This paper explores the implementation of the #eu's #digitalservicesact (#dsa) with a specific focus on the #riskassessment and #riskmitigation regime outlined in Articles 34-35"
"We show that past #operationalrisk losses are informative of future losses, even after controlling for a wide range of financial characteristics. We propose that the information provided by past losses results from their capturing hard-to-quantify factors such as the quality of operational risk controls, the #riskculture and the #riskappetite of the #bank."
We provide a #cyberrisk definition and classification scheme for #riskmanagement purposes, to be used as a data collection template for #financialinstitutions.
The #eu Digital Services Act (#dsa) establishes a #riskassessment and #riskmitigation regime to address issues like harmful content and structural discrimination, and codes of conduct are meant to guide interpretation of these obligations.
This paper introduces the application of Deep Reinforcement Learning (#drl) in #alm, addressing limitations of traditional methods reliant on human judgement. The findings highlight the potential of DRL to enhance #riskmanagement outcomes for #insurers, #banks, #pensionfunds, and #assetmanagers, providing improved adaptability to changing market conditions.
The article discusses the limitations of current #ai technologies such as #chatgpt, #largelanguagemodels, and #generativeai, and argues that we need to advance #researchanddevelopment beyond these limitations.
"We examine the impact of the U.S. withdrawal from the #parisagreement on the relationship between #climaterisk and #systemicrisk of #us #globalbanking. We find that after 2017, investors stopped pricing climate risk into U.S. systemic risk directly, consistent with domestic investors expecting climate risk #deregulation. However, climate risk still indirectly impacts the U.S. systemic risk through the internal capital markets of U.S. #global #banks operating abroad."
"This paper employs #computational #linguistics to introduce a novel text-based measure of firm-level #cyberrisk exposure based on quarterly earnings conference calls of listed firms. Our quarterly measures are available for more than 13,000 firms from 85 countries over 2002-2021. ... The geography of cyber risk exposure is well approximated by a gravity model extended with cross-border portfolio flows. Back-of-the-envelope calculations suggest that the global #cost of cyber risk is over $200 billion per year."
The authors use mid-quantile regression to deal with ordinal #riskassessments and compare their approach to current alternatives for #cyberrisk ranking and graded responses. They test their #model on both simulated and real data and discuss its applications to #threatlintelligence.
"... we find that the #bayesian approach outperforms the classical [#counterparty #risk #model] in identifying whether a model is correctly specified, which is the principal aim of any backtesting framework."