This study pioneers a stochastic process-based approach for analyzing and predicting residential property sales, treating the number of daily residential sales as a point process. Utilizing self-exciting point process and sinusoidal-based regression models, the results demonstrate the temporal Hawkes process’s effectiveness in predicting home sales, particularly addressing seasonality with only one known data cycle. Compared to linear regression, Poisson regression, and ARIMA models, our point process models with self-exciting features exhibit superior predictive accuracy without relying on external data. The study’s findings underscore the potential of point processes in refining temporal predictions in real estate, encouraging a new direction for future analytics.