How Meu Crediário helps large networks reduce defaults

[jisansorkar8990]

If you've come this far, you may be wondering: how can I reduce defaults in my store? We at Meu Crediário want to help you – just as we helped a large chain , with more than 50 stores, to reduce defaults and increase sales .

Identifying this need is the first step towards achieving better results.

Therefore, in this article, we will talk about the strategies we implemented to support a larger network. But remember that this is a method that works for any network, regardless of size.

It all starts with our Back Test. To understand how it works, you can continue reading or watch the video below. Let's go!

YouTube video
Back Test My Credit: the first step to reducing your store's defaults
The Back Test is the first step that will guide us towards this goal of reducing default. From there, we will conduct a long study, identifying the behavior and distribution of sales, in addition to understanding when to analyze the bureaus and how to define credit policies.

When we perform the so-called Back Test, the store facebook database gives us the database it currently has and we send this information to the credit engine . This is done so that we can identify how the behavior of these sales would have been with our own credit analysis.

This Back Test helps the store to have more security when closing with Meu Crediário , since it will know how assertive our score is in its operation.

From this assertiveness, we can also identify the distribution of our credit score , that is, we can verify, among the sales that our engine found in the Back Test, how the distribution of sales was, how many customers, for example, are of profile A, B, C, D or E.

With all the study we carried out, we also assessed whether this distribution is adequate.

Evaluating sales distribution
I know that in sales with risk profiles A and B, default rates must be around 3%. Profile C, in turn, will have default rates around 9%. And the closer to high risk, the default rate will be around 30%.

If I do the BackTest and identify that these indexes are not at values ​​close to those presented, I can recalibrate the credit engine.

It's like taking a car to a mechanic and asking them to resurface the engine so that it has better performance.

In this case, then, eventually, we will be able to increase the number of low-risk clients – or reduce it, it will all depend on the behavior of the portfolio.

And the same happens with the medium risk profile: I can expand it, while those with high risk, I can reduce it.

In practice, all this means that the store will implement a credit policy with a sales block for a certain profile . These are sales with such high risk that the store really doesn't want them to happen.

When we go to a demonstration meeting about the score distribution based on the Back Test data, we can have a more assertive view and have a very positive impact on the board.

Imagine that I go to a board of directors and say: there is 5% of your sales that have an average default rate of around 30%. Would you feel comfortable making the decision to cancel and block this type of sale? I imagine so.

When we consult credit bureaus
With the Meu Crediário Back Test, the store stops making decisions by checking whether the customer is in default or not. Now, it starts to set limits according to the risk profile, with a very low limit for customer D and blocking for customer E.

Based on this classification, we will then identify when we should seek information from credit bureaus.

Based on the Back Test, we were able to identify, for example, that in the case of a client with risk profile A, I can reuse SPC information for a period of 90 days. A client with risk profile B, for 60 days. And so on.

This type of information will help the store make the appropriate credit decision based on the risk profile and no longer looking exclusively at what I have in return from the credit bureaus .

The bureau is still important information, but it becomes secondary data in the operation , while the primary data is the risk profile.