TY - JOUR
T1 - The use of context-sensitive insurance telematics data in auto insurance rate making
AU - Ma, Yu Luen
AU - Zhu, Xiaoyu
AU - Hu, Xianbiao
AU - Chiu, Yi Chang
N1 - Funding Information:
This research is supported by Federal Highway Administration (FHWA) Broad Agency Announcement “Pay-As-You-Drive-And-You-Save (PAYDAYS) Insurance Actuarial Study” project. Contract award number DTFH61-13-C-00033. We thank Allen Greenberg from FHWA for his insight and expertise that greatly assisted the research.
Publisher Copyright:
© 2018 Elsevier Ltd
PY - 2018/7
Y1 - 2018/7
N2 - Historically, auto insurers use various socio-demographic underwriting factors to differentiate driver risks. With the invention of GPS devices, information such as mileage, traffic condition and driving habits can be incorporated into auto insurance premium calculation. Several major auto insurance companies have offered usage based insurance (UBI) programs where auto insurance premiums are sensitive to actual GPS readings combined with driver's driving behavior. However, given that telematics data are proprietary to the insurance companies that collect such data, the accessibility of UBI is extremely limited. In this research we showcase how second-by-second GPS data can be integrated into existing or new auto insurance pricing structures. We use two types of data: real-time GPS trajectory data collected using a traffic app, as well as survey data. We incorporate vehicle trajectories and accident data to quantify the relationship between driving hazard and accidents with the goal of establishing the linkage between driving risks and accident costs. GPS data considered in this study include not only tradition UBI factors, but also the unique contextual-based risk measurements that compares the driving speed of the vehicle with other drivers on the same road segment. Although smartphone data is used in this study, the methodology developed can be applied to GPS trajectory data collected from other devices such as on-board diagnostics (OBD) or black box solutions. We find that hard brakes, hard starts, peak time travel, speeding as well as driving at a speed significantly different from traffic flow are highly correlated with accident rate. We illustrate the potential underwriting loss an insurer may incur resulting from adverse selection if omitting pertinent risk factors. The results of our study can help insurance companies that are interested in getting into the UBI area set up their auto insurance premium rates.
AB - Historically, auto insurers use various socio-demographic underwriting factors to differentiate driver risks. With the invention of GPS devices, information such as mileage, traffic condition and driving habits can be incorporated into auto insurance premium calculation. Several major auto insurance companies have offered usage based insurance (UBI) programs where auto insurance premiums are sensitive to actual GPS readings combined with driver's driving behavior. However, given that telematics data are proprietary to the insurance companies that collect such data, the accessibility of UBI is extremely limited. In this research we showcase how second-by-second GPS data can be integrated into existing or new auto insurance pricing structures. We use two types of data: real-time GPS trajectory data collected using a traffic app, as well as survey data. We incorporate vehicle trajectories and accident data to quantify the relationship between driving hazard and accidents with the goal of establishing the linkage between driving risks and accident costs. GPS data considered in this study include not only tradition UBI factors, but also the unique contextual-based risk measurements that compares the driving speed of the vehicle with other drivers on the same road segment. Although smartphone data is used in this study, the methodology developed can be applied to GPS trajectory data collected from other devices such as on-board diagnostics (OBD) or black box solutions. We find that hard brakes, hard starts, peak time travel, speeding as well as driving at a speed significantly different from traffic flow are highly correlated with accident rate. We illustrate the potential underwriting loss an insurer may incur resulting from adverse selection if omitting pertinent risk factors. The results of our study can help insurance companies that are interested in getting into the UBI area set up their auto insurance premium rates.
KW - Context-sensitive data
KW - Insurance rate making
KW - Insurance telematics
KW - Smartphone data
KW - Usage based insurance
UR - http://www.scopus.com/inward/record.url?scp=85046363249&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85046363249&partnerID=8YFLogxK
U2 - 10.1016/j.tra.2018.04.013
DO - 10.1016/j.tra.2018.04.013
M3 - Article
AN - SCOPUS:85046363249
SN - 0965-8564
VL - 113
SP - 243
EP - 258
JO - Transportation Research Part A: Policy and Practice
JF - Transportation Research Part A: Policy and Practice
ER -