Search

Pay-as-you-go contracts reduce minimum purchase requirements, which may increase market participation. This paper randomizes the introduction and price(s) of a novel pay-as-you-go contract to the California auto insurance market, where 17% of drivers are uninsured. The pay-as-you-go contract increases take-up by 10.8 p.p. (89%) and days with coverage by 4.6 days over the 3-month experiment (27%). Demand is relatively inelastic, and pay-as-you-go increases insurance coverage in part by relaxing liquidity requirements: most drivers' purchasing behavior is consistent with a cost of credit in excess of payday lending rates, and 19% of drivers have a purchase rejected for insufficient funds. [ABSTRACT FROM AUTHOR]

This study examines how telematics variables such as annual percentage driven, total miles driven, and driving patterns influence the distributional behaviour of conventional rating factors when incorporated into predictive models for capturing auto insurance risk in rate regulation. To effectively manage the complexity inherent in telematics data, we advocate for the adoption of non-negative sparse principal component analysis (NSPCA) as a structured approach for data dimensionality reduction. By emphasizing sparsity and non-negativity constraints, NSPCA enhances the interpretability and predictive power of models concerning both loss severity and claim counts. This methodological innovation aims to advance statistical analyses within insurance pricing frameworks, ensuring the robustness of predictive models and providing insights crucial for rate regulation strategies specific to the auto insurance sector. Results show that, to enhance auto insurance risk pricing models, it is essential to address data dimension reduction challenges when integrating telematics data variables. Our findings underscore that integrating telematics variables into predictive models maintains the integrity of risk relativity estimates associated with traditional policy variables. [ABSTRACT FROM AUTHOR]

By 2030, the global chatbot market in the fintech industry is expected to reach 6.83 billion USD (Vailshery 2022), and while using chatbots, consumers might be unable to tell whether they are interacting with a human. As chatbots are increasingly being used to provide a more human-like service experience, understanding the factors affecting trust, adoption, and how chatbots drive consumer experience in insurtech is important. This study investigates key factors that encourage (practicity, enjoyment and personalization) and inhibit (privacy concerns, creepiness) trust in and intention to adopt a chatbot and the moderating role of technology anxiety. To do so, 430 respondents applied a simulated auto insurance quote involving a textual-based chatbot and responded to a questionnaire. The results highlight practicity and enjoyment as the key drivers of trust and adoption intention, while the positive impact of personalization is marginal. Creepiness decreases trust in chatbots and their adoption intention, whereas privacy concerns have little effect. Almost half of these relationships are moderated by technological anxiety; for instance, trust does not translate into stronger adoption intention for technology-anxious users, but positively impacts less anxious users. Managerial implications are provided for the successful implementation of chatbots. [ABSTRACT FROM AUTHOR]

We leverage telematics data on driving behavior variables to assess driver risk and predict future insurance claims in a case study utilising a representative telematics sample. In the study, we aim to categorise drivers according to their driving habits and establish premiums that accurately reflect their driving risk. To accomplish our goal, we employ the two-stage Poisson model, the Poisson mixture model, and the Zero-Inflated Poisson model to analyse the telematics data. These models are further enhanced by incorporating regularisation techniques such as lasso, adaptive lasso, elastic net, and adaptive elastic net. Our empirical findings demonstrate that the Poisson mixture model with the adaptive lasso regularisation outperforms other models. Based on predicted claim frequencies and drivers' risk groups, we introduce a novel usage-based experience rating premium pricing method. This method enables more frequent premium updates based on recent driving behaviour, providing instant rewards and incentivising responsible driving practices. Consequently, it helps to alleviate cross-subsidization among risky drivers and improves the accuracy of loss reserving for auto insurance companies. [ABSTRACT FROM AUTHOR]

Distracted driving is responsible for nearly 1 million crashes each year in the United States alone, and a major source of driver distraction is handheld phone use. We conducted a randomized, controlled trial to compare the effectiveness of interventions designed to create sustained reductions in handheld use while driving (NCT04587609). Participants were 1,653 consenting Progressive® Snapshot® usage-based auto insurance customers ages 18 to 77 who averaged at least 2 min/h of handheld use while driving in the month prior to study invitation. They were randomly assigned to one of five arms for a 10-wk intervention period. Arm 1 (control) got education about the risks of handheld phone use, as did the other arms. Arm 2 got a free phone mount to facilitate hands-free use. Arm 3 got the mount plus a commitment exercise and tips for hands-free use. Arm 4 got the mount, commitment, and tips plus weekly goal gamification and social competition. Arm 5 was the same as Arm 4, plus offered behaviorally designed financial incentives. Postintervention, participants were monitored until the end of their insurance rating period, 25 to 65 d more. Outcome differences were measured using fractional logistic regression. Arm 4 participants, who received gamification and competition, reduced their handheld use by 20.5% relative to control (P < 0.001); Arm 5 participants, who additionally received financial incentives, reduced their use by 27.6% (P < 0.001). Both groups sustained these reductions through the end of their insurance rating period. [ABSTRACT FROM AUTHOR]

In previous decades, seminal theories were advanced addressing the problems of asymmetric information. These scholarly works confirmed a significant positive correlation between insurance coverage and risk in insurance markets. We showed that when regression models were used to analyze the asymmetric information problem, the sample selection bias and chosen regression methods could produce distinct results. Moreover, samples at different quantiles may lead to considerably different levels of information asymmetry. The present research suggests that quantile regression models be employed to better reveal the situation of the insurance market and reduce estimation errors. [ABSTRACT FROM AUTHOR]

This research delves into the fusion of spatial clustering and predictive modeling within auto insurance data analytics. The primary focus of this research is on addressing challenges stemming from the dynamic nature of spatial patterns in multiple accident year claim data, by using spatially constrained clustering. The spatially constrained clustering is implemented under hierarchical clustering with a soft contiguity constraint. It is highly desirable for insurance companies and insurance regulators to be able to make meaningful comparisons of loss patterns obtained from multiple reporting years that summarize multiple accident year loss metrics. By integrating spatial clustering techniques, the study not only improves the credibility of predictive models but also introduces a strategic dimension reduction method that concurrently enhances the interpretability of predictive models used. The evolving nature of spatial patterns over time poses a significant barrier to a better understanding of complex insurance systems as these patterns transform due to various factors. While spatial clustering effectively identifies regions with similar loss data characteristics, maintaining up-to-date clusters is an ongoing challenge. This research underscores the importance of studying spatial patterns of auto insurance claim data across major insurance coverage types, including Accident Benefits (AB), Collision (CL), and Third-Party Liability (TPL). The research offers regulators valuable insights into distinct risk profiles associated with different coverage categories and territories. By leveraging spatial loss data from pre-pandemic and pandemic periods, this study also aims to uncover the impact of the COVID-19 pandemic on auto insurance claims of major coverage types. From this perspective, we observe a statistically significant increase in insurance premiums for CL coverage after the pandemic. The proposed unified spatial clustering method incorporates a relabeling strategy to standardize comparisons across different accident years, contributing to a more robust understanding of the pandemic effects on auto insurance claims. This innovative approach has the potential to significantly influence data visualization and pattern recognition, thereby improving the reliability and interpretability of clustering methods. [ABSTRACT FROM AUTHOR]

Abstract.Given that insurance companies often operate across multiple lines of insurance business, where claim frequencies on different lines are often correlated, it often becomes advantageous to employ multivariate count modeling where dependence between lines of business can be included in the modeling. Due to the operation of bonus-malus systems where there is a reward to the insurance policyholder for not claiming, claims data in automobile insurance often exhibits an excess of common zeros, a characteristic known as multivariate zero-inflation. In this article, we propose two approaches to address this feature. The first approach involves utilizing a multivariate zero-inflated model, where we artificially enhance the probability of common zeros based on standard multivariate count distributions. The second approach applies a multivariate zero-modified model, which separately handles the common zeros and the number of claims incurred in each line given that at least one claim occurs. We present several models under these frameworks, along with detailed inference procedures. In the applications section, we conduct a comprehensive comparative analysis of these models using data from an automobile insurance portfolio. [ABSTRACT FROM AUTHOR]

Fraudulent activities especially in auto insurance and credit card transactions impose significant financial losses on businesses and individuals. To overcome this issue, we propose a novel approach for fraud detection, combining convolutional neural networks (CNNs) with support vector machine (SVM), k nearest neighbor (KNN), naive Bayes (NB), and decision tree (DT) algorithms. The core of this methodology lies in utilizing the deep features extracted from the CNNs as inputs to various machine learning models, thus significantly contributing to the enhancement of fraud detection accuracy and efficiency. Our results demonstrate superior performance compared to previous studies, highlighting our model’s potential for widespread adoption in combating fraudulent activities. [ABSTRACT FROM AUTHOR]

I argue that the idea of pursuing a strong personalization of the motor liability insurance conflicts hopelessly with the need to refer to homogeneous groups of policyholders large enough to be correctly priced. I furthermore discuss the fact that, due to purely algebraic reasons, it is a priori impossible to obtain the right individual premium by the usual method of multiplying a common base one by a series of factors (as many, as the personalization variables one has chosen to consider). Finally, I point out that the choice of the said factors, not always made with sufficient respect for the statistical evidences, not only gives rise to serious problems concerning the system's fairness and transparency, but puts theoretically at risk the Companies management balance. [ABSTRACT FROM AUTHOR]

The analysis of driving behavior and vehicle dynamics using telematics technology paves the way for current insurance underwriting and disruptions in automobile insurance. This study aims to understand the behavioral intentions of actual users toward adopting telematics-based insurance products. Relevant constructs from extant literature were combined with the Technology Acceptance Model (TAM) to improve its explanatory power and identify meaningful linkages among the variables. A cross-sectional design with a quantitative survey examined how individuals perceived insurance telematics technology. The results suggest that the modified TAM fits the data well. The constructs explain the high variability in the behavioral intentions of telematics users. The results indicate that two-thirds of the respondents were willing to change their driving style towards safer and more balanced driving if they could reduce their car insurance premiums. Telematics-based motor insurance represents a radical shift in how motor insurance will work in the future. Apparently, this transition would cause dissonance among users. Insurers need to make this transition seamless for users to achieve scalability and profitability. The present study is one of the first in the Indian subcontinent to survey actual users of telematics-based insurance. It provides documented research on the development of behavioral intentions in Indian telematics-based insurance users using TAM or other prevalent technology acceptance models considering the mediating role of perceived usefulness. [ABSTRACT FROM AUTHOR]

The traditional modeling of dependence among branches in non-life insurance has historically rested on the assumption of independence between risks. However, recent studies have brought to light that overlooking the interdependence between risks is not prudent. An illustrative example is found in automobile insurance, where a discernible connection exists between the number of claims in the Automobile Damage branch and those in the Automobile Civil Liability branch. This paper aims to delve into the utilization of copula functions for modeling the interrelationships among claims charges across diverse branches within a non-life insurance portfolio. Moreover, it seeks to assess the consequential impacts of these dependencies on the capital requirements of AXA Insurance in the Moroccan market. Our research concentrates on four pivotal branches within AXA Insurance, namely "work accidents", "automobile civil liability", "disability", and "fire, marine, aviation, and transport". Through a thorough analysis encompassing dependency measurements, independence tests, and various graphical representations, we have substantiated the existence of interdependencies among the claim costs associated with these four risk branches. Importantly, our findings underscore that assuming independence leads to a slight underestimation of the capital requirement needed for effective risk accumulation management by the insurer. [ABSTRACT FROM AUTHOR]

Insurance pricing systems should fulfill the auto-calibration property to ensure that there is no systematic cross-financing between different price cohorts. Often, regression models are not auto-calibrated. We propose to apply isotonic recalibration to a given regression model to restore auto-calibration. Our main result proves that under a low signal-to-noise ratio, this isotonic recalibration step leads to an explainable pricing system because the resulting isotonically recalibrated regression function has a low complexity. [ABSTRACT FROM AUTHOR]

The objective of this paper is to assess the behavior of policyholders and insurance companies in the presence of adverse selection by accounting for costly search and selection efforts, respectively. Insurers seek to stave off high-risk types, while consumers are hypothesized to maximize coverage at a given premium. Reaction functions are derived for the two players giving rise to Nash equilibria in efforts space, which are separating almost certainly regardless of the share of low risks in the market. Empirical evidence from the Australian market for automobile insurance is analyzed using Structural Equation Modeling. Convergence has been achieved with both the developmental and test samples. Both consumer search and insurer selection are found to be positively correlated with risk type, providing a good measure of empirical support for the theoretical model. [ABSTRACT FROM AUTHOR]

This research paper investigated multivariate risk assessment in insurance, focusing on four risks of a singular person and their interdependence. This research examined various risk indicators in non-life insurance which was under-writing for organizations with clients that purchase several non-life insurance policies. The risk indicators are probabilities of frequency claims and correlations of two risk lines. The closed forms of probability mass functions evaluated the probabilities of frequency claims. Three generalized linear models of four-variate Sarmanov distributions were proposed for marginals, incorporating various characteristics of policyholders using explanatory variables. All three models were discrete models that were a combination of Poisson and Gamma distributions. Some properties of four-variate Sarmanov distributions were explicitly shown in closed forms. The dataset spanned a decade and included the exposure of each individual to risk over an extended period. The correlations between the two risk types were evaluated in several statistical ways. The parameters of the three Sarmanov model distributions were estimated using the maximum likelihood method, while the results of the three models were compared with a simpler fourvariate negative binomial generalized linear model. The research findings showed that Model 3 was the most accurate of all three models since the AIC and BIC were the lowest. In terms of the correlation, it was found that the risk of claiming auto insurances was related to claiming home insurances. Model 1 could be used for the risk assessment of an insurance company that had customers who held multiple types of insurances in order to predict the risks that may occur in the future. When the insurance company can forecast the risks that may occur in the future, the company will be able to calculate appropriate insurance premiums. [ABSTRACT FROM AUTHOR]

The pricing of insurance premiums in the non-life insurance sector remains a challenging and complex task. It demands a delicate balance between accurately estimating risk exposure and ensuring profitability for insurers. Generalised Linear Regression Models (GLMs) have become the preferred methods for premium price modelling in the motor insurance sector. While the approach of using a single superior model on which predictions are based ignores the use of robust estimator models. This paper examines various methodologies and sheds light on superiority of twenty-two models compared to each other for pricing automobile insurance. These methods vary from traditional actuarial methods to the modern statistical models such as machine learning algorithms. By using archival technique, their inferiority and superiority are explored, considering the ever-changing landscape of risk factors and market dynamics. Furthermore, it highlights the potential benefits of leveraging these methods and the mechanism for pricing short-term insurance, particularly in motor vehicle insurance. It also develops a framework that can be used in pricing to cater to risk analysis constituents to mitigate uncertainties and provide good services to clients. Our findings show that ANN, NN, XGB, random forest (RF) are superior models, and we conclude that the modern statistical methods can accurately estimate the risk exposure as compared to traditional methods such as the GLMs. [ABSTRACT FROM AUTHOR]

Ratemaking and risk‐based pricing revolve around the value at risk, involving a tendency of higher values implying higher premiums. Traditionally, vehicle values are set at "list price" and depreciate over time on a one‐size‐fits‐all factor model. If values change over time, so should premiums, but a single factor table that only goes downward to flat is out of sorts with today. In today's transactional data streams, vehicle features, and values can be easily observed—of note, a shift in product mix that has more options and higher retained values creates a gap in actual versus traditionally expected insurance values. This gap most acutely appears at the vehicle "as built" level, which is more granular than current segmentation levels of insurance ratemaking vehicle valuation analysis. The result is leaving companies to simply raise base rates universally until they can adapt more flexible price‐to‐value methods at an "as built" price accuracy sophistication level. [ABSTRACT FROM AUTHOR]

A "seat belt defense" allows a defendant in a third‐party auto liability event to present evidence of seat belt nonuse to reduce a plaintiff's recovery. When successfully applied, the defense has a direct effect on reducing liability payouts by insurers for accidents. Our analysis reveals that seat belt defense laws are associated with a 10% reduction in auto claims costs, which equates to a reduction in loss costs of approximately $14 per insured vehicle. We find little evidence that the defense reduces accident frequency and no evidence that the defense affects seat belt usage. The lack of a deterrent effect on driving behavior suggests that the availability of the defense is not salient to drivers. The reduction in loss costs supports the notion that defense lawyers are aware of the defense and make use of it. The resulting effect of lowering insurance payouts may alleviate increasing auto insurance costs. [ABSTRACT FROM AUTHOR]

This study delves into a critical examination of the Size of Loss distribution patterns in the context of auto insurance during pre- and post-pandemics, emphasizing their profound influence on insurance pricing and regulatory frameworks. Through a comprehensive analysis of the historical Size of Loss data, insurers and regulators gain essential insights into the probabilities and magnitudes of insurance claims, informing the determination of precise insurance premiums and the management of case reserving. This approach aids in fostering fair competition, ensuring equitable premium rates, and preventing discriminatory pricing practices, thereby promoting a balanced insurance landscape. The research further investigates the impact of the COVID-19 pandemic on these Size of Loss patterns, given the substantial shifts in driving behaviours and risk landscapes. Also, the research contributes to the literature by addressing the need for more studies focusing on the implications of the COVID-19 pandemic on pre- and post-pandemic auto insurance loss patterns, thus offering a holistic perspective encompassing both insurance pricing and regulatory dimensions. [ABSTRACT FROM AUTHOR]

Motivation: With rising unemployment in South Africa, new forms of digital work transcend legal conceptions and discourses on work. Uber's digital labour platform (DLP) has the potential to reduce unemployment and improve the livelihoods of South African households. Purpose: We examine the nature of employment through digital platforms to assess how such employment conforms to labour law and regulation in South Africa, the responsibility of the Department of Employment and Labour (DEL). Methods and approach: We review the literature on the gig economy, decent work, and labour relations. We examine Uber DLP in South Africa to assess its potential to create decent work given DEL's current labour regulations. We analyse factors that harm the relationship between Uber DLP and DEL. Findings: Despite numerous constitutional provisions governing DLP and DEL, Uber DLP workers in South Africa are considered independent contractors not entitled to employee benefits. Uber DLP employees in South Africa make less than the national minimum wage after the platform deducts its fees. Few professional drivers own the cars they drive; they rent them from owners, known as "partners," and split the earnings, meaning that the drivers earn very little. Policy implications: DEL needs to establish a balance between flexibility and labour standards. DEL should preserve workers' rights and ensure financial stability in the digital age. Labour laws should be updated based on reliable data while considering the effects of digitally enabled employment on society and the economy. Platform workers need the same protection at work as other workers. [ABSTRACT FROM AUTHOR]

In the assessment of car insurance claims, the claim rate for car insurance presents a highly skewed probability distribution, which is typically modeled using Tweedie distribution. The traditional approach to obtaining the Tweedie regression model involves training on a centralized dataset, when the data is provided by multiple parties, training a privacy-preserving Tweedie regression model without exchanging raw data becomes a challenge. To address this issue, this study introduces a novel vertical federated learning-based Tweedie regression algorithm for multi-party auto insurance rate setting in data silos. The algorithm can keep sensitive data locally and uses privacy-preserving techniques to achieve intersection operations between the two parties holding the data. After determining which entities are shared, the participants train the model locally using the shared entity data to obtain the local generalized linear model intermediate parameters. The homomorphic encryption algorithms are introduced to interact with and update the model intermediate parameters to collaboratively complete the joint training of the car insurance rate-setting model. Performance tests on two publicly available datasets show that the proposed federated Tweedie regression algorithm can effectively generate Tweedie regression models that leverage the value of data from both parties without exchanging data. The assessment results of the scheme approach those of the Tweedie regression model learned from centralized data, and outperform the Tweedie regression model learned independently by a single party. [ABSTRACT FROM AUTHOR]

The article focuses on the complexities and current state of underinsured motorist (UIM) coverage in Pennsylvania, particularly in light of the Pennsylvania Motor Vehicle Financial Responsibility Law (MVFRL). It delves into the statutory framework provided by the MVFRL, which governs car insurance disputes, and highlights key policy exclusions and recent case law impacting UIM coverage.

Studying driving records (DR) and assessing their risk relativity is crucial for auto insurance rate regulation. Typically, the evaluation of DR involves estimating risk using empirical loss cost or modeling approaches such as Generalized Linear Models (GLM). This article presents a novel methodology employing Generalized Additive Models (GAM) to estimate the risk relativity of DR. By treating the integer level of DR as a continuous variable, the proposed method offers enhanced flexibility and practicality in evaluating the associated risk. Extending the linear model to GAM is a critical advancement that harnesses advanced statistical methods in actuarial practice, providing a more statistically robust application of the proposed approach. Moreover, the integration of functional patterns with Class or Territory enables the investigation of statistical evidence supporting the existence of associations between risk factors. This approach helps address the issue of potential double penalties in insurance pricing and calls for a statistical solution to overcome this challenge. Our study demonstrates that utilizing the GAM approach yields a more balanced estimation of DR relativity, thereby reducing discrimination among different DR levels. This finding highlights the potential of this statistical method to improve fairness and accuracy in auto insurance rate making and regulation. [ABSTRACT FROM AUTHOR]

The history of Islamic banking in Pakistan parallels the global rise of Islamic banking. An essential part of State Bank of Pakistan (SBP) strategic vision has continued to be advancing Islamic finance as a competitive and robust part of the financial system through supportive legal, regulatory, and supervisory frameworks. Under Islamic banking, there are many products which are Shariah compliant such as current accounts, saving accounts, auto finance, fixed deposits etc. Car Ijarah - a car financing mode which stipulates that the Bank would buy the automobile and rent it to the consumer for an agreed-upon time--usually between one and five years--after the contract is signed, the car will either be given to the customer or sold for a reasonable amount at the end of the Ijarah period. This study defines how Islamic banking product Car Ijarah affects the growth, profitability and promotion of Islamic banking division of one of the reputed private banks of Pakistan Bank Al Habib Limited. Special customers of Islamic Banking who are well aware of Car Ijarah were selected as a sample to fill an online survey questionnaire along with the Head of auto finance Department to collect required data. Also, the data is gathered from previous research journals and annual reports of different Islamic banks. The results from the analysis confirmed that the product of Car Ijarah has proven to be an innovative and effective product and it positively effects growth of Islamic banking. [ABSTRACT FROM AUTHOR]

The article presents empirical evidence for a significant role of stated credit purpose for the level of interest rates paid by the end users of non‐mortgage credit (households). Data for the analysis was drawn from two waves of the Household Finance and Consumption Survey. The results of median regression show that, with the exception of credit for consolidation of other debts, remaining credit purposes were associated with lower interest rates compared to the interest rates of credit for car purchases. Especially low interest rates were evidenced for educational debt. Quantile regression analysis has shown that households which, given their socioeconomic background, experienced the highest interest rate costs, paid additional premiums if their debt was related to consolidation of previous debts or consumption. In the case of households paying relatively low premiums, consolidation of other debts and consumption motives were associated with lower interest rates than loans devoted to car purchase. [ABSTRACT FROM AUTHOR]

This paper explores some of the basic similarities and differences between fault-based and no-fault systems of automobile accident insurance from the perspective of a U.S. state and the laws of Macedonia. The majorities of U.S. states have adopted an at-fault system of compensation and therefore share more similarities with Macedonia than the laws of a dozen U.S. states that have adopted a no-fault system. Whereas Macedonia employs a system of nearly universal health care, such is not the case in the United States. The respective states have adopted either a fault-based system or a no-fault system, but in either case issues of coverage for automobile accidents is more complicated for the lack of a nationalized health care system that at minimum acts as a safety net for treatment of injuries related to automobile accidents. It is posited that complications relating to the lack of universal health care coverage in the United States has led to significantly heightened regulatory intervention by the several states such that the similarities between the free-market approach of the states and of civil law (code) systems are greater than the differences. The paper is not based on empirical data of the legal systems or the medical systems, nor is the undeniable interrelation between a medical system and the corresponding legal system with regard to insurance coverage of primarily health-based issues considered other than to highlight how both may play a part in regulation of the automobile accident insurance industry in light the involvement of public health, safety and welfare. [ABSTRACT FROM AUTHOR]

Anomaly detection plays a crucial role in data security and risk management across various domains, such as financial insurance security, medical image recognition, and Internet of Things (IoT) device management. Researchers rely on machine learning to address potential threats in order to enhance data security. In the financial insurance industry, enterprises tend to leverage the relation mining capabilities of knowledge graph embedding (KGE) for anomaly detection. However, auto insurance fraud labeling strongly relies on manual labeling by experts. The efficiency and cost issues of labeling make auto insurance fraud detection still a small-sample detection challenge. Existing schemes, such as migration learning and data augmentation methods, are susceptible to local characteristics, leading to their poor generalization performance. To improve its generalization, the recently emerging Decentralized Federated Learning (DFL) framework provides new ideas for mining more frauds through the joint cooperation of companies. Based on DFL, we propose a federated framework named DFLR for relation embedding aggregation. This framework trains the private KGE of auto insurance companies on the client locally and dynamically selects servers for relation aggregation with the aim of privacy protection. Finally, we validate the effectiveness of our proposed DFLR on a real auto insurance dataset. And the results show that the cooperative approach provided by DFLR improves the client's ability to detect auto insurance fraud compared to single client training. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new credibility model based on heterogeneous integer-valued self-exciting threshold autoregressive time series, in which the SETINAR(2,1) process is used to fit the claim numbers of policyholders for consecutive periods, and the unobservable heterogeneity is assumed to follow Gamma distribution. We obtain the Bayesian pricing formula for the proposed model and present some numerical examples to illustrate how the claim history affects the future premiums. We also apply the proposed model to a real panel dataset from the Wisconsin Local Government Property Insurance Fund. By comparing with some existing models, we find that our model can exploit the past information more efficiently and has better predictive performance. [ABSTRACT FROM AUTHOR]

Insurance is a pivotal element in modern society, but insurers face a persistent challenge from fraudulent behaviour performed by policyholders. This behaviour could be detrimental to both insurance companies and their honest customers, but the intricate nature of insurance fraud severely complicates its efficient, automated detection. This study surveys fifty recent publications on automobile insurance fraud detection, published between January 2019 and March 2023, and presents both the most commonly used data sets and methods for resampling and detection, as well as interesting, novel approaches. The study adopts the highly-cited Systematic Literature Review (SLR) methodology for software engineering research proposed by Kitchenham and Charters and collected studies from four online databases. The findings indicate limited public availability of automobile insurance fraud data sets. In terms of detection methods, the prevailing approach involves supervised machine learning methods that utilise structured, intrinsic features of claims or policies and that lack consideration of an example-dependent cost of misclassification. However, alternative techniques are also explored, including the use of graph-based methods, unstructured textual data, and cost-sensitive classifiers. The most common resampling approach was found to be oversampling. This SLR has identified commonly used methods in recent automobile insurance fraud detection research, and interesting directions for future research. It adds value over a related review by also including studies published from 2021 onward, and by detailing the used methodology. Limitations of this SLR include its restriction to a small number of considered publication years and limited validation of choices made during the process.

In this paper, we propose a new credibility model based on heterogeneous integer-valued self-exciting threshold autoregressive time series, in which the SETINAR(2, 1) process is used to fit the claim numbers of policyholders for consecutive periods, and the unobservable heterogeneity is assumed to follow Gamma distribution. We obtain the Bayesian pricing formula for the proposed model and present some numerical examples to illustrate how the claim history affects the future premiums. We also apply the proposed model to a real panel dataset from the Wisconsin Local Government Property Insurance Fund. By comparing with some existing models, we find that our model can exploit the past information more efficiently and has better predictive performance.