A SIMULATION-BASED APPROACH TO CALIBRATING TARGET FUND LEVELS FOR INVESTOR COMPENSATION SCHEMES

Abstract

Purpose: Investor compensation schemes (ICSs) provide a safety net for retail investors when investment firms fail. However, the optimal level of ex-ante funding remains largely undefined. This paper proposes a quantitative framework to calibrate target fund size, based on operational risk logic.

Methodology: The study applies a Monte Carlo model incorporating firm-level exposure data, supervisory risk classifications (SREP), and stochastic loss distribution (Beta-distributed loss given default). The simulation estimates expected ICS liabilities across 100,000 simulations, capturing low-frequency, high-severity loss patterns.

Results: The model confirms a highly skewed distribution of losses. Based on data from the Croatian market, while most simulations result in no payout, the 99th percentile loss reaches EUR 18.7 million, with a maximum payout exceeding EUR 96 million. The results validate the use of risk-sensitive approaches over fixed-percentage benchmarks and highlight the importance of confidence intervals in reserve planning.

Conclusion: This paper contributes a novel, adaptable methodology to size ICS funds based on probabilistic modelling. It supports regulators in transitioning from static to evidence-based reserve targets. Future refinements could involve dynamic monitoring, enhanced estimation of loss given default, and integration of SREP scores into contribution mechanisms.