Commentary

By Lars Jensen

Will the hub-focused product of Gemini be more or less prone to delays than MSC ’ s direct product ?

As we enter the final months of 2024 , there is increased focus on the question of whether the Gemini Cooperation alliance of Maersk and Hapag-Lloyd — launching in February — can deliver on its promise of 90 % reliability .

And running slightly below the surface is the question of whether Mediterranean Shipping Co . ( MSC ) will be in a better place to deliver reliability given its much larger array of direct services in contrast to Gemini ’ s hub-centric network .

Ocean Alliance and Premier Alliance are somewhere in between the two outer positions of MSC and Gemini .

Of course , time will tell , but what can we foretell from an analytical perspective ? To do that , it is instructive to focus on the two extremes represented by MSC and Gemini .

The simplest approach is to look at past performance . The problematic issue , obviously , is that the three carriers involved have not operated in these new constellations .

However , we can look at the existing 2M Alliance consisting of Maersk and MSC . Across all their alliance services , they had an average on-time performance of 84 % in 2019 , which was the last “ normal ” year before the major disruptions presented by the COVID-19 pandemic and Red Sea vessel diversions .

Breaking it down to a trade lane level , 2M was 73 % on time on the Pacific headhaul , 92 % on Asia to Europe and 74 % on the Atlantic westbound trade . 2M was the best-performing alliance on all headhaul trades , with the only exception being the trans-Atlantic , where THE Alliance performed better .

Hence , at least on the Asia – Europe trade , we have indeed seen 90 % performance previously , meaning that a 90 % target is not out of reach for carriers intending to do this .

Then we come to the core question : Will the hub-focused product of Gemini be more or less prone to delays than MSC ’ s direct product ? First , some simple mathematics . Take a service calling just two ports , one in Asia and one in Europe . Assume the carrier runs a tight ship ( pun intended ) and there is only a 2 % risk of the vessel getting delayed due to external circumstances such as bad weather .

In this case , there is a 2 % chance of getting delayed prior to loading in Asia and a 2 % chance of getting delayed prior to arriving in Europe , leading to an average reliability of 96 % by the time the vessel arrives in Europe .

But what if we had two ports in Asia and two ports in Europe ? Then the reliability in the last port in Europe would be 92 %. With a total of five ports in the rotation , reliability would be 90 %.

Of course , this is a theoretical approach and does not include the option of a carrier speeding up to get back on schedule or having a buffer of time to use .

If we look at Asia – Europe , Gemini has six ports on all services , except a single service with nine ports . Using the hypothetical 2 % risk of delays , this would lead to a Gemini reliability of 89 %, except for the one service with many port calls that would be at 83 %.

For MSC , all but two of its Asia – Europe services have 10 to 12 port calls . In the hypothetical of 2 % delay risk per port , this would imply a service reliability of 78 % to 82 %.

Hence , from the perspective that each port call is itself a risk factor that can reduce reliability , it is clear that Gemini has a better chance of staying on time throughout the full rotation than MSC .

Then there is the question of reliability for ports served directly by MSC but via shuttle by Gemini . In this case , there is an added risk that if the Gemini shuttle is not timely , for whatever reason , it further degrades the reliability .

Typically , the shuttles call one to two outports apart from the central transshipment hubs . This means an additional one to four ports in the rotation depending on whether a shuttle is needed in both ends or just in a single end . This would bring the port calls for the Maersk network up to seven to 10 , implying a reliability of 82 % to 87 %.

This means that if the focus is solely on the risk related to being delayed in each port , the Gemini network ought to , statistically , have a slightly higher core reliability than MSC when the need for shuttles is taken into account .

But it also shows that for Gemini to deliver on the promised 90 % reliability , the risk of being late in each port — when also taking the shuttles into account — must be less than 2 %. In this hypothetical scenario , it would require the Gemini network to operate with only a 1 % risk of a vessel getting delayed in each port .

But , of course , there are two additional tools that can also be used : having larger buffers to absorb smaller delays and speeding vessels up when necessary .

email : lars . jensen @ vespucci-maritime . com

44 Journal of Commerce | November 18 , 2024 www . joc . com