Down rounds and recaps
When a company is worth less than its investors paid, the preference stack decides who gets the money. Learn what a down round triggers, how an overhang works, and how carve-outs and recaps reset the table.
In this lesson you will learn
- How to identify a down round and what it triggers in the cap table
- What a preference overhang is and how it cuts the founders’ payout
- How a management carve-out works and who pays for it
- How a recapitalization, pay-to-play, and a cram-down change ownership and the stack
A high valuation is a price, not a promise. Some startups that investors valued at more than $10B in 2021 and 2022 sold, or agreed to sell, in 2026 for approximately one tenth of that. When a company is worth less than its investors paid, the terms in its charter decide who gets the money. This lesson shows four of these terms.
Down rounds
A down round sells shares at a lower price for each share than the last round. A quick test: if the new pre-money valuation is less than the last post-money valuation, the round is probably a down round. A change in the option pool can move this result. To make sure, compare the prices for each share.
Last round post-money $40M
New round pre-money $30M (less than $40M)
Result a down round
What a down round triggers
A down round dilutes every current holder, because the company sells more shares for the same money. It can also start anti-dilution protection. This term gives earlier investors more shares, so they pay a lower effective price. The founders pay for those shares with more dilution.
Go back to Northwind before its Series B. Suppose that the company does not get to its plan, and the next round prices at $3.00 a share, against $5.00 in the Series A.
Worked example · step 1 of 4
A down round with broad-based weighted average anti-dilution
- 1
Set up the numbers
Before the new round, Northwind has 11M shares: founders 6M, employees 1M, Seed 2M, and Series A 2M. The new round raises $9M at $3.00.
CP1 (Series A price) $5.00 New shares issued (C) $9M ÷ $3.00 = 3,000,000 Shares before the round (A) 11,000,000
The founders owned 54.5% before the round. The type of anti-dilution decides what they keep:
| Anti-dilution | Total shares after the round | Founders | Series A |
|---|---|---|---|
| None | 14,000,000 | 42.86% | 14.29% |
| Broad-based weighted average | 14,187,500 | 42.29% | 15.42% |
| Full ratchet | 15,333,333 | 39.13% | 21.74% |
The Seed paid $1.00, less than the new $3.00 price, so its conversion does not change. In 2025, WilmerHale found weighted average anti-dilution in 100% of first priced rounds and 97% of later rounds. Full ratchet was in 3% of later rounds. The lesson on anti-dilution shows the formula in detail.
The preference overhang
A preference overhang is a preference stack that is large compared with the price that a buyer will pay. The founders then get little from a sale, even with a large percentage.
Go forward again. Northwind raises a $100M Series C at $60.00 a share, an $880M post-money valuation. The Series C gets 1,666,667 shares, a 1x non-participating preference, and seniority over every other series. The stack is now $142M, and the founders own 40.9%. Then the market falls, and the best offer is $150M.
Predict first
Northwind, with its $142M stack, sells for $150M. The founders own 40.9%. What part of the money do they get?
Type a guess before you continue. A guess first, even a wrong one, makes the answer easier to remember.
The founders get 4.4%, or $6.67M. The Series C gets $100M, the Series B $30M, and the Series A $10M. The last $10M goes to 9M shares: the founders, the employees, and the Seed, which converts. Each share gets $1.11. For the founders to get their full 40.9%, the sale price must be $880M or more, the Series C post-money valuation.
Move the price in the widget. Watch how long the founders’ line stays near zero, and where it starts to increase.
Simulator
Northwind after a $100M Series C
The Series C paid $60.00 a share and is senior to every other series. Move the sale price and watch the founders.
- Founders get
- $6.67M
- 4.44% of the money
- Founders own
- 40.91%
- of all shares
- Per common share
- $1.1111
- Preference stack
- $142M
- paid before common
Who gets the $150M
- Founders
- Employees
- Seed
- Series A
- Series B
- Series C
| Holder | Owns | Choice | Gets | Share of the money |
|---|---|---|---|---|
| Founders | 40.91% | Common | $6.67M | 4.44% |
| Employees | 6.82% | Common | $1.11M | 0.74% |
| Seed | 13.64% | Converts | $2.22M | 1.48% |
| Series A | 13.64% | Preference | $10M | 6.67% |
| Series B | 13.64% | Preference | $30M | 20% |
| Series C | 11.36% | Preference | $100M | 66.67% |
| Total | 100% | $150M | 100% |
- Founders
- Employees
- Seed
- Series A
- Series B
- Series C
The Series A, the Series B, and the Series C keep their preferences, because conversion pays less. The Seed converts to common. Each common share gets $1.1111.
The founders own 40.91% but get 4.44% of the money: $6.67M of $150M.
The next change comes at $185M, where the Series A converts.
Change the cap table
A company in this position can become a zombie startup. It operates, but it cannot raise money at its last price, and no buyer pays enough to clear the stack. The team works for years for common stock that has little value.
Management carve-outs
Sometimes the sale price is so low that the preferences get all the money. The managers then have no reason to stay through a hard sale, but the board needs them until the deal closes. The board can set aside part of the price for the management team and key employees. This plan is a management carve-out. The company pays it before the waterfall, usually as a cash bonus.
Take Northwind before its Series C, with its $42M stack, and a $36M sale with a 10% carve-out:
Carve-out 10% × $36M = $3.6M (to the team)
Waterfall divides $36M − $3.6M = $32.4M
Series B (senior) $30.0M (no change)
Seed 2/12 × $2.4M = $0.4M (was $1.0M)
Series A 10/12 × $2.4M = $2.0M (was $5.0M)
Common stock $0 (no change)
The carve-out does not come from the buyer or from the founders’ shares. It comes from the holders who get that money in the waterfall without the plan. Here those are the Seed and the Series A, because the senior Series B still gets its full $30M. Thus the investors must agree to the plan.
Simulator
Who pays for a carve-out
Northwind sells, and the board sets aside part of the price for the team. Start below the $42M preference stack, then move the numbers and see whose money it is.
- Carve-out for the team
- $3.6M
- Paid by preferred holders
- $3.6M
- Paid by common holders
- $0
- Founders' shares get
- $0
- $0 with no carve-out
Who gets the $36M, with the carve-out
- Seed
- Series A
- Series B
- Carve-out
| Holder | No carve-out | With carve-out | Change |
|---|---|---|---|
| Founders | $0 | $0 | $0 |
| Employees | $0 | $0 | $0 |
| Seed | $1M | $400K | −$600K |
| Series A | $5M | $2M | −$3M |
| Series B | $30M | $30M | $0 |
| Carve-out | $0 | $3.6M | +$3.6M |
| Total | $36M | $36M | $0 |
The carve-out removes $3.6M from the top, and the waterfall divides the other $32.4M. The money comes from the Seed ($600K) and the Series A ($3M).
The Series B pays nothing, because it is senior and $32.4M still covers its $30M preference. The common holders got $0 before the carve-out, so all of it is investor money. Thus the investors must agree to it.
Move the carve-out to 20%. Now $36M × 80% = $28.8M is less than the $30M Series B preference, so the Series B pays too. Then move the sale price to more than $42M. The common holders now get money in the waterfall, so they pay part of the carve-out.
How the plan pays the money and how the government taxes it can be different from a payment for shares. Speak to a tax professional before you agree to a carve-out plan.
Work it out
Northwind sells for $30M with a 10% carve-out. How much does the senior Series B get?
The carve-out is 10% × $30M = $3M. The waterfall divides $27M. That is less than the $30M Series B preference, so the Series B gets all $27M. Without the carve-out, the Series B gets all $30M, and the other series get $0. Thus at this price, the senior Series B pays the whole $3M.
Recapitalizations, pay-to-play, and cram-downs
When the stack is much larger than the value of the company, a new investor often does not invest until the cap table changes. Three tools change it:
- A recapitalization, or recap, is a new structure for the cap table. Frequently, the earlier preferred stock converts to common, and the earlier preferences stop. The new investor gets new senior preferred stock, and the team gets new options.
- Pay-to-play makes current investors buy their part of the new round. An investor that does not buy loses rights. Frequently, its preferred stock converts to common, and its preference stops.
- A cramdown is a round with terms so harsh that current holders who do not invest get very large dilution. A round that leaves the earlier holders almost nothing is a washout round.
Here is a recap of Northwind with its $42M stack, at a time when the company is worth only $20M.
Worked example · step 1 of 4
Recapitalize Northwind, step by step
- 1
Convert the earlier preferred
All 6M shares of Seed, Series A, and Series B preferred convert to common. The $42M stack goes to $0.
Earlier holders: 13,000,000 common shares Stack: $42M → $0
The founders’ percentage falls from 46.15% to 23.85%. But the stack above them falls from $42M to $10M. In a $40M sale, the founders got $0 before the recap. After it, the new investor converts, because 33.33% of $40M is more than its $10M preference. Thus the founders get 23.85% × $40M = $9.54M. This example counts the whole new pool as shares that the team holds. Try it.
Simulator
Recapitalize Northwind
Northwind has a $42M preference stack, but the company is worth only $20M. A new investor will put in money if all earlier preferred converts to common. Compare the founders before and after.
- Founders own
- 23.85%
- 46.15% before
- Preference stack
- $10M
- $42M before
- Founders get
- $9.54M
- in a $40M sale. $0 with no recap
- New investor owns
- 33.33%
- $10M ÷ $30M
Ownership after the recap
- Founders
- Employees
- Earlier investors
- New investor
The founders' payout at each sale price
- No recap
- With the recap
The recap cuts the founders from 46.15% to 23.85%. It cuts the stack above them from $42M to $10M. In a $40M sale, the founders get $9.54M with the recap and $0 without it.
Above approximately $62M, the cap table without the recap pays the founders more. But without the new money, the company can close before it gets to that price.
Look at the two lines in the chart. In small sales, the recap pays the founders much more. In large sales, the cap table without the recap pays more, because the founders own a larger part of it. But that comparison has a limit. Without the new $10M, the company can close before it gets to a large sale. Then move the founders’ part of the new pool. New grants to the team are often the reason that founders agree to a recap.
Work it out
After the recap, Northwind sells for $60M. The new investor holds 33.33% with a 1x non-participating $10M preference. The founders own 23.85%. How much do the founders get?
33.33% × $60M = $20M, more than the $10M preference. Thus the new investor converts, and every holder gets its ownership. The founders get 23.85% × $60M = $14.31M. Without the recap, a $60M sale gives them $13.33M.
Quiz · question 1 of 5
Test what you learned
The last round had a $40M post-money valuation. A new investor offers $8M at a $30M pre-money valuation. What is this round most likely?
Key takeaways
- 1.A down round sells shares at a lower price than the last round. It dilutes current holders and can start anti-dilution, which gives earlier investors more shares.
- 2.A preference overhang is a stack that is large compared with the likely sale price. Founders who own 40.9% of Northwind get 4.4% of a $150M sale after a $100M Series C.
- 3.A management carve-out pays the team before the waterfall. The money comes from the holders who get it in the waterfall without the plan, usually the investors, so they must agree.
- 4.A recapitalization cuts the founders’ percentage but can raise their payout, because it removes most of the stack.
Sources
- 1.Q3 2026 PitchBook-NVCA Venture Monitor. PitchBook and NVCA, October 2026.
- 2.Q2 2026 Venture Financing Report. Cooley, August 17, 2026.
- 3.Q4 2025 Venture Financing Report. Cooley, February 9, 2026.
- 4.2026 WilmerHale Venture Capital Report. WilmerHale, April 2026.
- 5.2026 M&A Deal Terms Study. SRS Acquiom, March 2026.
- 6.NVCA Model Certificate of Incorporation (Section 4.4.4). National Venture Capital Association, October 1, 2025.