Share Dilution: How New Stock Issuance Shrinks Your Percentage

You own 1,000 shares of a company with 100 million shares outstanding. That is 0.001% of the business — a rounding error, but it is yours. Then one morning the company announces it is selling 20 million new shares to raise cash, and by the time you read the headline the deal is already priced.

Your 1,000 shares are still sitting in your account. But there are 120 million shares now, so your slice is 0.000833%. You lost 16.7% of your proportional claim on the company’s earnings without selling anything, and nobody asked you first.

That is dilution. Your share count is the numerator and the company controls the denominator.

The number is never the number you expect

Here is the first thing that trips people up. A 20% issuance does not cost you 20% of your stake. It costs you 16.67%.

The reason is which count you measure against. The company issued 20 million shares, which is 20% of the old 100 million. But your ownership is now measured against the new 120 million. Twenty million of those 120 million belong to someone else, and 20 ÷ 120 = 16.67%. That is the fraction of the company that changed hands, and therefore the fraction of your stake that evaporated.

The general shortcut is n / (1 + n), where n is the new shares as a fraction of the old count. Run a few through it and the shape becomes obvious. A 10% issuance costs you 9.09%. A 25% issuance costs you 20%. A 50% issuance costs you 33.33%. Double the share count — n = 1 — and you lose exactly half, permanently.

The gap between the headline percentage and the real loss widens as the raise gets bigger. At 1% issuance the loss is 0.99%, close enough that nobody needs the formula. At 200% issuance, the naive shortcut would tell you that you lost 200% of your stake, which is not a thing that can happen. You lost 66.67%.

Where the new shares come from

Common stock is a residual claim on a company’s assets and earnings, and that claim is strictly proportional to how many shares exist. FINRA’s overview of stocks lays out the structure: you own a piece, and the size of the piece depends on the total outstanding relative to what you hold. Change the total and you have changed every holder’s piece without touching a single brokerage account.

Companies expand the count in five main ways. A secondary or follow-on offering creates new shares and sells them to the public or to institutions in one transaction. An at-the-market program does the same thing quietly, dripping shares into the open market over weeks — easy to miss unless you read the filings. An acquisition paid in stock prints shares and hands them to the target’s owners. Employee stock compensation turns options and restricted units into real shares as they vest, year after year, on a schedule nobody announces. And convertible debt or preferred stock turns into equity at a set ratio when the holder decides.

All five raise the denominator. Only the first two put cash in the company’s hands at the moment the shares appear. That distinction is the whole story of whether dilution costs you money.

Ownership always falls; value does not

Percentage ownership is arithmetic. It falls every time, with no exceptions. Value per share is a separate question, and the answer depends entirely on what the company got in exchange.

Take the same company: 100 million shares at $40.00, a $4 billion market cap. It issues 20 million new shares. Assume for the moment that the market values the raised cash at exactly the dollars received — no premium for a good plan, no discount for a bad one. That is a simplification and a big one, but it isolates the mechanic.

If the shares go out at $40.00, the company banks $800 million. Market cap becomes $4.8 billion across 120 million shares, which is $40.00 per share. Flat. Your ownership dropped 16.67% and your per-share value did not move at all. The slice got thinner and the pie got proportionally bigger.

Now discount the offering. At $28.00 a share — a 30% discount, which happens when a company needs money more than it needs a good price — the raise brings in $560 million. Market cap is $4.56 billion, divided by 120 million shares, so $38.00. That is a 5.00% hit to per-share value.

Push it to the extreme. Shares issued for nothing at all, as employee grants or a stock-funded acquisition where you think the target was worthless: no cash comes in, market cap stays at $4 billion, and 4 billion ÷ 120 million = $33.33. Down 16.67%, which is exactly the ownership loss.

That is the pattern worth holding onto. The cash raised is the cushion, and the issue price sets how thick the cushion is. At full market price the cushion covers the entire blow. At zero there is no cushion and value loss equals ownership loss precisely. Everything else falls between.

The slow version nobody notices

A single 20% offering makes headlines. A company issuing 5% more shares every year for employee compensation makes no headlines at all, and over a decade it does more damage.

Say the company earns $200 million in year zero and grows profits 8% annually — genuinely good performance. Start with 100 million shares, so EPS is $2.00. Now run three issuance rates alongside.

With no issuance at all, ten years of 8% growth takes net income to $431.8 million and EPS to $4.32. That is the benchmark.

At 3% annual issuance, the count reaches 134.39 million by year ten and EPS lands at $3.21. Still growth, noticeably less of it.

At 5% annual issuance, the count compounds to 162.89 million — 100 million × 1.05^10 — and $431.8 million divided across those shares gives $2.65. Net income rose 115.9%. Per-share earnings rose 32.5%. The difference went to shareholders who did not exist ten years ago.

And at 8% annual issuance, EPS is $2.00 in year ten. Exactly what it was in year zero. A decade of compounding profit growth, absorbed completely by the share count, leaving existing holders with nothing per share to show for it. When issuance grows at the same rate as earnings, shareholders capture zero.

The arithmetic behind the general case is simple: EPS growth is roughly earnings growth divided by count growth. At 8% earnings and 5% issuance, 1.08 ÷ 1.05 − 1 = 2.86% a year. Over ten years that compounds to the 32.5% above.

A split is not dilution

This one comes up constantly and the answer is unambiguous: a stock split costs you nothing.

A 2-for-1 split turns your 1,000 shares into 2,000 and turns the company’s 100 million into 200 million. Your ownership: 2,000 ÷ 200,000,000 = 0.001%. Identical to before. The price halves from $40.00 to $20.00, so your position is worth $40,000 either way. Nothing has changed except the size of the unit you count in — like breaking a twenty into two tens.

Compare that to the 20% issuance. Your 1,000 shares stay 1,000, the total goes to 120 million, and your ownership drops to 0.000833%. If those shares were sold at the $40.00 market price your position is still worth $40,000, but you own less of the company. If they were granted for free, your position is worth $33,330.

A split changes the unit of measurement. An issuance changes who owns the company.

Where the arithmetic stops matching the filing

Everything above is clean math on stated assumptions. Real companies are messier in five specific ways, and each one moves the answer.

The proceeds change what the company earns. Every calculation above holds net income on its own path regardless of the raise. In practice a company that banks $800 million and builds something with it should earn more later. If the new capital earns a better return than the existing business, EPS can end up higher than the no-issuance case. The 8% column is only bad news if the money earns nothing.

Buybacks run the other direction. Plenty of companies issue shares to employees and repurchase shares on the market in the same quarter. Net change in the count is what matters. A company issuing 4% and buying back 5% has a shrinking denominator, and the gross issuance number tells you nothing on its own.

Options and convertibles dilute on a delay. A convertible note outstanding today is not in the basic share count. It appears in diluted EPS through an assumed-conversion calculation, and it becomes real shares only if someone converts. That is exactly why companies report basic and diluted EPS as two separate lines. Estimate your ownership off the basic count at a company carrying heavy convertible debt and your estimate is too generous.

Anti-dilution provisions can make it worse than the headline. Some preferred stock carries ratchet clauses that adjust the conversion ratio downward when new shares go out below a threshold price. Trigger one and common holders absorb more dilution than the announced issuance size implies. The terms are in the filing.

Rights offerings hand you a way out. In a rights offering, existing holders get first claim on the new shares, pro rata. Take up your rights in full and your percentage does not move. Ignore them and you dilute exactly as the arithmetic says, less whatever the rights themselves sold for if they were tradeable.

One more, on timing: reported EPS uses a weighted average share count, not the year-end count. Shares issued in November barely register in that fiscal year’s average and then hit the following year in full. The year-by-year figures above apply issuance at year end, which understates first-year dilution for anything raised mid-year.

Finding the actual number

Share counts live in company filings on EDGAR, the SEC’s public filing database. The cover page of a 10-K or 10-Q states shares outstanding as of a recent date, usually within a few weeks of the filing. The income statement gives you the weighted average basic and diluted counts used to compute reported EPS. Registration statements for new offerings are filed separately, which is why an at-the-market program can run for months without appearing in the numbers most people look at.

Reading the trend matters more than reading one number. Pull the weighted average diluted count from five consecutive annual reports and the pattern shows up immediately — one year of issuance tells you almost nothing, five years tells you what management does with the share count as a matter of habit. Then look at the gap between basic and diluted, and at what sits inside that gap. Options struck well above the current price behave very differently from convertible notes trading in the money.

What this does not tell you

Every figure here is arithmetic on stated assumptions. None of it is a forecast.

It cannot tell you what the market will actually pay per share after an issuance. The theoretical values above assume the market prices the raised cash at face value and revalues nothing else. Real announcements move prices before a single share is sold, sometimes far more than the dilution math implies and sometimes less, because traders are pricing what the raise signals about the company’s cash position — not just the change in count.

It cannot tell you whether the proceeds will be spent well. That is the actual question, and it is not a math question.

It ignores share classes entirely. A company with dual-class stock can issue non-voting shares, diluting your economic claim while leaving voting control untouched, or issue voting shares that dilute control specifically. Every calculation here treats all shares as identical. Many real capitalisation tables are not.

It ignores taxes, transaction costs, and your cost basis. Your ownership percentage falling has no tax consequence by itself, because nothing was sold.

And it assumes the count only goes up. Buybacks, reverse splits, and the retirement of treasury shares all push the denominator back down.

FAQ

If a company issues 10% more shares, how much of my stake do I lose?

9.09%. The total goes from 100 to 110 in relative terms, and 100 ÷ 110 = 0.9091, so you keep 90.91% of your original percentage. If you owned 0.001000% before, you own 0.000909% after.

Does dilution reduce the value of my shares?

Only if the company receives less per share than the shares were worth. Using the $40.00 example: issuing 20% more shares at $40.00 leaves theoretical per-share value at $40.00. Issuing them at $28.00 drops it to $38.00, a 5.00% decline. Issuing them for free drops it to $33.33, a 16.67% decline.

How many shares would I need to buy to keep my percentage after a 25% issuance?

25% more than you hold. Own 1,000 shares out of 100 million, and after a 25 million share issuance you would need 1,250 out of 125 million to stay at 0.001000%. The shares you need scale exactly with the issuance percentage — a different relationship from the ownership loss, which is 20% in this case.

What is the difference between basic and diluted shares outstanding?

Basic is the shares that actually exist. Diluted adds the shares that would exist if outstanding options, warrants, and convertible securities were exercised or converted under the accounting rules. A company reporting 100 million basic and 115 million diluted has 15 million shares of future issuance already committed — 13.04% of the diluted total.

If earnings grow 8% a year and share count grows 5% a year, what happens to EPS?

It grows about 2.86% annually: 1.08 ÷ 1.05 − 1 = 0.02857. Over ten years that compounds to a 32.5% rise in EPS, from $2.0000 to $2.6507, while total net income rises 115.9%. Most of the growth went to shareholders who bought in later.

Can a company issue shares without shareholder approval?

Often yes, within limits set by the corporate charter and exchange listing rules. Charters authorise a maximum number of shares and boards can generally issue up to that ceiling without a vote. Listing standards at the major exchanges typically require shareholder approval above a threshold, commonly around 20% of outstanding shares when the sale is below market price. The specific rule is in each exchange’s listed company manual.

This article is general information, not financial advice. See our disclaimer.

Also worth reading: Stock Splits: Why 4x the Shares Is Worth the Same

More in Market Mechanics · all topics · calculators · how this was checked