How do I calculate earnings per share (EPS) for a bank?

Divide net income available to common shareholders by diluted shares outstanding. For trailing twelve month (TTM) calculations between annual filings, sum net income from the four most recent quarters and divide by the latest diluted share count.

EPS equals net income available to common shareholders divided by diluted shares outstanding. Net income available to common shareholders is reported net income minus preferred stock dividends. For a bank with $40 million in net income, $2 million in preferred dividends, and 20 million diluted shares outstanding, EPS is ($40M - $2M) / 20M = $1.90.

Basic vs. Diluted Shares

Basic EPS uses the weighted average number of common shares that were actually outstanding during the period. Diluted EPS goes a step further, adding shares that would exist if all dilutive securities were exercised or converted. These dilutive securities include stock options, restricted stock units, warrants, and convertible preferred stock or debt.

Diluted EPS is always equal to or lower than basic EPS because the denominator is larger. It's the standard measure used in valuation ratios like P/E and the Graham Number, and the number most investors focus on.

Adjusting for Preferred Dividends

Many banks, particularly larger ones, have one or more series of preferred stock outstanding. The dividends on those preferred shares reduce the income that belongs to common shareholders, so they must be subtracted from net income before calculating EPS.

Preferred dividend amounts appear in the income statement, sometimes as a separate line item between net income and net income available to common shareholders. They can also be found in the notes to the financial statements. Skipping this adjustment overstates EPS, and the error can be material for banks with large preferred stock issuances.

Building TTM EPS from Quarterly Reports

The 10-K annual filing reports full-year EPS directly. Between annual filings, you need to assemble the trailing twelve months (TTM) figure yourself from quarterly data.

Sum the net income available to common shareholders from the four most recent 10-Q filings, then divide by the diluted share count from the latest quarter. If a bank has filed Q1, Q2, and Q3 of the current year plus the prior year's 10-K, TTM earnings equal current Q1 + Q2 + Q3 + prior Q4 net income.

An alternative approach sums the diluted EPS figures reported in each quarterly filing. This is simpler but can introduce minor rounding differences compared to calculating from aggregate net income and share counts.

Where to Find the Inputs

Net income and EPS (both basic and diluted) are reported on the consolidated statements of income in every 10-Q and 10-K filing. Preferred dividends declared appear either as a line item on the income statement or in the equity section of the notes. The diluted share count is disclosed in the EPS footnote and in the per-share data table on the income statement.

Banks that file call reports with the FDIC report net income on Schedule RI. Call reports don't include per-share data, so SEC filings are the better source for EPS calculations.

What Makes Bank EPS Different

The formula itself is the same for any public company, but several factors cause bank EPS to behave differently than EPS in other industries:

Provision for credit losses can swing EPS dramatically between periods without any change in the bank's revenue or operating expenses

Securities gains and losses create one-time noise in quarterly results that probably won't repeat

Share buybacks reduce the share count and boost EPS independently of earnings growth, and for the largest banks, buyback programs require Federal Reserve approval through the stress testing process

The provision for credit losses is the biggest driver of bank EPS volatility. This is the expense banks record to build reserves against anticipated loan losses. In a strong economy, provisions may be minimal, and EPS runs high. During economic downturns, provisions can surge and cut EPS by half or more. A bank reporting $3.00 EPS in a good year might report $1.00 in a stress year, with the entire difference driven by the provision line.

Securities gains and losses also add noise. Banks hold large investment portfolios, and realized gains or losses on bond sales flow straight to net income. A bank that sells bonds at a gain in one quarter gets an EPS bump that likely disappears the next quarter.

Buyback effects are more gradual but worth watching. When a bank repurchases shares, EPS rises because the same earnings are spread across fewer shares. This can make earnings growth look stronger than it actually is if net income itself is flat.

Common Calculation Mistakes

The most frequent error is comparing raw EPS numbers across banks of different sizes. EPS reflects net income per share, not total profitability. A bank earning $2.00 per share with 50 million shares outstanding ($100M in total net income) is far more profitable than a bank earning $5.00 per share with 10 million shares ($50M total). For size comparisons, use net income, return on equity (ROE), or return on average assets (ROAA) instead.

Treating a single quarter's EPS as representative of the bank's earning power is another common mistake. Quarterly EPS for banks can be lumpy because of provision timing, securities transactions, and seasonal patterns in loan demand. The TTM figure smooths out some of this noise, and comparing the current TTM to prior years gives a clearer picture of the trend.

Watch for one-time items that distort EPS in either direction. Goodwill impairment charges, litigation settlements, gains from branch sales, and tax law changes can all create temporary spikes or drops. Many banks report an adjusted or core EPS figure that strips out these items, and this can be helpful for gauging run-rate earnings. Always check what's being excluded and whether the adjustments seem reasonable.