The lack of data: a problem to be solved to help the credit manager
A major problem in the analysis of credit risk and of the work that each day has to be done by the credit manager is to address the lack of financial data in many countries of the world.
In fact, unfortunately, not all the world is the same in the distribution of public information companies:!
Below the table is (roughly) what we are facing:
As we see, the situation is extremely varied: it goes from countries where all the data are present and therefore the company’s data looks like this:
In countries where some parts can miss and then the company’s data is as follows:
A situation where there is nothing and then it is so:
Then you understand the many difficulties facing the credit manager: for him it is still crucial to get the definition of a credit limit and a credit scoring, regardless of whether the information is complete or not.
Therefore the problem is reversed, I would say rightly !, to companies like ours which are working on the assessment of credit risk: we have to help the credit manager in its credit management.
And how can we do? The only way is to give to the credit manager all necessary tools to deal efficiently every day his work.
This means, provide an assessment of the credit risk that is as efficient as possible: the models of credit limit and scoring / rating must be efficient and accurate even in the absence of data!
For this reason, companies working on risk credit assessment must be very attentive to the methodologies they use and then evaluate very well the methods used: in modeFinance for example we are very careful with the use of machine learning methods as they are very sensitive to the availability of data (in this post I discussed in detail the problem).
This does not mean that many times the big data model / machine learning can go very well!
Returning to the table above we could definitely say that in some countries (Italy, Spain, France, etc.) statistical models can be definitely used efficiently, but, unfortunately, in most Countries, those models are lacking.
The big problem is that the credit manager works almost entirely with credits distributed all over the world!