In October 2024 compared to the previous month, the producer price index of industry excluding construction has increased by 0.5%. A split of this figure in energy and the rest shows a rise of 2.7% for energy and a rise of 0.03% for total industry except energy. A further breakdown of total industry except energy obtains -0.4% for intermediate goods, except energy, +0.1% for capital goods, +0.3% for durable consumer goods and +0.5% for non-durable consumer goods.

When the general producer price index is subdivided according to destination, a rise of 0.6% on the domestic market and a rise of 0.4% on the non-domestic markets can be observed.

Compared to October of the previous year, the producer price index of total industry excluding construction has decreased by 1.8%. For the domestic market it fell by 1.2%, for the non-domestic markets there was a fall of 3%.

As part of the change of base year (2021 = 100), the methodology for the output price index has been revised. The major changes made are:

• the creation of an intermediate classification of enterprises in the sample into strata
• the inclusion of new enterprises from year to year
• the control of chain drift
• the definition and application of a strategic outlier exclusion method adapted to the data we receive
• the imputation by the highest NACE level available

Enterprise strata

The enterprises in the sample are weighted according to their turnover. Before the revision, the weight was a ratio between the enterprise and the sum of the turnovers for each NACE and market. The new method introduces an intermediate level. We create three strata: small, medium and large enterprises. The weights are then rebalanced, avoiding an over-representation of large sampled enterprises and an under-representation of small enterprises.

Inclusion of new enterprises

Firstly, we took advantage of the creation of the new indices to include prices that had been collected but not yet used. Secondly, in the new method we have added an annual review of the enterprises used, enabling us to allocate a weight to the new enterprises in the survey and to include their prices into the index from year to year. In this way, the representativeness of the sample can be guaranteed over time.

Control of chain drift

Chain drift can result from price variations, pulling the index up or down. When product prices return to their initial level without the index also returning to its base level, there is a chain drift. This effect is mathematically corrected in the new method.

Exclusion of outliers

Previously, outliers were only detected based on an abnormal month-to-month ratio or a change in description. In the new method, other cases are added, such as prices that have been stable for more than 18 months, or identical prices entered more than 25 times. This new strategy enables us to improve the quality of the data received via our survey. This change makes the indices more representative from month to month.

Imputation by higher NACE level

In the old method, when no price was available for a certain NACE market, the evolution used was that of another market for the same NACE or it was imputed by 1 (constant). In the new method, the change from month to month is imputed by that of the higher NACE level available for the same market. This method makes it possible to take into account the difference in reality between the various markets in the same activity sector. For example, if we do not receive any prices for NACE 1011 on the Belgian market, we will use the evolution in NACE 101.