n order to test the hypotheses, we estimate the CCHM model (Eqs. ((1), (2), (3) and (4))) and subsequently derive price elasticity figures using Eq. (5). The SUR estimation procedure yields a satisfactory fit with a McElroy R2 of 66% and 56% for cardigans and sneakers respectively. Table 2 summarizes the corresponding results. We find significant cross-price elasticities in 28 of 842 brand relationships. This translates to a quota of about 3.3% and thus represents only a small share of all possible effects. In support of Hypothesis 1, this value is well below the two-digit values found in offline settings.