Price, sales, and the business cycle : a time series principal component analysis

The main contribution of this work consist on studying sales behaviour and their relationship with local market conditions like labor market indicators through a time series principal component analysis. We study the correlation structure of a large database on prices and found that all product sectors share a common correlation structure and the highest correlation and significance is achieved between employment variation and the first principal component, mostly in the second week of the following month. Sales or promotions, are a channel for price exibility because firms can use them to change effective prices keeping sticky reference prices. We use a rich database of retail prices from Uruguay to characterize prices' exibility, the behavior of sales, and their relationship with local market conditions like labor market indicators. Finally, we find a positive and significant relationship between sales and unemployment and perform a time series principal component analysis to study these relationships.