While prior research employs linear stock returns as a proxy for economic losses, this study uses piecewise-linear stock returns to separate positive and negative stock returns. It also examines the relationships between non-current asset impairments and changes in sales and cash flows from operations, which can be viewed as short-term indicators of economic impairments. Consistent with prior research, I find a negative relationship between non-current asset impairments and negative stock returns in five years before their recognition. Contrary to prior research, I also find such a relationship in the recognition years. The results indicate that the relationships are stronger one or two years before their recognition than in the recognition years and three years before their recognition. These results suggest that the non-current asset impairment losses reported by Japanese firms are consistent with the Japanese accounting standard, although such losses are not necessarily reported in a timely manner. In addition, I find evidence suggesting that changes in sales and cash flows from operations in recognition years are short-term indicators of non-current asset impairments. Overall, incorporating piecewise-linear variables improves the empirical model of non-current asset impairment timeliness.