In the past, your credit score was the all-important metric that decided if you could access credit. Of course, it had its limits, but there were no alternatives for decades. Today, things are changing. Large banks and rising fintech startups alike now have an abundance of alternative data sources they can tap into. At the same time, advances in technology now allow us to process vast quantities of data to gain better insights quickly.
You may also like
Hence, financial institutions can now make credit decisions quickly with greater accuracy. The trend has resulted in better experiences for consumers. But it also comes with its own set of challenges. Speaking to Arteculate, Co-Founder of Taran, Martin Chudoba, shared with us the different approaches the banking industries of Europe and South-East Asia are taking in embracing decision intelligence for credit scoring.
A quick history of calculating credit scores
In the past, if you approached a bank to get a loan, your credit score decided everything. Back then, financial institutions could only rely on this figure provided by credit bureaus to gauge the likelihood of a loan being repaid. But, despite its prominence, it wasn’t a perfect measure. Firstly, there is a cost that banks have to bear with obtaining the score itself. On top of that, unless you paid for a full credit report, there’s no way to view the information that went into calculating the figure. Even then, you’d have found this calculation was very narrow as it focused mostly on credit history.
In contrast, data unrelated to your credit history could be used by the bank to better predict your ability to repay the loan. But the traditional process ignores this critical possibility. Yet, even considering this possibility is essential by fintech companies, banks, and other financial institutions. Not everyone has a credit history needed to obtain credit scores, particularly in emerging regions. By locking them out of the financial system, these societies are now dealing with unfortunate consequences.
For example, according to the Global Findex Database, only 27% of adults have bank accounts in Southeast Asia, and 67% of businesses lack credit scores. So even if a company enjoys steady recurring revenue, they’re unable to access traditional financing to grow, which hinders economic growth. Further, at an individual level, it means that only risky sources of credit like payday loans are available when a financial emergency occurs. Being unable to serve this populace has resulted in several social issues and proved to be a lost opportunity for banks and financial institutions.
Hence, banks have been trying to reduce their reliance on traditional credit scores. They sought data from alternate sources, which would serve to complement the existing traditional data. Thereby, allowing them to calculate their own credit scores. Already we see the largest banks in the U.S. reducing their reliance on the FICO credit score and utilising more data from internal and external sources for credit decisions. But this phenomenon didn’t happen overnight.
Instead, it’s the result of gradual progress over decades. Martin describes this journey of progress, saying, “Traditionally, it was a slow process where banks only had a few data points to decide with. Then came new tools that introduced a degree of automation based on rules and maybe a simple scoring model. Eventually, as technology advanced and new data sources became available, banks could take complex credit decisions in real-time. Thus, allowing financial institutions to offer a better experience to their customers.”
Data and the competitive edge it offers
In the 21st century, humanity lives in a vast digital ocean of data. Today, organisations can tap into several internal and external data sources. Some examples of these alternative data sources within banking and finance are internal transactional data, information from external parties such as telecom providers, and customers’ digital footprint on a lender’s website or mobile app. The latter refers to a broad range of information, which is particularly valuable as lenders can use it for credit scoring and anti-fraud purposes. For example, lenders could use general device data to compare an applicant’s address with their current timezone, identify their income level based on the device model, or confirm they read the full terms & conditions.
From those examples alone, we can see that banks and fintech companies can gain an invaluable competitive advantage by utilising all data from non-traditional sources. Most notably, it allows them to implement their proprietary credit scoring systems. In addition, automated data analysis from multiple sources allows for near-instantaneous responses to credit applications. Thus, improving the consumer experience dramatically as a process that used to take days is now done in seconds.
“Anyone seeking credit will always go for the easiest option,” explains Martin, “They don’t want to waste time filling out 20 forms and waiting days for a response. Therefore, financial institutions should ensure their application process is as seamless as possible.” So naturally, a bank or fintech company with a quick and simple application process will see consumers flocking to it en masse. Yet, this is only the beginning. The ability to rapidly analyse data from several sources means these companies can now better identify consumers who will repay their loans. Moreover, the technology also reduces the cost of this entire screening process. Thus, financial institutions can offer credit on better terms to the right consumers.
In contrast, the financial institutions that ignore these trends are setting themselves up to fail. The best applicants with solid credit profiles have already gone to wherever they can obtain credit quickly with minimal hassle. Thus, the remaining lenders are now relegated to being the second choice. In this position, the only consumers they get will be those with risky credit profiles rejected by those lenders with data-driven credit decisioning processes. Therefore, financial institutions must limit themselves to only asking for essential information. Instead, obtain data from other sources as must as possible to guarantee a hassle-free experience. By not utilising alternative data sources in the credit decisioning process, a financial institution would be giving up an invaluable competitive advantage.
Unlocking the potential of data from new sources
While the availability of new alternative data sources offers several benefits, they come with their own set of challenges. Previously, a bank might have monitored only ten attributes for a consumer. Today, that figure has risen to hundreds or even thousands of data points, including internal and external sources. Integrating these data sources is a complex process, and some, like traditional credit scores, have a cost element. Hence, Taran’s flagship product, the Taran Decision Manager, was designed from the ground up to integrate several data sources cost-effectively.
Once the data has been collected, the next step is to process it. With consumers now expecting instant results, banks have to process a mountain of data in a flash. But many existing decisioning systems struggle to do so. These systems were designed to deal with small volumes of data with rules or a simple logistic regression model. But now, with so many data points, a single consumer could be a few megabytes, which can be beyond the capabilities of such traditional systems.
The solution Martin notes is to utilise machine learning instead. “Technologically, we first need to process all this data, and secondly, we need to make sense of it. When dealing with a handful of predictors, a logistic regression model is fine. But when you have hundreds of predictors with non-linear dependencies and larger datasets, it’s more efficient to utilise machine learning. Granted, it involves taking new factors into account such as how to train the model, how to parametrize it, and so on but there is unmistakable value to be gained by adopting this approach.”
The infrastructure needed to harness the power of data
These events have meant financial institutions have to invest in data science talent and expand their IT infrastructure. The latter is a costly, time-consuming endeavour. But the limited capabilities of existing systems mean it can take months of coordination between data scientists and developers to implement new models. It’s here that the Taran Decision Manager steps in and serves as a bridge between the two parties, which speeds up the entire credit decision making process.
Martin elaborated, “A bank’s existing infrastructure can easily integrate our engine to complement and extend their capabilities. The Taran Decision Manager lets them process vast amounts of data while also rapidly implementing new models in a matter of hours!” In doing so, Taran enables financial institutions to make better-informed credit decisions in a short manner of time. Thus, allowing the organisation to enjoy the competitive advantage of giving its consumers a better experience.
The differences between Europe and Asia
Since its launch, the Taran Decision Manager has enjoyed strong traction in Europe and Asia. The system has found a home in both large banks and rising fintech companies alike in both territories. Sharing their experiences, Martin notes that organisations of equal size in both regions have similar technology infrastructures in place. However, he comments that smaller fintech companies are more enthusiastic about adopting cloud technologies.
By and large, financial organisations in both Europe and Asia agree upon the importance of alternative data sources over merely relying on traditional credit scores. Where they differ is in their approaches to utilising these new data sources.
“Asia is home to a fast-growing young population. But, unlike in Western economies, they lack formal credit scores. So financial organisations in Asia, particularly fintech companies, place greater emphasis on alternative data sources,” explains Martin, “In Europe, the importance of these alternatives aren’t ignored either. However, owing to good credit coverage and tighter regulation, banks still place a heavy emphasis on traditional credit scores.”
Today, leading financial institutions are moving away from simply obtaining credit scores from alternative sources. Instead, these companies are now investing in building their digital data collection capabilities. Looking at this trend, Martin observes, “Initially, this will allow them to reduce costs as they can freely obtain the data they need. But this is only the beginning. The 3rd party credit scores, even with alternative data, can be obtained by competitors as well. Whereas possessing the raw underlying data means a company can create their unique competitive advantage to become a market leader.” He adds that the Taran Decision Manager was designed from the ground up to help banks and fintech companies obtain the data they want from multiple sources efficiently.
Financial organisations everywhere agree that they can no longer rely solely on traditional data. They are using more alternative data sources in their credit decisioning processes despite differing approaches between regions. Yet, these new data points won’t replace any traditional data sources. Instead, the two will co-exist and complement each other. However, the deciding factor between them will be the infrastructure they have in place. Whether it’s a rising fintech company or a large established bank, financial institutions everywhere must invest in building a flexible and robust risk management infrastructure. Only then can the actual value of this data be unlocked to react quickly to market changes and gain a valuable competitive advantage ensuring long-term success.
