So let's consider other components one by one.

And we then go to the computation of the the index itself.

So the first component as I just mentioned is the number of

print shipments required to get from country I to country J.

As we have seen before,

we know that having a direct connectivity

could actually be a big booster for your exports.

The second component is the number of common direct connections.

So this is an indicator that should reflect

actually the capacity to get from country I to country J.

Considering all the possible alternatives.

So the higher the number of

common direct connections the easier it should be to move from country I to country J.

Even if there is actually a common direct connection between this two countries.

The third component is actually made of

two indicators of centralities of the countries making a specific country pair.

And that is the geometric mean of the number of direct connections.

So the number of direct connections a country has is a possible indicator of centrality.

And by putting together two countries with a very high centrality in the network of

international trade maritime connections is

likely to reflect a higher capacity to trade with each other.

The fourth component is

an indicator of the level of competitions on services that connect country pairs and that

correspond to the number of

companies providing a service between country I and country J.

The fifth component is the size of

the largest ships on the weakest routes and that is an indicator of infrastructure.

So these last two components are also part of the LSCI.

And they were explained extensively by Jan previously.

So once we have these five components,

we have to decide how to put them together.

So the units of all these components is not the same,

so we have to establish the unit free component and

unit free index based on components that are also are unit free.

So we most adopt normalization formula to get to these results.

The formula we consider is the following.

So we take the raw value of the components that we observed for a specific country.

We subtract to it a minimum value that correspond

actually to the minimum observed over the whole period under

investigation and we divide

these difference by the difference between corresponding maximum value.

So the maximum component can take over

the whole period under investigation minus the minimum value.

So we opted this for this formula rather than

the perhaps more standard formula looking simply at

the ratio between the raw value of the component and the maximum of the later.

Simply because the existence of minimum

values which are different from zero impose this choice.

And that's our case here.

If all minimum values of all our components were zeroes,

the two formulas would be perfectly identical.

So the LSBCI once we have done these normalization is computing by

simply taking the average of the five normalized components.

In other words, we will obtain values lying between zero and one.

So we can multiply these values by one hundred to get something between zero and

one hundred but that is simply a type of scale decision in terms of scale.

So we'll stick here to the range zero one.

This type of normalization allows us to

compare the values that we obtained for the index,

not only across countries but also across years.

As we have decided to fix maximum and minimum values,

otherwise if we decided to choose

maximum and minimum values corresponding to each of the years during that period,

well no comparison would be possible across years.

So now our normalization and the decisions we took in terms of

maximum and minimum value once again will allow us a direct comparison across,

not only across countries but also across time.