Earnings quality is the degree to which reported earnings reflect the underlying economic reality of the business. High quality means the numbers will continue, repeat, and convert to cash. Low quality means the numbers are technically correct but not informative — or worse, are designed to mislead.
Accruals — where most quality issues hide
Net income is computed on accrual accounting: revenue when earned, expenses when incurred, regardless of cash timing. The gap between net income and cash from operations is mostly accruals. High accrual ratios — net income much bigger than CFO — are a classic earnings-quality warning sign.
Accrual ratio = (Net income − CFO) / Average total assets
Studies have shown high-accrual firms underperform low-accrual firms over multi-year holding periods. The intuition: when reported earnings are systematically running ahead of cash, something will give.
Non-GAAP — proceed with care
Companies are allowed (and encouraged by Wall Street) to report 'adjusted' earnings that strip out items management deems non-recurring. Some adjustments are legitimate (a one-time legal settlement). Many are not. Three rules of thumb:
- If the same 'one-time' item appears every year, it isn't one-time
- Stock-based comp is not non-cash — it dilutes shareholders. Never accept its exclusion
- Acquisition-related amortization is real — the company paid for it; pretending it doesn't exist is wrong
Common red flags
- Net income growing while CFO declining or flat
- Receivables growing faster than revenue (booking sales but not collecting)
- Inventory growing faster than COGS (building up unsold product)
- Unusual jumps in 'other income' or 'other comprehensive income'
- Frequent restructuring charges that never seem to settle the business
- Aggressive use of pro-forma or 'adjusted' metrics in the headline number
- Auditor changes, especially after a disagreement
- CFO turnover unrelated to a CEO change
Beneish M-Score and Altman Z-Score
Two academic models that combine ratios into a single score. Beneish M-Score flags potential earnings manipulation (above -1.78 = suspicious). Altman Z-Score predicts bankruptcy (below 1.8 = distress zone). Neither is a verdict — both are useful screens for prioritizing where to dig.
The single best earnings-quality test
Compare cumulative net income over 5-10 years to cumulative free cash flow over the same period. They should be roughly equal — say within 20%. If reported earnings massively exceed FCF over a long stretch, the earnings are not real cash, and at some point will get reset.
When low quality is okay
Early-stage companies legitimately have high accruals (deferred revenue, growing receivables) because growth itself absorbs cash. The same pattern at a mature stable business is alarming. Context matters.
Exercise
An analyst is reviewing a Kenyan-listed corporate's financials and spots these patterns over the past three years: Year 1 — net income KES 1.0bn, operating cash flow KES 1.1bn, DSO 45 days. Year 2 — net income KES 1.3bn (+30%), operating cash flow KES 1.0bn (-9%), DSO 65 days. Year 3 — net income KES 1.7bn (+30%), operating cash flow KES 0.7bn (-30%), DSO 95 days. The CEO publicly highlights 30% earnings growth in each annual report. (1) What are the three earnings-quality red flags here? (2) Which is most concerning and why? (3) What questions would you ask management in their next investor call? (4) Senior analyst's frame: how would you summarise this in a one-sentence research note?