Common non-GAAP EPS measures are imperfect indicators of true earnings
Companies experience big earnings-per-share (EPS) swings during recessions and recoveries – which cyclically distorts growth rates from trend – so we use normalized EPS estimates to gauge observed price-to-earnings ratios (P/Es) in such periods. Normalized EPS estimates are typically expressed as non-generally accepted accounting principles (non-GAAP) figures, as is typical for all future estimates. But the performance of GAAP EPS during the last recession and full prior cycle helps to assess the quality of a company's non-GAAP EPS history and its truly normal EPS as appropriate to base its current steady-state value. Asset impairments, restructurings and other costs are usually elevated during recessions, which can greatly depress GAAP EPS from normal EPS during recessions, but these full cycle costs should be considered when estimating true mid-cycle normal EPS.
True EPS, as fair to capitalize, sits in-between GAAP and non-GAAP measures
GAAP EPS includes all items except discontinued operations and the cumulative effect of accounting changes. Non-GAAP EPS for non-financial companies typically excludes asset impairments and gains/losses on asset sales and other lumpy items like restructuring charges and litigation. Non-GAAP EPS figures are sometimes misleading and often inconsistent on items included/excluded. In our opinion, non-GAAP EPS helps to evaluate current period performance, but it usually overstates EPS appropriate for capitalization. We find that true EPS, fair to capitalize, usually sits in-between GAAP and non-GAAP EPS.
S&P 500 GAAP/non-GAAP EPS ratio 63% in 2020, better than last two recessions
The GAAP/non-GAAP EPS ratio improved to 66% in the fourth quarter of 2020 from 38% in the first quarter of 2020. The very depressed first-quarter ratio was due to negative GAAP earnings at energy and financials. Financials recovered in the fourth quarter, but energy is still negative and industrials still depressed. Health care had a low fourth-quarter ratio at 26% because of large legal and acquisition expenses at two companies. Since 1990, the average GAAP/non-GAAP EPS ratio is 80% and 86% outside of recessions. In past recessions, the GAAP/non-GAAP ratio fell below 50% and was even negative when certain sectors took huge losses, such as tech in 2001-2002 and financials in 2008. In this recession, energy and financials took the biggest hits on loan losses, asset-sale losses, asset/goodwill impairments and restructuring costs. In our S&P 500 intrinsic valuation model, we make a 7% accounting quality adjustment (12 U.S. dollars) to our 2021 estimated (E) EPS.
GAAP net margin is far more cyclical than non-GAAP net margin
S&P 500 GAAP net margin sank to 8.0% in 2020, and non-GAAP net margin to 10.5%. Peak GAAP and non-GAAP net margins were 11.0% and 12.1% respectively before the pandemic. Quarterly GAAP net margin troughed at 3.7% in the first quarter and non-GAAP net margin troughed at 9.0% in the second quarter. GAAP net margin is far more cyclical than its non-GAAP counterpart. We expect the two measures to both rise and get closer to each other in 2021.
Another way to assess earnings quality is relative to cash flow, but be careful
Free cash flow vs. net income is more an indication of investment activity than earnings quality. But comparing cash flow from operations (CFO) to net income + depreciation & amortization (D&A) over several years cumulatively helps to reveal earnings quality. Consider: (net income + D&A) / (CFO - stock option expense). We subtract stock-option expense from CFO because stock option grants do not reduce aggregate FCF, but reduce expected future FCF per share. We add D&A to net income to reverse the provision for maintenance capital expenditures (capex), to better compare net income to gross cash flow before investment. Adequacy of maintenance capex and the expected returns on investment capex must be considered separately. We make this comparison on a multiyear rolling basis owing to many timing differences between accounting provisions and cash payments, such as with taxes and pensions. Non-GAAP net income plus D&A is usually 0-10% higher than CFO less option expense. So non-GAAP EPS overstates gross cash generation. But GAAP net income + D&A understates it, being 0-10% below CFO less option expense.