Replication code for "Residual Value Risk Management in Automotive Leasing Using Offer Timing Score and Value-at-Risk"

### DescriptionUpdated R script with refined out-of-sample evaluation framework and asscoiated output files. ### AbstractThis study applies the Offer Timing Score (OTS) proposed by Ko (2026) to real-world data from the Korean automotive finance market and combines it with Value-at-Risk (VaR) to examine the dynamic structure of residual value risk. Unlike prior studies that primarily assess risk based on loss levels at a single point in time, this study focuses on the temporal evolution of asset value distributions and the rate of option value decay, thereby providing an analytical framework for quantitatively evaluating the timing of intervention by financial institutions. The empirical results show that used-vehicle residual value distributions are well characterized by a Weibull distribution and that OTS can serve as an operational indicator for identifying periods when residual value risk intensifies. Furthermore, by reinterpreting lease contracts as asset replacement options from a real options perspective, this study extends customer retention decisions into an asset value–based decision-making framework. In addition, the OTS-based outreach strategy provides a framework that integrates retention marketing with residual value risk management by serving as a managerial tool for evaluating the urgency of intervention. However, the analysis is based on data from which observations associated with extreme-loss scenarios were removed. Consequently, while the proposed framework demonstrates strong explanatory power in the automotive finance sector, its applicability to automobile insurance, where tail risk must be explicitly incorporated, remains limited. Future research should therefore develop extended models that incorporate such extreme-loss observations and their associated risk dynamics. ### Contents of the Repository- `OTS.Rproj`: RStudio project file to manage the working directory.- `scripts/`: An R script containing data preprocessing, outlier detection (EIF, LOF, IQR), Weibull MLE parameter estimation, put option pricing, and CCF-based robustness validation.- `data/`: Filtered datasets to replicate the empirical results.- `output/`: High-resolution figures (JPEG) and final summary tables (CSV) generated by the framework. ### Computational Environment & Required R PackagesThe analysis was performed using R. The following packages are required to execute the replication code:- conflicted- data.table- dbscan- dplyr- fitdistrplus- ggplot2- glue- gridExtra- here- isotree- lubridate- openxlsx- patchwork- readr- stringr- tidyr- zoo ### UsageOpen `OTS.Rproj` in RStudio and execute the main R script inside the `scripts/` directory. Ensure that the required data files are placed within the `data/` folder before running the pipeline.