Cyber Risk Assessment and Mitigation (CRAM) Framework Using Logit and Probit Models for Cyber Insurance

Malicious external attackers commonly use cyber threats (such as virus attacks, denial-of-service (DoS) attacks, financial fraud, system penetration, and theft of proprietary information), while internal attackers resort to unauthorized access to compromise the confidentiality, integrity, and availability (CIA) of the data of individuals, organizations, and nations. This results in an opportunity cost, a loss of market capitalization, and a loss of brand equity for organizations. Organizations and nations spend a substantial portion of their information technology (IT) budgets on IT security (such as perimeter and core security technologies). Yet, security breaches are common. In this paper, we propose a cyber-risk assessment and mitigation (CRAM) framework to (i) estimate the probability of an attack using generalized linear models (GLM), namely logit and probit, and validate the same using Computer Security Institute–Federal Bureau of Investigation (CSI–FBI) time series data, (ii) predict security technology required to reduce the probability of attack to a given level in the next year, (iii) use gamma and exponential distribution to best approximate the average loss data for each malicious attack, (iv) calculate the expected loss due to cyber-attacks using collective risk modeling, (v) compute the net premium to be charged by cyber insurers to indemnify losses from a cyber-attack, and (vi) propose either cyber insurance or self-insurance, or self-protection, as a strategy for organizations to minimize losses.