyer's renewal period can exaggerate a persistent national difference.

Volume and shape also matter. A contract for a fixed quantity may leave additional production exposed, while a standard delivery profile can differ from a plant's operating hours. The relevant hedge is therefore the coverage of actual consumption under its contractual conditions. A percentage labeled fixed should not be assumed to protect every possible production schedule.

National industrial costs also depend on which activities are being aggregated. An economy with a larger share of electricity-intensive production can devote more of its industrial revenue to electricity even when comparable plants pay the same unit price. Conversely, a lower aggregate electricity share can reflect a different production mix rather than superior energy efficiency.

A second hypothetical exhibit holds unit prices constant. Let electricity-intensive production spend 30 percent of its revenue on electricity, while another industrial group spends 5 percent. Economy A derives 40 percent of the combined revenue from the intensive group; Economy B derives 20 percent. These labels do not correspond to measured German or French sector weights.

Economy A's aggregate electricity expenditure share is 15 percent: 0.40 × 0.30 + 0.60 × 0.05. Economy B's share is 10 percent: 0.20 × 0.30 + 0.80 × 0.05. Giving both economies a common 30 percent intensive-sector weight produces the same 12.5 percent expenditure share.

This standardization isolates a composition effect under the assumptions. It does not erase the actual economic importance of industrial structure. A country with more exposed production really does face a different aggregate adjustment problem. The distinction is between explaining that problem and interpreting the aggregate share as a measure of plant efficiency.

Combining the two exhibits requires another explicit step. Procurement coverage belongs to particular consumers and activities. Applying one national average contract share to every sector can conceal the concentration of exposure in a few firms. A useful assessment therefore links sector weights to sector-specific electricity intensity and contract timing, rather than multiplying unrelated national averages.

Electricity expenditure per unit of revenue contains both a physical and a price component. Physical electricity use per unit of output reflects the process, product specification, operating conditions and utilization. The purchase price reflects the defined procurement and delivery terms. Revenue also moves with the product's selling price.

A higher expenditure share can consequently arise even if the plant uses no additional electricity per physical unit. Its output price may have fallen, its electricity contract may have renewed or its product mix may have changed. An efficiency claim requires physical input and output measures with a sufficiently comparable production boundary.

The margin effect is another question. If a producer can raise its selling price by the additional cost without losing sales, its immediate margin exposure differs from a producer facing a fixed selling price. Neither outcome follows simply from being located in a country with a particular generation mix. Customer contracts, competition and the availability of substitutes determine how much cost can be passed on.

Energy-intensive intermediate products can also transmit costs to other industries. A firm with modest direct electricity purchases may buy materials whose prices embody substantial electricity use. Direct exposure and supply-chain exposure should be identified separately. Adding the full supplier bill to the purchaser's electricity bill would otherwise risk counting the same cost more than once in an aggregate account.

A unit-price series does not measure all aspects of the electricity service. Some processes require continuous supply, while others can shift operating hours. Interruptions, restart costs and product losses can make the same average tariff economically different across plants. Those effects need operational evidence rather than an assumed premium assigned to a national energy source.

Flexibility can reduce exposure when production can move to cheaper hours, but it may require storage, different equipment or changes to labor schedules. The saving should be assessed after those costs and constraints. A calculation that assumes an industrial process can stop whenever prices rise can misstate the feasible response.

Investment can change physical intensity over time. Assessing it requires the cost of equipment, expected operating life, maintenance and the electricity saved under a specified utilization rate. A high current price increases the potential value of savings, but does not independently establish that every proposed investment has a positive return.

For a cross-country comparison, these operating conditions should accompany the matched price band. Otherwise a nominally cheaper supply arrangement may be compared with a different service, different flexibility or a different production process. Matching the consumer category is the beginning of comparability, not its completion.

Recoverable taxes are excluded from the selected price series because they differ from a final cost retained by the purchaser. That does not establish that every included consumer has identical cash-flow timing. A recoverable payment may still need to be financed until it is recovered, while a nonrecoverable charge remains part of the cost. The statutory invoice, final economic burden and temporary financing requirement should therefore remain separate concepts. When comparing a particular pair of plants, analysts should retain the statistical tax boundary and then identify any documented differences in actual treatment, rather than substituting an assumed national exemption for the measured price.

The Eurostat observation establishes a substantial delivered-price difference for a specific non-household consumption band and historical half-year. The generation evidence supplies context about the systems in which those consumers operate. Keeping the periods and boundaries explicit prevents either observation from being asked to explain more than it measures.

The two hypothetical decompositions then identify distinct sources of industrial exposure. Contract coverage determines how much of a market-price shock enters the current invoice. Industrial composition determines how much electricity-intensive production contributes to an aggregate cost measure. Physical efficiency, selling-price adjustment and indirect input costs add further dimensions.

An industrial competitiveness assessment should therefore begin with matched delivered prices, follow actual procurement and renewal conditions, and apply them to comparable production activities. It can then distinguish a price disadvantage, a timing difference and a composition effect. That sequence supports an economically interpretable comparison within the common market without treating generation shares or a single national average as a complete account of industrial energy costs.