The five term sheet clauses that move money

The valuation is the price of the round. The clauses are the price of the exit. Founders negotiate the first for weeks and sign the second in an afternoon, which is exactly backwards, because at the sale of the company the clauses usually move more money than the valuation ever did. I price private companies for a living. I have priced more than a thousand of them, and I sat on the buy side while documents like these got drafted. I have read these clauses from both chairs, and the pattern never breaks: the number gets the press release, the pages get the money.

Here are the five clauses that move real dollars, ranked. For each one: the mechanism, a worked example, the standard versus the red flag, and the counter. Every example is illustrative, round numbers chosen to make the math visible. And to be plain up front: this is education, not legal advice. Norms cited are US venture conventions as of 2026 and vary by sector, stage, and geography. A qualified startup lawyer reads your actual documents. My job is to show you where to point them.

1. Liquidation preference: the multiple and the stack

What it says. On a sale or wind-down, each preferred series gets [N]x its money back before junior stock sees a dollar, in order of seniority.

Why it moves money. Because the headline valuation never appears in the exit waterfall. The preference stack does. Every dollar of preference comes off the top of the sale price before common splits anything, and it compounds across rounds. If you raised on SAFEs, they convert into preferred carrying a 1x preference too, so the stack starts building before your first priced round.

Walk the math with me. Illustrative: a company raises $30 million in total, a $10 million seed and a $20 million Series B, then sells for $60 million. At a clean 1x across the stack, $30 million comes off the top and common splits the rest. A founder holding 30 percent of common takes $9 million. Now give the Series B a 2x preference. The stack becomes $50 million, common splits $10 million, and the same founder takes $3 million. One digit moved $6 million, more than most valuation negotiations move in total.

Standard versus red flag. The standard is 1x. Cooley's Q1 2026 Venture Financing Report puts 1x preferences in 98.2 percent of deals. The red flags are any multiple above 1x, and "senior" stacking where each new round primes the last.

The move. Hold 1x. At later rounds, push for pari passu so the series share pro rata instead of stacking. Before you sign, model the waterfall at a modest, a base, and a strong exit. And if an investor offers a preference multiple as the price of defending your headline valuation, take the honest lower price instead. A high valuation propped up by structure is a down round wearing a costume, and the costume is expensive.

2. Participating preferred: getting paid twice

What it says. After taking its liquidation preference, participating preferred also shares in whatever is left, as if it had converted to common. Non-participating preferred has to choose: the preference or conversion, whichever pays more. A participation cap limits the total take to [N]x, after which the investor must convert to get more.

Why it moves money. This is a different mechanism from the multiple. The multiple thickens the top slice. Participation double-dips on the residual. Non-participating creates a clean either/or; participating takes both.

Walk the math with me. Illustrative: an investor puts in $10 million for 25 percent, and the company exits at $80 million. Non-participating: the investor takes the greater of $10 million or 25 percent of $80 million, so $20 million, and common splits $60 million. Participating, uncapped: $10 million back plus 25 percent of the remaining $70 million, so $27.5 million, and common splits $52.5 million. Identical exit, identical ownership, and $7.5 million moved by one adjective. A founder holding 40 percent of common just paid $3 million for it. Note what a 3x cap would have done here: nothing. The $30 million ceiling does not bind at this exit. Caps only protect on large outcomes; on the modest exits where participation hurts most, capped and uncapped pay the investor the same.

Standard versus red flag. The standard is non-participating, in 96.4 percent of deals per Cooley's Q1 2026 report. Red flags: any participation at seed, and uncapped participation anywhere.

The move. Strike it first. Cap it second, at 2x to 3x total return including the preference, only if striking fails. Never sign it uncapped. Quote the market out loud: an investor demanding participation is asking for a term roughly 96 percent of the market does not get.

3. Anti-dilution: one word, half a million shares

What it says. If a later round prices lower than yours did, the earlier preferred's conversion price adjusts downward, handing it more common shares on conversion.

Why it moves money. The extra shares come at the direct expense of common. This is the largest single-event ownership transfer a signature can trigger, and it is conditional: it costs nothing until a down round, then it costs everything at once. Broad-based weighted average adjusts the price proportionally to how much money was actually raised at the lower price. Full ratchet resets the conversion price all the way to the new price, no matter how small the new round is.

Walk the math with me. Illustrative, and I walk this same example in full in my down round essay: founders hold 2,000,000 common shares, and Series A bought 1,000,000 shares at $10.00, a $10 million round. A $12 million down round prices at $6.00, issuing 2,000,000 new shares. Under full ratchet, the Series A conversion price resets to $6.00 and their 1,000,000 shares become about 1,666,667. That is 666,667 free shares, cutting the founders from 66.7 percent to about 54.5 percent before the new money even lands. Under broad-based weighted average, the price resets only to $8.40, roughly 190,000 extra shares. Same round, same price, and full ratchet hands the investor about 476,000 more shares. At the round's own $6.00 price that gap is roughly $2.9 million transferred by one phrase, and it scales with every dollar the exit clears above it.

Standard versus red flag. The standard is broad-based weighted average, and the market has essentially finished this argument. In Cooley's Q3 2024 Venture Financing Report, 100 percent of reported deals used broad-based weighted average and zero used full ratchet, a first in the report's history. The NVCA model documents treat it as the default form. Red flags: full ratchet, and narrow-based weighted average offered as the "compromise" that is not one.

The move. Insist on broad-based weighted average with the standard carve-outs for the option pool, equity compensation, and conversions. This is the clause you win by knowing it is nearly free to win. The market already abandoned full ratchet; do not let one term sheet resurrect it on your cap table.

4. The option pool inside the pre-money

What it says. Before closing, the company will create or expand an unallocated option pool of [X] percent of the post-money fully diluted capitalization, included in the pre-money valuation.

Why it moves money. This one moves money at signing, not at exit, and it is the most invisible of the five because it does not look like an investor term at all. A pool created inside the pre-money is carved entirely out of existing holders; the incoming investor is diluted by none of it. It is a price cut wearing an HR costume.

Walk the math with me. Illustrative: $20 million pre-money, $5 million round, and the investor requires a 15 percent post-close pool inside the pre-money. That pool is 15 percent of the $25 million post-money, or $3.75 million carved from existing holders before the wire hits. Your effective pre-money is $16.25 million, not $20 million. Now suppose a real 18-month hiring plan only needs 10 percent. The extra 5 points is 5 percent of the company given away for nothing, and at a later $100 million exit that oversize is roughly $5 million of founder-side value before later dilution. Dilution is already the largest quiet cost of venture: Carta's Founder Ownership Report 2026 puts the median founding team at about 56 percent fully diluted after the seed round and about 36 percent after Series A. An oversized pool accelerates that slide for no return.

Standard versus red flag. Standard is 10 to 15 percent at seed, sized role by role against a real hiring plan. The red flag is 20 percent or more dropped into the pre-money "to be safe."

The move. Bring a written 18-month hiring plan. Recompute the effective pre-money with the pool inside it, and counter on that number, not the headline. Push to size the pool post-money, or to share the cost. A point of pool moved is worth more than a point of headline valuation.

5. Cumulative dividends: the meter that runs while you build

What it says. Preferred stock accrues a dividend, typically 6 to 8 percent per year, whether or not the board declares it, payable on liquidation or conversion, in cash or shares.

Why it moves money. A cumulative dividend is a liquidation preference that grows. Every year the exit takes, the top slice of the waterfall gets thicker. It converts time into investor dollars, and slow-but-good outcomes, which describes a lot of real companies, are exactly where it bites. This is the sleeper of the five.

Walk the math with me. Illustrative: an 8 percent cumulative dividend on a $15 million round, exit in year six. That is $15 million times 8 percent times six years, or $7.2 million added to the preference stack, more if it compounds. At any exit where the preference is taken rather than converted, common's pool just shrank by $7.2 million. A founder holding 40 percent of common paid about $2.9 million for a clause that got thirty seconds of attention.

Standard versus red flag. The standard is non-cumulative, payable "when, as and if declared," and in venture practice declared means never. The NVCA model documents treat non-cumulative as the base case. Red flags: cumulative accrual, compounding accrual, or payment-in-kind dividends, especially stacked on top of a preference multiple.

The move. Ask for the standard non-cumulative language and you will usually get it. If forced to accept accrual, keep it non-compounding and have it vanish on conversion or IPO. Then do the multiplication at signing: rate times check size times realistic years to exit. That product is the real price of the clause. Pay it knowingly or not at all.

The two I cut, and why

Pro rata rights move money between investors, not out of your exit check. Redemption rights are leverage, rarely exercised, mostly a forcing device. Both matter, and both are in my Term Sheet Decoder along with board seats, protective provisions, vesting, drag-along, and no-shop.

Price the terms, not the headline

Here is the discipline, the same one I apply from the buy side. Read every clause as money. Build the waterfall before you sign, not after you sell: my cap table essay walks the full exercise, and my free SAFE calculator will rough out your dilution in minutes. Model three exits, not one. A clause that is invisible at the dream number can eat half of a merely good outcome, and merely good outcomes are the ones that actually happen. Then total the money moved across all five clauses and compare it to the valuation gap you are fighting over. The clauses win that comparison more often than founders expect.

The decoder's three questions are the short version: if we sell for less than we hoped, who gets paid and in what order; who can make a decision I disagree with; and what does my ownership actually look like fully diluted. Answer those with real numbers and these five clauses stop being boilerplate.

Then have a qualified startup lawyer review the actual documents. Not because the concepts are hard, but because the drafting is specific, and small wording differences change what a clause does. Your lawyer negotiates the language. Your job is to know which language is worth paying for.

Educational reference only, not legal advice and not an offer of any kind. Market norms cited are US venture conventions as of 2026 and vary by sector, stage, and geography. All worked examples are illustrative. Always retain a qualified startup lawyer to review any term sheet before you sign it.

Frequently asked questions

Which term sheet clause costs founders the most at exit?

Liquidation preference, including participation. It is first in line at every sale: preferences come off the top before common sees a dollar, and a participating preference takes its money back and then shares in the rest too. The standard to hold is 1x non-participating: Cooley's Q1 2026 report shows 1x preferences in 98.2 percent of deals and non-participating structures in 96.4 percent.

Is capped participation an acceptable compromise?

Better than uncapped, worse than none. The cap only binds on larger exits, so on the modest outcomes where participation hurts most, capped and uncapped pay the same. Strike it first, cap it at 2x to 3x as a fallback, never sign it uncapped.

Is full ratchet anti-dilution ever market standard?

No. Broad-based weighted average is the standard, it is the default in the NVCA model documents, and in Cooley's Q3 2024 report 100 percent of reported deals used it with zero full ratchet. A term sheet carrying full ratchet is asking for a term the market has effectively abandoned.

Should I ever accept cumulative dividends?

Treat them as a growing liquidation preference, because that is what they are. An 8 percent cumulative dividend on a $15 million round adds $1.2 million to the preference stack every year, out of common's share of the exit. The standard is non-cumulative, payable only if declared, which in venture practice means never. If forced, keep it non-compounding and have it disappear on conversion.

Why does the option pool count as a money clause when it is not an investor payout?

Because of where it sits. A pool created inside the pre-money is carved entirely out of existing holders: founders pay for all of it, the new investor for none. The gap between a 15 percent pool demanded "to be safe" and the 10 percent a real hiring plan supports is 5 percent of the company, often more than the whole valuation negotiation moved. Size it to a written plan and negotiate it as part of the price.

Related reading

• How to value a pre-revenue startup, from someone who prices them.
• What acquirers really pay for when they buy your company.
• How to handle a down round.
• How to read a cap table before it reads you.

Tomasz Felpel is an investor, founder, and advisor in private markets and healthcare, based in New York. He is a three-time founder of Value Alpha, an AI-powered private-markets valuation platform, Sonnerie VC, an early-stage healthcare venture firm, and Pond. Previously he led corporate development and M&A at Fortune 500 scale, pricing more than 1,000 private companies. Columbia Business School EMBA. Read the full story.