Page 509 - ISC PROCEEDINGS 21.4
P. 509

investment portfolios creates new revenue streams beyond the traditional banking model.
                  However, for AI to become a genuine driver of economic transformation, data
                  infrastructure and cybersecurity must reach sufficient maturity - a threshold currently
                  achieved only by the enterprise-oriented cluster, while small and medium-sized banks
                  remain substantially behind.
                        3.7. Three channels linking digital infrastructure to economic transformation
                        Synthesising the empirical evidence, this paper proposes three main conceptual
                  channels through which banking digital infrastructure investment is associated with
                  broader economic transformation in Vietnam.
                        Channel 1 - Operational efficiency: Back-office automation (RPA, hyper-automation)
                  and AI-assisted decision-making reduce operating costs and accelerate processing.
                  According to SBV (2024b), VPBank automated 332 internal processes; VietinBank saved
                  over one million working hours per year through workflow automation. These efficiency
                  gains create competitive pressure across the industry to raise productivity, thereby
                  reducing financial intermediation costs and supporting lower-rate credit expansion for the
                  wider economy.
                        Channel 2 - Financial inclusion: eKYC and mobile banking infrastructure lower
                  barriers to financial access. Mobile Money reached 10.2 million accounts by end-2024,
                  with 72% in rural and remote areas (SBV, 2024b) - demonstrating digital infrastructure’s
                  ability to bring financial services to previously excluded populations. Digital financial
                  inclusion promotes saving, investment, and economic participation by households and
                  small businesses - the actors generating the majority of economic activity in Vietnam
                  (World Bank, 2022).
                        Channel 3 - Innovation and ecosystem: Open banking APIs, fintech partnerships, and
                  financial super-platforms create new value beyond traditional banking services.
                  Techcombank has developed a digital ecosystem integrating banking, insurance, and SME
                  services (SBV, 2024b). Experience from DBS Bank with its DBS Marketplace and Tinkoff’s
                  super-app model      shows that open innovation ecosystems amplify the economic
                  transformation impact of digital infrastructure far beyond the banking sector boundary.
                        4. Conclusions and policy recommendations
                        This paper has documented substantive structural transformation in the digital
                  infrastructure, data capabilities, and technology capacity of Vietnam’s banking sector over
                  2018–2024. Total IT expenditure reaching VND 32,437 billion (14.85% of operating costs)
                  in 2024, NAPAS transaction CAGR of 35%, and a 23-percentage-point increase in adult
                  bank account penetration are indicators reflecting a deep and broad digital economic
                  transformation process in the financial sector (SBV, 2024b).
                        However, analysis also reveals that the benefits of digital infrastructure are
                  unevenly distributed. The enterprise-oriented cluster leads both in investment intensity
                  and business performance, while state-owned and small-to-medium banks lag
                  significantly in digital transaction share and CIR - reflecting structural limitations in IT
                  spending composition, disclosure transparency, and investment-outcome alignment (MIC,
                  2018–2024). Several robustness considerations should be noted when interpreting these
                  findings. First, the observed performance differentials across clusters are based on
                  descriptive comparison rather than causal inference; confounding factors - including bank
                  size, ownership structure, loan portfolio composition, and macroeconomic conditions -
                  may partially explain the patterns observed. The cross-cluster performance differentials
                  documented here are consistent with, but do not conclusively demonstrate, a causal


                                                                                                      508
   504   505   506   507   508   509   510   511   512   513   514