Actuarial Science and Quantitative Risk Management Seminars – Spring 2025
For online sessions, please use the following Zoom meeting room:
https://osu.zoom.us/j/91543120239?pwd=cCcAXbYqQI9VQyhkO7AaIAbYphyrEm.1
Meeting ID: 915 4312 0239
Password: 809706
| Date | Speaker | Title | Location/Time | Host |
| 2/14/2025 | Yumin Wang
(University of Manitoba) |
Risk-Sharing Pricing of Variable Annuities within a Principal-Agent Framework | MA 105/
1:30 PM – 2:30 PM |
Zhang |
| 2/21/2025 | Jianxi Su
(Purdue University) |
Some results of the multivariate truncated normal distributions with actuarial applications in view | MA 105/
1:30 PM – 2:30 PM |
Zhang |
| 2/28/2025 | Xing (Clara) Wang
(Illinois State University) |
Asymptotic Normality of the Tilted Density Estimator with Applications to Heavy-Tailed Data | MA 105/
1:30 PM – 2:30 PM |
Zhang |
| 3/21/2025 | Xiaochen Jing
(University of Illinois at Urbana-Champaign) |
Can Market Effectively Price Cyber Risk? Evidence From a Blockchain Insurance Platform | MA 105/
1:30 PM – 2:30 PM |
Ng |
Abstracts
Speaker: Yumin Wang, University of Manitoba
Title: Risk-Sharing Pricing of Variable Annuities within a Principal-Agent Framework
Abstract:We propose a new risk-sharing pricing approach for variable annuities within a principal-agent framework where an insurer (principal) is the contract provider and a policyholder (agent) is the follower having the surrender option. While the risk-neutral pricing approach adopted in the existing literature leads to significantly higher fees and more frequent surrendering than market observations, this new risk-sharing pricing approach reconciles the misalignment between theoretical results and market observations. We also find that a surrender penalty or a lower insurance fee can make the insurer’s expected profit more robust.
Speaker: Jianxi Su, Purdue University
Title: Some results of the multivariate truncated normal distributions with actuarial applications in view
Abstract: The multivariate normal distributions have been widely advocated as an elegant yet flexible model, which uses a simple covariance matrix parameter to capture the intricate dependence involved in high-dimensional data. However, insurance loss random variables are often assumed to be non-negative. Thereby, the multivariate normal distributions must be properly truncated to be adopted in insurance applications. In this presentation, we are going to review some fundamental statistics properties of the multivariate truncated normal distributions, including their independence, non-steepness and maximum likelihood estimation properties. For actuarial applications, we propose an efficient numeric algorithm to compute the tail-based risk functionals for the multivariate truncated normal distributions.
Speaker: Xing (Clara) Wang, Illinois State University
Title: Asymptotic Normality of the Tilted Density Estimator with Applications to Heavy-Tailed Data
Abstract: This project investigates the asymptotic properties of the nonparametric density estimator constructed using the tilted method. Under suitable regularity conditions, we establish the asymptotic normality of the tilted density estimator. Additionally, we conduct simulation studies to evaluate its finite-sample performance, particularly in the context of heavy-tailed data. The results highlight the effectiveness of the tilted estimator in improving density estimation for challenging distributions.
Speaker: Xiaochen Jing, University of Illinois at Urbana-Champaign
Title: Can Market Effectively Price Cyber Risk? Evidence From a Blockchain Insurance Platform
Abstract: As Bitcoin price reaches its new high in 2024, cyber and crypto risk management has become an increasingly concern for both investors and policymakers. This paper investigates the pricing efficiency of these risks via decentralized insurance. Unlike traditional insurance mechanisms, decentralized insurance introduces a novel approach by allowing investors to directly participate in risk assessment and pricing. Using empirical data from the largest decentralized insurer platform, we examine whether investors can effectively price crypto risks by analyzing the relationship between insurance premiums and the likelihood of claims. Our findings reveal that higher premiums are associated with a greater probability of claims, suggesting that investors incorporate private information and risk characteristics into their pricing decisions. We identify the channel primarily stems from investor attention, as historical information of the insured products, Google Search Indices, and crypto market returns are found to significantly influence the identified effect. Our results contributes to the growing literature on decentralized finance (DeFi) by providing evidence from blockchain-based insurance that market-driven mechanisms can effectively price emerging risks in the cryptocurrency ecosystem and address actuarial challenges in underwriting cyber risks.
Speaker’s Bio: Xiaochen Jing is an assistant professor in the Actuarial Science program at the University of Illinois Urbana-Champaign. Prior to joining UIUC, he obtained his PhD in the Actuarial Science, Risk Management & Insurance program at University of Wisconsin-Madison. His research interests include actuarial pricing, insurance product innovation, decentralized insurance, and textual analysis.