EPS & Dividend Cover Explained: The Two Numbers That Reveal Whether Your FTSE 100 Income Is Real or a Maths Trick

EPS is arithmetic, not magic. Take a company's net profit, subtract any preference dividends, and divide by the number of ordinary shares in issue. That is it.

AstraZeneca posted £7.2 billion in net profit in its most recent full year on roughly 1.5 billion shares — giving an EPS near £4.94. That £4.94 is the raw per-share earnings that fund everything else: the dividend, the buybacks, the R&D budget. When the share price is £139, the market is paying 28 times those earnings — the P/E ratio.

Barclays, by contrast, earned the equivalent of about £0.43 per share last year while trading at £4.28. A P/E of 9.9. The market is pricing Barclays at less than 10 times earnings while AstraZeneca commands nearly three times that multiple. Same index, same currency, vastly different expectations.

What actually moves EPS. Three things matter. First, profit growth — or shrinkage. If Barclays doubles earnings without issuing new shares, EPS doubles. Second, share buybacks. If a company buys back 10% of its shares, EPS rises 11% even if profits are flat. Third, dilution. If a company issues new shares to fund an acquisition or pay executive bonuses, EPS gets spread thinner across more shares.

Buybacks are the invisible EPS booster. Shell bought back roughly £10 billion of its own shares in 2025. That mechanically boosts EPS without the business selling a single additional barrel of oil. Whether that is smart capital allocation or short-term cosmetic management is the debate — but the EPS number does not distinguish between the two.

Companies report EPS figures in their annual reports filed with Companies House. For UK-listed firms, the FCA's Listing Rules require disclosure of basic and diluted EPS. Diluted EPS accounts for all potential new shares — options, convertibles, warrants — and is almost always the lower, more honest number. When a company trumpets its basic EPS growth but buries the diluted figure in note 37 of the annual report, that is not an accident.

Shell data page: see live fundamentals including EPS history.

Dividend cover is EPS divided by dividend per share. British American Tobacco earned £3.49 per share and paid out £2.45 in dividends — cover is 1.42x. For every £1 paid to shareholders, the company retained £0.42 for reinvestment, debt reduction, or a rainy-day buffer.

What the cover ratios actually tell you.

Barclays at 5.0x cover has enormous headroom. Earnings could fall 80% before the dividend is theoretically uncovered. That is why banks that slashed dividends in 2020 are now rebuilding payout capacity — the regulator forced them to conserve capital, but the underlying cover was never the problem.

British American Tobacco at 1.4x is inside the danger zone. A 29% earnings decline wipes out the dividend at current payout levels. Tobacco volumes decline structurally — roughly 3-4% per year in developed markets — so the question is not whether cover tightens but when. The company has maintained its progressive dividend policy for over two decades, but a progressive policy on a shrinking earnings base is a countdown, not a commitment.

Glencore at 0.2x is the cautionary tale. The company paid out more in dividends than it earned. That is not sustainable — it is a return of capital funded by debt, asset sales, or reserves. Glencore's payout is not really a dividend in the going-concern sense; it is a commodity windfall distribution that disappears the moment copper and coal prices turn.

The threshold that matters. A cover ratio below 1.5x is a warning flag — not a sell signal, but a prompt to dig deeper. Below 1.0x, the company is borrowing or selling assets to pay you. Above 2.0x, the dividend is well protected and has room to grow even if profits dip.

Income investors who only screen by yield walk straight into the BATS trap: a 5.0% headline yield with cover so thin that a mild downturn turns it into a dividend cut. The Dividend Yield Explained article covers how the yield number itself can deceive. For guidance on reading company financial statements yourself, the FCA's InvestSmart campaign provides free educational resources on understanding corporate disclosures.

UK dividend tax rates jumped materially in April 2026. The basic rate rose from 8.75% to 10.75%. The higher rate moved from 33.75% to 35.75%. Only the additional rate stayed frozen at 39.35%. The dividend allowance stands at just £500 for 2026/27 — a trivial buffer that roughly £5,000 worth of FTSE 100 stocks at a 5% yield will blow through.

That changes the arithmetic for income investors holding dividend-paying stocks outside a tax wrapper. If you hold British American Tobacco in a general investment account and are a higher-rate taxpayer, that 5.0% yield becomes 3.2% after tax. Suddenly the 4.8% available on a gilt — with no credit risk and zero capital gains tax — stops looking dismissible.

The interaction between weak dividend cover and a 35.75% tax rate creates a double penalty. Outside an ISA, you are paying top-rate tax on a dividend the company might not be able to sustain. The cover ratio becomes a tax-risk multiplier — a fragile payout taxed at nearly 36% is worse than a fragile payout taxed at 0% inside a wrapper, and far worse than a well-covered dividend in either location.

Our ISA hub covers the allowances in detail. For a fuller picture of how dividends fit into total return, see What Is a P/E Ratio.

1. EPS without context is noise. A company reporting 15% EPS growth sounds impressive until you check whether it came from buybacks or actual profit growth. Shell's buyback programme adds roughly 3-5% to EPS annually without the business improving. That is not fraudulent — it is disclosed — but treating it as earnings growth is self-deception. Always check the share count alongside the EPS trend. Companies report their weighted average share count in every annual report — compare year-over-year to see whether buybacks are real or a rounding exercise.

2. Dividend cover below 1.5x demands an explanation. Sometimes the explanation is benign. A utility like National Grid runs predictable, regulated cash flows — the board can safely operate at 1.3x cover because earnings are not cyclical. A commodity miner at 1.3x cover is a different story entirely: copper at $9,000 produces generous cover; copper at $7,000 wipes it out. Sector context is everything. The Bank of England's Financial Stability Report makes this same distinction when stress-testing UK bank dividends — sector matters more than the absolute number.

3. The dividend cover trend matters more than the snapshot. A company moving from 2.5x to 1.8x cover over three years while the share price is flat is telling you earnings are declining relative to the dividend commitment. The board may hold the payout for reputational reasons — but the maths is tightening. A company moving from 0.8x to 1.5x cover is recovering; earnings are growing into the dividend rather than the dividend outrunning earnings. Direction is the signal. The absolute number is just the starting point.

For the other side of the income equation — how a company funds its operations and what the balance sheet reveals — read How to Read a Company Balance Sheet. See also our investing hub for more fundamentals guides.

This article is for informational purposes only and does not constitute financial advice. You should seek independent financial advice before making any investment decisions.