Thursday, November 15, 2018
[Short excerpt from my submission to the European Commission in the context of its public consultation on the evaluation of the consortia block exemption regulation].
Please cite this document as:
Haralambides, H.E. (2019) "Gigantism in Container Shipping, Ports and Global Logistics: A time-lapse into the future". Maritime Economics & Logistics, (21):1, pp 1-63, February 2019].
I have often remarked that gigantism in container shipping has been induced by both port competition and shipping alliances. Indeed, without the possibility to use each other’s ships, no carrier by himself alone would be able to achieve a capacity utilization high enough to justify the use of present day mega-ships, while at the same time offering the frequency that shippers require.
But carriers have gone a step too far: At the time of writing, three alliances carry 80% of global trade. Such consolidation, in an industry that is already highly concentrated, is bound to finally attract the scrutiny of the regulator who, with the final consumer in mind, is likely to encourage more competition rather than further consolidation. If this happens, i.e., if container shipping becomes more open and competitive in the future, and if alliance agreements regarding vessel sharing, investment planning, etc. are scrutinized more closely for their compatibility with competition law, as I expect, the joint filling of the ship will become more difficult and ship sizes shall by necessity decrease, together with an increase in the number of ports of call. Low prices would then be achieved through more competition rather than big ship sizes. This is the more so when it is doubtful if the economies of scale in shipping are passed on to the final consumer, as required by the consortia block exception from the provisions of competition law in Europe.
A voice from the past: the ‘second scenario’
There are a number of macro-trends that, in addition to the above, might advocate for smaller ships and more port calls; particularly the latter. In a nutshell: (a) Transshipment costs and if they can help it shippers prefer to have their goods as close to them as possible; (b) Consolidation and distribution use land infrastructure without paying full costs for the private use of a public good; (c) The external costs of hub-and-spoking (congestion; pollution; accidents) may at times be as high as 2% of European GDP.
I thus argue that transshipment, warehousing and distribution don’t come cheap, as our enthusiasm with logistics often assumes. It is good to keep this in mind and thus make sure that the costs (internal and external) of logistics operations are paid in full, including the costs of using public infrastructure. The latter, because (to a large extent) infrastructure is no longer a public good and thus the user-pays principle should in principle apply.
 2M: (MSC, Maersk, HMM); Ocean Alliance: (CMA-CGM, Cosco Group, OOCL and Evergreen); and THE Alliance: (Hapag Lloyd, NYK, Yang Ming, MOL, K-Line).
 Haralambides, H.E. (2000) ‘A second scenario on the future of the hub-and-spoke system in liner shipping’. Latin Ports and Shipping 2000 Conference, Lloyd’s List, 14-16 November 2000, Miami, FL., USA.
Saturday, November 3, 2018
Mega-ships and mega-ports are the two faces of the same coin: the one ‘feeds’ and reinforces the other, and the one cannot exist without the other. I prepared this infographic in an effort to explain to a wider audience, in as simple a manner as possible, the ‘concept of hub-and-spoking’ (HS), in other words the intrinsic relationship between mega-ships and mega-ports.
The HS idea is quite enticing as a start: Simply put, “it is cheaper to ‘shuttle’ between hubs with a bigger ship and then distribute, rather than call directly at smaller ports, with smaller ships, serving a smaller demand”. A mega-ship can realize significant economies of scale as long as a) it sails full; b) spends most of her time at sea. To achieve both objectives, she must limit her ports of call to a minimum number of hubs, such as ports 1, 2, and 3 (calling also, however, at some additional ports such as 4 and 5). Port 2, let us say Singapore, is a ‘consolidator’: It attracts cargo, destined for Europe (or North America), from places as far away as Australia, Indonesia, Philippines, India, etc., and it has it ready, at its modern terminals, waiting to be expediently picked up by the mega-ship as soon as the latter arrives. Remember: such a ship cannot wait much and if it has to, before too long she may be looking for another, more efficient hub.
Port 1 in Northwestern Europe, let us say Hamburg, is a ‘distributor’ and the process is reversed: Our mega-ship arrives, drops its ‘call size’ (i.e. the number of containers destined for that particular port) within a tight time-window, and departs. What happens next, i.e., how Port 1 shall manage to distribute to its hinterland, as efficiently as possible, a huge inflow of thousands of containers, is not our ship’s concern; rather, our ship looks at this challenge as a NIMBY (not-in-my-back-yard) question.
Leaving it at that, the HS freight system is preferable to direct calling, reducing notably transport costs, in spite of the substantial feedering operations which are required (too many studies have convincingly shown this and I won’t attempt to repeat them here). In addition, the ‘consolidation-distribution’ operations around the hub ports create substantial economic activity for transport operators (road; rail; inland waterways; short-sea-shipping).
When it comes to feedering operations, there are two systems in daily practice, depending on the amount of freight volume available: If the latter is adequate, a shuttling system is often used, like the one between hub 1 and regional ports 9-12. When regional cargo traffic is limited, however, a cyclical feedering system is preferred, such as the one between hub 2 and feedering ports 6, 7 and 8. Finally, we also have two types of feeder companies: Dedicated feeders, belonging in other words to major carriers like COSCO, and common feeders, i.e. shipping companies that offer their services to all major carriers indiscriminately.
But we should not leave it at that; not any longer. As I have argued many times since 2000, starting with my “second scenario”:
- Transshipment costs: a bicycle manufactured in Vietnam and ordered in Madrid may be handled four or five times;
- Shippers do not like too much transshipment and long distances, preferring to have their containers as close to them as possible;
- Consolidation and distribution use land infrastructure without paying for the private use of a public good;
- External costs of hub-and-spoking (congestion; pollution; accidents) may at times be as high as 2% of European GDP;
- It is doubtful if the economies of scale in shipping are passed on to the final consumer, as required by the exception of consortia and alliances from the provisions of competition law;
- Mega-ships are becoming an increasing headache to most ports and distribution centers, and a NIMBY approach is no longer acceptable to them and to the taxpayer who finances them;
- Large ships reduce loop frequency and increase the inventory costs of traders, thus defying the very same principles of supply chain optimization;
- HS penalizes the legitimate development plans of other ports, particularly as major hubs, now claiming from others efficiency and market-driven port investments, have been financed with public money for most part of their economic life.
In the last quarter of a century, economies of scale in shipping, distribution and logistical systems have totally changed our lives to the better. But transshipment, warehousing and distribution don’t come cheap, as our enthusiasm with logistics often assumes. It is good to know this and thus make sure that the costs (internal and external) of logistics operations are paid in full, including the costs of using public infrastructure. The latter because (to a large extent) infrastructure is no longer a public good and thus the user-pays principle should apply.
One might counter-argue on the above that, in this way, higher transport and logistics costs would be passed on to the final consumer, as it usually happens with privatization. This may or may not be so, depending on how competitive transport and logistics markets are. But even if it is so, what is certain is that the final consumer will now be paying less taxes to develop ‘private’ infrastructure.
On balance, he should be indifferent.
 Haralambides, H.E. (2000) ‘A Second Scenario on the Future of the Hub-and-Spoke System in Liner Shipping’. Latin Ports and Shipping 2000 Conference, Lloyd’s List, 14-16 November 2000, Miami, FL., USA.
Sunday, October 21, 2018
The first time I expressed my concerns about the mega-containership phenomenon was in 2000, in my “second scenario”. I could then see 6 macro-trends, reinforcing each other, which could potentially halt the gigantism in container shipping, as well as in mega-hub-port development. These trends were: i) worldwide port development; ii) regionalization of trade; iii) infrastructure development in southern Europe; iv) road pricing in Europe; v) the future of liner shipping alliances; and vi) the impact of information technology.
I was laughed at then, as “the professor with the different opinion”. You see, those were the days when everybody was talking about the Malacca-Max (18,000 TEU) ships and similar creativities, which were listened to, unfortunately impulsively, by many ports. And I was laughed at again, because ships continued to grow unabated, in spite of the fact that my six trends had conspicuously materialized in these 18 years since 2000.
This, however, was not due to a failure of the underlying trends to influence ship-size development, but to failure in regulatory policy; both in terms of our inability to develop a coherent port policy in Europe, and our ‘eyebrow-raising’ leniency towards increasing concentration in liner shipping, in the form of global shipping alliances.
CMA-CGM, Drewry, OECD, and Fairplay now believe that things may have started to change. Adding a note of personal gratification, I am happy to take note of their recent conviction. However, things have started to change 20 years ago; we only didn’t know it; or did we?
 Haralambides, H.E. (2000) A second scenario on the future of the hub-and-spoke system in liner shipping. Latin Ports and Shipping 2000 Conference, Lloyd’s List, 14-16 November 2000, Miami, FL., USA—(you can download the paper from my Academia and ResearchGate profiles).
Tuesday, October 16, 2018
In liner shipping, the pursuit of Economies of Scale (EoS) through bigger and bigger ships is something like the pursuit of the Holly Grail. Thirty years ago, the largest containership would just pass through the locks of the Panama Canal and it had a size of roughly 4,700 TEU (known as a Panamax-size ship). Today, the size of the largest containership is five times that, i.e. about 22,000 TEU. What are therefore those EoS, and are they as important as carriers tend to believe?
Economies of Scale (EoS) refer to the situation where unit costs (i.e. cost/dwt or cost/slot), in other words the costs relevant to pricing and competitiveness, decline as ship size increases. This decline is more pronounced in the case of shipbuilding costs, manning costs and fuel costs. Simply put, you don’t need twice the amount of steel in order to build a ship twice the size, nor twice the crew to sail it, or fuel to move it. In the case of shipbuilding (capital) costs in particular, the above figure speaks for itself when it comes to the relationship between cargo-carrying-capacity (CCC) and ship dimensions: The almost doubling of CCC, from 3,500 to 8,000 to 15,000 TEU requires only moderate increases in ship dimensions and, most importantly, draft. This shouldn’t surprise us: The amount of steel I need to build a ship increases linearly while CCC increases in the cube of dimensions.
The bottom part of the figure (borrowed with many thanks from the Port of Hamburg) is even nicer: The larger ship (March 2018) is the CMA CGM 'Antoine de Saint Exupéry': 400 meters long, 59 meters wide, with a container capacity of 20,776 TEU. In her gut (May 1968), like Jonah in the gut of the whale, is the "American Lancer": 213 meters long, 26 meters wide, with a container capacity of 1,200 TEU. The Antoine is just twice as long and about twice as wide as the Lancer, but it can carry 17.3 times more cargo (20,776/1200)! What better example could one need to explain economies of scale? The question however is: can our ports continue to cope with this kind of gigantism? Should they? My answer is clear and unswerving: They can (could) but they shouldn't.
 The size of a panamax ship was determined by its beam which could not naturally exceed the 33.53 meters width of the Miraflores locks of the Panama Canal. As the ‘beam to length’ ratio of a ship is fairly fixed ( in heavy weather, a ‘sausage’ type ship could easily break in two if, when loaded, it found itself at the top of two wave crests), the maximum size of a panamax containership was about 4,700 TEU. At a beam of 33.5 meters, the ship can carry 33.5/2.5=13.4, or 13 containers across (counting containers astern is an easy way to impress your interlocutor, when asked about the size of a ship…).
Thursday, March 22, 2018
Liner rates are complex structures that include various, often obscure, components, such as BAFs, CAFS, etc. Regarding public announcements on the other hand, and as the argument goes, if carriers only wanted to inform their clients of future price increases, they could do so directly (e.g. by letter or email), without recourse to public announcements in the Press. Economic theory suggests that, in a cartelized market, advance price announcements by one producer (carrier) might lead to actual price increases by its competitors.
Recently, the European Commission and the Russian Competition Authority (Federal Antitrust Service -FAS) have instigated investigations, adopting “remedies” that virtually annul any possible impact of GRIs on actual freight rates.
But was this effort necessary?
New econometric research appearing in MEL shortly shows that the impact of GRIs on actual prices, if any, is minimal and if it exists it exists only under specific market conditions. Moreover, assuming GRIs could have some effect on competition, they constitute explicit collusion and economic theory says that such collusion is attempted only when the ‘conspirators’ are unable to sustain for long tacit collusion (which is the real concern of competition authorities).
Much ado about nothing therefore? Shouldn’t we be spending our time instead on the grave consequences of alliances and of shipping industry concentration?
I (and many shippers and consumers) think so.
Friday, February 23, 2018
Following a long legal dispute with the concessionaire (DP World of Dubai, UAE), the Government of Djibouti decided today to unilaterally terminate the concession and seize the Doraleh Container Terminal (DCT), operated by DP World. From what it is known, the dispute concerns the government’s wish to renegotiate the concession agreement; and no doubt, DP World understands what this means… But there’s more to this dispute that meets the eye.
In view of its geographical location on the Horn of Africa (Gulf of Aden; see map), and its function both as a gateway port to Ethiopian trade and as transshipment hub, Djibouti is a very strategic port and a modern day Cerberus, watching the entrance to the even more strategic Red Sea.
As a result, Chinese interests in this port-city are also very strong: China Merchant Holding International (CMHI) has developed and financed the new Doraleh Multipurpose Port (DMP), next to the defunct old port, so as to relieve city congestion and turn the old port and its waterfront into a business and recreational district. A Free Trade Zone is also being planned by CMHI next to DMP.
As competitors, DP World and CMHI do not quite see eye to eye, and doubtful if they could both ‘fit’ in the tiny state, even if they wanted. Moreover, it is becoming increasingly known that UAE is deploying naval forces in the neigh of Djibouti, while China has already been promised a naval base there.
In conclusion, let me say for one more time: “trade and infrastructure” go a bit further than just counting containers and making port rankings; quite a bit actually...