Rising non-performing loans (NPLs) have forced many banks to significantly increase provisioning, placing substantial pressure on business performance. In this context, digital transformation in debt management is emerging as a strategic direction for many credit institutions to control risks from the very beginning of the loan lifecycle.
NPL pressure and the management challenge
According to official data, the on-balance-sheet NPL ratio of Vietnam's banking system increased rapidly during 2022-2023, then gradually stabilized by late 2024-2025, but remains at a high level. Specifically, the State Bank of Vietnam (SBV) reported an on-balance-sheet NPL ratio of 1.9% at the end of 2021 (compared to 1.69% at the end of 2020). According to an IMF report citing SBV data, this ratio rose to approximately 2.3% in 2022 and peaked at 5.9% in September 2023 before slightly declining to 5.3% in March 2025.
To address rising NPLs, one of the common measures adopted by credit institutions is to significantly increase provisioning. For example, in 2025, Sacombank proactively set aside an additional VND 11.3 trillion (double the previous year) to strengthen its risk buffer. Similarly, ACB made provisions of approximately VND 3.3 trillion in 2025, twice that of 2024, in compliance with new regulations under Decree 86/2024/NĐ-CP, which mandate provisioning for risk handling rather than deferral. Smaller banks have also shown determination to sacrifice short-term profits to build stronger buffers: Eximbank increased provisions by 57.5% to VND 1.526 trillion (2025) to ensure asset transparency.
However, these measures increase the cost of credit and reduce profitability. Moreover, rising NPLs heighten credit risks for credit institutions, putting pressure on equity and liquidity.
It is evident that NPL resolution has become a significant burden for the banking sector and a race among banks.
The trend of debt management in banking
In reality, at many credit institutions, debt recovery activities are still primarily carried out manually, relying on the personal experience of debt collection officers and fragmented processes across departments. This approach poses considerable legal risks. Communications via phone, email, or in-person meetings, if not properly recorded and controlled, may lead to disputes over commitments and even complaints related to debt recovery procedures.
Not only that, manual debt recovery also entails high operating costs. Fragmented file handling and a lack of automation require organizations to allocate significant human resources to repetitive tasks such as consolidating loan information, updating case statuses, or monitoring customers' repayment commitments. As credit portfolios continue to expand, this model struggles to ensure economic efficiency while limiting the ability to analyze data for optimal debt resolution strategies.
In this context, digital transformation in debt recovery is becoming an inevitable trend for credit institutions. Instead of fragmented and manual handling, banks and fintech companies are increasingly moving toward building centralized management platforms where loan data, customer status, and case processing progress are integrated and monitored in real time.
In addition to digitizing existing debt management processes, some solutions introduce new approaches to debt management. For instance, the credit lifecycle model embedded in the Debt Collection solution used by several joint-stock commercial banks views a loan not merely as appraisal and disbursement, but as a closed-loop process comprising: customer assessment – credit limit approval – disbursement control – collateral management – post-disbursement monitoring – debt recovery management (including on-time collection and monitoring/handling of problematic loans) – loan settlement – credit limit renewal (if any).
Interface of a feature in Debt Collection
“When the entire loan lifecycle is monitored as a closed-loop process, data on customers, collateral, repayment history, and financial fluctuations are continuously updated, creating a foundation for early detection of anomalies," a representative of TNTech, the developer of Debt Collection, said.
At the same time, the close linkage between stages from credit granting to post-disbursement monitoring and debt handling helps standardize processes, reduce reliance on manual operations, enhance risk control, and optimize capital efficiency.
From a governance perspective, the credit lifecycle model also enables management to gain a comprehensive view of loan portfolio quality, thereby supporting timely and effective decisions on credit policy adjustments, allocation of debt recovery resources, or loan restructuring.
In practice, effective debt management not only helps reduce NPL ratios but also improves capital allocation and enhances the quality of credit portfolios. As technology tools enable real-time data monitoring and process automation, debt recovery is gradually shifting from a reactive approach to proactive management. This is also considered a critical step for the banking sector to adapt to a new phase of credit growth, where risk management becomes a key determinant of competitiveness.