Abstract
Earnings asymmetric timeliness captures both accrual and operating cash flow (CFO) asymmetric timeliness. Because recognition of operating cash flows does not reflect differential verification thresholds for recognizing unrealized gains versus losses, CFO asymmetry adds noise or bias to tests of conditional conservatism. We show that CFO asymmetry is predictable in the cross-section, and varies systematically with life-cycle characteristics. Removing CFO from earnings and using accruals-based measures of asymmetric timeliness eliminates several biases that prior studies have attributed to other sources. Moreover, accrual asymmetric timeliness varies in the cross-section as theory predicts. Going forward, we recommend researchers use accruals-based asymmetric timeliness measures when testing for conditional conservatism.