You raised your Series A at $2.00 a share, and eighteen months later the market turns. To keep the company alive you raise a Series B at $1.00 a share — half the price. You expect everyone to share the pain equally. Then your lawyer explains the anti-dilution clause you signed without much thought: your Series A investor's conversion price just dropped, their one million preferred shares now convert into well over a million common shares, and every one of those extra shares comes out of your slice and your employees' slice. The investor put in no new money. You paid for their protection with your ownership.
This guide explains how that happened, the two formulas that decide how much it costs you, and the carve-outs and negotiation levers that keep a down round from rewriting your cap table against you.
What an Anti-Dilution Clause Actually Protects
Start with what it does not do. When a company issues new shares, every existing holder's percentage falls. That is ordinary dilution, and no anti-dilution provision prevents it — everyone gets diluted when new shares are created.
Anti-dilution protection is narrower: it protects an investor against a lower price, not a lower percentage. Preferred stock converts into common stock at a conversion price that starts out equal to the price the investor paid. If the company later sells stock below that price — a down round, defined by price per share regardless of how much money is raised — the anti-dilution formula lowers the conversion price. Each preferred share then converts into more than one common share, so the protected investor recovers part of the value the cheaper round took away. The extra shares are created without new cash, so the total share count grows more than it otherwise would, and founders, employees, and other unprotected holders end up with a smaller percentage.
The clause typically lives in the charter alongside the rest of the preferred stock terms — term sheet, stock purchase agreement, and certificate of incorporation — and it triggers on any new issuance below the protected series' conversion price that is not carved out. Nobody gets a guaranteed ownership percentage. The clause only changes who takes the bigger hit when the price falls.
Down rounds stopped being rare years ago. Carta's private-markets data shows at least 15 percent of new priced rounds were down rounds for eleven consecutive quarters, including three quarters at or above 20 percent, before the share finally fell below 14 percent in late 2025 — the lowest reading since the start of 2023. If you raise venture capital, there is a meaningful chance you will one day live under this clause. Price that downside today, while you can still negotiate it.
Full Ratchet: the Punitive Reset
Full ratchet is the most investor-favorable and founder-unfriendly version. If the company issues any new shares at a lower price than a protected investor paid — even a handful of shares — that investor's conversion price resets all the way down to the new lowest price. The size of the down round does not matter. One cheap share reprices the entire earlier investment.
Walk through a simplified example. A founder holds 8,000,000 common shares. A Series A investor puts in $2,000,000 at $2.00 per share and receives 1,000,000 preferred shares. Before any down round, the founder owns 8 million of 9 million shares (88.9 percent) and the Series A investor owns 1 million (11.1 percent).
Now the company raises a Series B down round at $1.00 per share, and a new investor puts in $1,000,000 for 1,000,000 shares. With no anti-dilution at all, the total becomes 10 million shares: founder 80 percent, Series A 10 percent, Series B 10 percent. Everyone shares the dilution.
Under full ratchet, the Series A conversion price resets from $2.00 to $1.00. On conversion, the Series A preferred behaves as if the investor had originally bought at $1.00 — so the original $2,000,000 now converts into 2,000,000 common shares, a full extra million created from nothing. The new total is 11 million shares: the founder holds 8 million (72.7 percent), Series A holds 2 million (18.2 percent), and the Series B investor who actually funded the round holds 1 million (9.1 percent).
Compare the two outcomes. Ordinary dilution took the founder from 88.9 percent to 80 percent. Full ratchet took the founder to 72.7 percent — and handed the earlier investor nearly double the stake of the investor who just wrote a check. That is why full ratchet is considered aggressive enough to poison future financings: a new investor looking at that cap table sees earlier investors collecting windfalls from the new money, and founders too diluted to stay motivated. Full ratchet is rare in mainstream venture deals, and it is worth refusing unless the company genuinely has no other source of capital.
Broad-Based Weighted Average: the Market Standard
Weighted average anti-dilution is the compromise nearly every venture deal uses. Instead of resetting to the new low price, it blends the old price, the size of the existing capitalization, and the size of the cheap round into a new conversion price that lands between the old price and the new low. A small down round produces a small adjustment; a large one produces a bigger adjustment — but never the full reset of a ratchet.
The standard formula, the same one in the NVCA model documents, is:
CP2 = CP1 x (A + B) / (A + C)
Where CP1 is the conversion price before the down round and CP2 is the new conversion price. A is the fully diluted share count before the round. B is the number of shares the new money would have bought at the old price — the dollars raised divided by CP1. C is the number of shares actually issued in the down round.
Because B is always smaller than C in a down round (the same dollars buy fewer shares at the old, higher price), the fraction (A + B) / (A + C) is always below one, so CP2 always lands below CP1. And because A — the entire existing capitalization — dominates both sides of the fraction, the adjustment stays modest unless the down round is large relative to the company.
Run the same example through it. Take A as 10,000,000 fully diluted shares, CP1 of $2.00, and a $1,000,000 down round at $1.00 per share. B equals $1,000,000 divided by $2.00, or 500,000 shares. C equals 1,000,000 shares actually issued. CP2 equals $2.00 times (10,000,000 + 500,000) divided by (10,000,000 + 1,000,000) — $2.00 times 10.5/11 — which comes to about $1.91. The Series A conversion ratio becomes $2.00 divided by $1.91, so the 1,000,000 preferred shares convert into roughly 1,047,600 common shares: about 47,600 extra shares, versus the full extra million full ratchet would have granted. Same down round, same dollars — one-twentieth the damage to the founder.
Broad-Based vs. Narrow-Based: the Denominator Is the Fight
Weighted average comes in two versions, and the difference is what counts in A. Broad-based counts essentially everything: outstanding common, preferred on an as-converted basis, plus options, warrants, and other convertible securities. Narrow-based counts fewer shares — typically only the actually outstanding stock, excluding the option pool and unexercised convertibles.
A smaller A makes the fraction (A + B) / (A + C) smaller, which pushes CP2 lower and hands the investor more shares. Narrow-based is therefore more investor-favorable with every down round, and the gap compounds across rounds. Since broad-based weighted average is the default in both the NVCA and Series Seed documents, the negotiation is rarely about the method — it is about the denominator. When a term sheet says "weighted average," confirm the base is broad and that the outstanding option pool is inside it. Founders who skip that sentence can discover years later that their "standard" protection was the expensive variant.
The Carve-Outs: Issuances That Never Trigger the Clause
Anti-dilution applies only to non-excluded issuances below the conversion price, and the list of exclusions — "Exempted Securities" in the NVCA charter — is where founders get most of their practical relief. The policy logic is straightforward: the clause exists to reprice investors when the company admits, by selling stock cheap, what the shares are really worth. Issuances that are not such an admission should not trigger it. The standard carve-outs cover:
- Employee and contractor equity. Grants under the approved option plan, and the shares issued on exercise, are excluded. Without this, every option grant at a low strike price would reprice the preferred.
- Conversions and exercises. Shares issued when preferred stock, SAFEs, convertible notes, or warrants convert or are exercised do not count — the economics were already priced when those instruments were sold.
- Stock splits and stock dividends. Mechanical adjustments that change share count without changing value are excluded, with the conversion price adjusting proportionally instead.
- Approved strategic issuances. Shares issued to lenders, landlords, equipment lessors, or strategic partners in arm's-length transactions the board — and often a required investor vote — approve.
- The shares of the current round itself. The financing that triggered the adjustment does not recursively trigger further adjustments.
Review this list against your actual plans before signing. A company that expects to issue meaningful equity to a strategic partner, or that runs a large contractor-equity program outside the formal plan, should make sure those issuances are inside the carve-outs rather than discovering the gap during a down round.
What About SAFEs?
Post-money SAFEs carry no anti-dilution provision, and that is by design. A SAFE fixes the investor's ownership at the valuation cap; if the priced round comes in below the cap, the SAFE converts at the lower round price, which is itself the protection. Once converted, the holder receives the same preferred stock — with the same anti-dilution terms — as the priced-round investors.
Some investors now ask for side letters adding weighted average anti-dilution to a SAFE. That layers a price adjustment on top of an instrument that already adjusts by converting at the lower price, and it makes conversion modeling genuinely confusing. It is not standard. Decline it and keep the SAFE clean.
From Formula to Cap Table: Modeling the Adjustment
On the cap table, an anti-dilution adjustment appears as extra shares for the protected series — often labeled anti-dilution adjustment shares — or as a new conversion ratio on the preferred. No cash changes hands. If you keep your books in Beancount, the financing itself is still worth recording precisely, because the proceeds, the issuance costs, and the fully diluted share count are the inputs every down-round model needs:
2026-03-15 * "Series A close - 1,000,000 preferred at $2.00"
Assets:Checking 1,940,000.00 USD
Expenses:Legal-Financing 60,000.00 USD
Equity:Preferred-Series-A -2,000,000.00 USDWith the rounds on the books, build a fully diluted cap table model — a spreadsheet is fine — that lists every series, its conversion price, its protection type, and the option pool, and that can simulate a financing 25 to 50 percent below the last round's price. Run that simulation before you sign the term sheet, not after the down round prices. The model answers the only question that matters: how many extra shares does each protected series collect, and whose percentage pays for them? Founders who model a 40 percent down round in advance negotiate very differently from founders meeting the formula for the first time in a crisis. The Beancount documentation covers structuring equity accounts if you are setting up your ledger for the first time.
Negotiating Levers Worth More Than the Formula
Accepting broad-based weighted average is market, and fighting it spends credibility needed elsewhere. The real negotiation is everything around it:
- Refuse full ratchet outright. It is the single most destructive term in this family, it signals to every future investor that the founders can be pushed, and it makes subsequent rounds structurally harder because new money subsidizes old investors.
- Pin down the denominator. Confirm broad-based in writing, with the option pool and outstanding convertibles inside A. A one-word change to narrow-based quietly reprices every future down round against you.
- Broaden the carve-outs. Make sure your equity plan, SAFE and note conversions, and any planned strategic issuances are Exempted Securities. The default list is good; your situation may need additions.
- Get a majority waiver, not a unanimous one. Anti-dilution lives in the charter, and most down rounds get done because earlier investors waive the adjustment so the financing can close. If the waiver requires every series — or every investor — to agree, one holdout can block the round or extract a payoff. A waiver by majority of the preferred keeps one difficult holder from holding the company hostage.
- Ask for pay-to-play. The mirror image of a waiver: investors who buy their pro rata share of the down round keep their protection, while investors who sit out lose it or convert to common. It concentrates the adjustment on free riders instead of founders, it rewards the investors actually funding the company, and it has become common enough — appearing in a large share of later-stage deals — that asking is unremarkable.
- Cap the scope if you have leverage. Some founders limit anti-dilution to a time window after the round, to rounds below a price floor, or to the first financing only. Each limitation shrinks the scenarios where the clause can bite.
Two final cautions. First, early terms set precedents: grant aggressive protection in the seed round and every later investor will demand at least as much, compounding across the stack. Second, stacked protections interact — overlapping series with different conversion prices and different formulas produce cap tables that surprise even experienced counsel, which is one more reason the down-round model should exist before the term sheet is signed.
Anti-dilution only activates in a down round, which makes it easy to file under boilerplate during good markets. Do the opposite: model a financing at half your current price, read the denominator and the waiver vote, and negotiate the carve-outs while the company is fundable. The clause you barely notice in an up round is the clause that decides how much of the company is still yours after a down one.
