The 2023 annual reports of 58 listed banks took stock: net interest income grew negatively for the first time since 2017, accelerating the layout of large models

Author | Liu Fen ed Jiang Shizhou

With the disclosure of the 2023 annual reports of 58 listed banks, the progress of digital and intelligent transformation, as well as the latest changes in operating performance, asset quality, and business model, have surfaced.

According to the 2023 Review and Future Outlook of China's Listed Banks (hereinafter referred to as the "Report") released by Ernst & Young, listed banks are also facing challenges in 2023, when the macro economy is under pressure. From the perspective of the revenue structure of the 58 listed banks, net interest income accounted for the highest proportion, reaching 75.40%, a decrease of 1.55% from 2022, mainly due to the narrowing of net interest margin. In 2023, the average net interest margin of listed banks will be 1.69%, which has been declining for four consecutive years. Net interest income declined year-over-year for the first time since 2017; Net fee and commission income decreased by 8.05% compared to 2022, marking the second consecutive year of decline.

It is worth noting that although the listed banks will achieve a total revenue of 5.87 trillion yuan in 2023, a year-on-year decrease of 0.98%, they will still achieve a total net profit of 2.17 trillion yuan, a year-on-year increase of 1.43%, by reducing costs and increasing efficiency, strengthening risk prevention and control, and reducing credit costs. Among them, the operating performance of different types of banks is significantly differentiated, and compared with large banks, small and medium-sized banks are more affected by changes in the macro environment.

1. Listed banks are under pressure, and their performance is significantly differentiated

In EY's view, the slowdown in the net profit growth of listed banks is mainly due to the continued pressure on the operating income of listed banks in the environment of slowing economic growth and intensified market competition. Among them, the net profit of 44 listed banks increased year-on-year, and the net profit of 14 listed banks decreased year-on-year.

In terms of categories, the total net profit of the six large banks was 1.39 trillion yuan, an increase of 2.47%, of which the net profit of the Agricultural Bank of China and the Bank of China in 2023 increased by more than 4%, and the net profit of the remaining four large banks increased by 0.83%-2.34%.

The net profit of the 10 national joint-stock banks totaled 502.087 billion yuan, down 3.18%. Among them, the net profit of 6 banks increased, and the net profit of Zheshang increased the most, at 10.75%. The net profit of 4 banks declined, among them, the net profit of SPD decreased by 28.02% year-on-year, and the net profit of Bohai, Industrial and Everbright decreased by 16.80%, 15.97% and 8.80% year-on-year respectively.

The net profit of 29 urban commercial banks totaled 223.872 billion yuan, an increase of 6.09%. Among them, the net profit of 7 city commercial banks declined, and the net profit of Jiujiang Bank and Jiangxi Bank fell significantly, down 55.65% and 32.92% year-on-year respectively. The net profit of 22 city commercial banks increased, of which the net profit of Harbin Bank, Bank of Hangzhou and Bank of Luzhou increased by 24.54%, 23.15% and 23.02% respectively, and the net profit of Qilu Bank, Bank of Suzhou, Bank of Chengdu, Bank of Qingdao, Bank of Jiangsu and Bank of Ningbo also increased by more than 10%.

The net profit of 13 rural commercial banks totaled 49.645 billion yuan, an increase of 1.47%. Among them, the net profit of 3 rural commercial banks declined, and Jiutai Rural Commercial Bank fell the most, at 89.35%. The net profit of 10 rural commercial banks increased, of which the net profit of Jiangyin Rural Commercial Bank increased by 24.92%, and the net profit of Changshu Rural Commercial Bank, Suzhou Rural Commercial Bank, Ruifeng Rural Commercial Bank and Qingdao Rural Commercial Bank increased by more than 10%.

The report pointed out that in terms of asset scale, the total assets of listed banks at the end of 2023 were 293.91 trillion yuan, a year-on-year increase of 11.14%. Among them, the loan balance (net value after deducting impairment provisions) increased by 10.91% year-on-year. At the same time, in 2023, listed banks will increase their lending to the manufacturing industry, and the total amount of loans to the manufacturing industry will increase by 18.71%; The balance of loans in the real estate sector increased by 3.08%, but the growth rate decreased by 2.05% compared to 2022. The growth rate of loans to the real estate industry in personal loans continued to decline, and the balance of personal housing loans of listed banks in 2023 turned from an increase of 1.49% in 2022 to a decrease of 1.59%.

In terms of liabilities, at the end of 2023, the total liabilities of listed banks totaled 270.48 trillion yuan, a year-on-year increase of 11.37%. In terms of asset quality, at the end of 2023, the total balance of non-performing loans of listed banks was 2.16 trillion yuan, an increase of 149.095 billion yuan from the end of the previous year, and the weighted average non-performing loan ratio decreased to 1.29% from 1.33% at the end of 2022. The weighted average provision coverage ratio was 240.10%, an increase of 2.39% from the end of 2022. The weighted average capital adequacy ratio was 15.80%, unchanged from the previous year.

According to the "Overview of the First Quarter Performance of China's 42 A-share Listed Banks in the First Quarter of 2024" released by EY, the revenue of the 42 A-share listed banks in the first quarter of 2024 decreased by 1.73% year-on-year and net profit decreased by 0.81% year-on-year. Among them, the net profit growth rate of large banks was -2.19%, down 4.61% from the same period last year, and the net profit growth rate of Postal Savings Bank fell the most, down 6.28%. The net profit growth rate of national joint-stock banks was -0.34%, down 2.15% year-on-year, and Ping An Bank had the largest decline, down 11.37%. The growth rate of city commercial banks was 6.83%, down 5.09% year-on-year, and the net profit of Bank of Jiangsu, Bank of Zhengzhou, Bank of Suzhou and Bank of Xiamen decreased by 14.47%, 21.39%, 10.53% and 11.63% respectively. The growth rate of rural commercial banks was 1.56%, a year-on-year decrease of 10.28%, of which the growth rate of Chongqing Rural Commercial Bank decreased the most, down 18.11%.

(Image source: Ernst & Young report)

Second, to speed up the layout of large models, what is the difficulty in landing?

It is worth noting that in 2023, the banking industry will pay more attention to large models, and more and more listed banks will deploy large model technology, make breakthroughs in basic capacity building and business scenario construction, move from "digital transformation" to "digital and intelligent transformation", and innovate existing products, processes, operation methods, and business models, including urban and rural commercial banks.

According to the Ernst & Young report, a total of 25 banks will disclose the amount of fintech/information technology investment in their annual reports in 2023, totaling 197.012 billion yuan. There are 27 listed banks that disclose the number of fintech/information technology personnel in their annual reports, with a total of more than 144,200 related scientific and technological personnel.

In terms of quantity, ICBC has the largest number of fintech personnel, which continues to maintain at 36,000, while CCB, Agricultural Bank of China, Bank of China and China Merchants all have more than 10,000. In terms of growth, the fintech talent of some banks increased by more than 30%, among which Bank of Chongqing grew the fastest, reaching 57.62%; Bank of Communications and Agricultural Bank of China ranked second and third, with 33.30% and 31.22% respectively; Bank of Qingdao was 30.20%. In terms of the proportion of scientific and technological personnel, Industrial Bank was the highest, reaching 13.91%, followed by Bohai Bank with 11.30%, and Bank of Shanghai third with 10.75%.

At the same time, more banks are embracing the implementation of large models with an open attitude. For example, according to the annual report, Fang Heying, chairman of China CITIC Bank, bluntly said that "the emergence of ChatGPT and Sora is reshaping the banking service ecosystem". In 2023, ICBC established the industry's first fully self-developed AI large model technology system with 100 billion parameters, and achieved innovative applications in diversified financial business scenarios. IB released ChatCIB, a 10-billion-level model, focusing on vertical fields such as wealth, investment, and reporting. Ping An Bank uses large-scale model technology to improve the availability of financial services by applying large-scale model technology to retail loan approval, digital and intelligent upgrade of operation management, consumer protection and complaint reduction, and AI auto finance vehicle inspection.

"At present, digital intelligence capabilities have become the core competitiveness of listed banks, and the exploration and practice of large-scale model technology will further innovate existing products and services, business processes, operation methods and even business models, bringing new opportunities for the transformation of the banking industry." Jiang Changzheng, EY North China Financial Services Audit Leader, said that although large models have broad application prospects in the banking field, they also face some challenges in the actual implementation process, such as data privacy and security, model interpretability, prediction accuracy, and ethical and legal protection.

In EY's view, in terms of scenario construction, listed banks need to form scenario maps, use case maps and implementation roadmaps suitable for their own development, so as to avoid seeking perfection from the big. In terms of platform construction, listed banks should rationally apply different tools and platforms based on different scenarios to avoid blind use and waste of technological resources. In terms of model application, factors such as cost, infrastructure, and model performance should be comprehensively considered, and finally the invocation or direct procurement authorization of the model should be clarified. For some small and medium-sized banks, due to their small customer base and low marginal benefits, it may be difficult to afford the investment of technology resources required for digital and intelligent transformation, and industry differentiation may be further intensified. Therefore, under the trend of digital intelligence, small and medium-sized banks can consider forming alliances to achieve a mechanism for joint construction of capabilities and cost sharing.