Loyalty programs (LPs) have become crucial tools for businesses seeking to cultivate customer relationships while maintaining profitability. Among the important factors influencing the profitability of LPs is the expiration of points, especially for high-value, low-frequency (HVLF) programs with high financial liabilities. While expiration induces complex dynamics, its long-term effect has been underresearched in the literature. Point expiration can incentivize point redemption and purchases but may affect customer engagement. Thus, determining the overall long-term impact of expiration on profitability is challenging. To evaluate the role of expiration in the profitability of an LP, managers carefully monitor breakage, which corresponds to the fraction of accrued reward points that expire before being redeemed. Although breakage provides a succinct indicator of the impact of point expiration in the long term, its estimation presents several challenges associated with the projection of customer behavior over a long period of time. In this research, we propose a new methodology for estimating breakage, which can assist businesses in making informed decisions regarding the design of their LPs. Our proposal considers models of point redemption and purchases and a detailed description of point expiration according to the internal rules of the program. We apply our methodology to a large regional airline and find that (1) our approach provides better recovery of the breakage than other metrics used in the industry, (2) there are significant variations in the estimated breakage among customers and across markets, (3) customer heterogeneity and dynamics are important for calculating long-term metrics, (4) expiration can be quite inelastic to some strategies commonly used in the industry to equalize breakage, and (5) to decrease the breakage rate, it is more effective to promote more frequent redemptions than to favor rewards that require a greater number of points.