Peter Ulrich of RMS.
Tianjin multi-billion dollar cargo loss: an aberration or the new normal?
By James Brewer
Containerisation has greatly increased cargo catastrophe risk. This challenges the insurance industry to change the way it captures data to solve the problem of how to manage port-based accumulations of risk, such as hit insurers when a huge area of the Chinese port of Tianjin was devastated by explosions in August 2015.
This was the message from Peter Ulrich, senior vice-president of Risk Management Solutions, when he addressed a session of the International Union of Marine Insurance at its 2016 conference, in Genoa. Silicon Valley-headquartered RMS describes itself as the world’s leading catastrophe risk modelling company.
Mr Ulrich warned that “cargo cat” losses will continue to grow along with the development of shipping, increasingly risky port locations and rises in sea level.
Average daily exposure to cargo in transit in major global ports could be massive, and exposure in “free ports” was a great unknown.
Current industry data practices do not support detailed quantification of port exposure – but some unwelcome outsiders are on the job: Mr Ulrich cited a Business Insiderarticle which began: “High tech pirates hacked a shipping company to figure out the perfect ship to plunder, and located specific high value containers.”
Mr Ulrich entitled his talk Tianjin multi-billion dollar cargo loss: an aberration or the new normal?
Cargo accumulation risk in ports must be actively managed, he stressed, and his company was working on appropriate risk modelling.
Mr Ulrich said ports had evolved geographically – for instance, the Roman port of Londinium was 30 miles from the sea, and so was relatively safe. Later the port of London had thousands of individual warehouses and the many locations diversified the risk.
Formerly there were numerous key river ports with manual labour to load and discharge cargo from ships with 600ft by 60ft beam, to non-contiguous warehouse areas. Nowadays of course you cannot get a 18,000 teu ship (1,300ft by 194 ft) up the Thames, exemplifying that ports are moving closer to the sea coast for automated loading and unloading into vast container parks.
Superstorm Sandy which hit the east coast of the US in 2012 was “a wake-up call.” The storm caused a record for the time of $3bn-plus marine loss and $19bn property losses a part of which was borne by the marine market. Effective risk management enabled the Port of New York and New Jersey to return to action quickly, but while marine mitigation had focused on windstorm, the marine losses were driven by surge.
Tianjin, the third largest port in the world, in August 2015 suffered numerous explosions equivalent to 21 tons of TNT (twice the size of a large bomb) and fire at a hazardous chemicals storage facility. Total damage at the Chinese port could reach between $3bn and $5bn. Involved were large quantities of cargo containers, 70,000 cars, 176 deaths, 800 injuries – the loss estimate continues to grow.
Of the exposures, Mr Ulrich said that in general terms the most vulnerable cargo type was cars in open lots. “The most vulnerable is almost the most valuable thing. If you take one thing home today [from this conference workshop] it is that accumulations of cars are going to cause problems. Earthquake, flood, winds, hale – cars are extremely susceptible to these.” Automobiles in the open were twice as risky as autos in a container.
Of the Kobe 1995 earthquake, Mr Ulrich said that the Japanese port was built on two artificial islands of landfill, which meant disaster caused the sinking of buildings and flooding of cargo, also damaged by liquefaction. Amplified shaking destroyed piers and cranes: “it is as though you are on a trampoline, and everything gets shaken further.” He said that the disaster rendered 90% of Kobe berths inoperable, and other ports could not handle Kobe traffic, meaning that the shipping cost increased significantly. Most roads, railroads and lifelines connecting the port suffered severe damage. It took infrastructure longer to recover than the port.
In Japan’s Tohoku earthquake, tsunami and nuclear plant scare of 2011, many adjacent ports were impacted. Operational risk planning at ports needed to incorporate risk to adjacent ports and connecting infrastructure.
In response to those who said the costs of the Sandy event “could not happen anywhere else,” Mr Ulrich’s company looked at 100 ports around the world. Of these, 13 came up with potential losses higher than hit the Port of New York and New Jersey.
Until a short time ago, catastrophe modelling had never addressed the cargo fall-out. “This year, we have built a specific cargo model,” said Mr Ulrich, showing that under certain circumstances there could be similar losses to Tianjin.
Current industry “best practice” was to model cargo and specie as “warehouse contents.” In fact, vulnerability varied widely depending on the type of goods (auto, iPad, food, fine art etc), and on the way it is stored (container, warehouse, open lot, tank etc. The solution was to develop cargo-type specific vulnerability curves, and implement secondary modifiers to fine tune the curve, to give underwriters the ability to account for their specific knowledge, and account for different storage facilities. Cargo and specie exposures change over time, in location and value, and the type of catastrophe peril as a ship takes items across the world.
Current data capture practices were “sub-optimal,” Mr Ulrich asserted. The solution involved developing “hi-res port exposure databases” and performing detailed market share analyses of key ports.
Ports should incorporate risk information into their planning, such as avoiding storing high value items in high risk locations (for example no autos in low elevation open storage lots, and preparation for an approaching wind or surge event.
Catastrophe risk could constitute a material portion of the cargo risk for underwriters, and cargo data capture can enable accurate risk assessments: insurers will do the work to track more data on their cargo exposure (cargo type, storage, etc). “The benefits outweigh the extra effort – increased transparency and more accurate modelling of catastrophe load and thus improved risk selection.”
Typically a lower catastrophe risk led to better reinsurance decisions, and a more efficient use of capital.
Matthew Cocchiaro. Photo courtesy IUMI.
In a presentation at an earlier conference session, Matthew Cocchiaro, director for product management global trade at IHS Markit gave some examples of daily throughput in terms of value for districts in China. This sort of statistic helped in understanding the potential accumulation of cargo in case of a disaster. Noteworthy ports included Tianjin at $477m a day, Shanghai at $1.6bn, and Guangzhou at $311m.
“Looking specifically at Tianjin, we can see their top 10 commodities make up almost $200m of daily throughput,” said Mr Cocchiaro. In the event of another disaster, sectors that would be hardest hit were: vessels other than railway or tram stock, and parts $74m; nuclear reactors and machinery $59m; iron and steel $46m; mineral fuels, oils and products $34m; electrical machinery and equipment $30m; and iron or steel articles $25m.
Mr Cocchiaro said: “Global trade data allows us to understand cargo values and volumes across different trade lanes. The ranges in cargo value differ greatly based on different factors including geography, vessel size, shipper mix, etc.”
