Basic EPS divides income available to common shareholders by the weighted average number of common shares actually outstanding during the period. Diluted EPS adds in the shares that would be created if all dilutive securities — in-the-money employee stock options, restricted stock units, convertible bonds, convertible preferred stock, and warrants — were converted into common stock. Because the denominator grows, diluted EPS is always less than or equal to basic EPS. U.S. GAAP (FASB ASC 260) requires public companies to report both figures on the face of the income statement, and analysts almost always use diluted EPS for valuation because it reflects the dilution a current shareholder is exposed to.

GAAP EPS follows the rules in FASB ASC 260 and uses net income exactly as reported on the audited income statement. Non-GAAP or adjusted EPS is a company-defined measure that strips out items management considers non-recurring or non-cash — typically stock-based compensation, restructuring charges, acquisition-related costs, amortization of acquired intangibles, and certain tax items. Adjusted EPS is almost always higher than GAAP EPS and is the number most consensus estimates and earnings beats are measured against. SEC Regulation G requires companies to reconcile any non-GAAP measure back to the closest GAAP figure, so always check the reconciliation table in the press release before treating adjusted EPS as comparable to GAAP.

Buybacks reduce the weighted average share count in the EPS denominator, so even with flat net income reported EPS rises. A company earning $1 billion with 1 billion shares posts $1.00 of EPS; if it spends excess cash to retire 50 million shares evenly through the year, the weighted denominator falls to roughly 975 million and EPS rises to about $1.026 — a 2.6% boost with no operational improvement. That mechanic is one reason large-cap S&P 500 EPS growth has historically outpaced underlying earnings growth: roughly a quarter to a third of long-run EPS growth comes from net buybacks. When evaluating buyback-heavy companies, look at total net income growth and free cash flow alongside EPS so you can tell organic earnings power from financial engineering.

The price-to-earnings ratio is just stock price divided by EPS, so EPS is one of the two inputs to P/E. If a stock trades at $40 and reported $2.00 of trailing diluted EPS, the trailing P/E is 20x; if analysts forecast $2.50 for next year, forward P/E is 16x. EPS tells you how much profit you own per share; P/E tells you how many years of those earnings you are paying for at the current price. The two move together mechanically — when EPS rises and the price stays flat, P/E falls — but they answer different questions. Pair them with EPS growth rate (the PEG ratio is one way) to judge whether a high P/E is justified by future earnings power.

Public companies miss the analyst consensus EPS when reported earnings come in below the average of sell-side estimates tracked by services like Refinitiv, FactSet, or Bloomberg. Common causes include weaker-than-expected revenue, gross margin compression from input costs or pricing pressure, higher operating expenses (often headcount or marketing), one-time charges, a higher effective tax rate, FX headwinds, or share counts that didn't shrink as fast as the buyback model assumed. Misses typically trigger an immediate sell-off — often 5% to 15% on the day — because the consensus is priced into the stock; the magnitude depends on whether forward guidance is also cut, since investors weight the future outlook even more heavily than the printed quarter.

Net income is earned over the entire reporting period, so the denominator has to reflect how many shares were entitled to that income on average. If a company starts the year with 100 million shares and issues 20 million more exactly halfway through, only 50% of the year benefited from the extra capital those new shares brought in. The weighted average is 100 + (20 × 6/12) = 110 million, not 120 million. Using the period-end count would understate EPS by attributing a full year of earnings to shares that were only outstanding for part of it. FASB ASC 260 specifies this time-weighting precisely, including the treatment of stock splits and stock dividends, which are applied retroactively to all prior periods presented.

Yes — when a company reports a net loss, EPS is negative and is often labeled "loss per share." Negative EPS is common for early-stage growth companies, biotech firms before commercialization, and cyclical businesses in a downturn. It does not automatically signal a bad investment, but it makes the standard P/E ratio meaningless (negative P/E ratios are not used by professional analysts). Instead, value loss-making companies using revenue multiples (price-to-sales), gross profit multiples, EV-to-EBITDA when EBITDA is positive, or path-to-profitability metrics like the rule of 40 for software. Watch the trend: a narrowing loss per share quarter over quarter often matters more than the absolute level.