Where is the cargo right now? Storage, accumulation and the limits of the spreadsheet

Cargo insurance is written as though goods are moving. Most of the loss history says otherwise.

Goods sit; in a warehouse waiting for a container, in a port waiting for customs, in a consolidation facility over a weekend, in a yard because the onward leg was cancelled. Transit is punctuated by storage, and storage is where value concentrates. A single shed can hold more insured value on a Tuesday night than a vessel does, and unlike the vessel it does not appear on anyone’s screen.

This has been a theme in marine market commentary for some years, usually framed as an underwriting problem. It is also a broking problem, and the broking side of it gets discussed far less.

The question a client will eventually ask

Sooner or later a risk manager asks a version of this: how much of our insured value is currently sitting in one place, and where?

It is a fair question. It is also, for a broker running a cargo book across multiple clients, open covers and countries, surprisingly hard to answer inside a working day.

The information exists. It is in the declarations. But if declarations arrive as email attachments and spreadsheets, and get keyed into a system that treats each one as a document rather than as data, then the answer requires someone to go and read several hundred rows across several files and add them up by hand. By the time the answer is produced, the goods have moved.

Why this is getting harder rather than easier

Three things have changed the shape of the problem.

Supply chains hold more buffer stock than they did before 2020. That was a deliberate response to disruption, and it means more goods at rest for longer, in more places.

Routing has become less predictable. Diversions, congestion and rerouting mean that where goods pause is often not where anyone planned for them to pause, so a static schedule of locations no longer describes the exposure.

And capacity providers have become more specific about what they want to see. Requests for exposure information have become more granular, and the broker who can answer quickly has an easier renewal conversation than the broker who cannot.

What answering it actually requires

Not a model. A structure.

If each declaration is captured as data, with its locations, values, dates and currency held as fields rather than as text in an attachment, then the accumulation question becomes a query. If declarations are captured as documents, no amount of analytics on top will fix it, because the data was never there to begin with.

That is the unglamorous version of the answer, and it is the one worth acting on. Before any conversation about exposure modelling, ask whether the declaration data in your own system could support the question at all.

A short test. Pick one client and one open cover. How long would it take, today, to produce a list of every declared consignment currently at rest, by location, with values converted to one currency? If the answer is measured in days, the problem is structural rather than analytical.

A note on where we sit

tigerlab BMS holds open cover and declarations as structured records rather than as attachments to a policy document, which is the precondition for questions like this one. It was built with a global cargo brokerage, in co-creation with the brokers using it, with first markets live and rollout continuing across more than 30 countries.

We would rather you tested your current system against the question above than take our word for it.