A cap table lists who owns what in a company. It shows every share class, every price paid and every vote. Private companies keep it out of sight, and the one moment a large fintech's cap table becomes public is the IPO prospectus, so I went looking for one that shows how far private prices can drift from public ones.
Chime's final prospectus is the clearest case I found. It is dated 11 June 2025. It sold 32,000,000 Class A shares at $27.00 each, raising $864m before fees, and it printed the full preferred stock history in the notes to its accounts.
I read the notes and the ownership tables twice. The gap between what the last private investors paid and what the IPO priced is wide enough to explain why anyone buying into a late-stage fintech should ask for the cap table before the valuation.
The research I keep beside any cap table is by Will Gornall and Ilya Strebulaev. It first appeared as NBER working paper 23895. It was later published in the Journal of Financial Economics.
They valued 135 US unicorns from the legal terms in their filings, and their finding is stark: "reported unicorn post-money valuation average 50% above fair value, with 15 being more than 100% above". The cause sits in the cap table. A headline valuation multiplies the newest preferred share price by every share outstanding, as though common stock were worth the same as the most protected paper.
Common stock is worth less. The paper lists what late investors usually get: "IPO return guarantees (14%), vetoes over down-IPOs (24%), or seniority to all other investors (32%)". After adjusting for those terms, "almost one-half (65 out of 135) of unicorns lose their unicorn status". I find that number hard to read past.
Chime's Note 12 lists ten series of redeemable convertible preferred stock. Series Seed was issued at $0.172 a share. Series A came at $0.272. Series D was priced at $5.2247, Series E at $17.2618 and Series F at $40.9208.
Series G was the last round. Its original issue price was $69.0698 per share, on 15,998,298 shares, with an aggregate liquidation preference of about $1.105bn. Across all ten series Chime had 258,464,156 preferred shares carrying $2.89bn of preference, ranked "on a pari passu basis" with each other and ahead of common stock.
That ranking explains a balance sheet line I nearly skipped. On 31 March 2025 Chime carried $2.89bn of preferred stock outside equity and reported a total stockholders' deficit of $1.91bn. The pro forma column, after conversion, shows positive equity of $689m. Nothing in the business had changed. Only the cap table had.
CNBC reported on 13 August 2021 that Chime "raised $750 million in a Series G funding round that values the fintech company at $25 billion". PitchBook, cited in the same piece, put the previous valuation at $15bn in October 2020.
Here is something I cannot explain. The prospectus shows $1.105bn of Series G preference, not $750m. Perhaps the round had later closings. Perhaps some Series G shares were issued in another deal. I have not found the answer in the pages I read, so I am leaving the gap open rather than guessing.
Put the IPO price next to the Series G price. At $27, a Series G share was worth 39% of its cost. That is a paper loss of about 61%. I multiplied the 15,998,298 Series G shares by $27 and got roughly $432m against a $1.105bn preference.
Series F holders lost about a third. Series E holders gained about 56%. Seed investors who paid $0.172 made roughly 157 times their money, on paper, on the same day and at the same price.
A headline valuation hides that spread. My own estimate of the IPO valuation is crude, and I offer it only as a rough check: the prospectus says 332,239,249 Class A and 32,182,289 Class B shares would be outstanding after the offering, which at $27 comes to about $9.84bn before options and restricted stock units, against the $25bn of 2021.
This clause surprised me. Chime's preferred stock converted automatically on an IPO "with not less than $17.26180 per share" and net proceeds above $200m. That is the Series E price. It is not the Series G price.
So the $27 IPO cleared the bar easily. Every series converted into common stock immediately before closing, producing 258,667,836 Class A shares, and $2.89bn of liquidation preference vanished from the balance sheet in one line.
The Gornall and Strebulaev sample shows about a quarter of unicorns gave late investors a veto over a down IPO. My guess is that Chime's Series G holders did not have one, because the conversion bar in Note 12 sits far below their price. That is a guess. A protection could sit elsewhere in the charter or in a side letter that I have not seen.
The other half of Chime's cap table is about control. Class B shares carry 20 votes each. After the offering, the cover page says, chief executive Christopher Britt "will hold approximately 34.7% of the voting power" and co-founder Ryan King "approximately 31.3%".
I did the share count myself. The 32,182,289 Class B shares are about 8.8% of the 364.4 million shares outstanding. They carry about two thirds of the votes.
The founders' share of the vote can still grow. The prospectus says Chime "may issue up to 16,345,526 shares of Class B common stock" in exchange for Class A shares under equity award exchange agreements, and that if all the founders' awards had been exercised or settled at the offering, Britt and King "would hold approximately 40.1% and 34.4%" of the voting power. Control was built to rise after the listing. It was not built to fade.
DST Global shows the other side of that trade. Entities affiliated with DST owned 52,268,715 shares, which the table puts at 15.7% of the Class A stock after the offering and only 5.4% of total voting power. Crosslink Capital held 8.8% of Class A and 3.0% of the votes. General Atlantic held 5.8% and 2.0%.
A short aside, and it has nothing to do with price. Before the IPO, Series A holders elected one director nominated by Crosslink, and Series C holders elected one nominated by Menlo Ventures. The prospectus says the voting agreement behind that "will terminate" with the offering. The old cap table chose the board; the new one hands control to two founders. Back to the money.
If I were valuing a late-stage fintech today, I would ask for the full cap table first. I would want each series' price. I would want its preference, its conversion trigger and its votes, because Chime's file shows one company, one IPO price and outcomes running from a 61% loss to a 157-fold gain.
The paper says common shares in its sample were 58% overvalued on average. Chime's earliest holders did far better than that warning suggests. I do not know whether the Series G investors had any protection outside the charter, and it still bothers me that a group paying $69.07 a share signed up to a conversion bar of $17.26.