Cap Table Modeling with AI: 3 Rounds of Real Dilution
What is a cap table?
A cap table (capitalization table) is a ledger that records who owns shares in a company, how many shares each party holds, and what percentage of the total that represents. It is updated with every funding round, option grant, or equity transfer.
TL;DR
- -After 3 rounds of 20% dilution each, a founder retains 51.2% — not 40%, because dilution compounds
- -Option pool is carved out before the investor enters, so founders bear all pool dilution
- -Pre-Seed example: $2M pre-money + $500K investment + 10% option pool leaves founders with ~40.5% each
- -SAFEs and convertible notes delay dilution until the next priced round — then hit all at once
- -AI can model full 3-round cap table scenarios in minutes using structured prompts with dilution formulas
After three funding rounds, a typical founder owns 20–30% of the company instead of the original 50%. Option pools, convertible notes, SAFEs — every instrument adds dilution that’s hard to calculate without a model. Most founders see the full picture for the first time only after they’ve signed a term sheet.
Here’s how to model a cap table three rounds out using AI, understand your real dilution before negotiations begin, and make decisions based on actual numbers.
What is a cap table and why model dilution
A cap table (capitalization table) is a ledger of company ownership: who holds how many shares and what percentage that represents. Every funding round triggers a recalculation.
Dilution kicks in when the company issues new shares. An investor puts in money, receives shares, and every existing shareholder’s percentage shrinks. The founder’s share count stays the same — but its percentage of the total drops.
Key terms:
| Term | Definition |
|---|---|
| Pre-money valuation | Company valuation before the investor’s money |
| Post-money valuation | Pre-money + investment amount |
| Price per share | Post-money valuation / total shares after the round |
| Dilution % | Investment / Post-money valuation |
| Option pool | Shares reserved for employees (typically 10–20%) |
| SAFE / Convertible note | Instruments that convert to equity at the next round |
The dilution formula
The core formula for equity round dilution:
Dilution % = Investment / Post-money Valuation
Post-money Valuation = Pre-money Valuation + Investment
New founder share = Old share × (1 - Dilution %)
Example. Pre-money valuation $4M, investment $1M:
Post-money = $4M + $1M = $5M
Dilution = $1M / $5M = 20%
Founder's share: 100% × (1 - 0.20) = 80%
Formula for multiple consecutive rounds (cumulative dilution):
Final share = Starting share × (1 - d1) × (1 - d2) × (1 - d3)
Where d1, d2, d3 are each round’s dilution expressed as a decimal.
With 20% dilution in each of three rounds:
Final share = 100% × 0.80 × 0.80 × 0.80 = 51.2%
Three rounds of 20% don’t produce 60% dilution. They produce 48.8%. Each subsequent round dilutes an already-reduced stake.
Modeling three rounds: Pre-Seed, Seed, Series A
A typical scenario for a two-founder startup (50/50) from Pre-Seed through Series A.
Round 1: Pre-Seed
Starting parameters:
- 2 founders, 5,000,000 shares each (10,000,000 total)
- Pre-money valuation: $2M
- Investment: $500K
- Option pool before the round: 10%
The option pool gets created first. It’s carved out of the founders’ stakes before the investor enters:
Option pool shares = 10,000,000 × 10% / (1 - 10%) = 1,111,111
Shares before investment = 10,000,000 + 1,111,111 = 11,111,111
Now the investor:
Post-money = $2M + $500K = $2.5M
Dilution = $500K / $2.5M = 20%
Investor shares = 11,111,111 × 20% / (1 - 20%) = 2,777,778
Total shares = 11,111,111 + 2,777,778 = 13,888,889
Cap table after Pre-Seed:
| Shareholder | Shares | Ownership (%) |
|---|---|---|
| Founder 1 | 5,000,000 | 36.0% |
| Founder 2 | 5,000,000 | 36.0% |
| Option pool | 1,111,111 | 8.0% |
| Pre-Seed investor | 2,777,778 | 20.0% |
| Total | 13,888,889 | 100% |
Each founder dropped from 50% to 36% — a 14 percentage point loss. About 5pp went to the option pool (50% to 45% before the investor entered), and another 9pp to investor dilution (45% down to 36%).
Round 2: Seed
Parameters:
- Pre-money valuation: $8M
- Investment: $2M
- Option pool expanded to 15% of post-money
Post-money = $10M. The current pool sits at 8.0% of total shares. To bring the pool to 15% of post-money, the company needs to issue additional shares into the pool.
Post-money = $8M + $2M = $10M
Investment dilution = $2M / $10M = 20%
Total dilution (including pool expansion) ≈ 28.6%
Cap table after Seed:
| Shareholder | Ownership (%) |
|---|---|
| Founder 1 | 25.7% |
| Founder 2 | 25.7% |
| Option pool | 15.0% |
| Pre-Seed investor | 14.3% |
| Seed investor | 19.3% |
| Total | 100% |
Each founder lost another ~10 percentage points. Cumulative dilution from the start: from 50% down to 25.7%.
Round 3: Series A
Parameters:
- Pre-money valuation: $30M
- Investment: $8M
- Option pool: hold at 15% (no increase)
Post-money = $30M + $8M = $38M
Dilution = $8M / $38M = 21.1%
Cap table after Series A:
| Shareholder | Ownership (%) |
|---|---|
| Founder 1 | 20.3% |
| Founder 2 | 20.3% |
| Option pool | 11.8% |
| Pre-Seed investor | 11.3% |
| Seed investor | 15.2% |
| Series A investor | 21.1% |
| Total | 100% |
Three rounds later, each founder owns ~20% instead of the original 50%. Total dilution: 59.4%.
AI prompts for cap table modeling
These prompts work with Claude, GPT-4, or any reasoning-capable model. For tips on structuring multi-step prompts like these, see our prompt engineering guide. If you’re running a multi-provider LLM setup, you can cross-verify results between models for higher confidence.
Prompt 1: basic three-round model
Build a cap table for a startup with two founders (50/50).
Starting share count: 10,000,000 (5M per founder).
Rounds:
1. Pre-Seed: pre-money $2M, investment $500K, option pool 10%
2. Seed: pre-money $8M, investment $2M, expand option pool to 15%
3. Series A: pre-money $30M, investment $8M, hold option pool at 15%
For each round, show:
- Price per share
- Number of new shares issued
- Cap table with all shareholder ownership (%)
- Each shareholder's dilution for the round
At the end — a summary table of founder ownership changes across rounds.
Prompt 2: scenario analysis with different valuations
Use the cap table from the previous calculation.
Model 3 scenarios for Series A:
Scenario A (conservative): pre-money $20M, investment $5M
Scenario B (base): pre-money $30M, investment $8M
Scenario C (optimistic): pre-money $50M, investment $10M
For each scenario, show:
- Final ownership for each founder
- Price per share
- Difference in dilution between scenarios
Which scenario is optimal for founders in terms of
capital raised relative to dilution?
Prompt 3: SAFE conversion modeling
Before the Seed round, the company raised $300K through 2 SAFEs:
- SAFE 1: $150K, valuation cap $3M, no discount
- SAFE 2: $150K, valuation cap $5M, 20% discount
Seed round: pre-money $8M, investment $2M.
Calculate:
1. The conversion price for each SAFE
2. How many shares each SAFE holder receives
3. The cap table after Seed including both SAFE conversions
4. How much additional dilution the SAFEs caused founders
compared to a round without SAFEs
Prompt 4: option pool sensitivity analysis
Cap table after Seed (from previous calculation).
Series A: pre-money $30M, investment $8M.
Model 4 option pool scenarios before Series A:
- 10% (current, no expansion)
- 15% (standard)
- 20% (aggressive hiring)
- 25% (very aggressive)
Table: founder ownership, Series A investor ownership,
and difference in founder dilution for each scenario.
Hidden dilution factors
Simple formulas miss several mechanisms that amplify dilution.
Option pool shuffle. Investors require the option pool to be created or expanded before the round closes. The pool is technically carved out of the pre-money valuation, so it dilutes only existing shareholders. With a pre-money of $8M and a 15% option pool requirement, founders’ effective pre-money lands closer to $6.8M.
Liquidation preferences. These don’t directly affect the cap table but they determine payout order at exit. A 1x non-participating preference means the investor chooses: either get their money back, or convert to common stock and take their pro-rata share — whichever pays more. A 1x participating preference is worse for founders: the investor gets their money back first and then takes their pro-rata cut of the remainder. At a modest exit, founders can receive far less than their ownership percentage would suggest.
Anti-dilution protection. If the next round closes at a lower valuation (a down round), investors with anti-dilution rights get additional shares at no cost. Weighted average anti-dilution is gentler; full ratchet can double founder dilution.
Convertible instruments. SAFEs and convertible notes don’t show up in the cap table until they convert. At the Seed round, they convert and pile dilution on top of the direct investment. $300K in SAFEs with a low valuation cap can cost founders an extra 5–8% in dilution.
Prompt for full analysis:
Analyze the cap table after Series A and identify hidden
dilution factors:
1. Calculate the effective pre-money for founders accounting for
option pool shuffle in each round
2. At an exit of $50M with 1x non-participating liquidation preference
for all investors — how much does each shareholder receive?
3. Model a down round: Series B at a 30% lower valuation than Series A,
with weighted average anti-dilution. How does the cap table change?
Validating the model: how to check AI calculations
AI models occasionally make arithmetic errors in multi-step calculations — a known issue with LLM output quality in general. Three checks catch most of them:
When building SOPs for recurring financial calculations, embed these validation steps directly into the template.
Check 1: ownership percentages sum to 100%. After each round, all shareholders’ percentages must add up to exactly 100%. If you get 99.7% or 100.3%, there’s a rounding error or a missing shareholder.
Check 2: investor’s share = Investment / Post-money. The most common error is a wrong post-money calculation when an option pool is involved. Post-money includes all shares, including the pool.
Check 3: price per share is identical for all round participants. The investor and option pool in the same round receive shares at the same price (exception: SAFEs with a discount).
Self-verification prompt:
Check the cap table calculation for errors:
1. Do ownership percentages after each round sum to 100%?
2. Does each investor's share equal their investment / post-money?
3. Is price per share correct and consistent within each round?
4. Is the option pool properly accounted for in pre-money (not post-money)?
5. Does share count × price per share = post-money valuation?
If discrepancies exist — identify the round and source of error.
Cap table tools
AI works well for quick modeling and scenario analysis. For maintaining an actual cap table, you’ll need a dedicated tool:
| Tool | Best for | Cost |
|---|---|---|
| Carta | Full cap table management, 409A valuations | From $3K/year |
| Pulley | Cap table management for early stages | From $50/month |
| Captable.io | Free modeling | Free |
| LTSE Equity | Cap table + compliance | From $200/month |
| Google Sheets + AI | Quick scenario modeling | Free |
Best combo: AI for modeling during negotiations, a dedicated tool for the legally binding cap table after the round closes.
Where to start
-
Lock down the current state. Share count, all shareholders, all convertible instruments (SAFEs, notes). This is your input data for the model.
-
Define the next round’s parameters. Target valuation (a range), amount to raise, required option pool. If you don’t know the parameters yet, use market benchmarks for your stage and geography.
-
Run the base model. Prompt 1 from this article. Get the cap table and dilution for each shareholder.
-
Test scenarios. Prompt 2 — how does the picture change with different valuations. Prompt 4 — the impact of option pool size on dilution.
-
Account for hidden factors. SAFE conversions (prompt 3), liquidation preferences, anti-dilution.
-
Validate. Check calculations against the three control points. Cross-verify critical numbers manually or with a second AI query.
Unit economics shows whether the business is profitable at the customer level. Cap table modeling shows how much of that profitability stays with the founders after raising capital. Both calculations complement each other and together give you the full financial picture of a startup.
FAQ
How much dilution is normal after 3 funding rounds? Founders typically retain 15–25% each after Pre-Seed, Seed, and Series A. The exact number depends on valuations, investment sizes, and option pool requirements. In the model above, each co-founder went from 50% to ~20% across three rounds.
Does the option pool come out of the founders’ share or the investor’s? Almost always the founders’. Investors require the option pool to be carved from pre-money valuation, which means it dilutes existing shareholders before the new investor enters. This is called the option pool shuffle.
Can AI accurately model a cap table? AI handles the arithmetic and scenario modeling well, but it can make errors in multi-step calculations — especially with SAFE conversions and option pool expansions. Always validate outputs using the three checks described in this article: 100% ownership sum, investor share = investment / post-money, and consistent price per share.
What’s the difference between a SAFE and a convertible note for dilution? Both convert to equity at a future round, but a convertible note accrues interest and has a maturity date. A SAFE has no interest and no maturity. For dilution modeling, the key variables are the valuation cap and the discount rate — these determine how many shares the holder gets at conversion.
When should I switch from spreadsheet modeling to a dedicated cap table tool? Once you close your first priced round. Before that, AI + spreadsheets are enough for scenario modeling. After the round closes, you need a legally accurate record — tools like Carta or Pulley handle vesting schedules, 409A valuations, and compliance that spreadsheets can’t.