Keeping price indices representative despite constant market changes using auxiliary information in index estimation formulae

Compiling price indexes includes many challenges. The CPI (Consumer Price Index) calculation, for example, is performed in different stages, from an elementary aggregate level up to the general index level. This paper is focused on the study of the aggregation structure of the CPI. In particular, it shows how auxiliary information (such as market shares) can be added to the aggregation structure of the CPI in order to enhance the quality of the final estimates, comparing new proposed methodologies of index compilation to the classic ones. Preliminary results confirmed that there are noticeable differences considering indexes by type of store (specialized and non-specialized). Thus this seems to suggest that a new aggregation structure can improve the robustness and the representativeness of the estimates of prices movements. The study is developed through a longer period of time in order to evaluate in the long-term the impact of the use of auxiliary information. The chaining perspective will be also assessed. Moreover the new proposed methodology and the traditional methods are compared to the indexes (defined sub-indexes) computed by type of store.