The differential predictive ability of accrual- versus cash-based accounting information for future cash flows

The Financial Accounting Standards Board (FASB) asserts in Statement of Financial Accounting Concepts No. 1 (1978) that current earnings information is a better indicator of a firm\u27s future cash flows than is its current cash flows. However, the FASB fails to substantiate the assertion with empirical evidence. The objective of this thesis is to empirically examine the differential predictive ability of current earnings versus current cash flows for a firm\u27s future cash fows, and thus to test the FASB\u27s assertion. The empirical results do not support the FASB\u27s (1978) assertion that current earnings is a better predictor of future cash flows than current cash flows. Instead, an intermediate measure of cash flow (between all-accruals earnings and no-accruals operating cash flows), current working capital from operations, has a higher predictive ability than current earnings for future cash flows. These findings are consistent with those of Bowen et al. (1986). Furthermore, current accruals predicts better than noncurrent accruals. Given that (1) working capital contains only current accruals and operating cash flows; (2) working capital predicts better than full-accruals earnings; and (3) current accruals forecast more accurately than noncurrent accruals, it follows that the noncurrent accruals component of earnings is the main factor that renders earnings to possess a lower predictive ability than working capital. Industry analysis also suggests that working capital from operations is an overall better predictor of future cash flows than earnings for both manufacturing and non-manufacturing companies. However, there is ancillary evidence that for firms in the manufacturing industry, earnings predicts better than working capital, whereas working capital predicts better than earnings for non-manufacturing firms