XP Inc.: The Cheapest Number on Our List Is Wrong Twice

There is a reflex that reliably costs investors money, and it looks harmless: you see a small number and read it as a price. A price-to-earnings ratio of 4. A price-to-book of 0.5. Or, as here, a price-to-free-cash-flow ratio of 0.6 — meaning the entire market value would be paid back out of current cash flow in seven months. Your head does the arithmetic and your gut shouts: the market must have made a mistake.

Sometimes it has. And sometimes the number itself is simply wrong. With XP Inc. (NASDAQ: XP) it is the latter — twice over. So we are going to do something we rarely do: take apart our own metric before we talk about the company. By the end you will know what is left of that 0.6, and you will be able to reproduce every step.

What XP actually does

XP is the largest independent investment platform in Brazil. Before XP, securities business there ran almost entirely through five large banks selling their own products. XP built an open marketplace alongside them: independent advisors bring in clients, the platform executes the trades, and it earns money on commissions, on the interest spread and on asset management.

The scale as of December 31, 2025: 4.76 million active clients, R$1,491 billion in client assets under custody, roughly 18,000 affiliated advisors and 8,069 employees of its own, based in São Paulo, Rio de Janeiro, New York and Miami. On top of that a loan book of R$35 billion, a card business with R$53 billion in transaction volume and retirement plan balances of R$95 billion.

One detail that matters in a moment: XP is legally a Cayman Islands company, listed in New York and priced in U.S. dollars — and it reports in Brazilian reais. The share costs dollars; the business earns reais. The annual report says so explicitly:

Where the stock came across our desk

Every day we run about 3,500 stocks through our in-house stock scanner. One of those lists is the price-to-free-cash-flow ranking: it sorts companies by how many times the market capitalization is contained in free cash flow, cheapest first.

The finding, measured on July 27, 2026: 545 stocks in the U.S. selection meet the criteria. XP sits at a displayed ratio of 0.6 in rank 10 — among the ten apparently cheapest names on the entire list, visible on the first page. Measured identically on both of our brands. These lists are recalculated daily.

That rank 10 is the reason for this analysis. Because it does not survive being recalculated.

Uncomfortable truth no. 1: the metric divides dollars by reais

A price-to-free-cash-flow ratio is a fraction. Market capitalization on top, free cash flow underneath. For the result to mean anything, both figures have to be in the same currency — otherwise you are measuring apples in pears.

With XP that is exactly what happened. Recalculated with the inputs behind our list:

• Numerator: market capitalization of about $8.29 billion (data as of July 27, 2026).
• Denominator: free cash flow for the four quarters to March 31, 2026, of R$15,049 million.
• Result: 8.29 divided by 15.05 gives 0.55 — displayed as 0.6.

Those two figures are in different currencies. The annual report supplies the exchange rate itself — the commercial selling rate published by the Brazilian Central Bank for December 31, 2025:

„Unless otherwise indicated, we have translated real amounts into U.S. dollars using a rate of R$5.502 to US$1.00, the commercial selling rate for U.S. dollars as of December 31, 2025 as reported by the Central Bank.“

— XP Inc., SEC annual report on Form 20-F for 2025, Item 5.A

Now the calculation in one currency. Convert the numerator into reais: $8.29 billion times 5.502 is roughly R$45.6 billion. Divided by cash flow of R$15.05 billion, that gives:

A price-to-free-cash-flow ratio of about 3.0 — not 0.6.

The check in the other direction lands in the same place: convert the cash flow into dollars (R$15.05 billion divided by 5.502 is about $2.74 billion) and divide the market capitalization by it, and you again get roughly 3.0. The displayed figure is therefore too low by a factor of 5.5 — precisely the exchange rate.

What that does to the rank: rank 10 is an artifact. At a ratio around 3.0, XP would sit well behind the 25 rows the detail page shows at all. Anyone opening the list and seeing XP near the top is not looking at a bargain, but at a conversion error.

And this is not a one-off. We found the same break in a second Brazilian name: the StoneCo analysis carries a price-to-cash-flow ratio of 1.1 that likewise sets a dollar market capitalization against a real-denominated cash flow. The pattern hits foreign issuers listed in New York that do not report in dollars. How many names are affected in total we have not yet measured — that it is more than one, we now know.

Uncomfortable truth no. 2: the money flowing through belongs to clients

Suppose the currency error were not there — would a ratio of 3.0 then be a buying argument? Still no. Because at an investment platform, operating cash flow measures something different from what it measures at an industrial company.

At a machinery maker, operating cash flow is what remains after wages, materials and taxes: money the company can dispose of. At XP, client money runs through that same line — securities balances, cash sitting in settlement accounts, retirement plan balances. When those balances rise it looks like cash inflow; when they fall, like outflow. None of it is distributable, because none of it belongs to XP.

You do not have to infer this from the balance sheet — XP writes it into the annual report:

„These balances may fluctuate substantially from quarter to quarter and were the key drivers to the net cash flow from operating activities figures.“

— XP Inc., SEC annual report on Form 20-F for 2025, Item 5.B

How violently those balances swing shows up quarter by quarter: in the first quarter of 2025 operating cash flow was negative at minus R$2,634 million; in the third quarter of 2025 it was positive at R$12,307 million; in the fourth quarter it turned negative again. A metric whose denominator changes sign is not a valuation anchor.

We have met this pattern twice before. At OneMain Holdings the operating cash flow contains the interest margin before credit losses; at Fidelity National Financial it is policyholder deposits. The underlying rule is always the same: where the core business sits in investing activities, operating cash flow measures throughput, not profit.

The numbers over the years — fairly credited

After two sections criticizing our own metric, a correction is due: XP is a profitable, growing company. That is the part the broken number actually obscures.

Revenue rose 8 percent in 2025 to R$18,399 million, net income 14 percent to R$5,169 million. The net margin stands at 28.1 percent — a strong figure for a financial services group. Earnings per share came to R$9.80.

Revenue in 2025 splits as follows: retail R$14,584 million, corporate and issuer services R$2,733 million, institutional R$1,373 million, other R$743 million. The business clearly stands on its retail leg.

But the momentum is fading. Net inflow of new client money fell from R$104.9 billion in 2023 and R$108.8 billion in 2024 to R$94.3 billion in 2025. And the take rate — the share XP earns on assets under custody — slipped from 1.29 to 1.25 percent. Client assets grew about 16 percent, revenue only 8. That gap is the real question for the coming years.

Uncomfortable truth no. 3: ten votes per share

XP has two classes of shares. What trades on Nasdaq are the Class A shares — one share, one vote. Alongside them sit Class B shares, which do not trade and carry ten votes each.

Directors and executive officers — 13 people — together hold all Class B shares and therefore 71.09 percent of the voting power. They are bundled in a vehicle called XP Control LLC.

That is not automatically bad: founder-controlled companies can sustain uncomfortable investments that a quarter-driven shareholder meeting would strangle. But it shifts the question of power. Buying Class A stock buys a share of the profits, not influence. On a takeover offer, a capital increase or a dispute about the dividend, the founder group decides alone.

Valuation: what you get for your money

For orders of magnitude — all figures as of July 24 to 27, 2026, no daily prices: the market capitalization is around $8.3 billion. Equity stood at R$23,548 million on December 31, 2025, roughly $4.3 billion. That puts the price-to-book ratio in the region of 1.9. The price-to-earnings ratio is around 8.5.

On the balance sheet. Total assets at year end were R$396,528 million against equity of R$23,548 million. The equity ratio is therefore 5.9 percent. That sounds thin, but for a broker and banking group it is normal: most of the balance sheet consists of client positions passing through, not proprietary bets.

Why we quote no Altman Z here. The Altman Z score estimates distress risk and was developed for industrial companies. It sets working capital, retained earnings and sales against total assets — quantities that mean something entirely different at a bank. Our data set carries a value of 2.72 for XP, which on the relevant scale would sit between the 1.1 and 2.6 thresholds. We deliberately do not use it, because it says nothing about this business model. Anyone judging XP's stability should look at the equity ratio, client assets, net inflow, and the question of how stable the take rate stays.

Opportunities and risks at a glance

What speaks for XP:

• A profitable, growing core business: net income of R$5,169 million in 2025, a net margin of 28.1 percent.
• Market position as Brazil's largest independent platform, with R$1,491 billion in client assets and 4.76 million active clients (December 31, 2025).
• A broad revenue base across commissions, net interest income, asset management, cards and lending — not dependent on a single fee.
• A valuation that is not expensive even on a single-currency basis: a price-to-earnings ratio around 8.5 with double-digit earnings growth.

What speaks against it:

• The hook does not hold: the displayed price-to-cash-flow ratio of 0.6 is a currency error; calculated cleanly it is about 3.0.
• Even adjusted, the metric measures client money rather than distributable surplus — the denominator changed sign twice during 2025.
• Fading momentum: net inflow down from R$108.8 billion to R$94.3 billion, take rate down from 1.29 to 1.25 percent.
• No say: 71.09 percent of the voting power sits with 13 people, and Class B shares carry ten votes each.
• Currency and country risk: the business earns reais while the share is quoted in dollars. A weaker real depresses any dollar result regardless of operating performance.

A human conclusion

We started with the reflex: a small number read as a price. We deliberately demonstrated it on our own list, because that is where it is most dangerous — on a figure you trust because it came out of your own tool.

What is left once the 0.6 is cleared away? A solid Brazilian financial company with nearly R$5.2 billion in annual profit, a strong market position, slowing growth and an ownership structure that gives outsiders no voice. That is neither a bargain nor a scandal — it is an ordinary company at an ordinary price.

The real lesson, though, is not about XP but about the number in front of it: with any foreign-listed name, check that numerator and denominator speak the same currency. We have now found the same break in two Brazilian stocks. They are unlikely to be the only ones.

What you make of it is your decision. And that is exactly as it should be.

Sources

• XP Inc., SEC annual report on Form 20-F for 2025 (filed April 29, 2026) — Item 5.A exchange rate, Item 5.B cash flow, Item 6.D employees, Item 7.A voting rights, Note 2 reporting currency
• Interim reports on Form 6-K filed May 6, 18, 26 and 29, 2026
• Ownership filings SCHEDULE 13G/A (May 14, 2026) and SCHEDULE 13D/A (July 7, 2026), prospectus supplement 424B7 (July 2, 2026)
• Source: fundamental data & SEC filings (20-F/6-K), data as of July 24 to 27, 2026
• In-house stock scanner, price-to-free-cash-flow ranking, U.S. selection, measured July 27, 2026

This analysis is journalism and editorial context, not investment advice. It contains no buy or sell recommendation and is not a solicitation to buy or sell securities. Equity investments carry the risk of total loss. All figures come from the sources named above and carry the stated cut-off dates; metrics drawn from data sets can contain errors — this article demonstrates exactly that on one example. The author holds no position in the stock discussed at the time of publication.