A Cap Table Is the Only Document That Says Who Owns What

What It Records

a capitalization table It lists all the shareholders of a company what kind of security they own how many units and what percentage it represents

For an early-stage company this is a spreadsheet. For a multi-round company it involves multiple preferred classes with different rights a pool of options with grants at various exercise prices and vesting statuses warrants and convertible instruments that have not yet been converted

The reason it carries so much weight is that it's the only place where ownership is stated in a single view. The underlying legal authority is found elsewhere spread across stock certificates board consents grant agreements and financial documents signed years apart. The table is the summary of it all and when the summary and the documents disagree the documents win

Issued Versus Fully Diluted

The most common misinterpretation is the difference between these two counts

Shares issued They are the ones that really stand out today. Fully diluted shares It includes everything that could be converted into a share: unexercised options the pool of unassigned options warrants and convertible instruments on a converted basis

The gap can be large. A company with 8 million shares issued a pool of stock options of 2 million and convertible notes that convert into another million has 11 million fully diluted. A holder of 800,000 shares owns 10 percent of the issued shares and about 7.3 percent fully diluted

There is one complication worth knowing before trusting the phrase. Fully diluted is not a standardized term with an agreed upon meaning. Some tables include the unallocated portion of the option pool and others only count the options granted. Some include warrants and convertibles on a converted basis some reserve instruments that are far out of the money and others cite a figure that quietly excludes a round that is still being traded. So two parties can each say fully diluted goodfaith and dividing by different denominators. The practical custom is to ask what's in the number rather than accepting the label especially when comparing one offer with a previous one

Note which number is the honest one. Each share in that fully diluted count exists or is a commitment the company has already made so it is not possible to evaluate the dilution. It is agreed upon and waiting to appear. To quote the percentage issued is to quote a figure that is already known to be wrong

The Option Pool Trap

Investors typically require an appropriately sized option pool for future hires and trade to create it. pre money that is before your investment is counted

The consequence is that all pool dilution falls on existing shareholders rather than being shared with the incoming investor. A round with a pre-money valuation of 20 million with a 15 percent pool created pre-money is economically a lower valuation than the headline suggests

This is sometimes called option pool shuffling. It's standard practice rather than a gimmick and founders who don't model it are agreeing to a different price than they think they negotiated

What the Pre Money Pool Actually Costs

That last sentence deserves a number because the gap between the headline and reality is greater than most founders expect. The figures are illustrative and round

Conduct the round described above. The company is valued at 20 million before the money and the investor puts up 5 million so the valuation after the money is 25 million and the investor owns 5 divided by 25 which is 20 percent

Now add the pool. The investor requires an option pool equal to 15 percent of the company after the round and requires it to be created before the money. Fifteen percent of the company's 25 million post-money is worth 3.75 million

Ask who paid that 3.75 million. Not the investor because the fund was created before his money was counted so his 20 percent is measured after the fund already exists. It came entirely from existing shareholders

So the effective pre-money valuation for the people who already owned the company is 20 million minus 3.75 million which is 16.25 million. The headline said 20. The price actually paid was 16.25 a reduction of 18.75 percent

There is nothing hidden here and no one is behaving incorrectly. The term sheet says what it says. The point is that the pre-money number in the headline is not the number that determines what the founders give up and two deals with identical headline valuations can differ substantially once you read the size of the fund and its location

The consequence of the negotiation follows directly. Arguing that the pre-money valuation increases by a million and at the same time accepting a larger fund can leave a founder worse off than accepting the lower title with a smaller fund. Founders should size the pool based on an actual hiring plan rather than accepting a round number because each point in the pool created before the money is a price point

What a Complete Table Shows

A cap table that shows only percentages without class rights and preference terms is not sufficient information to value a share. Two people with identical percentages in different classes may receive very different amounts

Why Identical Percentages Are Not Equal

That final sentence is the one worth demonstrating because a percentage feels like a complete answer and it isn't. Illustrative figures again

Suppose a company sells for $10 million. Previously an investor invested $8 million in preferred stock that had a unique non-participating preference giving them the right at exit to get their money back first or convert it to common and take their percentage whichever pays more. On a converted basis that preference represents 20 percent of the company and common shareholders own the other 80 percent

The preferred holder compares the two options. Taking the preference returns the 8 million invested. The conversion to common returns 20 percent of 10 million which is 2 million. They take the preference

That leaves 2 million for everyone else divided among the 80 percent who are considered common

Now compare like with like.Each percentage point of the common receives 2 million divided by 80 which is $25,000.Each percentage point of preferred receives 8 million divided by 20 which is $400,000

The same one percent of the same company in the same outlet is worth sixteen times more in one class than another. None of the holders are confused about what part of the company they own. The percentage was never what determined the payout

This is why a table listing names and percentages without the preference class and terms is almost useless for valuing a position. It is also why the preference stack becomes more important as the rounds build up. Each round adds another block of money that must be returned before the common sees anything and in a modest outing that stack can eat up the entire price

Where They Go Wrong

Cap table errors are common and costly. Option grants approved by the board but never documented. Advisor shares were promised in an email and never issued. Convertible notes with terms no one modeled. Vesting schedules that don't match employment agreements. Departing employees whose exercise windows were handled inconsistently

These arise during the diligence of a financing or an acquisition at a time when they are most disruptive and least fixable. Clearing up a disputed grant during a sales process is a deal the company will lose

Professional cap table management software exists largely because spreadsheets do not enforce consistency between the general ledger and the underlying legal documents

Why the Timing of Discovery Is So Punishing

The observation that the company loses that negotiation is worth explaining because it is a question of position rather than who is right

Consider when the mistake arises. The company has agreed to terms informed its staff and board and is spending money on lawyers ahead of a signing date. The buyer or incoming investor has done none of those things to nearly the same degree. Whatever the legal merits of a disputed grant the two sides have wildly different costs of walking away and everyone in the room knows it

The mechanics of how these problems are resolved reinforce this. A company signing a financing or sale represents that the cap table is accurate and complete. A known defect cannot simply be left alone as that would make that representation false. Therefore it must be fixed disclosed or covered up

Fixing it means going to the person with the disputed claim and asking them to sign something at the precise moment when their influence is greatest because everyone needs their signature by Friday

Covering it means a larger indemnity or escrow that is money withheld from the product against the possibility that the claim is real. That comes from the sellers which in practice means the founders and common shareholders

Neither of those outcomes costs anything if the same problem is discovered eighteen months earlier when the affected person is a regular employee having a regular conversation and no agreement depends on the response. The expense is generated almost entirely by timing which is the argument for ongoing reconciliation rather than pre-diligence cleanup

What Founders Should Track

Model the fully diluted position after each planned round including the pool expansion each round will require. Founders consistently underestimate cumulative dilution because they model one round at a time

Understand the preference stack as it builds up as it determines whether the common equity has value at a given exit

And keep the legal documents and the table reconciled continually and not diligently because reconstruction after the fact is where disputes come from

How Cumulative Dilution Actually Compounds

It is worth showing the underestimation because the error is structural rather than careless. Illustrative and round

Suppose a founder expects three rounds each selling 20 percent of the company to new investors. Thinking additively they subtract 20 three times and hope to retain 40 percent

Dilution does not subtract it multiplies because each round takes its share of what is left. Retaining 80 percent three times gives 0.8 times 0.8 times 0.8 which is 51.2 percent. That error works in favor of the founder and it is not what hurts him

Now add the pools which is where the real gap opens up. Each round usually comes with a pool upgrade so let's say each round also creates or reloads a pool that costs 10 percent. Survival per round is then 80 percent of 90 percent which is 72 percent

Over three rounds that's 0.72 x 0.72 x 0.72 or 37.3 percent

Compare that to the 51.2 percent a founder gets simply by modeling investor dilution. Pool upgrades cost the company almost 14 percentage points and are the part most often left out of the model because a pool feels like an internal administrative matter rather than an equity sale

It's a share sale. The shares go to employees and not investors and are diluted in exactly the same way. Modeling one round at a time hides this as each individual top-up in the pool seems small and the combination between rounds is where it gets big

The Bottom Line

The cap table records who owns what across classes options and convertibles and only the fully diluted view describes actual ownership. Pre-money option pools pass dilution on to existing holders and effectively reduce the price paid. Errors quietly accumulate and come to the surface during diligence which is the worst time to discover that a grant was never properly issued. And a percentage alone is not an answer because the class associated with it decides whether that percentage is worth anything