How mobile SDK works for lending decisions

Are we the only ones who find it alarming that there are digital lenders who give out loans to Nigerians without a credit history?

Where exactly is the confidence coming from?

Many Nigerians don’t have a formal credit history but this is no surprise. The credit gap in Nigeria is N109 trillion wide and the tedious loan process with banks discourages individuals from going through this credit route. Many resort to less formal credit channels like loans from families, cooperative contributions, etc.

A number of lenders who are more cautious and aware of the realities, use smart technologies called software development kits (SDK). With this, they are able to program their mobile apps and offer some functionality they may otherwise not have access to. A top provider of these SDKs is Pngme, and even soon, Lendsqr.

 

How do SDKs help lenders make lending decisions?

Mobile SDKs are special software components integrated directly into a mobile app that allows developers to enjoy functionalities without having to write from scratch.

Lending SDKs work by using alternative data from borrowers, usually from their phones, to help the lenders make better informed decisions. Typical data that can be made available via SDKs include SMS, list of downloaded apps, call records, and thousands of other data points.

For instance, with SMS alerts from banks, a lender can use an SDK to reconstruct a borrower’s financial record and then know how much they can offer as a loan. Using smart algorithms and sometimes machine learning, lenders can infer from this data, the customers that are most likely to pay back.

To avoid getting ahead of ourselves, we should state that these decision models are not gods; they are only able to help lenders determine the probability of outcomes. However, we can all agree that  in lending, nothing is 100% sure.

 

Ethical concerns around the use of SDKs

Given that SDKs are used for the collection and processing of data, there are a few ethical concerns around this:

The first, is the issue of customers’ data privacy and protection. Lenders in their use of technology, are bound by the Nigeria Data Protection Regulation to protect customers’ data. Customers’ data is only to be used in line with approved guidelines and given consent.

At Lendsqr we maintain that borrowers’ right to privacy is not negotiable and we don’t support unethical lenders or provide them with the technologies to violate borrowers’ privacy. Unfortunately, the reality remains that some lenders can be quite unethical in their dealings. They may even go as far as sending SMS blasts to contacts of borrowers’ who delay repayment (this is not news to anyone).

Secondly, algorithms, no matter how advanced, are still written by humans, hence, they may be prone to biases. This may create unfair outcomes for certain groups of borrowers. Borrowers’ loan requests may be treated differently based on arbitrary standards and preferences or even assumptions in the model which may be erroneous.

Some have argued that SDKs should be banned but with limited available data for borrowers, lenders are loath to lend blindly. A halt on SDKs could potentially hurt borrowers a great deal as there aren’t that many credit options for them. Without some level of reassurance for lenders which these SDKs provide, lending profitably becomes a challenge.

SDKs have proven to be useful and important in the loan decision process. However, the government should increase enforcement efforts when it comes to poor handling of customer data. Bad lenders should be sanctioned or barred from lending once found in violation of customers’ data privacy.