As with Japanese keiretsu, the member firms in a Korean chaebol own shares in each other and tend to collaborate with each other on what is often a nonprice basis. The Korean chaebol differs from the Japanese prewar zaibatsu or postwar keiretsu, however, in a number of significant ways. First and perhaps most important, Korean network organizations were not centered around a private bank or other financial institution in the way the Japanese keiretsu are.8 This is because Korean commercial banks were all state owned until their privatization in the early 1970s, while Korean industrial firms were prohibited by law from acquiring more than an eight percent equity stake in any bank. The large Japanese city banks that were at the core of the postwar keiretsu worked closely with the Finance Ministry, of course, through the process of overloaning (i.e., providing subsidized credit), but the Korean chaebol were controlled by the government in a much more direct way through the latter's ownership of the banking system. Thus, the networks that emerged more or less spontaneously in Japan were created much more deliberately as the result of government policy in Korea. A second difference is that the Korean chaebol resemble the Japanese intermarket keiretsu more than the vertical ones (see p. 197). That is, each of the large chaebol groups has holdings in very different sectors, from heavy manufacturing and electronics to textiles, insurance, and retail. As Korean manufacturers grew and branched out into related businesses, they started to pull suppliers and subcontractors into their networks. But these relationships resembled simple vertical integration more than the relational contracting that links Japanese suppliers with assemblers. The elaborate multitiered supplier networks of a Japanese parent firm like Toyota do not have ready counterparts in Korea.9