An important question is how - and when - regional markets develop in relation to global markets. A common assertion is that local financial markets in a country pre-date national and global markets, but is this accurate? Better knowledge of how regional and global markets grow can help policy makers seeking to promote financial development both internally and internationally.
In this working paper, The Eli Lilly chaired professor of innovation, business and society, Bruce Kogut and Pietro Urso of INSEAD, and Gordon Walker of the Cox School, Southern Methodist University, make a detailed study of the American venture capital (VC) market over a 44-year period.
The authors analyse 159,561 syndication transactions between VCs in the United States during the period 1960 to 2004, taken from the Venture Economies database of the Securities Data Corporation. Using graph analysis methods, they investigate the relationships between incumbents and new entrants, and the extent of diversification.
Venture capital is used to fund enterprises, and an investee company may receive several rounds of venture funding to support its growth. When large sums of money are involved, funding firms are naturally keen to spread the risk, so VC firms form syndicates with one or more additional investors.
Networks form as firms connect and work together and in the VC industry the formation of syndicates provides a good marker for the development of connections, and hence networks.
The study of networks helps us to evaluate whether the subjects tend to form new links with existing partners or to look for new partners and thus expand their reach, which has implications for strategies regarding risk and diversification into new geographies and sectors.
Surprisingly, the results for the VC network in the United States suggest that so far it has followed a unique evolution that distinguishes it from other industries. The automotive industry, for example, demonstrates the typical pattern of growth followed by a period of shake-out and concentration, but the VC industry has experienced massive and growing entry throughout the period covered by this study.
Despite this massive entry, and even allowing for exit, the results point to the very early development of a coherent structure in the VC network.
The VC network is characterised by a few firms that have a high number of connections, and an extremely large number of firms that have only a few connections. The authors determine that repeated ties are more important than ties demonstrating 'preferential attachment', i.e. ties based on prestige and reputation.
Although it appears that incumbent firms prefer transacting with past partners, they must also form links with new partners if they wish to diversify into new geographies and new sectors. The coherence of the network structure supports firms looking to expand through sector and geographical expertise because it promotes communication within and between regions and within and between sectors.
And because this coherence developed early in the VC network, this supports the idea that local, regional and national connections developed concurrently rather than consecutively.
These results suggest that it is not possible to understand local financial market developments without understanding inter-regional ties and connections. It seems that, currently, policies for financial and industrial regional development are often formed in opposition to, rather than in conjunction with, the dynamics of the global market.
This study implies that policy makers would do well to take a broader perspective if they are not to go against the grain of market development.