Big DataCorner OfficeExpert OpinionNewsletter

The Role Of Alternate Data in Fintech

data analytics

In 2020, global economies were rocked by a pandemic, leading to unprecedented uncertainties. This impacted almost every sector adversely and each sector is pushing hard to emerge out of this crisis strongly in 2021. The financial sector continues to work diligently to adapt to the new normal. The adoption of technology, digitization and use of data are not only critical to come out of the current crisis but also to build strength against any such future events.

Financial services have always been focused upon the use of technology and data as differentiated strategy. However, the recent crisis has expedited the adoption of technology and the search for newer ways to get more data and insight.

As the industry went through a crisis, many consumers were impacted financially due to job losses, income cuts and no business activities for a long period. This, in turn, had consequences on consumers’ ability to attend to their current financial obligations, repayment of loans etc. The challenge is that this data on consumers’ new financial standing might not be easily available in a structured format for usage. Therefore, financial institutions must find a way to gather relevant data for a better insight into the consumer financial health. While this has always been a key priority for financial services, the importance of this data increases manifold in the current scenario to support the risk buying process and to offer a customized product. This is where alternate data, which is more than traditional information of repayment, becomes important for lenders.

Alternate data has had many definitions over the years. The International Committee on Credit Reporting (ICCR) expands on alternate data as ways to collect and analyse data on creditworthiness, based on information readily available in the digitized form but alternative to conventional methods such as documented credit history. This data gives insight into consumers’ unique needs. The alternate data, coupled with advanced analytics, can also be used to assess the creditworthiness of many unbaked and new to the credit segment consumers. Alternate data is also considered to be information readily available in digitized form, collected through a technology/electronic platform.

For lenders, such alternate data provides a deeper understanding of a consumer, leading to a broader credit process to predict their financial behavior. Alternate data also enables smarter product offerings on adequate terms. Alternate data can be broadly categorized as follows:

  • Structured data: utilities, mobile phone, rental information, tax data etc.
  • Unstructured data: social media, e-mail, texts, internet usage etc.

While all this data originates from various sources and in different structures, the use of advanced analytics provides an opportunity to use this information for consumer insight and to build a score which can strengthen the underwriting process of the lending organization. This further builds the capability to appropriately match the consumer profile with the best suited offerings. One key aspect to consider before processing alternate data for underwriting purposes is to curate the information and then use AI/ML to identify hidden patterns and systematic biases.

For financial services, every data is a credit data as it facilitates a deeper understanding of a consumer for an informed decision on buying risk. The importance of alternate data has increased considering there is lack of traditional information on credit history. The adoption of various forms of alternate data is primarily done to evaluate consumer “Identity”, “Intent” and “Capacity”. Broadly, this data can be divided into three categories.

  • Tier 1: Financial account data – This is the most acceptable data for determining the consumers’ capacity to repay a loan and to establish their credentials. Though this is established and historical information on the credit file, it can be further used for a better insight on consumers by deriving predictive summary variables.  Examples of this type of data are transaction level banking, credit trade line information, demand deposit account etc.
  • Tier 2: Utility data – Financial services can gather additional insights on a customer by using the next level of alternate data, which would not have been possible by using only traditional data. This is the most reliable alternate data among other sources. Insights from these data points can be determined from monthly spending patterns, consistent behavior, and financial discipline. The use of this information can also be combined with other data points to determine the identity and the intent of consumers. Examples of this type of data are mobile bills, electricity bills, DTH bills, rent etc.
  • Tier 3: Non-financial data – This next level of data is indicative of consumer behavior, lifestyle choices and spending patterns. This includes use of information from the social media, strictly not be taken on face value. However, usage of this information, in combination with Tier 1 and 2 categories, can provide a deep insight into consumer behavior. Examples of this are employment details, public records, education details, mobile device information, retail purchase information etc.

As lenders are seeking additional ways to assess credit worthiness of consumers, these data points become very important to understand the risk and to make credit available to consumers when they need it.However, one of the key points to focus on is the way an organization processes and consumes this information. Usage of advanced analytics and machine learning is of utmost importance while analyzing and using this data.

Lending institutions need to realize the value of alternate data and must make an investment to obtain this information.Further, there must be an investment in building capabilities like AI and ML to leverage this information for a robust scoring model.

New age lending organizations have entered the space of small ticket unsecured lending and are paving the way for the use of this information. However, there is a long way to go before it becomes a key component of enhanced customer decisioning.

(The author is Pratyush Chandramadhur, Chief Business Officer at AuthBridge Research Services and the views expressed in this article are his own)

Leave a Response