Big data and AI are transforming the credit industry in Southeast Asia. This process is visible in Vietnam where credit markets undergo rapid expansion. With some 60 million citizens either unbanked or underbanked and short of credit history, digital credit scoring emerges as a magic tool to foster financial inclusion.
This technology combines big data gleaned from smartphones and machine learning algorithms to assess borrowers’ creditworthiness. But given that much is happening in this space in a legal void, regulations are critical to foster fintech innovation while safeguarding the public interest.
The role of startups in extending credit
Credit scoring startups capitalise on the high mobile phone and internet penetration rates and the rapid development of digital technologies in Vietnam. These technologies collect intimate data about borrowers based on their telco and mobile use, social networks, internet browser and e-commerce patterns. Scoring startups make an argument for efficiency by comparing their technology with traditional scoring based on economic data and simple statistical tools, an “archaic credit data and scoring technologies of the 70s”, typified by the official public credit, as stated by TrustingSocial, a scoring startup based in Singapore and Hanoi. According to startups, their technology offers greater accuracy because it leverages a larger number of data points relevant to risk.
Startups also promote speed and embark on a race for ‘real-time credit decision’ in consumer lending, especially for small loans. In Vietnam, banks and consumer lending companies thrive for providing a ‘fast and easy’ credit experience. FE Credit, the leading consumer lender and subsidiary of VPBank, promotes the power of big data to foster credit growth and financial inclusion. In 2018, it launched $NAP, an automated lending platform that helps FE Credit cut down bureaucracy, expands its customer base, maximise profit while reducing risk, and lead the race for a ‘fast and easy credit’ experience in Vietnam.
Challenges in digital credit scoring
However, digital credit scoring also raises concerns, in particular discrimination against vulnerable groups. These populations include the poor, women and racial, ethnic and religious minorities that have been excluded from credit markets, which is the case in Vietnam. The hidden and “black box” nature of machine learning algorithms exacerbates anxieties. A widespread fear is that ‘opaque’ algorithms will standardise past prejudices and biases into discriminatory rules that will reinforce inequality. This pattern may be emerging in Vietnam given that consumer lenders target mainly urban workers, especially from the middle class.
Privacy advocates call for minimising data collection and banning the gleaning of data considered ‘intrinsic’ to consumers’ identity. Public debate and compromise will be required to determine what data is intrinsic to privacy.
Another area where digital credit scoring stirs anxiety is consumer privacy. Lenders use credit pricing and customisation to discipline behaviour. In the US, credit scores assess borrowers’ creditworthiness and trustworthiness. They determine access to not only (sub)prime loans but also life prospects, as a good score is essential to purchase a home, seek higher education and secure a good job. In Vietnam, digital credit scoring does not impact people’s lives to that extent yet. However, banks and lending companies increasingly educate consumers about credit scoring and how to improve their chances of obtaining loans.
A last area of concern is cybersecurity. With rapid economic and digital growth and a growing number of internet users (66% of the population) and social media users (60%), Vietnam is an El Dorado for cyber-offenders. Data breaches are frequent. In 2019, the Government Information Security Commission reported 332,029 access attacks (using improper means to access a user’s account or network) and 21,141 authentication attacks (using fake credentials to gain access to resources from a user). These data are just the tip of the iceberg as cybercrime remains largely underreported.
Reforms and regulations vital
Big data and AI pose big challenges that require oversight and regulation. However, in its current state, the Vietnamese law is ill-equipped to regulate digital credit scoring. To facilitate fintech innovation, the Vietnamese government has issued the decision 999/QD-TTg for ‘promoting a sharing economic model’. This decision has paved the way for a regulatory sandbox programme in five key domains, including credit scoring. Fintech startups will be given two years to prove their new technologies and contribute to legal reform. Regarding digital credit scoring, two reforms should be given priority.
The first calls for reviewing creditworthiness assessment in credit law. The regulator could limit what data is collected, how it is used, how it can be transferred, stored and sold, and whether and how it should be shared with the credit bureau. It could also ban the use of sensitive proxies to limit discrimination and ensure consumers the right to oversee the scoring process, appeal for an explanation in case of rejection and correct their data.
An area where digital credit scoring stirs anxiety is consumer privacy. Lenders use credit pricing and customisation to discipline behaviour.
Another area that requires revision is privacy regulations, which are scattered. The law provides data owners with a comprehensive set of protections against companies that collect personal data. However, the law does not delimit what data companies can collect to protect privacy. Privacy advocates call for minimising data collection and banning the gleaning of data considered ‘intrinsic’ to consumers’ identity. Public debate and compromise will be required to determine what data is intrinsic to privacy.
Regulation is vital to support fintech innovation and positive social and economic change. Considering that digital credit scoring thrives in legal limbo in Vietnam, the regulator should address these new challenges and negotiate trade-offs between efficiency and public interest without further ado.