Comment of EUROFER on the "Commission services paper on Energy Intensive Industries exposed to significant risk of carbon leakage" from the 12th of September 2008

EUROFER notes in general an increased convergence between the views of the different stakeholders on this issue. The content of the recent paper of the Commission is a positive example. EUROFER identifies the intensive stakeholder cooperation as the driver for this development and thus encourages the Commission to continue in this way. EUROFER believes that the progress made so far indicates that eventually it will be possible to design a system which is robust as well as representative for the industries concerned.

Also at EUROFER work is progress to contribute to the development the most appropriate system for leakage risk identification. Although this internal analysis is ongoing, today the following main characteristics can be defined: The starting point should be trade intensity. If there already is (or was) a significant trade observable, this identifies leakage risk (the same is true for observed dynamics in trade). Historically observed trade signals that only small trade barriers exist and that any additional CO2-cost-element will increase trade. Only in case of trade not observed, significant trade barriers (most easily expressed in sum of production cost differences and transport costs) can be assumed and the unilateral CO2-cost is to be compared to these. Thus whenever trade is observed or by trade-barrier analysis expected, leakage risk is present. Now, in a second step the leakage mechanism is to be investigated, both for short term leakage mechanisms and for long term leakage mechanisms. The first represents the view of the customer, the latter the view of the investor. According to this investigation, a differentiation into leakage risk-groups might be feasible (still to be investigated).

In the light of the leakage risk concept as outlined above, the following comments can be made on the paper of the Commission:

1. It is very positive and important to take both direct and indirect emissions into consideration.

2. Using cross-classification methods in principle is an elegant way to approximate very complex issues. The suggested principle is in line with steel industry thinking. However, it is imperative not to set the significance thresholds as depicted and thus establishing equal-surface-quarters. Instead some consideration has to be applied on how to distribute the surface of the diagram to the different risk-categories. This applies especially for the weights of the trade axis and the cost-axis. Significant trade (with respect to leakage, see above) is established already with low values, whilst for the cost-axis more poignant thresholds can be used to divide risk-categories within the at-risk-cluster.

3. The criterion of "trade openness" is in need of detailing. There are several possible indicators. It might be necessary to use more than one. The non-EU trade intensity indicator uses a definition of the market size that includes exports and thus artificially inflates it. This definition should be changed into (IMP+EXP)/(prod-EXP+IMP).

4. Trade openness indicators should provide a dynamic view of the market. Ideally, the indicator should not be restricted to observed data over the last years, but should rather take into consideration useful information that could predict market changes. Example: Because of the recent increases in steel production capacity in China (exceeding by far a receding local consumption) and the declining domestic demand, exports from that region are expected to soar (Metal Bulletin, 18th August 2008). More generally, market features of recent years will poorly reflect its future aspects in quantitative terms. Products with a today's limited exposition to global trade may in the future, because of market developments, be exposed more widely to global competition. However, as described above, the mere existence of historically observable trade indicates risk and thus it is not imperative to calculate the exact dimensions of future trade flows (evidence that changes are possible easily, should be sufficient). Therefore, observed trade should have a relatively low significance threshold and in case of non-exceedance of this threshold, the trade barriers should be investigated. Thus EUROFER considers the differentiation made into “first stage” and “second stage” not yet in line with the model as currently discussed by EUROFER and it therefore should be further analysed.

5. The 'non-EU trade intensity' indicator values shown in graph 1 for rebar is very low (about 1%) and far from reflecting the reality. When calculated with the data provided by Eurofer, it reaches 18%.

6. Using "cost increases" as an additional indicator is in line with the steel-industry concept of leakage risk-evaluation. This parameter mirrors the behaviour of customers when confronted with price increases. However, this does take into account only ONE leakage mechanism, namely the short term mechanism of loss of market share which materialises due to arbitrage by trading sectors. It does NOT take into consideration the leakage mechanism of investment decisions, which are characterised ideally by EBIT (but also approximated by GVA or production cost). Any leakage assessment methodology must also be based on long term leakage mechanisms.

7. The paper states that profitability (EBIT or GVA) is volatile (as anyway prices are) and is hardly reached with a sufficient level of certainty and thus should not be taken into consideration. We understand the need for simple concepts, but they must anyway reflects the main aspects of the reality. Even if complicated to appreciate, the EBIT or GVA must not be dismissed as it is a key element in the production and investment decision process. Impacts on profitability may have mainly long term effects (relocation) on leakage but short term implications should not be neglected.

8. The assumption that "existence of significant price differences at the international level indicates that markets are most national or European" is not correct. Regional price differences are the visible results of regional cost-pass-through capability. Based on these, equilibrium is achieved. Any additional unilateral cost element can not be passed through, because the cost-pass-through ability is already exploited to the maximum (except for non-traded good and services, for which the cost-pass-through capability is only confined by the price elasticity of demand). Once again, a sector or sub-sector with no significant international trade may as well be exposed to carbon leakage as a price increase subsequent to cost-pass-through of CO2 may attract new non-EU producers in the market – producers that were until then kept out of the market because of the relatively low prices).

9. The most robust quantification of trade barrier is the sum of production costs and transport costs.

10. The case of soaring commodity prices can NOT be compared to the situation under unilaterally imposed cost. Raw-material prices rise globally and thus can be passed through (to a certain extent !).

11. Taking into account the expected allowance shortage from 2013 on, it is advisable to consider higher allowances prices.

12. Annex 2 does not bear reference to the production of FeCr. EUROFER suggests to approach this sector for possible inclusion in the investigation.