As discussed in chapter “The Statistical Model”, the Bank of Italy’s ICAS con-sists of two components: (i) the statistical model, covering a broad sample of Italian firms, and (ii) the expert assessment, applied to the subsample of the most significant firms for collateral purposes. The structure of ICAS reflects a common approach in the banking and finance industry. Statistical models, valued for their objectivity, scalability, and reliance on historical data, offer speed and consistency in credit assessment. However, they may fall short of capturing forward-looking variables or incorporating qualitative factors. The expert assessment, described in detail in this chapter, introduces a forward-looking perspective and qualitative judgement, helping to address the limitations of purely statistical approaches, albeit at the cost of increased subjectivity
Esposito, L., Guglielmi, M., Monterisi, F., Narizzano, S., Orlandi, M., The Expert Assessment, in Antonio Scali, A. S. (ed.), Corporate Credit Analysis and AI. Advancing the Rating System at a Central Bank, Springer, Cham, Svizzera 2026: 2026 111- 143 [https://hdl.handle.net/10807/333256]
The Expert Assessment
Esposito, LorenzoPrimo
;
2026
Abstract
As discussed in chapter “The Statistical Model”, the Bank of Italy’s ICAS con-sists of two components: (i) the statistical model, covering a broad sample of Italian firms, and (ii) the expert assessment, applied to the subsample of the most significant firms for collateral purposes. The structure of ICAS reflects a common approach in the banking and finance industry. Statistical models, valued for their objectivity, scalability, and reliance on historical data, offer speed and consistency in credit assessment. However, they may fall short of capturing forward-looking variables or incorporating qualitative factors. The expert assessment, described in detail in this chapter, introduces a forward-looking perspective and qualitative judgement, helping to address the limitations of purely statistical approaches, albeit at the cost of increased subjectivity
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
