§3.2.1.2 Evidence of past substitution

51. In certain cases, there may be evidence of substitution between different products following past structural changes, events or shocks in the market. Such evidence can be particularly informative for demand substitution, in particular when the substitution is caused by an exogenous (76) shift in relative supply conditions of the products of the undertaking(s) involved, such as an unexpected cost shock, or by another similar event. Launches of new products can also provide useful information, when it is possible to analyse precisely which products have lost sales to the new product (77). Similarly, changes in quantities in reaction to a supplier’s exit from the market, or resulting from (temporary) unavailability of certain products (for example due to production outages or supply-chain disruptions), may be informative for substitution patterns (78). However, reactions to a significant change, such as the complete unavailability of a product or the introduction of a new product, may not always indicate how customers would react to more limited changes in relative supply conditions. 52. In some cases, undertakings may also collect relevant information on demand substitutes during the ordinary course of business. For example, an undertaking may have data on the customers that it has won and lost and the identity of the competitors which lost/won those customers. By contrast, evidence of customers shifting away from a product as a result of factors unrelated to changes in relative supply conditions, such as a change in preferences or consumption patterns over time, are less informative for demand substitution (79). 53. Where there is sufficient data on past substitution, it may be possible to derive quantitative measures on the substitutability of different products. For example, it may be possible to derive diversion ratios between candidate substitute products. Diversion ratios estimate the share of sales volumes lost by the product of interest that is diverted to each candidate substitute product in response to changes in relative supply conditions. In addition, it may be possible to estimate own-price elasticities and cross-price elasticities (80) for the demand of a product using econometric techniques (81) (82). In order to be reliable, such quantitative evidence must be sufficiently robust.

← 3.2.1.1 · All articles · 3.2.1.3 →

Source: EUR-Lex (Cellar) · retrieved 2026-09-07