14 M.Soppe et aL/Joumal of Transport Geography 17(2009)10-20 100 ☐HPH 60 APMT ■PSA 40 DPW CMH South Fig.3.Terminal operators'throughput distribution over world's regions(2006 data, th or's throughput) Source:Containerisation International 2006 Soppe M.,Parola F..Fremont A. and PTOs.In the following sections our quantitative analysis will self-handling 13% disclose the relations between carriers and their port providers, in order to identify the strategies devised by SLs to secure reliable terminal operations. 4.Shipping lines'operations in ports:the selection of major other ITOs handling providers 15% local TOs Fig.4 provides a breakdown of the WCTC by type of terminal 44% operator.The diagram reveals that the local operators account for nearly half of global capacity of the shipping lines.This suggests that despite the inroads of the vertically or horizontally integrated global companies,local and independent terminal operators re- main an important part of global container handling activity.It may be in part a reflection of the geographical disconnection men- top 5 ITOs tioned in the previous section,whereby the needs of global carriers 28% in the lesser markets are not yet served by ITOs and therefore use the local firms. Fig.4.Overall distribution of weighted weekly containerised transport capacity The market share of the other types of terminal operators(Fig.4) (2006 data)Source:WWCTC database(2006).Soppe M..Parola F..Fremont A. reveals the relative weight of"TOP 5"and"other ITOs",which to- gether account for a similar market share of local TOs.The "TOP 5".as expected,are particularly important,representing 28%of total shipping capacity.In contrast,shipping lines'terminal handling more vertical integration are still concentrating their efforts in subsidiaries account for only 13%of the total capacity,a figure hid- ports of their home market(Hanjin in Asia,Cosco in China,MSC ing substantial differences among various carriers'behaviour. in Europe). In this respect,Fig.5 provides a breakdown of the shipping The second group(Gp2)is composed of CMA-CGM,MSC,NYK lines'capacity by type of terminal operator,self-handling,interna- and Yang Ming.Similarly to the previous cluster these shipping tional providers and local operators.The triangular graph reveals lines are characterized by their wish for independence through four groups of carriers with similar strategies for securing terminal portfolio diversification in ports.Their traffic is handled by a wide services.The first group (Gp1)is composed of Maersk,Evergreen range of terminal operators both local or international,without and Hanjin.In the stevedoring market they are quite'independent' relying on a restricted set of partners.Their use of local players carriers as they are not heavily committed to any third-party ITOs outside Europe and the Mediterranean,areas in which they have and use their own network of facilities.Maersk handles 30%of its invested in many port facilities,shows the typical approach of WCTC through its own subsidiaries and JV.while for Hanjin and MSC and CMA.This is due both to their desire for independence Evergreen this share is around 26%and 22%,respectively.In this from major PTOs and to the structure of their maritime network, relation Maersk,as a pioneer in integrated handling activity,holds often calling secondary ports where international stevedores are a notable strategic and economic advantage.Its network is not only not present.Also the other two carriers of the cluster,NYK and the widest among ISLs,but also the most diversified among all ter- Yang Ming,reveal a high WCTC share handled by local operators. minal operators (both ISLs and PTOs).Its maritime network relies The major characteristic of the two remaining SLs groups(Gp3 on hub ports whose control is of strategic importance.While Ever- and Gp4)is to have a high share of WCTC handled by ITOs.How- green follows the Maersk's pathway towards a global network of ever,they reveal different self-handling activities and resort to lo- almost exclusive terminals,other top shipping lines striving for cal operators.Hapag-Lloyd,China Shipping and MISC(Gp3)haveand PTOs. In the following sections our quantitative analysis will disclose the relations between carriers and their port providers, in order to identify the strategies devised by SLs to secure reliable terminal operations. 4. Shipping lines’ operations in ports: the selection of major handling providers Fig. 4 provides a breakdown of the WCTC by type of terminal operator. The diagram reveals that the local operators account for nearly half of global capacity of the shipping lines. This suggests that despite the inroads of the vertically or horizontally integrated global companies, local and independent terminal operators re￾main an important part of global container handling activity. It may be in part a reflection of the geographical disconnection men￾tioned in the previous section, whereby the needs of global carriers in the lesser markets are not yet served by ITOs and therefore use the local firms. The market share of the other types of terminal operators (Fig. 4) reveals the relative weight of ‘‘TOP 5” and ‘‘other ITOs”, which to￾gether account for a similar market share of local TOs. The ‘‘TOP 5”, as expected, are particularly important, representing 28% of total shipping capacity. In contrast, shipping lines’ terminal handling subsidiaries account for only 13% of the total capacity, a figure hid￾ing substantial differences among various carriers’ behaviour. In this respect, Fig. 5 provides a breakdown of the shipping lines’ capacity by type of terminal operator, self-handling, interna￾tional providers and local operators. The triangular graph reveals four groups of carriers with similar strategies for securing terminal services. The first group (Gp1) is composed of Maersk, Evergreen and Hanjin. In the stevedoring market they are quite ‘independent’ carriers as they are not heavily committed to any third-party ITOs and use their own network of facilities. Maersk handles 30% of its WCTC through its own subsidiaries and JV, while for Hanjin and Evergreen this share is around 26% and 22%, respectively. In this relation Maersk, as a pioneer in integrated handling activity, holds a notable strategic and economic advantage. Its network is not only the widest among ISLs, but also the most diversified among all ter￾minal operators (both ISLs and PTOs). Its maritime network relies on hub ports whose control is of strategic importance. While Ever￾green follows the Maersk’s pathway towards a global network of almost exclusive terminals, other top shipping lines striving for more vertical integration are still concentrating their efforts in ports of their home market (Hanjin in Asia, Cosco in China, MSC in Europe). The second group (Gp2) is composed of CMA–CGM, MSC, NYK and Yang Ming. Similarly to the previous cluster these shipping lines are characterized by their wish for independence through portfolio diversification in ports. Their traffic is handled by a wide range of terminal operators both local or international, without relying on a restricted set of partners. Their use of local players outside Europe and the Mediterranean, areas in which they have invested in many port facilities, shows the typical approach of MSC and CMA. This is due both to their desire for independence from major PTOs and to the structure of their maritime network, often calling secondary ports where international stevedores are not present. Also the other two carriers of the cluster, NYK and Yang Ming, reveal a high WCTC share handled by local operators. The major characteristic of the two remaining SLs groups (Gp3 and Gp4) is to have a high share of WCTC handled by ITOs. How￾ever, they reveal different self-handling activities and resort to lo￾cal operators. Hapag–Lloyd, China Shipping and MISC (Gp3) have Fig. 3. Terminal operators’ throughput distribution over world’s regions (2006 data, % of the terminal operator’s throughput). Source: Containerisation International 2006, Soppé M., Parola F., Frémont A. Fig. 4. Overall distribution of weighted weekly containerised transport capacity (2006 data). Source: WWCTC database (2006), Soppé M., Parola F., Frémont A. 14 M. Soppé et al. / Journal of Transport Geography 17 (2009) 10–20