Why can hierarchical clustering be computationally expensive on large datasets?

A. It cannot store labels B. It scans transactions for itemsets C. It requires many pairwise distances and repeated cluster comparisons D. It estimates one probability only