Originally published in International Journal of Emerging Markets
Artificial intelligence (AI) technologies such as machine learning and deep learning have gained popularities in emerging markets in the past decade (Bughin et al.,2017). AI finance, a concept that integrates AI technology and finance, plays an increasingly important role in modern emerging economies. Specifically, AI finance greatly enhances the efficiency of financial service sby creating readily accessible digital financial platforms and expanding the breadth and depth of financial services. Consequently, this benefits the whole society through better financial inclusion, a concept that has been promoted by the World Bank as “a key enabler to reducing poverty and boosting prosperity [1]”Relative to conventional finance, AI finance offers a wider range of products, financial services, software and novel communication channels (Gomber et al.,2017). At the core of AI finance are big data, cloud computing, blockchain and other high-tech and data-driven elements, which facilitate digital innovation, improvement of traditional financial systems and alleviation of information asymmetry. AI finance also helps financial institutions to remotely confirm customers’ identities through face recognition technology. Fuster et al. (2019) found that credit scores based on the digital footprint broaden information channels, improve information transparency –hence effectively shorten the process time for loans –and improve the service efficiency of financial institutions. With the rapid development of digital and mobile payment, Peer-to-Peer (P2P) lending, digital assets, crowdfunding and other emerging AI tools, it is reasonable to expect that AI finance will continue to build momentum in big emerging economies such as China.
Non-state-owned enterprises (Non-SOE) firms are essential to China’s market economy. As of 2019, non-SOE firms represent 90% of enterprises, produce 60% of Gross Domestic Product (GDP), account for 70% of technological innovation and employ 80% of urban workforce. Despite the importance of non-SOE firms, China’s conventional financial system, dominated by large state-owned banks, strongly favor SOE firms to non-SOE firms (Brandt and Li, 2003;Brandt et al., 2012). The latter are perceived as “less credible borrowers” due to small scale, lack of collateral and insufficient credit information (Beck and Demirguc-Kunt, 2006; Chen and Guariglia, 2013). Accordingly, the financing constraints faced by non-SOE firms under traditional financial system needs to be reduced to achieve economic efficiency.
Extant literature has documented that AI-based fintech [2] effectively mitigates information asymmetry through better information disclosure (Islam and Mozumdar, 2007; Cheng et al., 2014). Specifically, fintech uses big data to assess a borrower’s credit risk and hence can do a better job in less time (Jagtiani and Lemieux, 2018; Langley and Leyshon,2017). Big data also tailors loans at right amount to right borrowers at right price, making it easier for borrowers to obtain loans which are otherwise hard to get. However, while in general fintech and AI finance could ease financing constraints, little is known about whether this mechanism could work for non-SOE firms. We know many things that work for SOE firms do not necessarily work for non-SOE firms. This research hence investigates the relationship between AI finance and financing constraints of non-SOE listed companies in China from 2011 to 2018. We find that the development of AI finance can alleviate financing constraints of non-SOE firms. Furthermore, we document that the role of AI finance inalleviating financing constraints is stronger among smaller non-SOE firms, more innovative non-SOE firms and non-SOE firms in less developed areas.
This study contributes to the literature in two ways. First, it documents a new economic consequence of AI finance. That is, AI finance effectively eases financing constraints faced by non-SOE firms. Furthermore, this research expands the literature on determinants of firms’ financing constraints. Prior literature documented determinants such as bank market power(Ryan et al., 2014; Abadi et al., 2016), stock market liquidity (Beck and Demirguc-Kunt, 2006), financial liberalization (Laeven, 2003) and firm characters (Aristei and Gallo, 2021; Ahiadorme et al., 2018; Vermoesen et al., 2013; Beck et al., 2006), all from perspectives of traditional financial development.
This research complements existing literature by taking afresh perspective from the angle of AI finance. The remainder of this paper proceeds as follows. We review literature and develop hypotheses in section 2.Section 3 describes data and methodology. Section 4 reports empirical results. We conclude in section 5.2.
Read the full article at International Journal of Emerging Markets.